Hacker News new | past | comments | ask | show | jobs | submit login
Maybe today’s Navy is just not very good at driving ships (militarytimes.com)
287 points by smacktoward on Aug 28, 2017 | hide | past | web | favorite | 342 comments

  “We do not put a premium on being good mariners,” Hoffman
  said. “We put a premium on being good inspection takers
  and admin weenies.”
This is a problem across the DoD, not just the Navy. I've seen it first hand in both military and civilian sides of the USAF. It's comparable to the degradation of the US public schools with extensive testing. The tests that start as attempts to evaluate and quantify effectiveness become the goals themselves, rather than measures of the true goals.

I think it's a consequence of taking away individual judgement and becoming, it seems, terrified of letting a person make a decision, because sometimes they're bad at it, or have biases, or whatever. Organizations are afraid of individuals deciding things, and individuals are eager to push their decisions onto a "process" or "policy" because if they actually make a decision and anything goes wrong—whether or not the decision was a good one given the information they had—they're toast, fired if they're lucky, publicly demonized and blackballed and fined and jailed if they're not. The individual cost of a decision that is bad—even if it's only bad in hindsight—is too often way out of proportion to the harm it caused, so why risk it?

I don't know why it's happening, but it sure explains a lot of things across much of US society. Politics, business, education, parenting. I wouldn't be surprised if it's part of what's behind the frightening US "cost disease" phenomenon from the Slate Star Codex post that was discussed here a few months back.

I'm only speculating, but I think social media has only exacerbated this problem. Now every screw-up an employee makes (or a teacher, or an administrator, or whatever) is magnified times ten to everyone on the aggrieved party's Facebook feed. Have a teacher that said something a parent took offense to? An employee who misplaced an item you had ordered? An employee can always fall back on the policy and push the blame to the institution and the institution can say "we'll review our policies".

Funny thing is that as a bureaucrat, coming up with policies is my job. It tends to improve overall outcomes and satisfaction, but the same outliers described above are also magnified ten-fold (this policy is stupid because ...).

My best guess is that it's the result of 20-30 years of never accepting "shit happens, we looked and everything seems basically fine, we're not changing anything" as an answer when something bad happens and someone asks "so what are you doing to make sure this doesn't happen again?"—even when that answer's actually appropriate. The consequences are so bad if you make the do-nothing call and get it wrong (or are just unlucky) that everyone does a whole bunch of stuff to ensure they can say "look at all the things I did!" if lightning strikes twice (or if they were simply wrong and something did need to be done, but it wasn't clear until a pattern emerged)

Instead we always freak out and add a bunch of new policies/guidelines, some more mandatory educational videos or classes, and maybe bring down the hammer on some people as well.

[EDIT] in fact it may be yet another symptom of post-Nixon-era cynicism. We expect institutions to lie to us about everything (which, to be fair, they often do and often have), so we (rather, the media) dig into any problem expecting criminality and neglect and generally bad behavior (and a great story!), so institutions have to go nuts to make sure it never looks like there's anything there worth reporting as maybe-sort-of-kind-of any of those bad things.

The FAA, NTSB, and airlines have made air travel incredibly safe over the years by operating the way you described.

I am by no means an expert on this subject so correct me if I'm wrong but....

The exception that I've noticed from the airline-industry, is that they actually want to solve problems systemically and don't care who has to take the blame.

For something to go wrong, multiple failures have to happen, and there's not a lot of blame put on individuals, as much as "the process".

In most other organizations, IME, fixing "the process" is expensive. It takes time, and effort, and it's quicker and easier to blame "those idiots in IT/accounting/HR/etc". Most of those types of organizations don't actually want to solve problems, they just want the problem to go away. So they end up choosing the easier option, rather than dealing with the problem systematically.

/bitter rant

I am not an expert but I know many pilots, watch only documentaries, read research papers almost exclusively, and have stayed in a Best Western.

NTSB is, as far as I understand, actually very careful about not blaming anyone. They look for causes and solutions, not blame. They, as far as I know, pride themselves on this.

They try, again inasmuch as I know, to not blame a person - but to blame a process. This is not a distinction without difference. It fosters an attitude of cooperation and openness.

Again, I'm not an expert but I know a bunch of pilots and listen when they speak. If the NTSB blamed people, I suspect they'd have a lesser willingness to speak positively about NTSB.

Edit: Someone beat me to it. I will leave this as I think it offers a bit more of a comprehensive view.

> They try, again inasmuch as I know, to not blame a person - but to blame a process. This is not a distinction without difference. It fosters an attitude of cooperation and openness.

Because even when the cause was blatant human error, it can still be a process problem -- how did this type of human error slice through all the protections against it? What process can be put into place to prevent disaster even when some bonehead does that same thing again?

But air travel safety is atypical.

If the crash rate for planes was the same as it is for cars then air travel would have to be prohibited as a necessary measure to prevent the extinction of the human race.

Most endeavors are not of that sort.

And so in most other contexts we could use their methods to produce a process that will prevent a particular category of trouble, but suffering the trouble costs $2M (instead of $2B) and implementing the process costs $20M.

And then people will want to implement it anyway, even though they shouldn't, because "it solves the problem".

Or, seeing the obvious fallacy in spending $20M to prevent $2M in harm, a "compromise" is proposed to spend $1M to prevent 5% of the $2M harm, still with no one doing the math. And then, problem still 95% unsolved, more half measures are kludged on over time until the surrounding bureaucracy becomes politically powerful enough to be self-sustaining.

Because people don't want to admit that some diseases aren't worse than their cures.

> If the crash rate for planes was the same as it is for cars then air travel would have to be prohibited as a necessary measure to prevent the extinction of the human race.

Well, no, this is obviously false. Much more car travel is done than air travel; bringing air fatalities up to the level of car fatalities would have a negligible effect on the human population.

Bringing crashes per vehicle mile up to the level of highway crashes per vehicle mile according to https://www.rita.dot.gov/bts/sites/rita.dot.gov.bts/files/pu..., assuming that an average airplane carries 140 people, and assuming (very wrongly) that crashing it kills all 140, you could get deaths to airplane crashes up to 843,000 per year for the US. This would be significant -- it would be an increase of over 30% in the annual death rate -- but it is less than the existing surplus of annual births over annual deaths. A modest reduction in population growth just isn't going to drive the human race extinct.

You're assuming the existing expense of making air travel safe is not reducing the quantity of air travel done.

Imagine it was as easy to become a pilot as to get a driver's license, there were no flight plans or restrictions on where people could take off or land and a 500MPH jet could be purchased for $150,000 because there would be less regulatory overhead and more competition in aircraft production.

Affluent people would be commuting by jet. There could easily be ten times as many air miles traveled or more.

If you want to "reason" that way, take into account that (1) if people started dying from air travel at high rates, they'd do less of it; and (2) societies with much higher death rates have no problem growing anyway, as they compensate with higher birth rates.

> if people started dying from air travel at high rates, they'd do less of it

That doesn't seem to stop people from driving cars.

> societies with much higher death rates have no problem growing anyway, as they compensate with higher birth rates.

That's just survivorship bias. Why would one cause the other? There were also societies with high death rates and low birth rates which for the obvious reason no longer exist.

> There were also societies with high death rates and low birth rates which for the obvious reason no longer exist.

If you claim to be worried about extinction of the human race, you'll have to consider all societies, not just the ones with pathologically low birth rates. A tiny group of people committing suicide has zero effect on the overall human population.

>They look for causes and solutions, not blame. They, as far as I know, pride themselves on this.

If the cause is a person's behavior, that is the same thing as blame. Look at the report of the Pinnacle 3701. Probable causes from [1]:

1. the pilots' unprofessional behavior, deviation from standard operating procedures, and poor airmanship;

2. the pilots' failure to prepare for an emergency landing in a timely manner, including communicating with air traffic controllers immediately after the emergency about the loss of both engines and the availability of landing sites;

3. the pilots' improper management of the double engine failure checklist

It's hard to look at that and say that the pilots weren't being blamed.

1. https://en.wikipedia.org/wiki/Pinnacle_Airlines_Flight_3701

It does happen, but is the exception. Look at the aftermath section. You can also look at the aftermath sections of other flights.

Sometimes, a human is to blame and there's no getting around that. This doesn't change their goal.

The difference is the approach — they look for the cause and sometimes the cause is the blame.

When you only look for the blame, you tend to be satisfied once you reach that goal and might ignore the cause.

NTSB, i think, no one takes the blame. For example, if the error is something like "Mechanic didn't tighten bolt to appropriate torque."

So, what's wrong? maybe they forgot and need a checklist item? Maybe the reading on the wrench should be recorded in the log? should there be a second person that verifies?

They know they ask a lot of people, and it has to be right, every time. People get sick, tired, distracted, whatever. we're people. People are the least reliable part of the system.

I think blame only falls to a person when they've lied. Falsified a log, claimed to do something they didn't etc.

That's the thing. In the Air Force (at least)- there is no thought of cost/benefit analysis. A single machine fails once because a bolt wasn't torqued and causes $5000 of damage ... So they add the checklist item, and the second person, and an extra training item, and a log record that has to be maintained. The list of special rules and required documentation only ever gets longer. Eventually, you spend all your time on the checklists and maintaining the documentation, and none on actually doing whatever procedure you were originally messing with that bolt for.

As someone who was a mechanic in the Army, I feel that checklists more often helped then harmed. Many mechanical problems we saw in the shop were easily attributable to an operator skipping a step on their PMCS. Doing something relatively dangerous or confusing could be error prone, especially for people who haven't done it much, and a checklist can help ensure the safety of everyone in the shop. A good example is running an engine indoors. Before you did it, there were a serious of steps you had to walk through before you turned it on, that reduced people getting sick from carbon monoxide quite a bit.

"The Checklist Manifesto" is an -excellent- read on the helpfulness of checklists in preventing problems and mistakes (primarily from a surgeon and medical viewpoint but he covers airlines as well.)


As a submarine fire control tech, I always thought it was cumbersome to have three full size binders open in my lap during weapon firing drills, casualty drills, everything except calm open water transit. But flipping those pages and grease pencilling things even while in the hot of prepping weapons to leave the tube avoided a LOT of "hurry up and whoops!" mistakes. Sometimes crufty stuff needs a review and rewrite but, n the whole, I'd rather know what happen wrong the previous times and avoid it myself.

