Hacker News new | past | comments | ask | show | jobs | submit login
Much of the “science” used in design is bullshit (2014) (mjparnell.com)
330 points by spking 8 months ago | hide | past | web | favorite | 137 comments

My experience is that most people in business are 1) terrible at constructing trials, tests, or experiments, of any sort whatsoever, 2) terrible at interpreting the results of same, and 3) have a very limited tolerance for discussion of how bad they are at it and why their various tests are meaningless, no matter how gently it's presented.

I've not seen a place where this isn't broadly true, no matter where the business folks come from (top-tier B-school analysts, Bob's discount MBA emporium, doesn't matter) or how good the reputation of the firm they're at. This also extends to management and administration in the public sector—schools are rife with obvious pseudo-science bullshit, bad attempts to replicate results while skipping half the measures that were taken, et c., but dumb-ass superintendents (you would not believe it, seriously, "dumb-ass" was not chosen lightly) and principals eat it up.

[EDIT] which is to say I'm not at all surprised design "science" is full of BS, because it's sold to people who almost all suck at evaluating those kinds of things. Any "science" that largely exists to sell stuff to middle- and upper-management or "stakeholders" probably tends to be awful, because it doesn't need to be good.

Remember that academia has a longer history as an accreditation vehicle for any concept you can come up with as long as you write convincingly enough, than as a platform of science since academia predates science by centuries if not millenias.

I think this tradition of authoritarian bullshit has some traction in the modern world still. Which in part causes all the silly pompous pseudoscientific bullshit.

Completely agree. In my experience every client I've worked with that wanted 'User Testing' or 'A/B Testing' would have lofty goals of having an awesome product backed by science, but at the end of the day were unwilling or unable to actually execute it in a way that did anything other than smush the data into the shape they wanted. They already have results in mind, they just want some data to support their conclusion so they can put it up on a slide for 10 seconds in a meeting.

I would go much further and say almost all "science" is done poorly.

The scientific method, when applied rigorously (!), is deeply uneconomical in the vast majority of cases, because the market has already eliminated most glaring inefficiencies by pure trial and error. Any possible gains are likely marginal anyway.

Doing a bunch of half-assed and meaningless experiments just to slap the prestigious "science" label on something on the other hand is very economical. So that's what people do instead.

Also, if you've studied human psychology in depth, you realize that the scientific method is a mirror of a very specific set of traits "owned" by a subset of humans.

If you're on the outside of that subset, it turns out that not only is it easier just to slap the label on, but your ideal customer (relative to your psychology) doesn't really care to do much more than a simple label check anyway.

From zero to 'rm -rf science-budget' in one marketing exercise.

A sad truth but coming around to it can help us understand why scientists need to work on social policy and outreach messaging as well. Scientific values simply will not sell themselves the way we think they should, outside of our little sphere.

Fortunately our cultural messaging has vastly improved in this domain recently. Bridge psychologies like that of Bill Nye are great examples of how this can work.

I'm also interested in what motivates this behaviour. Anecdotally I feel that the current incentive structure is entirely financial. People may get into a field to do good, but they have to justify an ROI on everything they do. Society fetishizes financial success and universally condemns financial failure.

Live in that world as a struggling student scientist, juggling the prospects of a "marketable" thesis versus an ethical one, as debts pile up and staring down the barrel of your entire life amounting to being yet another mediocre drone / living on the streets / not being a scientist at all.

The prospect of being completely broke, unable to afford rent, unable to afford food, being an embarrassment to friends and family, these factors exist and must surely inform the judgement and behaviours of students in a position to have a study funded.

I know of phd's still struggling in their careers against the same thing, where an institute is more interested in publicity than the actual science.

The symptom may be half-assed and meaningless experiments but that is not to say the individuals responsible aren't capable or willing of so much more if the appropriate supporting social structures existed to allow ethical scientific thinkers to exist, or if institutions stopped taking in students for the money and instead tested applicants for profit.

It's easy to point out the problem but how can it be solved? To me, I think there needs to be publicly funded institutes that have no profit motive and exist purely to intake the best academic minds, this might seem anti capitalist but as a model it is an investment in the future of a society that is calculated in quality rather than quantity.

By and large, profit does represent societal good, this whole "anti-profit" attitude is misguided.

On the other hand, it is often possible to profit unethically and sometimes that is the most expedient way. There is a cost to this, which is reputational. That's why it's important to call out bullshit.

The problem with research is that the business risk is high so the potential rewards need to be equally high. So there is an argument to be made for large entities (up to and including the state) to fund research, but there need to be checks and balances. The ones we have may not be perfect, but that doesn't mean they're inadequate. It's all just overhead as far as I am concerned.

