Hacker News new | past | comments | ask | show | jobs | submit login
Soma Water Filters Are Worthless (tylermw.com)
149 points by dmitrygr on May 2, 2017 | hide | past | favorite | 102 comments



Many years ago, I had an argument kind of like this with my mom about the difference between decaf Diet Coke and plain Diet Coke. Since she was the Mathemetician in the family, I let her design the double-blind taste test.

When I could conclusively tell the difference every time, she was flabbergasted. But she did finally believe me that there actually was a taste difference between them, at least there was for me. So, she stopped trying to get me to drink decaf.


I performed a reasonably rigorous experiment for my AP statistics class in high school - I tested a ton of different sodas and their diet equivalents across my classmates, and of the 8 sodas tested almost all had a statistically significant bias in favor of the non-diet version. For coca cola, I also tested the non-caffeinated versions - the non caffeinated regular Coke was the top performer.

The only exception to this was diet SunKist - it handily beat regular SunKist, by a large margin.


This shouldn't be too surprising, as caffeine itself has a flavor. It is somewhat bitter, which is one reason why it's particularly common in root beer.


Did you mean that it's not common in root beer? Wiki says that root beer has little to no caffeine.

Root beer is so sweet that I would think it could mask the bitterness. I wonder if the plants that originally were brewed into it just didn't have any caffeine.

Do the ingredients to coke naturally have caffeine, or was that a substitute for the original alkaloid? I.e., is decaf coke because the removed the caffeine, or is caffeinated coke that way because it was added?


Coca-Cola lists caffeine as an ingredient in the UK.

I don't know why caffeine was originally added - maybe for flavour, maybe for the stimulant effects, possibly because of having to not have cocaine in it anymore...


The "original" Coca-Cola had caffeine as well - it's in the cola nut, which is where the "cola" part of the name comes from.


Root beer doesn't naturally have caffeine, but sometimes it's added for flavor.


(pedantic mode on)

That reminds me of my mom, always calling it decaf. There is a difference between decaffinated and caffeine free or non-caffeinated. Coke can be non-caffeinated, meaning that it can be made without adding caffeine. Decaffinated implies that there was caffeine present at some time, but was taken out (like coffee).


That's only half the test. The other half is to have them decide which is which.

The human taste is incredibly good at discerning two different flavors. Not so good at identifying them.


Decaf diet coke is the nost useless thing anyway. You get neither coffein, it is just water with some taste.


That people exist who drink decaf diet sodas boggles the mind.

Why not drink carbonated water at that point? Or just water, you know.

Or drink real soda in moderation.


A general drink typically has— a) a certain taste b) a certain mouthfeel (due to the CO2 in case of sodas) c) caffeine and / or other substances

People might want some or all of the above.

For eg. I sometimes drink decaf Starbucks blended drinks at night because I don't want to mess with my sleep cycle but I like their taste.


Decaf still has caffeine. Maybe not enough to affect you personally, but pay closer attention and maybe it does.


Yeah, switching to decaf coffee still set off my addiction, just not as strongly as full strength. It still has something like 10% as much as normal coffee. I can still detect the slight high.


It is low enough to not have bothered me personally. And I can attest to that because the level of caffeine in a regular coffee would definitely be noticeable (to me).


I don't understand your point in the slightest.

I don't (usually) drink cola for the caffeine, as far as I'm concerned that's just there as a legal way of getting customers chemically addicted to their product. I didn't think many people (consciously) bought it specifically for the caffeine at all.

As for diet - I also don't drink it because I want to ingest sugar. I have plenty of food energy available to me as a first world person living in 2017 and if anything, too much of it. The sugar is there purely to make the drink taste good. I didn't think anyone else really drank it for the purposes of ingesting sugar either. In some drinks (No for Dr. Pepper, but yes for Pepsi), I find the artificial sweetener tastes perfectly acceptable.

I drink cola because it tastes good and satisfies thirst. Not because I want caffeine or sugar. Those two things are bad for me.

Since I don't want the caffeine, and I'm satisfied with the taste of artificial sweetener, I drink caffeine-free diet cola. It still satisfies the goal of quenching thirst and tasting good.

What's your argument?


Of course it's carbonated water, with some artificial sweetener and usually a little acid bite. That's what people like and that's why they drink it.


It can be a really handy diet tool, depending on your situation. I lost 35% of my bodyweight and part of my reward system was a diet soda. Mostly caffeinated, but when I've had too much caffeine, like maybe I took a Jetalert or something, it's nice to have the no caffeine decaf option. Zero calories + sweet is very nice on the system when you're aiming to get the most out of the calories you are eating.