This kinda contradicts the idea further up that too much attention is paid to process, at the detriment of, for lack of a better term, "skill" (or "personal responsibility")

I think the focus on process can be terribly annoying, because it's so much easier to say "Bob fell asleep at the machine" than diving into all the events that led up to Bob falling asleep. Note that both can be true or, more accurately, the process problems actually need the statement "bob fell asleep" to be true to make sense.

The idea to prioritise processes is that it's repeatable. Individual behaviour isn't within your control, except for the processes that hired somebody, educated them, defined their tasks, and created the environment in which they perform these tasks.

In a small team, or family, or group of friends, it's perfectly fine to dock Bob's allowance for falling asleep. And the instinct to exact revenge, or punish people is strong, because those instincts are the result of evolution and norms of society, both of which until recently only knew small groups where everyone knew everybody. That's why it takes a lot of discipline to act against these instincts, and obviously, as you said, because focussing on the process may also uncover others who failed, and it is more tedious.

People get frustrated with bureaucracy. Sometimes they're obviously right. But the idea that all bureaucracy is wasteful, or (equally common) that governments have any particular talent for wasteful bureaucracy is somewhat misguided. Because there aren't many examples for organising teams of sometimes hundreds of thousands of people to work on shared goals without a wasteful bureaucracy, and governments and the private sector tend not to differ that much.

I've seen one interesting example of a different structure for organising a large number of people: capitalism. It arguably works, being excellent at delegating authority to exactly the right person for any possible decision. But I believe even that example isn't convincing, if you start considering all the "waste" that capitalism hides in plain sight: advertisement, competing efforts doubling work, the complete financial sector, non-monetary losses due to financial hardship, the losses from unemployment being essentially wasted potential etc.

But unlike the NTSB, the military still routinely blames "pilot error" to cover up problems with terrible aircraft designs.


Not accepting "shit happens" makes sense in aviation, and you're right about that

But taking out decision control and relying on automation has had its problems.

I think it's less about outside regulators than organizations/individuals optimizing to cover their own asses. I assure you, I'm not in the "gov'ment regulators ruin everything" camp.

> I think it's less about outside regulators than organizations/individuals optimizing to cover their own asses. I assure you, I'm not in the "gov'ment regulators ruin everything" camp.

The problem being that the ass-covering is a predictable response to known stimuli.

What do you propose to do to get them to stop doing that?

I believe you're being slightly unfair to those institutions. As one example: the US Government, as a policy, never lies to the American press (and, by extension, the people). They may sometimes do so to the foreign press, or they'll deny comment, or they'll give you an answer that doesn't actually answer your question. But there used to be a line they dared not to cross.

There are many reasons for this. Some will insist it's the morally right thing to do. But a better explanation is usually that it's in no one's interest: Career bureaucrats are simply never in a position where lying is a good option unless they're trying to cover up a crime they committed. They're risking their job, and possibly criminal prosecution, without any real potential upside. For elected officials, the same is usually true–with the potential exception of a lame-duck President who has nowhere to go but retirement. But at that point, they've probably gotten used to managing their office without lies.

Of course that's a description of the past, and the lower reaches of elected executive office (but not employees) have a worse track record, where many a town mayor with a coke habit has lied to cover his creative accounting. Nixon also comes to mind. But before you say "Iraqi WMDs", I'd say that was arguably a case of a government convincing themselves through wishful thinking.

Maybe they're just really good at lying? It's the old problem: is the revelation of a scandal proof of spreading corruption, or does it show that the system is working, and the bad apples are caught? It really seems to come down to personal judgement. I can only say that having interacted with governments on multiple levels, and seen a lot of, for example, the vast amounts of data they produce and decisions they make that are never the subject of the news, I have come away with the impression that the Government is usually trustworthy, and that decisions are often made with a surprising lack of emotions. For example, I once won a lawsuit against a local government. Yet a few months later, we won a tender that we deserved to win, even though the decision was made by the very same people we had previously fought in court.

If all else fails: even if government were corrupt to its core, saying so would be less effective than acting as if it weren't, and focussing on specific incidents. Because faced with a public that believes only the worst of every politician removes all incentives for them not to become exactly that sort of thoroughly corrupt politician. And if it's part of society's lore that all powerful people lie, that becomes a self-fulfilling prophecy when people new to government just follow the clues they got from popular culture.

> the US Government, as a policy, never lies to the American press

I'm sorry.. what?

If you speak any foreign languages try watching government-controlled media in 3rd world countries.

They just... make stuff up. Stuff that can be easily checked via a quick Google search.

Look at Russia's state-controlled TV, the current leader in this.

I agree other countries are worse but the US government makes things up and lies constantly. Obama wasn't perfect, but I don't think it's partisan to point out how Trump's admin is utterly blatant in this, comically so.

> As one example: the US Government, as a policy, never lies to the American press (and, by extension, the people).

Is this some kind of advanced sarcasm?

If you accept that the US Government didn't lie just told the least untrue thing it could then it fits. It just turn outs the former is identical to the latter in most cases.

I think you meant to say "US Government, as a policy, never tells the truth". If they did, we the people would have thrown them out on their asses before their terms are up.

Yeah, I think the big problem is trying to come up with the one big issue and change that, as if that will not have a whole bunch of issues.

People want one answer on what the problem is, because then the most amount of people can understand it in a tweet. They don't want complicated solutions. You say it yourself, it is one problem.

I think you're observing Pournelle's Iron Law of Bureaucracy, which I guess could be stated pithily as "The goal of every bureaucracy is to become self-sustaining. Those that don't are replaced by those that do":


It's interesting that another active comment thread on the HN front page today is about evolutionary game theory & the selfish gene. I'd posit that there's a whole class of "evolutionary systems", subject to the axioms of 1.) large numbers 2.) random variation 3.) selection bias and 4.) competitive pressure. And any such system will always tend toward survival and maintenance of the status quo as its primary imperative, for the simple, tautological reason that those organizations that don't are replaced by those that do. This operates on multiple levels, from the molecule to the gene to the cell to the individual to the firm to the bureaucracy to the society as a whole. Oftentimes, behavior that makes no sense on one level seems obvious if you consider the "actor" to be the level above or below what you're looking at: the gene instead of the individual, or the firm instead of the individual, or the society instead of the firm.

In your example, for instance, the primary actor whose survival that the bureaucracy optimizes for isn't the individual: it's the bureaucracy itself. Individuals who serve their own interests instead of the bureaucracy are rejected and fired, as are individuals who serve the other individuals that the bureaucracy supposedly benefits. Why? Because bureaucracies without this incentive structure are replaced by bureaucracies with it.

And there actually are plenty of areas within U.S. society where bureaucracies are not the primary actors. The individual is still valued over the firm or bureaucracy within therapy culture, artist communities, fandom, Internet forums, dinner parties, and many other social contexts. But because there is no central bureaucracy or organization to point to, these niches fall off our radar screen; niches where individual humans remain dominant are by definition areas without enough resources to support a bureaucracy.

Baumol's cost disease is different, and IMHO a more short-term ailment that's afflicting developed economies as they transition to the information age. This phenomena of growing ever more complex systems on top of individuals is pervasive, and will continue to happen as long as the population keeps growing.

The trouble with that analysis is that "self-sustaining" doesn't inherently imply large or bureaucratic or inefficient. It's possible for a large inefficient bureaucracy to be replaced by a more efficient self-sustaining system with less centralization.

The bureaucracy can fight but it can also lose. Blockbuster Video is no longer a thing that exists because of market pressure. Standard Oil was broken up because of political pressure. There have existed countries that, following the conclusion of a major war, actually disbanded the majority of their military forces, both voluntarily and due to external force. Countries have revolutions where the revolutionaries win and then exile or execute the previous ruling bureaucrats.

In theory the same thing can happen through the democratic process, although this is obviously less common because bureaucracies have a solid understanding of how to subvert electoral opposition.

Your examples are of bureaucracies destroyed from the outside, by stronger competitors for the same resources. It's quite impossible to curb a bureaucracy from the inside, e.g. through "democratic process". The system will not fix the problems with the system.

Perhaps, but so don't do that.

If the wasteful bureaucracy will not commit suicide then help some less wasteful one to commit fratricide.

Just need a competing Navy to conquer the US. Easy as pie.

If they have container ships, I wouldn't bet against them...

I suppose it's like cows that evolve ever larger horns to compete for mates, then one day a new predator comes in and kills all the cows because their horns are too heavy and they can't run fast enough.

I think that's slightly different, an "outside of context" problem, as Banks would say. The cows didn't leave enough flexibility to adapt to the unexpected, rather than just being inefficient.

> I think you're observing Pournelle's Iron Law of Bureaucracy

This sounds like a special case of the more general law of survival of the fittest (where 'fittest' means 'most self-sustaining').

A less cynical take on the taking away of individual judgement is that it arises from the need for individuals to be replaceable rather than solely as a means of deflecting blame ...

Coastguard ships for example cycle the entire crew every two years. I imagine it must be similar in the navy. There is very little institutional memory retention -- and so the "process" emerges to try to direct all the human cogs as they pass through ...

UK experience here. All military units (including ships) experience staffing ‘churn’ as individuals join and leave for career and other reasons. Some military units are periodically completely reset (e.g. a new ship or a ship that has been extensive refitted), and they then have extensive training regimes to become ready for operations again. The churn then starts again. Some units persist for long periods of time (e.g. a formed air squadron) and do not get ‘reset’ under usual circumstances, although they do have churn. Others (e.g. army formations) have repeating multi-year cycles of training, being ready for ops, and other tasks.

The institutional memory is notionally preserved by the instructors in the schools and training centres (as well as documented Standard Operating Procedures, standing orders, etc). New joiners will have undergone standardized training, and although they still have to establish their role (storming, forming, norming, etc) when they join an existing team, they do in principle understand what they are supposed to be doing. In some cases, new joiners bring valuable new knowledge, skills or attitudes, so churn isn’t always bad. Individuals who have joined up for the long term will themselves have cyclic periods of training, with initial trade training followed by time in a unit, then additional specialised training at a school, then more operational tours, etc. Some highly-trained individuals may be embedded in the units themselves (e.g. at a barracks) to provide standardized, decentralised on-the-job training / supervision to newbies who have joined from a military trade school.