Knowledge is a product, we should not only be a geek and leave the profit to the others voluntarily. All scientists should stand up to learn how to raise funds, market, sell, and refine. In fact, the business side of the effort is also very difficult and professional. The effort can further help reshape the research. Having a small team can reduce the bureaucratic requirements of prestige pseudo-scientific stamp but enforce the true scientific methology to solve the problems.

The business relationship among people is becoming flatter instead of deeper. Tooling is going to help individuals achieving more independently. At least it's the reality I would like to live in.

a depressingly accurate assessment.

> Any "science" that largely exists to sell stuff to middle- and upper-management or "stakeholders" probably tends to be awful, because it doesn't need to be good.

as long as it's reeaally expensive. that's really the only way to asses the accuracy and importance of these things.

Hell, I'll go a step further. Most people business (especially business schools) are absolutely abysmal at structured thought and reasoned analysis.

Simple shit like p -> q, orders of magnitude, independent probabilities, partitioning[0], and the classic correlation/causation go completely over their heads.

At the risk of sounding arrogant or elitist, there's a reason the peripatetic school inscribed "let none but geometers enter here". By and large, a mathematical education structures thought in a way that nothing else can.

[0] My personal pet peeves. People seem unable to break concepts down into component parts that both (a) cover the entire problem and (b) don't overlap.

So... my business school covered all of that. Probability, statistics, linear algebra, diffeq... and yes, boolean algebra in the context of gradient descent. Even Markov chains and neural nets. Needless to say, there was a lot of math.

Interviews at top business schools are nothing but partitioning and mathematical analysis. Google “business school case interviews”. Modern business is analytical and quantitative. Over 50% of my classmates came from software development or engineering backgrounds. This mix is not at all unusual.

I think your reaction may be to the PHB types — who aren’t good at any of that but they have soft skills (aka people like them) so they get promoted. Honestly the biggest truism in business is that it’s better to be liked than to be good.

>So... my business school covered all of that. Probability, statistics, linear algebra, diffeq... and yes, boolean algebra in the context of gradient descent. Even Markov chains and neural nets. Needless to say, there was a lot of math.

That's an incredibly rare thing. To date, the only systematic exception to the above rule I've encountered have been from certain European schools (mostly French: HEC, ESSEC, etc) that have a math-heavy prep-school curriculum to get in.

From what I've seen though, this is very much exceptional.

A subset of business schools are considered “quant” schools. These schools tend to be pretty high up in the rankings as their faculty often end up winning Nobel prizes.

What school?


Yeah just had to say something because it was 100% opposite from my experience. Like they literally hammered everything he claims business people lack into our heads ;)

What are one-dimensional business people actually good at? Besides just being a piece in the system that soaks up resources?

The good ones are good at convincing, negotiating, charming, manipulating, uniting people, turning people against each other etc. All very important skills out in the real world of human interactions and they translate to $$ quite directly.

Learning and exploiting unwritten rules of a game referred to as business.

Making deals with similar people at other companies.

They're good at talking to people and convincing them of things, which is enormously valuable in business.

I will never tire of reminding people of this: https://www.cbsnews.com/news/pepsis-nonsensical-logo-redesig...

The PDF linked in the article [0] is pure gold.

> C. The investment in our DNA leads to breakthrough innovation and allows us to move out of the traditional linear system and into the future

> BREATHTAKING is a strategy based on the evolution of 5000+ years of shared ideas in design philosophy creating an authentic Constitution of Design.

> B. Magnetic Fields: Magnetic fields exert forces on inner and outer surfaces of the Earth.

> B. Pepsi Energy Fields: Symmetrical energy fields are in balance.

> C. Magnetic Dynamics: Magnetic field are impacted by sun radiation and wind motion.

> C. The Pepsi Globe Dynamics: Emotive forces shape the gestalt of the brand identity.

Also, take a look at the deconstruction of the old Pepsi logos into arbitrary ellipses ("Perimeter Oscillations") on pages 8ff.

It gets increasingly surreal ("Light Path with Gravitational Pull" vs "Gravitational Pull of Pepsi", "Relativity of Space and Time" vs "Pepsi Proposition / Pepsi Aisle", difference between a "Pepsi Galaxy" and a "Pepsi Universe", ...) on the last pages.

Dimensionalize exponentially.

[0] https://jimedwardsnrx.files.wordpress.com/2009/02/pepsi_grav...

This PDF is hilarious. How can anyone get away with so much bullshit is beyond me.

I can't tell if this is real or not. It feels like the Turbo Encabulator of design.

It's one step away from VX junkies.


Because no one is really reading it. It's not there to be read. It's there to exist. You can't pay 7 figures for a simple image file. No. There must be "a reason" for that logo to look like that. So you get someone to drum up a few pages of bullshit to justify paying 7 figures for it.

And the sad thing is that everyone knows it's bullshit on some level. But I'm not paying 7 figures without getting handed that bullshit, and you're not charging 7 figures without being able to create it.