More importantly, ALL military needs to be ready for massive war. If another world war breaks out they need to quickly scale to 100x as many humans. That cannot be done unless everything is easy to teach. Follow a checklist exactly is easy to teach.

One extra consideration here is that as power centralizes, the costs of individual mistakes increases.

It seems interesting to contrast this with the explanation of the financial crises that decision makers had 'privatized benefits and externalized costs' (I remember hearing it from Nassim Taleb, but there were probably others making the same point)

aligning incentives between organizations and the people who make them up is hard

> Organizations are afraid of individuals deciding things, and individuals are eager to push their decisions onto a "process" or "policy"

But! but we're in IT and the Agile Manifesto is the exact reaction to that! We have a living proof that "another way" is possible. Hire champions and trust them on the task, it makes them more enterprising and crafters and makes them deliver systematic higher quality. Remember how IBM's IT used to be: Pathetic chain of specification-driven documents that "decision-makers" couldn't possibly read.

Let's just say for the moment that the rest of the industries haven't found an "Agile metgodology" that befits them. Let's hope it's just a matter of time before government-sized organizations have an epiphany about how to apply Agile to death-risking processes.

Looking at this issue more broadly, I definitely agree that non-IT industries need to take into account Agile methodology to improve their results.

I've been looking for the reason why IT has come up with Agile before anyone else (I don't consider Lean Industrial Management as Agile whatsoever) and I simply think it's because the first industry with so much information in it (it's in the title even...).

Now that other industries have to orchestrate huge processes with huge amounts of information, I hope they'll take into account what IT brought up to solve this issue instead of taking the "IBM route" further (more policies, more guidelines, more bureaucracy).

> Looking at this issue more broadly, I definitely agree that non-IT industries need to take into account Agile methodology to improve their results.

Non-IT industries are often ahead of the “Agile” movement in IT, having adopted Lean methodology and related approaches that actually applying engineering approaches to solve the problem identified, but not actionably addressed, in the Agile Manifesto.

Most so-called Agile methodologies (the exceptions mostly being adaptations from Lean methodologies, either in other industries or from the Lean Software Development movement) fail to do this, and Agile has largely become exactly the kind of cargo cult, situation-blind consultant-driven, value-blind nightmare that the Agile Manifesto was a response against.

I believe in Dan Carlin's recent Hardcore History, The Destroyer of Worlds* he talks about the changes in military commands and micromanagement that have occurred with advances in telecommunications.

He compared the discretion given to commanders in the American West vs. Civil War battlefields. The more realtime the communications, the more that higher commands interjected in lower level decisions.

With advances in telecommunications, we shift decision-making up the chain of command, leaving subordinates to manage inspections and admin-weenieness.

*I can't find a transcript to verify, and as much as I'd like to re-listen to the 6 hour episode, I just can't at the moment.

It also could have a Harvard Business Review episode, after reading this: http://www.leadingageoregon.org/assets/docs/Other-resources/...

In the Battle of France in WW2 some German commanders (pretty sure it was either Rommel or Guderian) would regularly have radio "problems" isolating them from their HQs when they found tactical opportunities they wanted to exploit.

Supposedly the British Commandos were in part formed so that the generals could get Churchill off their back.

This while over at the German side Hitler would increasingly micromanage as the war dragged on.

You could make the argument that WW2 would have turned out very differently if Hitler hadn't been allowed to meddle with operations on the eastern front (e.g. calling a position "fortified" and forbidding retreat, which led to horrific casualties and usually ended up with the generals disobeying orders and retreating anyway, so the result was often the worst of both outcomes). Germany was broken by Stalin and Hitler before the Allies even hit the beaches.

This is a problem which is commen in the US-Military. As I know a dozen Soldiers - mainly Army - of the Bundeswehr, they all recall their trainings against US-Army units stationed in Germany. They all said: "Take out the higher ranking officers, then they are done." Even though the German Units were completely inferior in the main strength of the US, their superior quality material-based approach, they mostly managed to rip them to shreds with a high degree of flexibility. Practice in the Bundeswehr is to tell Officers, Nomcoms and Enlisted only "What to Achieve" and "Why to Achieve it", not "How to achieve it", allowing for higher initiative, less micromanagement and greater flexibility. What every US-Officer most feared were the Hit-and-Run tactics of German "Wiesel"-Vehicles, something the armored platoons were completely unable to counter. Or, as one US-Officer at a beer-fueled post-post-debriefing said: "We may have the better Material, but you have the better Soldiers"

I don't remember this from the "Destroyer of worlds", but hey, 6 hours, it might have been there somewhere and I just missed it.

BTW - it's not the recent one anymore, there's a new one. Yay!

The new one ("Celtic Holocaust") is about the Roman conquest of the Celtic people. Definitely worth a listen if you're into history/Carlin podcasts.

Personally I'm hoping for a "Common Sense" podcast by him sometime soon !

I saw that when I looked for a transcript (a recent, not the recent I guess). I'm two months behind in my podcast feed, so I'll catch up to the next one eventually!

Goodheart's Law [1]: "When a measure becomes a target, it ceases to be a good measure."

[1] https://en.wikipedia.org/wiki/Goodhart%27s_law

Let's popularize this amongst the ranks of management, senior administrators and every other supervisor.... and especially the groups that help create the tests and stats and 'merit' systems.

> The tests that start as attempts to evaluate and quantify effectiveness become the goals themselves, rather than measures of the true goals.

I think you need a second factor to make flawed metrics a big problem: increased competition.

More competition strongly incentivizes all actors in a complex system to favor optimizing over the currently-measured metrics rather than the true goal.

You see this playing out across all parts of society today: academics that chase impact score rather than groundbreaking research, bankers that chase short-term profits to the detriment of actual sound banking, and now we have a Navy that can't do basic seamanship.

I think we really need to recognize as a society that if the thing you care about is inherently hard to measure (e.g. science, the overall stability of the financial sector, the ability to fight the next war while still at peacetime), there are no shortcuts, you just need to suck it up and keep trying to measure the real thing rather than only relying on proxies.

Throwing more and more effort at optimizing over bad metrics will only lead to long term disaster.

This is so common one of Murphy's Laws of Combat Operations is: "No combat ready unit has passed inspection. No inspection ready unit has ever passed combat."

A favorite saying for USMC infantry, and much truer than people realize. There is a required balance though, and the best units have officers that know how to walk that tightrope, and listen to and respect their NCO's. A great example of this would be, much fewer uniform inspections, much more combat testing. Less formations 15 mins early by BN CO + 15 mins early for company CO, +15 mins for platoon, +15 mins for squad leader... It's the type of thing that really affects morale which will in turn affects effectiveness.

Off-topic: Let me get an Oorah.

On-topic: I noticed the tendency towards formality when dealing with many, many municipalities. This is, in and of itself, inefficient in many areas.

Example: A hole opened up in my schedule and I can fit your meeting in sooner. By the time I got a response, that hole was long since gone. Worse, I left that hole open in hopes that they'd respond sooner. They will say that you should have let them know, but you left a half-dozen messages with their secretary.

Lesson learned: Get to know the secretarial staff. I'm not going to suggest you bribe them, but I understand if you do.

Compromise: You get an 'errr', cause I'm salty.

Always makes friends with the doc, supply, and admin. Usually through booze and not being an ass.

My default reply would get voted down, a lot.

I don't care. My verbiage is intentional.

We all salty, Marine. We all salty. It took me years before I could leave my rack without ensuring it was squared away. It took years before I could invite the missus to touch me while I slept.

Eight years is what I served. Worse, it was in two pieces and separated by time at MIT. Yeah, I was a smart Marine. Don't laugh, I can drive a broom like it is nobodies business.

Anyhow, I won't tell you I understand. I will tell you that I can relate. As for the karma, I ain't never scared.

But, you're right. Always be nice to the folks who decide your life. Strangely, that's frequently not the person who projects power. However, I speak to the choir. Hopefully someone else will pick up on this. Otherwise, I ain't never scared. Karma was meant for burning.

<we all salty

Methinks us that are out tend to be salty, those still in, seems like drawdown has bled a lot of salty knowledge out.

I would claim the core problem is an environment than incentivizes people to care more for their careers than for the actual work they are doing.

Sounds pretty much like Soviet Union, actually [0]. The chickens have come to roost, indeed. Managerial science driven process oriented organizations and what happened in the soviet countries share the same basic tenent: individuals don't matter. This is a cancer for performance and quality of life. The former should concern all economically minded people and the latter just everyone. It's not a lose-money, gain happiness proposition, it's a gain-money, gain happiness one.

The problem is that local optimization (career, prestige) causes the global system to reach an unoptimal state. In computer science we say - greedy algorithms seldom lead to globally optimal solutions[1].

Of course, no one is stopping from developing key-performance indicators for a person being a good person - but this seems really arse-backwards way to approach the problem as being a good person is what most people would prefer to do naturally given a healthy environment to begin with.

To slightly misquote Suzuki on his basic idea on how to maintain good character in a group: "Watch, but don't judge. But, do watch". [2]. Ie. make it understood everyone is responsible for their results but don't be an anal retentive accountant while you're at it.

[0] http://militera.lib.ru/research/suvorov12/index.html This is a damn fascinating read. A bit like autobiographic Catch-22.

[1] https://en.wikipedia.org/wiki/Greedy_algorithm

[2] https://www.amazon.com/Zen-Mind-Beginners-Informal-Meditatio...

In my view, hand-in-hand with measurement is the proliferation of stuff to measure.

It occurs to me that at some point in the history of navigation, the skipper was probably the only literate person on the ship, if that. His log, if he kept one, was the only source of data. And you couldn't get his account until he got all the way back to port, assuming he survived.

Today, the ship (or business) is a floating, real time data acquisition server. Folks all over the world are analyzing the performance of the ship from every conceivable angle. They are literally gathering and processing more data than the skipper herself can even comprehend.

Your culture is what you actually reward and punish, not the abstract ideals that you talk about.

There's a great aphorism with similar meaning:

People do what you inspect, not what you expect.

edit: Wikiquote attributes to Louis V. Gerstner, Jr. (IBM) but earlier instances can be found (see comments below).

Lovely bit of language.

That's an old nuclear-Navy saying from at least as early as the 1970s: You get what you INspect, not what you EXpect.