I would pay a million dollars for this as a piece of performance modern art.

They need to justify the spend for the execs with something that looks like it took more than 100 man hours to work out.

What I'm curious about is how much these design agencies recycle their stock bullshit for different clients. Just swap in different logos. Restyle with the new palette of pretentiously named colors. Apply the new commissioned font and voila, you have a rebranding document the C levels can feel chuffed about.


I wouldn't assume that. In these echelons, people are generally aware that they're all bullshitting each other.

I'm curious why your perfectly fine comments are being downvoted.

Money laundering? Tax breaks? Not in this case but as a general rule where there is a big, weird, transfer of money

I just skimmed through the PDF (not that it has much content). I can't believe someone paid for that. Clearly I'm in the wrong business.

Make no mistake, it's a lot of work to get to the point where you get to be the one to charge Pepsi a million dollars for a logo.

Creating the bullshit in the PDF is the easy bit, getting someone to pay for it is the hard bit.

That document is astounding. My jaw dropped, literally, on some pages.

> "Emotive forces shape the gestalt of the brand identity."

This is where I lost it.

And that's when they literally started building a "tiny brain, normal brain, enlightened brain, galaxy brain, universe brain" meme, out of freaking Pepsi logos.

I can't possibly be the only one here who understands what they were getting at, right? It seems silly but it's pretty obvious that they're just trying to get at the intentions and associations behind their aesthetic decisions. It's only as absurd as human pattern-seeking behavior is in general.

> I can't possibly be the only one here who understands what they were getting at, right?

It was both accurate and profound.

Unfortunately the parts that were accurate were trivial and the parts that were profound were incorrect.

The pdf author thinks a light year is a speed and the universe expands at f(x) = e^x. Someone was copying stuff they didn't understand off the internet.

Oh, I definitely get what they're getting at. They just wrapped it up in way, way more window-dressing than was remotely justified. It makes the whole thing look farcical.

Well, you could take that literally, or you could take that as a marketing draft, where every idea written in the PDF would later be used as part of advertising.

Bullshit is what makes good video adverts rememberable. It just has to be unique bullshit.

My favorite is:

    C. The Pepsi Globe Dynamics
         Emotive forces shape the gestalt of the brand identity.

I didn't even know those words were allowed to be together.

Apparently that branding company is also active on the Wikipedia page for the logo:

> The subliminal advertising involved with the Pepsi Globe logo is also extensive. The different logos and packaging designs are purported to represent the human body, rediscovery of the Vitruvian principles and their publication, Chinese art of placement and spatial arrangement and many other representations that may not seem clear or obvious from just a glance at a Pepsi Bottle. The most famous visual representation is the Pepsi Globe logo’s representation of The Earth. The swirling horizontal stripe running through the center of the globe is claimed to provide a visual representation of the earth’s constant movement around its own axis and around the sun. The stripe also represents a naturally occurring electric generator in fluid motion generating and sustaining the magnetic field of the Earth. This marketing has resulted in an extremely recognizable logo and an aid to a profitable venture.

This sounds like someone just cribbed off their syllabus from design school to write up a brief.

It still amazes me that they're persisting with the fat-belly logo. It looks like a fat dude with his belly hanging out. It's not something that, once seen, can ever be unseen, even years later. Makes me chuckle on the inside, every time, and plays its part in reminding me of dangers of pop over-consumption.

I never thought of it as a fat-belly logo, the mockup is pretty funny. I always thought of the logo as similar to a tennis ball. In my country, I remember Pepsi giving away blue/red coloured 3x tennis ball packs with a cases of Pepsi.

Every abstract logo can be interpreted in many ways if people culture jams it.

I was literally stunned how MS Office team was announcing icon redesign for their suite and how some part of the community was mindlessly applauding this breathtaking change as improvement in their lives: https://www.youtube.com/watch?v=YplAU5myNP4

Sure, it's something that looks new and good - minding how current icons are having a rather simple style to put it mildly, but that's not a technological advancement like introducing Ribbon interface in Office 2007. And sadly, that's how MS, Apple and others are trying to portray all these barely significant UI changes - as features.

Don't get me wrong, good looking UI is important but I'd be happy if cosmetic changes would remain cosmetic changes and be treated as such. I can appreciate visual improvements and work by myself and I don't need, want to be instructed by marketing teams to feel "enthusiastic" about such changes.

That icon video was nothing.

They, literally, went full-on action movie for Office 2010.


Oh my.


Incidentally, that video is very nice for showing off global illumination in 3D rendering.

Visio and Access: Kicked to the curb.

I can imagine the thinking on Pepsi's procurement end after all said and done:

"Oh yeah, looks like a modern version of our old logo - good! Seems like there's some golden-ratio-ey stuff behind it too which will have subliminal effects - bonus! Price is abit steep but only exceeds our budget by 10% - don't wanna cheap out either, so let's take it!"

Seems like money well spent in the butt covering department. If their new logo flopped and someone in charge said "I paid some designer $1000 to draw up this new logo" he'd be out of a job and blackballed from the industry. If he says "we spent a million dollars building this logo and had a study done to back it up!" then who can blame him we never saw this coming.

That's almost certainly what's happening. The pricing of logo design and branding is about how much it's worth to the client and how much risk the designer is taking on [0]. This design firm put its name on a logo for an international corporation worth billions of dollars and people made fun of it for looking like a fat dude; that's a major embarrassment whose potentiality was probably considered in the pricing.

The same people chuckling at how nonsensical the design document is supposed to be (and it really isn't) are probably the same people who could talk your ear off about the esoteric particulars of a given programming language or management strategy, which are also, to some degree, subjective BS couched in jargon and seeming non sequiturs.

0 - https://youtu.be/RKXZ7t_RiOE

Oh man, it reads just like timecube: http://timecube.2enp.com/

Per the last 'update' of Gene's:

"In 1884, meridian time personnel met

in Washington to change Earth time.

First words said was that only 1 day

could be used on Earth to not change

the 1 day marshmallow. So they applied the 1

day and ignored the other 3 days.

The marshmallow time was wrong then and it

proved wrong today. This a major lie

has so much boring feed from it's wrong.

No man on Earth has no belly-button,

it proves every believer on Earth a liar."


If any of you enjoy brand critics tearing apart corporate brand and identity redesigns, check out Brand New, the reader comments section is gold:


Thank you @andybak, you made my day. Truly admirable document! This is the highest concentration of BS I have ever seen. I have to check who was the VP Marketing and CEO of Pepsi were at the time. They deserve a place in history for having bought that pipedream.

Someone took the Post-modern Essay Generator[1] to a whole new level it seems...


Pepsi is a global brand, $1M is on the low end for a redesign of their primary logo.

The amount of money is fine; the quality of services rendered for that money is a joke.

I dunno. Not only did they get a logo, they got a pretty great historical document.

Thanks for sharing, didn't know the story. Absolutely astonishing.

What in the world? This can't be real!

And just a couple of days ago there was a paper and discussion on "Bullshitters" on HN.

This always comes to my mind when I have to speak to someone in marketing.

Marketing as a discipline is mostly about numbers (reach, sales, etc.). Ad buys and campaigns. This... this is the work of "Creative Professionals".

I immediately CTRL+F'd pepsi to look for this.

> Everyone bullshits at some point, but there seems to me to be a particular desire to show design decisions as being based in some empirical facts. In some ways this is admirable, and forces designers to justify their design decisions. Analytics have certainly been one cause of this.

As a designer I've always felt that the larger and more important chunk of design work is purely intuitive. Analytics implies time-dependence which is counterproductive to design when included upfront.

The only time quantifiable metrics are used is when a design is married to user experience in the context of a user interface (in the end product). At that point, patterns and practices dictate the baseline from which feedback takes place. This is commonly referred to as a design system and is done on a larger horizontal scale across interfaces (web, mobile, print, and so on).

Company branding, medium | message, target audience, color schemes/themes and other aspects of functional design are comprised more of intuition than raw analytics, in my opinion. Apple provides a great example of design choices that focus on the human aspect first. [0]

Added to that, the fact that market share is happily split amongst vastly different UI approaches is testament to the non-linear nature of design.

While designers are plenty, good design is not. Science based models as cited in the article are there to bring up the rear in a standardized manner but don't provide avenues to true "novelty" that defines great design. Just my two cents.

[0] https://developer.apple.com/design/human-interface-guideline...

> As a designer I've always felt that the larger and more important chunk of design work is purely intuitive.

It's a highly trained intuition. From watching artists work, they have a tremendous amount of experience with color, form, different materials and composing those visual elements, and then with understanding the emotional impact they have on an audience.

> As a designer I've always felt that the larger and more important chunk of design work is purely intuitive.

Quite true. However there is some value in applying rational analysis to design, and doing it can help to be a good professional:

If your design is (partially) based on rules (like following exact ratios in proportions, applying color palette increments and color complements...), these rules create a design space that you can explore by changing the parameters to those rules in a systematic way.

This exploration allows you to generate * a lot * of slightly different possible solutions for the design, many of which you wouldn't have created spontaneously. Ultimately, you select the right choice by intuition; but seeing a lot of possibilities can help you to find details that you wouldn't have considered otherwise.

I totally understand the desire to quote some exterior authority in the face of idiot managers whose nephew/niece once thought about going to art school and thinks that the font choice is wrong...

After all, if it's just your opinion then everyone has one of those, and why is yours so much better? Just because you've spent decades working on and obsessing about design, that doesn't make your opinion worth anything more, right?