Earliest citation I could find is a management book from 1962:


I've thought about not saying this, I intend no disrespect to anyone. Are there two militaries now? There is JSOC, with a budget, toys and gear, their pick of everything and they are a smallish, nimble ultra trained force designed to engage these terrorist groups and do special tactical missions, fighting dirty if needed. Then there is the greater navy and army and, respectfully, it is sort of a conservative voter friendly rural jobs program. I'm not saying this about the volunteers that sign up, but national policy. It only makes sense that they'd squeeze every dime from it like everything else. To top it off, there are the private contracting outfits that do stuff too, where does that fit in?

The idea that they're supposed to learn on the job is insane, you don't get a lot of second chances on aflight deck, you don't get to make a lot of mistakes on a boat twice.

There sort of are two militaries:

- One we maintain to prove to our potential adversaries that we maintain it, in order to preserve stalemate.

- One we maintain to actually use.

I'm going to venture a guess that a well-functioning defensive military is inherently unstable, as anyone who understands how powerful it is will realize that it never actually has to function.

> well-functioning defensive military is inherently unstable

This is why it makes sense to have a purely defensive military work on a reserve and draft basis. You don't really need the defensive military - until you do - in which case you want to magic a really big and reasonably competent one into existence quickly.

It is not as simple as that e.g. JSOC doesn't have its own ships, limited aviation, etc. It is totally reliant on the "regular" military to support its operations. And of course it gets most of its recruits from there too - AFAIK only the SEALs do direct entry for civilians.

I think the "second military" you describe is more of a "minimal readiness for SHTF situation" group.

Because no politician wants to go down in history books as the person who let the military shrink and lose readiness for a WW3. Sure, maybe the American military has indeed lost readiness, but it's more nuanced and can't be pinned on one person. So just like other parts of this thread suggest, it's a giant CYA operation.

I hate to be so grand but this seems to be a problem across many sectors of the entire western world.

Over and over, in many different guises, I hear stories like this one. What characterizes them all is excessive faith in reductionist measurement-driven processes, and then the attendant army of humans-of-stock-photography consultants that inevitably lurks in the vicinity of such processes.

Goodhart's Law, "When a measure becomes a target it ceases to be a good measure."

Juking the stats ala The Wire, gaming statistics about crime becomes the objective over good police policy or reporting good, valid data on crime. https://www.themarshallproject.org/2015/04/29/david-simon-on... http://www.motherjones.com/crime-justice/2012/06/fbi-crime-s...

When I worked at DoD, it was explained to me with an analogy of monkeys and ladders. You have a ladder, and at the top is a bunch of bananas. You rope the ladder with electrical tape, so that any monkey who attempts to climb is shocked. Eventually, the monkeys will socialize and prevent each other from climbing the ladder.

What's interesting is that this will continue even after you remove the electrical wire.

Partially mythical, there were no bananas and if I remember well there were only two monkeys. [1]

[1] https://www.scribd.com/doc/73492989/Stephenson-1966-Cultural...

What was that quote? If you measure success by testing, then you will get results that improve performance against the test at the detriment of all else?

Sometimes (often) involving cheating where existential considerations for organizations or leadership is involved.

"When a measure becomes a target, it ceases to be a good measure."


not in military, but the joke among colleagues has been that we are very in following our processes, not so good in getting anything done. sadly this is excused because the process was followed and the auditors are happy

Is the degradation of US public schools really due to extensive testing?

I believe the idea is that the entire curriculum is then centered around teaching the test, versus what kids actually need. As to whether that's the primary issue, no idea.

Do you really think eliminating standardised tests would improve the education system?

I'd be far more okay with standardized tests if the requirement also banned homework until at least Junior High. Otherwise it's a race to the bottom to see which schools can suck up as much of the children's and families home time to help improve the schools scores.

How would you propose that students practice learning and studying on their own, outside of a classroom setting, without homework?

I don't propose that children do any of that. That's what high school is for, and 4 years of "practice" is more than enough.

Some of the reason schools start homework so early is that parents demand it. I've seen parents complain that a first grade class isn't academic enough because the teacher didn't assign enough homework.

There's an excluded middle, in that statement. De-emphasizing the importance may be an option.

Fair point. Let me rephrase the question:

Do you really think reducing standardised tests would improve the education system?

If so, what objective evidence can you provide for this belief?

If you measure it less, you'll have a weaker argument for how the education system is doing!

Your comment is precisely the problem: objective metrics will always support their proponents' arguments for more measurement, while any upsides to a lessened objective grading will be difficult to show because they'll involve subjective assessment or less assessment.

Do not confuse the investigator's convenience in accessing a truth value for the truth value itself.

That argument has been made many times and in great detail. See for instance the work of Diane Ravitch, former US Assistant Secretary of Education (https://en.m.wikipedia.org/wiki/Diane_Ravitch)

Edit: she was Assistant Secretary, not Secretary

School-funding and even whether a school closes or not is tied to these tests, so there would be some work involved in replacing that aspect of standardized testing in US schools as well.

No, the opposite. We didn't understand the decline of US public schools until we started trying to systematically measure their performance.

Standardized testing has been around for decades, but was a tool for sorting students. It is only within the last decade or so that we have tried to measure the performance of schools, districts, and states, as systems.

It always blows my mind when I see people on HN complaining about school testing and measurement. If you tried to run a web application or a business with zero ways to measure its performance, how do you think that would go?

The thing is, what do you gain by knowing that school X scores 13% lower on standardized tests than school Y? That school X is worse than school Y? How does that help you provide a better educational experience for the students?

My mom's been a teacher for ~40 years, and one thing to realize about standardized tests it that they don't provide any new information to the teachers---the teachers already know which students are struggling and which students aren't. In fact, standardized testing reduces teachers' ability to respond to variation in student ability, because all students need to be prepped for the same test. Preparing lesson plans is one of the most time-consuming aspects of being a teacher, and standardized tests make that task more difficult by adding more requirements. IMO, it would be better to focus on ways to reduce teachers' administrative burdens---perhaps by making it easier to collaborate and share lesson plans with other teachers[1].

[1] And I mean genuinely make it easier to collaborate, not forcing teachers to spend 90 minutes a day trying to enter their plans into the latest and greatest ed-tech website. If you want the information standardized, hire someone specifically to transcribe paper plans into whatever system the school administrators pick.

It is possible (even likely) that our current standardized tests need some adjustment to fit the needs. However the purpose of standardized tests is there are standardized needs. As a member of society I want everybody to know some things.

If a kid passes first grade without getting the correct answer to 2+2 there is a problem, it isn't acceptable to say your school doesn't teach math until second grade. In a single school that is just fine, but people move and so on a national scale that does not work.

Now I understand that not all kids have the same abilities or backgrounds, and both make a big difference in how far those kids can go. That needs to be managed, but it needs to be managed in the standardized process.

It does seem to me that teachers should have a few common lesson plans to choose from. There is no reason a teacher should have to create their own plan. Being a good teacher within a plan should be easier because they can prepare to use a working plan.

  Now I understand that not all kids have the same abilities
  or backgrounds, and both make a big difference in how far
  those kids can go. That needs to be managed, but it needs to
  be managed in the standardized process.
If our society were prepared to manage the fundamental issues that plague our educational system, then your rationale makes perfect sense, assuming that the standardized lesson plans were designed to facilitate the assessment of meaningful programs. But we're not prepared. Rather, the standardization efforts were an excuse to avoid addressing the fundamental issues, and have become a distraction that causes authorities to focus on the wrong aspects of educational reform.

The overwhelming factor in educational achievement of a population is the quality of home life, and in particular the participation of adult family members. Poor families have neither the time, money, nor enculturation to facilitate meaningful improvements in home life. Without a social commitment targeting the root causes, such as by providing free day care, the standardization efforts achieve nothing, and unnecessarily constrain local control.

Since the Bush-era standardization reforms, the achievement gap has only _grown_. And we're still matriculating kids without the requisite skills as the alternatives, like holding kids back year-after-year, are neither desirable nor practical.

"How does that help you provide a better educational experience for the students?"

It helps you identify which schools are in need of special attention and require structural changes.

With some competitive pressure from school choice at the margins, coupled with per-student funding, it would also provide incentives for schools to improve.

In practice, it seems to create a black hole where a school that serves a population that isn't good at tests (e.g disabled kids, non English speakers) gets low scores so everyone who could get a good score leaves and then the school does even worse and gets less funding and then the school is failing and the "fix" is to fire the administration, spend money bringing in new staff who don't know the area or the students, and expect that to solve the systemic issues revealed by poor test scores. To avoid this, schools are indeed incentivized to improve - for instance, by directing low performing students to stay home on the day of the test, or by discouraging students with learning difficulties from attending at all.

This effect is visible in urban centers as well.

Public education is supposed to be a leg up into the world for everyone, not the start of the rich getting richer and the poor getting poorer.

That means we are not using the tests results correctly though. Knowing there is a problem is the first step in fixing it - but it doesn't tell you what the next step is.

The idea that competitive pressure improves school performance is one of those simple, obvious, and wrong Econ 101 ideas.

Even if the bad schools can't respond to incentives to improve, allowing students to choose good schools over bad ones naturally increases utility.

No it doesn't. It could easily result in less social utility by promoting self-segregation and the concentration of funding into high achievement schools where the marginal utility of each additional dollar is substantially diminished at the outset.

The reason why kids from poor families do poorly in school is also the reason why those families are _least_ likely to be selective in their choice of schools. These are not independent phenomena.

So unless the only utility function that matters is freedom to choose your school, approaches like vouchers are unlikely to improve things and could easily not only perpetuate but exacerbate achievement gaps and derivative social issues, like crime.

The evidence indicates that voucher systems improve test scores, graduation rates, and other aspects of both public and private schools, not just the latter:


  Scaled-up voucher programs like those previously advocated
  by Secretary DeVos show the worst effects. There have
  recently been four statewide voucher programs: Florida,
  Indiana, Louisiana, and Ohio. The Florida study is
  inconclusive, and the others show large negative effects. In
  some respects, the Louisiana results are more convincing
  because the results have been corroborated by two different
  sets of researchers and students were assigned to vouchers
  by lottery—the most rigorous way to evaluate vouchers. In
  terms of providing convincing results, the Indiana and
  Ohio programs, are somewhere in between, but these show
  negative results as well.