But if it's backed up with SCIENCE, well that's totally different. You can't argue with SCIENCE, can you?

I basically agree, I think that most good design work is done based on intuition, but designers should be able to justify design decisions, and criticize them, based on objective criteria, especially for aspects of a design that further some functional purpose of a product rather than being simply aesthetic or decorative in nature.

"the minimum ratio of positive-to-negative emotions humans need over time to “flourish” to be 2.9013"

"It reflects the highest positivity ratio (observed ratio = 5.6) and the broadest range of inquiry and advocacy. It is also the most generative and flexible. Mathematically, its trajectory in phase space never duplicates itself, representing maximal degrees of freedom and behavioral flexibility. In the terms of physics and mathematics, this is a chaotic attractor.”"

They needed someone to write a paper to point out this was pseudoscientific BS? Not only that, but the paper was cited over 1000 times?!

Absolutely terrifying.

> Absolutely terrifying.

Indeed. We seem to be increasingly awash with evidence that the standards of rigour in academic journals across a number of fields are... quite poor.

Peter Boghossian, James Lindsay and Helen Pluckrose set out to demonstrate this in the field of grievance studies by submitting and, in some cases, successfully publishing a series of fake papers:


I mean, in some sense this is highly entertaining for the rest of us, but it is also utterly horrifying.

That's not the first time fake papers have been published and even cited.

No researcher has the time to read all the papers published in their field.

There are no penalties for publishing bad science, and no rewards for debunking it.

Recent moves to stop p-hacking by removing the need to show statistical significance will (if they get through) only make this worse.

Something has to change.

Some people will read this and say to themselves: "I'm in the wrong field". Others will say to themselves: "I'm sure glad I'm not in that field".

> Not only that, but the paper was cited over 1000 times?!

With a bit of mining, you could probably identify other likely BS papers by looking at what they cite. Has that been done before?

> With a bit of mining, you could probably identify other likely BS papers by looking at what they cite. Has that been done before?

Yep. http://www.citnetexplorer.nl/

> Mathematically, its trajectory in phase space never duplicates itself, representing maximal degrees of freedom and behavioral flexibility. In the terms of physics and mathematics, this is a chaotic attractor.”"

Someone at Paramount is kicking themselves for not putting that in a Star Trek script.

I expected bad execution and missed confounders here. I also expected an impossibly-high effect size, the same way random priming effects were claimed to override basically all other factors in our lives.

I didn't expect outright gibberish.

I honestly don't think there is anyone who would claim that the current standards in social sciences meet the same degree if scientific rigor found in the physical sciences.

I can't find it now, but I read a crticism of recent physics papers that analyse partical collisions, because partical collisions are not deterministic and therefore have to be analysed with statistics. The criticism is that the manipulation of statistics to p-hack results is creeping into physics.

I've been involved in UX design for nuclear power plants, air traffic control systems, gambling machines, and more recently lots of web sites. None was "science" but all were very important for their intended audience. I the case of the nuclear power plants and air traffic control systems, there was much more thought and analysis put in by user interface specialists, but I doubt that any of those specialists would consider themselves scientists or their practice a "science".

I would definitely call that engineering though.

As a field, I think UI/UX has a ton of anxiety about it's place in the cannon of design. There's a pretty clear path from that anxiety to these hollow attempts to justify itself as a "serious discipline."

Look at the other paths that visual designers could take, from "pure art" illustration to a much "harder science" like architecture, UI/UX sits pretty square in the middle. So many of the people I've worked with want it desperately to be both art and science. I'm sure we've all had conversations with a UX/UI designer who will ping-pong between space "feeling better" with different padding, and something like Millers Law, or perhaps a misapplied statistical inference from Optimizely.

I'd love to see UI/UX split into a more art focused design field and a more science focused HCI field. This might stop the navel-gazing and impulse toward faux-science, and publications like this: https://www.invisionapp.com/inside-design/why-designers-shou...

I'm a little confused by the thesis of the article, and how it relates to the title.

Unless I'm reading it wrong, this is the thrust of the facts:

* The Android design team describes their design process. * In their description, they cite a paper that suggests giving people at least 3 positive experiences for every negative experience they have. * That paper was debunked.

How does that translate to "much of the 'science' used in design is bullshit"? That bogus paper they cited doesn't affect the actual content of Android's design; it just influenced the team's design process. How is that bad? How does that discredit all the other things they talk about? The FastCompany article[1] linked from this blog post says:

> the CliffsNotes version is that Google creates design mantras from the point of view of the user, like “keep it brief,” “delight me in surprising ways,” and “it’s not my fault.” Each time an Android feature lives up to these expectations, they get a single marble in the good emotion jar. But every time they fail, that bad feature produces three marbles in the bad emotion jar. The marbles illustrate that bad ideas stack up quickly.

Even if this heuristic isn't scientifically proven, does it really result in worse UX than if you don't use the heuristic? I just don't get the vitriol.

[1] https://www.fastcompany.com/1672657/google-s-dead-simple-too...

Wrong premises are still wrong even if they lead to correct result.

The habit of relying on dubious techniques is under question here, as is our collective ability to find better guidance.

Later in the article the writer discusses how the "science" the designers say they were using probably had little to no effect on their design. The design is good, but not because of this principle. What the writer takes issue with is the attempts to justify design decisions (which may be fine on their own) with bullshit science.

"The veneer of mathematics and complexity science – a culturally-powerful yet poorly-understood science – allowed many people to believe a very improbable thing: that a simple model from fluid dynamics could explain the influence of love, hate, anger, sadness, grief, joy, culture, time, geography, war, famine, birth, etc. etc. on human behaviour. It also suggests a degree of gullibility in the academy: researchers were willing to accept the claims of the papers because they didn’t understand the maths. Worse, of course, is the implication that people simply don’t critically read the papers that they cite."

Pretty much sums up the reproducibility problem of studies in the field of Psychology.

No, it doesn't at all. The reproducibility issue in psychology has mostly to do with p-hacking, which is a very specific abuse of legitimate statistical analysis in which statistical significance is brute-forced and cherry-picked (rendering it no longer significant), and is often applied to otherwise sound experiments and hypotheses. The paper in question is literally gibberish.

> [I]t’s better for us to accept that we don’t always have good reasons for our design decisions and that much of our work is intuitive.

My intuition tells me that designers wish they could do that without having their designs rejected. If the absence of bullshit empiricism is punished, who can complain when people start providing it?

Agreed and I I think a lot of it comes down to financial factors. In other words, why are we paying someone so much to make design decisions and how can we be assured they are doing it correctly? Because of science.

Speaking of bullshit designs...

Why does that blog have a load screen that takes about 15 seconds to disappear? If you open up debug tools and delete the overlay, the blog entry is perfectly readable underneath it. There's no network activity during those 15 seconds of "loading". It's just ... bullshit.

Pretty interesting article. The only bit I'd have something to say about it is:

> Indeed, it would mean that the designers would have to add a bad experience in for every three good ones in order to get the positivity ratio right!

Which appears to be incorrect as the positivity ratio was supposed to be a minimum as according to that same article. Presumably having more positive experiences doesn't make the situation somehow worse.

Nevertheless, maybe it's an interesting or even useful heuristic that you could get away with one bad moment for every three good moments you provide, even if it is about as scientific as an analysis of unicorns on a flat earth.

The article is short of specific examples. But as far as UI's, unless you really are genius of design, keep it mostly the same and only do incremental improvements. I get used to most bad UI's by shere memorization of where and how to go. Move the cheese, and I'm lost and cussing.

The MS Ribbon comes to mind. It still don't think it's notably better than the original toolbar. But the problem is that shuffling everything around confused many for roughly a year as they had to re-learn where everything is.

MS traded one randomness for a different randomness. Maybe they were thinking a 5% to 10% improvement is worth it over the long run even if they piss off existing customers during the learning curve. So, either they are idiots, or they actually sat around and did "piss off accounting": intentional jerks who willingly sock existing customers in exchange for future news ones. So we have the idiots theory and the jerks theory.

Another thing, redundancy is not necessarily bad in UI's. Some get caught up in the "keep the tree clean" mantra. However, having more than one way to do or find an option often improves the UI. Perhaps an ideally designed UI could avoid the need for repetition, but most designers are not good (or constrained by other factors) such that they should indeed fall back on some redundancy. Rules for ideal conditions (such as great UI) often don't apply to typical conditions (average UI).

For bigger applications, perhaps put all the options in a table(s) and let the users search for options Google-style. That may be faster than digging around in menu trees. You still have menu trees, but ultimately they are tied to the table. Use synonyms and allow bookmarks to improve look-ups. I'd like to see more experiments in Table Oriented Programming. To me, that looks the future. OOP can't handle complex relationships well.

I disagree - the ribbon was a massive improvement for feature discoverability and usability.

1. Lots of options visible up front. Instead of reading tons of text, or guessing at hieroglyphics, you could find the tool to accomplish your desired task visually - if you want to change text color, look for the color picker.

1. Big improvement to mouse navigation. Navigating nested menus is a bit like playing Operation - you position your cursor on the item you want, then slide it horizontally to the next menu that expanded. But if you drift too far vertically, which was easy to do of you had a disability or a crappy trackpad, you have to start again. With the ribbon, there was no destructive effect to moving the mouse around without clicking. You could always take the shortest path to whatever you wanted to click on.