  Voucher supporters argue that the results have been worse in
  the recent statewide programs because they have been
  “heavily regulated,” by which they mean the requirements
  that students be tested, that these results be made publicly
  available, and that schools must let in any student who is
  eligible for the voucher. The fact that this fairly minimal
  oversight is considered controversial or heavy-handed,
  however, only reinforces that private schools are designed
  to be exclusive and have little interest in external

15 years later, http://news.stanford.edu/2017/02/28/vouchers-not-improve-stu...

I wouldn't expect charter schools to actually change much because the evidence is very clear that the most important factor in achievement, overwhelmingly, is the child's home environment.

Do you have evidence to back up your claim that it is "wrong"?

If your metrics for your web application aren't a good proxy for user experience, those metrics can drive you to make your application worse.

Focusing schools on testable performance leads to unintended consequences. My kids had heavy homework loads starting in first grade. They were still kids! and got robbed of much of the joy of childhood that should been afternoons playing at the playground and running around the neighborhood. All so the schools could pump up their test scores, by forcing their "customers" to carry the load.

There's some relevant business wisdom: "what gets measured gets done".

By aligning our educational institutions towards testing we created educational institutions that teach the test and not the rest.

Standardized testing is super important. I agree, though, it's a pretty poor proxy for how well schools are performing and who needs what... By way of example: schools with low-scoring test takers likely have environmental issues that impact school performance. The conclusion should likely be beefing up the budget and expanding the schools role to being more "community centers" full of social workers.

Poverty is a cycle, individual disruptive elements in school hurt all students, and that little 'war on drugs' has removed a lot of the support structure we expect in those families... Reducing the budget of schools with students in bad situations who can't test properly because their students are in bad situations and therefore can't test properly seems like a self-fulfilling prophecy. A prophecy that happens to align very nicely with the ideology and stated goals of the group that pushed "no child left behind".

> By way of example: schools with low-scoring test takers likely have environmental issues that impact school performance. The conclusion should likely be beefing up the budget and expanding the schools role to being more "community centers" full of social workers.

This is a great idea I would be fully in favor of. But we can't start having these types of conversations unless we know how schools are actually doing.

Reminds me of something I read on HN the other day, I'm not sure if it's true but one commenter said he got a driver's license without ever actually demonstrating he can drive.

If it's true then it's outrageous that this can happen in the US.

A good admin weenie would have laid down the law and forced the ships in their fleet to step up operational awareness and training. Astonishingly whatever attempt was made to address the problem was useless.

Emphasis is put on didactics rather than common sense.

> The tests that start as attempts to evaluate and quantify effectiveness become the goals themselves, rather than measures of the true goals.

Is this a dank meme?

sounds like the tests are bad.

sounds like US Army in 1975.

>“We put a premium on being good inspection takers and admin weenies.”

such a shift is natural and necessary in the context of the future warfare being completely robotic/autonomous. People role would be exactly inspection and admin. And as preview of the future - we've just had 2 people driven destroyers lose to the most probably autopiloted a cargo ship and a tanker.

>After 2003, each young officer was issued a set of 21 CD-ROMs for computer-based training — jokingly called “SWOS in a Box” — to take with them to sea and learn. Young officers were required to complete this instructor-less course in between earning their shipboard qualifications, management of their divisions and collateral duties.

I'm trying to find the "seaman's schedule" someone in the navy posted on reddit, but it basically demonstrated that between watches and other ship duties, there was on average 5 hours of sleep a day a seaman could get, usually at random hours, and not always consecutively. Curious when in that schedule they could slot in "watching video lectures," and whether they'd be able to even stay awake during them, let alone focus.

I served on both a FFGs and DDGs as a BM. This is pretty much a normal routine out at Sea. Doesn't matter if it's a week, month, or year. It's fairly similar sleep and work schedule. Combine it with other collateral duties, Flying Squad, SAR Swimmer, LSE, Boat Coxwain, etc and it just slowly erodes your time away.

It's not all doom and gloom though. If you properly communicate with your chain of command, and slowly train junior personnel into the proper qualifications, the problem starts to go away. Though, you're not looking at more than 5-6 hours of sleep straight on a good day.

If peace time has 5-6 hours of sleep on a good day as the bar, what are they going to do if an actual war breaks out? It sounds like there is no margin left for them to use and it would only take a few casualties to make the boat unable to operate

They're probably already damaged badly enough from the lack of sleep in peacetime that they'd underperform on any less during an active operation.

This is like giving your sailors food poisoning every week to make them tougher. You just make them weaker instead. And when the time comes, you don't get soldiers capable of eating rotten food. You just get weak people.

Agreed. I could see putting people in that situation for 1-2 week every few months just so they know what it's like and can handle it when it comes up, but running everyone ragged is just going to lead to lots of preventable problems for no gain that I can see

> If you properly communicate with your chain of command, and slowly train junior personnel into the proper qualifications, the problem starts to go away. Though, you're not looking at more than 5-6 hours of sleep straight on a good day.

This, to me, is a major problem. Many (most) young humans do not do well on 6 hours of sleep or less for a long time. One could probably load up on stimulants, but this I suspect will bite you later.

Given that this happens in peacetime, do you think it can be improved with better technology? That is, instead of, say, three people watching screens 18 hours/day you get them to do the work as well (or better) in 12 hours/day if you give them better displays? I suspect it is not that simple, but would still like to know where the snag is.

In most cases, better technology is just as expensive, if not more expensive. You need better trained and better staffed crew on ships, regardless of the systems in place.

As an example: Deck division will have 5/6 people on watch for a smaller ship: Port & Starboard lookouts, an aft lookout, a phone talker, someone actually steering the ship, and then a boatswain mate of the watch that organizes those individuals, keeps the ships schedule, and other various tasks. None of these individuals will actually be looking at a screen, but instead scanning the water for other ships.

Another division down in CIC may have just as many people looking at all the radar and managing the various data streams coming into the ship.

So, if you have 12 to 15 people in your division, which does happen, you'll be on Port & Starboard watch ( 5 hours on then 5 hours off, with a 4 hour mid-shift ).

Now, take an individual that is trained on Ship A, then after 3 years goes to Ship B. Ship B is a completely different class and has new(ish) technology systems. That sailor is then going to have to be completely retrained on those systems. After another 3 years, that sailor can seriously be shipped to a completely new platform where they then have to repeat the cycle.

Adding new technology would just complicate the matter.

The snag is funding. The level of resources they have is inadequate for the mission they are carrying out.

I understand that it looks this way, but in my experience, human time is EXPENSIVE. Even a low pay grade soldier/sailor costs us a lot in benefits, insurance, medical, support, etc.

That cost may be coming from a different bucket and thus not visible to the folks who decide whether to invest $X in a systems upgrade, but in the grand scheme of things, getting humans to do the job that machines today do well is a big waste.

When you hear arguments that a badge reader is expensive and instead we will just put two soldiers at this or that door (24x7, btw) and no one jumps and calls BS on it, it is a strong indication of a broken system.

That's hard to believe, considering how much the US spends on the military.

The ships are understaffed (see the discussion here about people having 16+ hours of duty per 24 hours) and under maintained.


Add crew and duty hours would go down. Reduce hours of operations and maintenance would be better. More crew obviously costs more. Reducing hours of operation means a smaller mission or more ships.

Crashing and repairing ships is expensive, too :-P

I'm genuinely curious how there's so much work sailors have to abide by a schedule like this. Is this only in active war zones or is it always like that?

There is an enormous amount of work to do on warships (in my experience); the following doesn't apply everywhere always, but all of it applies sometimes and some of it applies all the time.

If you're in engineering, kit is constantly breaking and it is never all working; you'll never catch up, so it's just a question of what you decide to leave broken right now and what you need to fix. The things that are working need constant attention to make sure they stay working; you'll never get them all checked, so there's always something that definitely needs checking really quite soon.

In warfare branch(es), as in all branches, the manning situation is never right; there are always gapped billets, requiring other people who don't quite have the right skills and experience having to share that extra workload amongst themselves and to cover extra watches. Going to defence watches or the like (six on, six off, or even four on, four off) is a permanent mindfuck that leaves everyone as a kind of zombie, skating through on routine.

The kit is often old, and it's all pretty custom, and so much of it is painfully delicate. This ship doesn't like warm water, that ship has a habit of not being able to power the radar all the time, damn comms constantly up and down. With relatively new classes of ship, every one of the first dozen is basically an experiment, so the new kit refuses to work properly with each other, or maybe just plain at all, and the original purpose the ship was designed for is long gone and what was a cold water submarine hunter meant to last a week on its own before being back in port is now a warm water patrol vessel out for months at a time, constantly stressing systems that were meant to have a steady supply of cold water cooling and were never meant to be active for weeks at a time.

Every ship and crew will habitually be asked to do more, with less. Harmony time vanishes, billets get gapped, refits get put off, maintenance cycles become less frequent. It stops when ships basically cease to function.

Sometimes it's a wonder they don't sink upon leaving harbour :)

At the risk of imaginary internet points for being too reddit-like, this is my favorite software engineering GIF and is appropriate here:


EDIT: Here's an article with a full video... Which is basically the same as the GIF:


What is actually happening during watch? Does that literally mean someone is posted in the modern equivalent of a crow's nest, looking out for ships?

Yes most surface warships have at least four lookouts on every watch to visually look for vessels and other hazards. That system works well as a safety backup for radar as long as they pay attention and communicate effectively.

It's always like this, or worse. Active war-zones are a bit different if we are sailing through "hazardous" waters. Usually on top of work and normal watches, you'll man up some different watches since the crew will have to man extra guns.

Some divisions are different, but a majority of sailors have a pretty solid stream of work. Maintenance of a ship that is close to 20-30 years old is a constant battle. Hell, even a ship that is 5 years old requires meticulous care.

Deck division is responsible for maintaining a majority of the "skin" of the ship. So, a lot of painting and preservation. We are also responsible for all life-saving equipment on the ship: Life preservers, life-boats, any number of RHIB (Rigid Hull Inflatable Boats), life-lines, etc. A lot of this time is spent supervising extremely under-qualified and unmotivated individuals.

It's almost akin to teaching an unwilling 5 year old to read, and then every single day constantly monitoring his progress. As soon as your comfortable that he or she can read without supervision, they are transferred to another division, or are taken off the ship.

Surprised we don't see more accidents. How would any of these ships actually work in an actual war?