1. Big improvement to keyboard navigation. Tap the alt key, and every affordance in the ribbon displays its keyboard shortcut. This allowed the same exploratory navigation as the mouse, but in a way that built muscle memory and avoided RSI.

Of course it was a large change, but the ribbon was an amazing improvement.

Many of these features could have been added to the existing tool bar and/or menu. Anyhow, everyone likes things differently and no approach will make everyone happy. If given my way, I would have made it easy to create a "favorites tray" tool bar section because I find myself using a relatively small number of features quite often, and then use the google-esque tabular search approach mentioned for the non-frequent ones.

No surprise.

Much of the "science" used to support business decisions is bullshit, made up long after a decision was made by someone powerful.

There's a David Foster Wallace short story on this subject. It's called "Mister Squishy". Here's an especially relevant excerpt from an overview/interpretation of it that I wrote[1]:

Now it seems that the purpose of the focus group, rather than gathering information about consumer response to the snack-cake, is to improve the design of future focus groups. Schmidt, however, informs the reader through glimpses we are given to his thoughts during the session, that the focus groups have no material impact: rather than using the collected data to make inferences about consumer preferences, it is desirable to end with a nebulous analysis which could conclude one outcome or another based on which direction the client company is already planning on moving in: the focus groups can only confirm a decision which has already been made: a deviation on this will result in the termination of the marketing firm.

[1] http://westoncb.blogspot.com/2012/12/interpretation-of-david...

Focus groups are the "Uber" of management consulting -- low wage independent contractors doing the same work

I'm reminded of this The Thick of It quote:

"The thing is, you’ve been listening to the wrong expert. You need to listen to the right expert. And you need to know what an expert is going to advise you before he advises you."

Which is about politics, but I think high level business decisions are often as political as anything else.

Humans seem to be comforted when there's an explanation for something, even if it's just plain wrong, it seems we don't care that much.

Our whole lives consist of rationalizing after-the-fact what we observe ourselves to be doing.

And rationalizing because we actually have no idea why we are doing it.

There are also two other things at play. Media hype over small scientific announcements that need more study[0], and misrepresentation of scientific findings by researchers or their sponsoring party (e.g. a university's PR department).

[0]: https://xkcd.com/882/

Science still carries significant cultural cachet. It may be the only remaining sources of authority that does. How much longer that cachet can be abused before it is lost is a concern of mine. The replication crisis and general low quality of academic journal articles are especially worrisome in this regard.

Unscientific people are repulsive and serve as tragic human reminders that there is no alternative to science.

This reminds me of how Apple fans used to love to quote Fitts Law to convince themselves that the Macintosh OS UI was any good. Meanwhile, Apple never really followed Fitts Law very well with the Mac OS or iOS and they quite obviously can't even get the most basic functionality right, like Window or File management.

While I can't tell Apple they did the same for classic MacOS, for the latter they did provide guidelines for developers about their design philosophy and while OSX changed visually over the years, for most of the part it works in same way - controls, widgets, buttons do the same. Microsoft tried doing same thing around Vista but they never managed to keep following own rules and OS, Office, Windows Live software releases were always split apart in terms of how they look and work; this still happens even today - Metro interface still shares the ground with classic elements. Not mention that deep beneath all visual changes they made in last 18 years there are still resources dating 9x/NT times.

Fitts's law is why the menu bar belongs at the top of the screen. But not even Apple gets everything right, or applies the psychological principles that lead to the easiest, most efficient UX. If they did, they would've used pie menus a long time ago.

As for file management, that's one area where Apple had it right and then buggered it up. Everyone agrees that spatial Finder was a revolution in working with files visually, then Apple changed to navigational Finder for OS X. Boo.

I remember the spatial finder making my life difficult in the MacOS 7 days. For example, if I had a parent and child directory open at the same time in list view and then expanded the child to drill down to another directory, the finder would obnoxiously close the child directory window. I found the navigational approach in OS X to be a breath of fresh air.

I switched to Mac a year ago after more than 25 years using only using windows.. there are some glaring holes in mac os, but it's 1000x better than windows (and I'm not even going to get into the Linux argument).

So I've used Mac, Windows and Linux since the late 90s. With the amount of experience that I have, I'd bet real money that you're not very familiar with any of them since you have that opinion.

I'd love to hold a contest where we test people's productivity on each platform because whenever I watch someone who "likes it better than anything else" and supposedly knows even how to use a Mac, they're obviously incompetent when you watch them try to use it. It's absolutely hilarious watching someone trying to quickly find the application window that they want to switch to on a Mac, especially if they've full-screened some windows. (You aren't even allowed to have a real maximize function that always works the way you expect it to in every application lmao!)

I find it somewhat difficult to work at a company with a UX team. It is just another stakeholder you have to take care of. One more obstacle to get anything done as a developer.