99% of the time, you're doing nothing except watching and waiting. Every once in a while, you are overwhelmed by sheer terror, or at least obvious, immediate threat to life. During those moments, adreneline keeps you going. The trick is realizing you are in a situation that warrents adrenaline.

Wars last for years, though. They're marathons, not sprints.

Battles last for hours, days max. Most of war is boring.

At the individual level they're sprints.

Right. And the context for that:

> For nearly 30 years, all new surface warfare officers spent their first six months in uniform at the Surface Warfare Officer’s School in Newport, Rhode Island, learning the theory behind driving ships and leading sailors as division officers.

> But that changed in 2003. The Navy decided to eliminate the “SWOS Basic” school and simply send surface fleet officers out to sea to learn on the job. The Navy did that mainly to save money, and the fleet has suffered severely for it, said retired Cmdr. Kurt Lippold.

Six months of training vs a box of CD-ROMs. Amazing.

And now we are 10 sailors down and two major crashes... What is the ROI?

17 sailors down in two months, and much more than that in total since the elimination of the basic training course.

It's penny wise, pound foolish. We're perfectly willing to spend many billions of dollars per large ship, yet we eliminate the basic training programs for the sailors that actually run said large ships to save a few million. It'd be like buying a Ferrari, but skipping the seatbelts to save ten bucks.

Or, in tech, imagine if we eliminated CS educations and just expected all programmers to learn skills on the job. I don't think that'd work out too well.

CS educations and just expected all programmers to learn skills on the job

Or replaced 4-year degrees with 4-month "bootcamps"...

Your granddaddy OP for this thread did a bootcamp and is operating just fine in the wonderful world of programming ;)

I don't exactly know the history of these courses, but SWOs now attend BDOC,though sometimes not for a few months after reporting. It's pretty heavy on seamanship.

Out of curiosity, I looked up who was Secretary of the Navy in 2003. It was tumultuous year, with 3 different people holding the office. Only one of them had any military experience (in the Air Force).

Gordon R. England held the office at the beginning of the year and was re-instated at the end of the year. He held the office for about 4 years total. He had no Navy experience. His background was in the defense industry. What a crock.

So another Bush appointee fuck-up, just this one took much longer to fester. Great. I wonder what kind of lingering effects we'll be seeing from the current rash of unqualified people (like Betsy DeVos) in the decades to come.

What's wrong with civilian control of the armed forces? Why should service secretaries have served?

We do have civilian control of the armed forces. The question of past experience in the service, though, is the same as experience in any other leadership position.

Also, in England's case, the big red flag was his close ties to the defense contracting industry.

What's wrong with PMs who can't program?

More like what's wrong with VPEng who can't program? Answer: plenty.

Nothing, if their job is to color code spreadsheets; everything if their job is to actually put the content into those spreadsheets.

Because you end up cutting crucial corners because you think you understand whether the two options are equivalent when they are not. Case in point: the topic we're discussing.

Meanwhile, they probably take 100 hours a year of stupid compliance training about drug use, not killing sea turtles, etc.

All delivered in Adobe Flash...

At the very least they should have included an "America's Navy" simulation game.

12 hours of World of Warships and you're good to go.

An officer's schedule shouldn't be compared to an enlisted persons's

An enlisted person's schedule that leaves only 5-6 hours of sleep per day is also fucked, regardless of what the officers schedule is.

An officer's schedule can be worse.

Wow, that's concerningly little time to sleep. I know that when I'm getting that little sleep I can't properly focus on things and my memory is impaired. That mustn't help.

And/or maybe Navy higher-ups, having resolutely learned nothing from the airline industry and crew resource management, pretend that there is no such thing as sleep deprivation.

Culture of self flagellation. Somehow the finance industry, Japan, and Taiwan all still haven't learned from this. From the way my investment banker roommate talks about it, they almost seem to get off to it.

Same thing happens with doctors in the US. Doctors should know that after the first 12 hours, the doctor's not as good at doctoring anymore.

Yes. Although in talking with doctor friends, it turns out there's an interesting counter-force. A major source of errors is handoffs.

I've certainly seen this when in the hospital with family; new doctors would frequently ask me for historical information, including about what other doctors had said or done. I was fine giving it, but what if I hadn't been there?

Do they work with checklists? That seems to be the number one reason why crews in aviation can work with such low error rates. Handoffs with strict checklists would probably cause lower error rates.

Yes, they do have a lot of checklists, but it's one thing to create a checklist for an airplane with very strict requirements for validation and maintenance and another to create a checklist for a human being.

As others mention, there's a big push towards checklists. (Note: they only work if they are bottom up, not top down.) But aviation mechanics work on something that is heavily standardized, and they get to shut down the thing they are working on, which eliminates a time dimension to the problem. Neither is true for people, so checklists aren't really the solution to the information-handoff problem in medicine.

Can you expand on top down vs bottom up? Do you literally mean reading from the bottom of the page, or starting with the most granular items and decreasing granularity?

Sorry, I mean in an organizational sense. If managers tell workers "Here are checklists! Now follow them!" then they aren't as effective in improving quality as workers saying, "Let's collaborate to make sure we are doing procedure X well. How about a checklist?"

Checklists are an easily observable symptom of an team drive for improvement of outcomes. But you can't reverse that and plop down checklists that are used under threat.

I don't know much about how hospitals work, do doctors even have "managers?" I would figure the doctors would be the ones coming up with and using the checklists.

It's complicated. Doctors employed by hospitals report to two sets of managers. There are other doctors who serve as medical department heads and the chief of medicine. But those doctors are also accountable to non-physician business managers for administrative and financial issues. And then independent self-employed doctors may have privileges at the hospital and see some patients there but don't have managers in the traditional sense.

Crap, I listened to a podcasts sometime last year about how using checklists in a medical setting decreased errors by a ridiculous amount.

I'm going to try to find it. But, if I don't, you got the gist of it:

Have a checklist, and stick to it. Don't expect one person, or even multiple people, to remember all of the steps off the top of their head.

Possibly talking about "The Checklist Manifesto" - http://atulgawande.com/book/the-checklist-manifesto/

It's an -excellent- book about the application of checklists to error reduction etc.

That's exactly it. I knew it was something based on a book. I've updated my comment below, if anyone is looking for the specific radio show about it.

If you find it, post it so I can listen.

I'm having a bear of a time trying to figure out what podcast it was. It might even have been a YouTube video. I've been going through different permutations of search terms to find it.

I'm worried it might have been .NET Rocks![1] They have a great podcast, but their episodes aren't indexed by Google because of their website structure.

I'm gonna keep looking, and even check my home computer's history when I get home tonight, but I wouldn't hold your breath. Someone a bit above linked to a book that I'm pretty sure was the impetus behind the podcast I listened to, so it might be worth giving that a read.

[1] https://dotnetrocks.com/

EDIT: HA! I found it! It was actually in the press section on the book's web page.


The discussion about work hours here typically seems to put effectiveness after 4-6 hours as dropping off for programming, for pilots regulations allow 8 work hours (with 10 required hours off, and at least 8 of those consecutive uninterrupted sleep). It makes one wonder at the working conditions of doctors and sailors.

Toss in architects, visual effects artists, kinda of whatever job you want in the US, especially if there's no history of a union.

As someone else said, there's a lot of ego involved too. Given that a lot of it is "brainwork", it's the only physical way they might be able to feel they're testing themselves against something.

Most of it in those jobs is just production-churn and an attempt by the higher-ups to save money on things that a) were bid low and b) have shitty margins.

Medicine as well. Medical residents can have shifts that are 24 hours long.

Recently enough a group of doctors went on strike with perhaps one of the most reasonable demands I have ever heard. They didn't want to continue to do shifts that were over 24 hours long!

Well, as long as they only let people with jobs that can kill people be sleep deprived, it'll be ok

I understand that medicine is tricky because the handoff from one doctor to the next can be dangerous for patients. You have to balance continuity with exhaustion. That's not to say that they necessarily have the right balance now, but it's a hard problem.

I've heard doctors say this, but they never cite evidence. But there is a wealth of evidence that says that doctor performance drops significantly with motor performance being harmed so much as to increase the number of errors and excess movements. I'll add the citation at the end, but it's not about the single study; it's about the fact that there is little evidence to the contrary and lots of evidence saying it's the case. If this were a drug, we'd be prescribing it every which way.

And most importantly, know that this is not necessary. American doctor supply is restricted by the AMA. America could suck the best doctors out of many other nations if it so desired and its healthcare practitioners would be better rested, more mentally well, and far less error prone.


> I've heard doctors say this, but they never cite evidence.

Seems pretty well researched by a quick google. Maybe they're eliding the support not because it's an assumption, but because it's so thoroughly established they take it for granted?




> But there is a wealth of evidence that says that doctor performance drops significantly with motor performance being harmed

That's true too, but I worry you're trying to rebut (X & Y) with evidence for Y.

If we find ourselves slipping into a battle of emphasis, maybe we face a problem of optimization?

From what I'm seeing, your hypothesis about this being a question of optimisation seems well supported. I worry that medical professionals resist changes here out of cultural belief or a resistance to process. That wouldn't be without precedent. Even something as well-established as washing hands between patients isn't done without explicit process (compliance is near 13% without explicit reinforcement for something that everyone learns the legend of Semmelweis for), and surgeons refuse even _private_ outcomes tracking.

Agreed, I can't know from first principles if the right shift length is 4 or 8 or 30. I'm sympathetic to your suggestion that it would be a surprising coincidence if doctors landed on the perfect number through historical accident. Maybe they have a feel for safety in shift length, but they don't in other areas like washing, so maybe not.

It probably varies by specialty and even individuals, probably a monster of a problem.

Do you think tech in Silicon Valley is somehow above that, though?

Maybe? I've only been here about 2 years. First half was a bootcamp, which while intense certainly allowed for enough sleep (I would leave the bootcamp at 8pm, get home ~9, gives me an hour to relax, wake up at 7, and commute for an 8am arrival).

Second half has been at a company with a very pro-family-time CEO, and as a bachelor that just means a lot of time to explore this awesome new city :) So I haven't yet experienced the "utter madness startup" I guess?

They forget their own history, too.

A battle in 1942 was lost, badly, because the USN crews were zonked from a routine that gave them almost no chance for sleep.