Most knowledge is inarticulable. It’s not only learned by doing, it’s intimately indexed by our lived relationship with tools.

A culture of evidence and objectivity, well exhibited by this very blog post, has had two implications: (I) actual knowledge that was not based on explicit, technically appealing principles, has been demoted — witness the come-and-go of architecture, for example and (II) people where explicit and technically appealing principles didn’t exist fell for “objectivity strawmen” or, where they were able to socially impose their deeper inarticulable knowledge, they have used strawmen to get the culture of objectivity off their sense.

What’s perverse about this is that people who are cynical about the socially inflated need for objectivity end up making other, more naive, people buy bullshit like the 3:1 ratio; and people who debunk bullshit like this blogger further perpetuate the misconception that insufficient objectivity and bibliographical references are a relevant problem, even while going about the motions of disavowing the role of science (a heightened mode of objectivity discourse) on lowly design-engineering.

Is design even engineering discipline? To its core? Is the world better off when Google does some ersatz research and we start trusting “Material Design” more than the deep civilizational background that artists (industrial or not) are supposed to have at their disposal?

This is a consequence of subjective choices (even when beneficial) being perceived as easy to perform (hence, cheap and unworthy of value) and likely to be incorrect, whereas choices borne out of systematic processes are seen as reliable and accurate.

A hunch isn't always a trustworthy source, but too often it is taken as an always untrustworthy source.

Someone needs to explain to me what is gender neutral typography:


Wow. That somehow explains why we end up having very dull monotonous icons in many things today. One of the GUI desktop applications in my company had a redesign for a new release, and they opt to go for the Material Design[1] icons. I remember that being a Google thing for their Android UX, so I thought it couldn't be that bad.

When the icons are actually used in the application, they look so... bland. They also look mostly similar to each other, basically being rectangular outlines with rounded corners, that it takes slightly longer to recognise which one is which, compared to the more traditional, colourful icons. (I wonder if colour blind people have always felt this way)

I should've prefaced that our application uses Qt5 Widgets as the UI, with hardly any styling. That may have contributed to the icons looking so out of place. If we had used a more metro-looking theme (like the ones used in Adobe PS) they would be a better fit. But our main users are engineers and scientists, who I assume would prefer more traditional-looking UI (see for example Paraview).

Meanwhile, I have no problems with the same icons being used on my smartphone. I guess this is because on smartphones, icons are larger and more easily distinguished even if they're monochrome. And they can afford to be larger because they're mostly used in main menus or swipe menus. In contrast, in desktop applications they always appear on screen (in the toolbar), and monitors are usually further from your eyes compared to smartphones, and you can't afford to sacrifice precious screen estate just to accommodate distinguishable icons.

[1] https://material.io/design/iconography/system-icons.html#des...

While I have no problem with the linked article I think the headline is a bit unfair. How often is BS-ey pseudoscience like this really used and cited by UX designers in the field? It would seem to be mainly an issue among the most bloated, coddled, and self-important UX teams.

And was the BS paper in question really used to design the Android UI rather than justify the design decisions? (The article's author basically makes this same point.)

Except that you don't need to publish journal articles anymore. Instead, combine 1 cup pseudo-psychology, 1 cup vague design principles and 1 cup quotes from Donald Norman and Massimo Vignelli. Sprinkle with Kahnemann on thinking fast and Pinker on evolutionary strategies. Publish it on Medium and watch it go viral - pop design science and Medium's business model being perfectly aligned right now.

It's noteworthy that Alan Sokal was involved in debunking the paper discussed in the article.

He's the guy who caused an uproar in academic circles (in mainstream media too) with his notorious 1996 hoax in which he submitted a paper loaded with comically absurd assertions to a postmodern journal called Social Text and it got published. Google "Sokal hoax" for a good laugh.

Once after I had a bad experience with adding a "FAB" button to an app I got the chance to ask a Google Material Design designer if they had AB tested or user tested the design and they said no, obviously, because users don't know what they want. So. Much. Bullshit. in design.

That's just a bad designer. The foundation of UX is doing actual user research.

No, I'm talking about the Material Design guidelines that Google has been pushing for years. They publish their design system without testing.

Just a few hours ago in the frontpage: Bullshitters: Who Are They and What Do We Know about Their Lives?


The problem with Android and UX is that they think human experience can be treated like a stat.

The fact that there is no better mobile UX than that provided by Google's Android 9 (Pixel phones, not Samsung) suggests that whatever they're doing is working.

Seems like an opinion rather than a fact.

“Design by committee” is always a bland, sterile and soulless experience.

"Yea, can you just give it a bit more 'zing'?"

Does this explain why Google went from the tasteful white-on-black status bar of Android 8 to the enormous blue balloon PlaySkool buttons on Android 9?

It wouldn't be an android release without Google making random nonsensical to the status bar.

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