Speaking of WW2 I've seen claims that one of the factors that helped the RAF win the Battle of Britain was there culture of pilots taking time away from the battle to rest up, relax and generally recover in between periods of intense fighting. e.g. "Talking shop" was definitely frowned on in a culture of apparent gentlemanly amateurism.

The Luftwaffe was more "professional" and worked their crews much harder and didn't take time to relax and had a culture of focusing 24 hours a day on the job.

Interesting. I'm not sure I could do that: I'd want to talk shop, badly.

I read - years ago - that the most active combat was in the area around London. Not a small area, but it allowed RAF to move squadrons to lower threat areas for reconstitution and rebuilding before rotating them back to a hot area.

My experience in the Navy made me lose all respect for the "leadership" there. It's a place where careers are established by doing everything except that which involves any actual thought or problem solving.

The culture is risk-averse to a fault, but at the same time have nearly zero capability to accurately asses risk or fix problems beyond punishing people when things inevitably go wrong.

Official "risk assessments" are performed before every dangerous undertaking, with numbers that are pulled out of someone's ass, along with additional "risk mitigation" numbers that are also pulled out of someone's ass so that leaders have cover when an accident happens. "We performed all the necessary risk assessments and were still a go."

Let me give you an example. Underway replenishment is when one ship will pull alongside a supply ship and transfer fuel and supplies from the supply ship. It's very dangerous because the ships are within 300 feet of each other for hours on end.

A risk-assessment is done and briefed the night before. Those are a PowerPoint presentation with a grid of things that can go wrong on the Y-axis. There are two columns, "risk" and "mitigation," and in line with each disaster are the probability conflated with the severity of that happening. Not actual numbers, just words like, "severe," "marginal," "likely," etc. Those are color-coded between red and green.

One of those things is always "crew preparedness" or similar. The risk of having an unprepared crew that didn't get a lot of sleep is shown as, for example, "severe" and colored red. The briefer will say, "well this is a severe risk, so we're making an announcement that everyone needs to get plenty of sleep tonight and that mitigates that risk down to "tolerable."

Actually this is more technical than what actually happens most of the time, which is that you'll hear things like, "we've mitigated the yellow risk down to a green." The entire ship's company of NCO's and officers will watch that presentation and nod their heads along. There is never one of these that results in a decision not to do something.

And there you have it. I've taken you through the entire risk mitigation strategy for the Navy, that happens nearly every day.

Does anyone know the actual probabilities of disaster? If so, they're not saying. Are they changing existing procedures or spending additional money to fix problems based on actual numbers? Hell no.

Really severe problems are dealt with in the following manner; a "zero tolerance" policy for that problem existing will be created, and when disaster strikes, it's a career-ending event for whatever scapegoat is standing there. Problem solved!

By the way, don't join the Navy.

EDIT: Also, the "too much technology" portion of the article is laughable. Part of the problem is that technology that the rest of the world uses to avoid these sorts of things is seen as anathema to the luddite SWO community, and rather than embrace anything that could help automate error-prone tasks, they pay lip service to it and continue doing extraordinarily stupid things for tradition's sake.

Aircraft carriers are driven by issuing voice commands to (mostly undertrained) helmsman in an environment where any additional noise could result in a misunderstood command and disaster. Or the helmsman gets tired of standing for a five hour watch every day and accidentally turns left instead of right like she's told. Imagine driving a car by telling the driver what to do from the passenger seat. Not just a "turn left up there," but "turn the wheel 15 degrees to the left, ok, now turn the wheel back to center, ok, now turn it back three degrees to the right because you oversteered..." It's insane.

> Aircraft carriers are driven by issuing voice commands to (mostly undertrained) helmsman in an environment where any additional noise could result in a misunderstood command and disaster. ... Imagine driving a car by telling the driver what to do from the passenger seat. Not just a "turn left up there," but "turn the wheel 15 degrees to the left, ok, now turn the wheel back to center, ok, now turn it back three degrees to the right because you oversteered..." It's insane."

Former aircraft carrier OOD (and nuke SWO) here. I agree with your comment about fatigue, but did you get much bridge time? At least back in the day, the conning officer's voice commands weren't so micro-managey. They were along the lines of, "Right 15 degrees rudder [0], steady on course 090," which the helmsman would repeat back to make sure the command was correctly understood. The helmsman would then turn the wheel as needed to make the turn and steady up on the ordered course.

[0] The conning officer would specify the rudder angle in part to control how much the ship would heel in the turn. That can be important when you have armed jet aircraft driving around on the flight deck.

I spent nine months as the conning officer on an aircraft carrier and qualified as OOD. On numerous occasions we narrowly averted disaster because the helmsman misunderstood the command, or repeated back the correct command but did the opposite.

In many cases those helmsmen were undertrained or tired, as a direct result of cuts in manning/funding and additional requirements.

The "old" people in the navy, including many of the admirals at the time (ten years ago) had many of the same opinions as those expressed in the article (they need to stop looking at radar!!!)

What those people didn't seem to realize is that we were doing the same things they were doing 20 years prior with about half the crew they had.

Spot on. In the Coast Guard we have a similar operational risk assessment model called General Assessment of Risk (GAR)[0], which uses six categories and a score of 1-10 in each to get a cumulative score, which corresponds to either Green (low risk), Amber (moderate risk), or Red (high risk). In a best-case scenario, all involved in the operation will conduct a GAR brief as a group, and the person doing the brief will solicit the group for what number they would give for each category, and the highest number anyone calls out is the one the whole group goes with. Typically, if somebody calls out a 5 or higher, they are asked to explain why they feel the risk in that category is so high. Once the reasons are identified, the group then "mitigates" the risk by discussing the identified reasons so that everyone understands the risks. The score is tallied, and the color (risk level) is identified.

That is the best-case, textbook way to run the model. Still very subjective pseudo-science, masquerading as objective risk management, but at least has something of a method to it.

In practice, the repetitiveness of the GAR model results in many crews blowing it off, and giving a vague, arbitrary cumulative score without any discussion around how they got there. This sounds bad, and by policy it is bad, but in practice I have observed no discernible difference in how crews approach risky missions and operations when they conduct a full GAR brief or just give a somewhat random score and move on. In other words, the GAR model does not seem to provide any tangible risk management benefit, and largely seems to serve as a bureaucratic CYA solution.

GAR was introduced in an attempt to reduce the number of mishaps occurring due to what was deemed to be excessive risk-taking. The statistics may demonstrate that it had that effect, though I would argue from my perspective that other training programs introduced to address problems related to risk assessment have far more deckplate-level impact and effectiveness. The problem with GAR is that it tries to objectively standardize something which is, by its very nature, dynamic and subjective. No two people, in the exact same situation and having the exact same capabilities and experience, assess the risk the same way. Assigning numbers to a series of broad categories and giving a color-coded risk level to the situation does not inform anyone of anything very useful. Discussion of risk factors is more helpful, but due to the way the system is structured, is a step frequently skipped.

What matters far more is focusing on continuous, dynamic training and education of those in billets for whom risk management is a critical part of the safe completion of their mission, and emphasizing clear communication unfettered by rank or positional authority to ensure that everyone has full situational awareness. Be respectful, but make sure that information can move freely between all involved.

[0] https://www.uscg.mil/hq/nsfweb/foscr/ASTFOSCRSeminar/Present...

Shout out to a fellow Coastie.

I think what I learned from my superiors is that it doesnt matter so much what system is used for risk assessment, but whether a conversation about risk was had in a meaningful way. Sometimes quantification helps this, where complex systems can be analyzed and the consequences assessed. Usually if you are going to do that it has to happen well in advance of an operation. Othertimes the desire to get a number leads to a real ham fisted attempt to "quantify" things like how fatigued the crew is on a scale of 1-10, or to rate the environmental conditions. When the GAR model, which stands for Green - Amber - Red, was used to facilitate an honest conversation about how people felt about an operation as opposed to just checking a box so that it could be put in the logs. When I saw it being used as the former it absolutely made things safer, but that was very dependent on the attitude of those participating.

Agreed, and a more succinct way of saying what I was getting at. Attempts to quantify inherently subjective attempts aren't very productive in their own right. What matters far more is having a substantive conversation about risk, and work to promote a culture of open and honest communication.

Semper P

If you want to know how tired your crew is, wouldn't it suffice to pluck a random few of deck and ask them if they're tired as fuck?

How would you make them answer truthfully? Admitting to being tired either means you're showing personal weakness, or you're calling your superiors incompetent at scheduling. Neither possibility sounds good for your career. I think it would be better to measure reaction times, eg. by making them catch a falling ruler immediately after it's dropped and seeing how far it falls.

This is why good leaders create an environment where you can be honest about such things. You are basically talking about an inhumane system led by fear. Then again we are talking about the navy here...

This is part of the reason why they institute fatigue standards, to try to keep people from being over tired and still working, which led to people being hurt.

The hard part is that no one wants to admit that they are tired. Everyone wants to be "that guy" or "that girl" who is always ready to go. After all, most people Ive met in the military have a hard time putting themselves to bed because they have the next watch when its 0200 and we're pulling migrants off rafts or running helicopter ops to interdict $100 million worth of cocaine. During lots of operations you will have a lot of the crew stood up to run things, so balancing crew fatigue with surge operations is no trivial task. The SWO community could take a lot of pointers from the aviation folks - they are way better about making sure people get mandatory rest.

In your opinion, are high level commanders well educated? field experienced ?

Well-educated...it depends. Many have impressive sounding degrees. However, an extreme anti-intellectual bent is very common.

Do you know the people who studied for and aced all exams through brute force memorization, promptly forgot all the material or never understood how it would be applied, but now have the certificate that says they know X? That was 90-95% of Naval officers.

I had a one week class in digital communication, as part of a nine-week introductory class for new officers. They go over basic information theory as it applies to naval communication systems and encryption. The valedictorian of that class, who had just aced the test on digital communications the week before, raised her hand the following week to ask a clarifying question about a follow-on topic: "When you say everything is just ones and zeros...what does that even mean?"

That sums up naval training. I realized then that this person, despite all appearances of knowing the material and being an excellent student, had NO clue what anyone was talking about and hadn't for the entire class despite having the highest grade of anyone.

So you'd have an aircraft carrier CO with a Master's degree in chemistry who didn't really understand the concept of vector addition as it applied to calculating relative wind. He'd be yelling at subordinates for not being able to calculate (improperly) wind vectors and distances. All subordinates were clearly confused by this, felt like the CO, being highly educated and the CO of the ship, must know what he's talking about. If you'd ask the CO to explain what he meant he'd become irate and tell you you were stupid.

The nuclear officers/enlisted knew their shit, and those were some of the smartest and best people I've ever worked with. They'd nod their heads at the aforementioned CO and then ask, "he really doesn't understand vector math, does he?" when he was out of earshot. The rest of the crew would assure the nuclear guys that they must be wrong.

The valedictorian of that class, who had just aced the test on digital communications the week before, raised her hand the following week to ask a clarifying question about a follow-on topic: "When you say everything is just ones and zeros...what does that even mean?" That sums up naval training.

That sums up a big chunk of computer science. What answer did she get from the instructor?

It's a good question, coming from a non-technical person, and I'd definitely give a dirty look to anyone who laughed at it.

> I'd definitely give a dirty look to anyone who laughed at it.

Perhaps that would be appropriate before said person had taken a digital communications class and aced a test on it. After that, however, I think a wry chuckle is very much appropriate.

It's not a good question from someone who aced a class in digital communications. What did the guy think was being communicated?

So the guy accumulated all the pieces of the puzzle before he could assemble them? It's not efficient,but it works as long as you don't get laughed down for finally grounding your facts in something. Sometimes you just have to keep making forward progress, and put things together later.

He just unpacked the black boxes in a different order.

not OP but gonna answer from my experience...

1)...at what? Many of them are very bright and well trained, but seriously, at what? People who go into leadership are often strong managers but that doesn't make them good captains or good seaman. They are incredibly skilled, just at the wrong thing. You gotta remember that at the end of the day it is likely someone under the age of 22 with their hands on the wheel and people of similar age running all the major systems. The promotion path to captain is not exactly hands on.

2)...oh yeah. They have been serving on ships for a long time. But in the Navy, serving on ships does not in any ways equate to seamanship. The Navy (and obviously IMHO) has a little bit of a chip on its shoulder about 'blowing shit up' that likely comes from feeling underutilized in our last to endless wars, and this is especially prevalent in the surface fleet. They tend to prioritize things like warfare officers and sometimes even ex-pilots for promotion up towards captain ranks. Those people are absolutely deployed but doing things totally unrelated to navigation and seamanship as a core mission. They should, but often don't have the time or space, have impeccable seamanship skills.

The article puts focus on Officers, but same is true with Enlisted Sailors. I was an Electrician who was sent out to the fleet with 1 week of electrical training. When I arrived to the fleet I did know the names of the tools I would be using, but Navy was very confident that I could be trained onboard to solve complex problems.

Luckily, I had a Chief who had gone through the old Electrical school and was a capable Electrician. I got to learn a lot from him and by the time I finished my contract I felt like I had become a good enough Electrician.

I am afraid when these guys who had proper Electrical training retire, what will happen to the fleet ? This problem is very rampant in the Navy for almost every discipline.

That's a really good read.

If the Navy really tried to replace helm and watch time with DVD courseware, they are in trouble. How do you teach a cadet how to get used to fading light and natural night vision via a 30min lesson?

Kind of related: http://www.yachtingworld.com/blogs/elaine-bunting/comment-ho...

A professional navigator put a Volvo Ocean Race boat onto a reef, due mostly to not looking at paper charts.

They could of did what the US Army did in coming up with a game simulating real-life conditions...you could simulate fading light, navigation going out,and whole host of things requiring exercise of good judgment..might even up numbers of recruits for the US Navy as a whole..

But the thing about fading light is that it fades on a predictable schedule every day for free. There's no opportunity for anyone to get paid!

Replacing training in natural light with a set of DVDs solves this problem nicely.

> fading light and natural night vision

Have instructions in the video to dim the screen?


Crew resource management is non-existent: Listen to the officer of the watch shout down the other person in the lead up to the USS Porter collision. http://gcaptain.com/intense-bridge-conversation-porter/

This is an area where I think the Coast Guard does a decent job. We have a program called Team Coordination Training[0] which, while monotonous and annoying to those who have to take it every year, aims to instruct students on the importance of clear, non-punitive communication between all members of a crew or team to ensure maximum situational awareness. Some implementations of the training are better than others, but the intent is good, and is a more useful tool than generic risk assessment models that are frequently blown off in practice.

[0] https://www.uscg.mil/safety/cg1131/tct.asp

I was a Junior Officer on a Coast Guard cutter for a year and a half. Without a doubt it comes down to command climate. Our Captain had "Standing Orders" that every deck watch officer was required to follow. This is required for all Coast Guard vessels, and I imagine for all Navy vessels as well. Included in these were reasons to call the Captain, where if certain parameters were met you had to call the Captain (and if its late at night that meant waking him/her up, explaining a situation over the phone, and providing a valid recommendation in accordance with the navigation rules of the road). The JOs I knew who made shitty and dangerous deck watch officers were those who would try to skirt the rules in order to avoid calling the captain. If the standing orders said to call the captain and provide a solution to avoid coming within two nautical miles of another vessel, and the radar was providing closest point of approach solutions that bounced between 1 nm and 2 nm, they would interpret that as not being necessary to call the captain.

So why would a deck watch officer not want to call the captain? Well if the captain refuses to ever admit to being wrong, chews people out for tiny things, tells people not to disturb them, or overall doesnt support their junior members then people become less worried about safety and more worried about covering their ass. A lack of training, crew fatigue and bad command end up with bad decisions being made, and when bad decisions are made at that scale people can die.

Totally. In the audio you can hear in the junior guys voice that he has no respect for his superior officer but is also terrified of him. The officer doesn't have a feel for how his ship handles and doesn't seem to fully understand the COLREGS either.

Wow, you can hear the toxicity in that chain of command within the first 30 seconds of audio. Thanks for sharing.

Plug for David Marquet's book "Turn the Ship Around" about his time improving training / operations on a submarine he commanded. Every one had to announce their intention to do an action before actually doing it, this gave others the chance to identify when someone was going to take an incorrect action before it happened.

It's a really great, short read that I'd highly recommend and gives some more background into the operational issues with the Navy.

That approach works even if you are doing an action alone.

Stating clearly: "I am pushing my commits to production", or even trivial stuff like "I am feeding the cats".

You (a) get a chance to realize if that's a stupid thing to do,(b) get to notice if your action doesn't match your stated intent, and (c) have a better chance of remembering that you did it. It really helps me, particularly when there's a lot going on to distract you.

See also the Japanese train guards and drivers, who point at stuff they are attending to, even when no one is looking.


Reminds me of rubber duck debugging: https://en.wikipedia.org/wiki/Rubber_duck_debugging

Where I work, we'll ask each other to be our "talking duck". Usually just talking through the problem, with no response from the person playing the "duck", is enough to figure out what's wrong.

Seems almost like explicit mindfulness training.

I am going to type "cd /; rm -rf *<cr>"

And having said it aloud, we can fix the mistake:

     cd /; sudo rm -rf *<cr>

haven't got root?

I wonder if he got the idea from Japan Railways, which use a technique called "Pointing and Calling" to ensure every action is intentional.


Seconded; this was a good one. It was on the 2014 U.S. Navy reading list.

Totally seconded. One of the principles he had that I really like is, "Don't move information to authority. Move authority to information."

I'm doing this in my pilot training as well. Makes a difference, sure learn faster/better.

The article makes a lot of sense. If you know ships than it's not difficult to imagine how timidity - an unwillingness to literally "rock the boat" by making a course or engine revs change, waking up a senior officer, etc. - could lead to a situation where two vessels are converging past the point where anything can be done about it. In the merchant fleets, and in yachting circles where more competent sailors are involved, the people in control of vessels are taught that the best way to avoid a collision at sea is to make big, early changes to course and speed that the other vessel can clearly recognize and act on. It's possible that's the one sort of action the current culture discourages.

>the best way to avoid a collision at sea is to make big, early changes to course and speed that the other vessel can clearly recognize and act on.

Longtime USN Surface Warfare Officer here: I can assure you that this is Rule #1 in the Navy as well.

It's rather bizarre isn't it? Not just the failure to maneuver safely in the presence of another ship, but also the fact that whatever sensor technology they have didn't alert them or was ignored. Pretty much every modern commercial vessel these days has AIS and collision avoidance, I'd imagine naval vessels have _at least_ that. Just an odd thing all around, and to have it happen twice is either a hell of a coincidence or there is indeed something systemic at work.

A factor in the McCain collision will likely be the merchant ship's over-reliance on AIS and radar and not getting a visual check before turning into the path of fast-moving ship that a.) doesn't use an AIS transponder, and b.) is purposely hard to see on radar. There will be certainly enough blame to go around to hit McCain's leadership, though. ("They should have known that the merchant couldn't see them well and been prepared for yadda yadda.")

The Fitzgerald collision, however, is a real mystery...

Good points. I would also have expected that they'd be in radio communication maneuvering that closely to a large merchant vessel in a busy seaway.

FWIW, US Navy ships don't broadcast on AIS, though they monitor it. That means the other vessels around them have to look out the window to notice them (or get hailed on radio.)

Is it possibly good navy "opsec" not to telegraph maneuvers?

If they're not actively engaged in a naval exercise or actual warfighting requiring them to maneuver outside the Navigation Rules[0], they are subject to the rules the same as any other vessel - and even then, that's a pretty hard argument to make without actual shots fired in anger. Considerations are made in the rules for special requirements for lighting of naval vessels that, by the nature of their design, are unable to comply with the rules. However, the rules regarding avoiding collision apply to everyone, at all times.

[0] https://www.navcen.uscg.gov/pdf/navRules/CG_NRHB_20151231.pd...

I can't imagine that would be the case in the context of at least this latest collision. On a civilian waterway. They don't appear to have been on maneuvers. And by all accounts one of the most crowded waterways you can sail through.

That does not seem like the context in which you would want to not telegraph maneuvers. Either way it's still a training and culture issue if they can't recognize the difference between the two contexts.

It's worse opsec to needlessly crash into other ships...

The Trump school of opsec: Never reveal your plans to anyone, not even your own allies or subordinates. This way your enemy will be maximally confused. So will your allies, but... oh, look over there!

Registration is open for Startup School 2019. Classes start July 22nd.

Guidelines | FAQ | Support | API | Security | Lists | Bookmarklet | Legal | Apply to YC | Contact