I'm one of those who prefer decaf diet sodas. I have sleep issues so I don't like taking any amount of caffeine after a certain time, and once you get used to diet sodas, regular ones don't taste as good (on top of the ridiculous amounts of sugar that each one has).


Same. Also, real soda is poison.


I agree, one of the reasons I drink coke is for the sugar and caffeine hit that it provides. Take that away and I may as well be drinking fizzy water. Coke was an energy drink for me before red bull ever existed. For the record however I don't drink energy drinks, I used to mix my vodka with redbull when I was clubbing but the stomach cramps it gave me got old quick.


You could say the same thing about decaf coffee. If the caffeine delivery is the only reason to drink it, that is.

But then if that's the case, you can also ask why everyone doesn't just drink water and take a caffeine pill.


Caffeine triggers seizures for me, so un-caffeinated is a must. But yeah, mostly what I drink is just fizzy water from a sodastream.


Maybe I didn't do a good job scanning the comments, but I'm surprised nobody (not even the article) mentions the NSF certification body.

Taste is important but very subjective. I don't buy a water filter unless it shows up as certified by NSF[1].

A few years ago I almost fell for a "water purification system" on Amazon, with stellar reviews that was supposed to filter out everything. With a baby on the way I figured it was worth the price tag. Good thing I did my research. The water filtration system was not certified or verified by anybody.

edit:

Same thing applies to humidifiers, air cleaners, air conditioners, etc. Don't buy unless it has been certified[2].

[1] http://info.nsf.org/Certified/DWTU/

[2] http://ahamverifide.org/


Well, maybe it's a bit more complicated than that. It's certainly true that if you're using a water filter to remove something dangerous rather than remove "hardness" or "chlorination", then you're going to want a certification (and regularly test to see if the RO filter is still working). I'd put ship, cistern, and catchment on that list of critical filters.

However, if your decision is primarily based on flavor (as the article's is) or mineral build-up, then it's not clear that it's so important since you're drinking the water anyway!

I'd note that one brand you may be referring to (although it's true for others) has both "residential" and "professional" series. The latter are WQA Certified (NSF/ANSI 58 & 53)... although you won't find them listed on the NSF website, they are both ANSI and SCC (Canada) certified.

http://yourwaterfilterguide.com/difference-apec-water-reside...

https://www.wqa.org/learn-about-water/faqs


Thanks!!! Safety above taste seems obvious for a water filter.


Tap water has more stringent safety standards than bottled. So unless you live in flint, then safety shouldn't really be an issue.


Not just Flint.

https://www.wsj.com/articles/pittsburgh-tries-to-avoid-becom...

"A total of seven U.S. water systems, which each serve more than 100,000 people, had lead concentrations above the federal action level of 15 parts per billion in recent months, according to Environmental Protection Agency data. They include Portland, Ore., and Providence, R.I., which both exceeded the limit at least one other time in the past five years."


Or you could get a conductometer and measure the purity yourself.


What is the equivalent certification called for air cleaners?


Apparently the author is married to a water filter, but at least she's stylish.

> This experiment was inspired by an argument with my wife, a stylish but atrocious water filter, and the explosion of start-ups attempting to turn everything you purchase into a subscription service.


It's got the Oxford comma. What more do you want?


That would explain this:

  > How I Used R To Win An Argument With My Wife
I don't know about author, but the set of arguments with my wife that can be won using R is pretty much empty. :)


Aren't we all just complicated water filters given a particular perspective?


i'm rethinking my whole life now


If only English had precedence rules!


This; right here; is why I replace all of my punctuation with only semicolons;


Hahaha. But I think the author is just listing the things that inspired the argument.


Yes, like an atrocious wife


Very nice setup, but he didn't show he couldn’t tell the difference between tap and the Soma water.

A biased subject who can pick out both untreated water and Soma-filtered water from the other variants could lie about that, and claim that untreated water and Soma water taste the same.

As a thought experiment, assume the variants are plain water, orange juice, and beer. A biased subject could easily claim orange juice and water taste the same and look the same, and are easily distinguished from beer.

Also, but minor: this is about taste. Such experiments typically are done in the dark, to rule out an effect of how the water looks on taste judgments.


I agree about the risk of bias. If he cheated, even subconsiously, on the tap vs Soma test, the results are consistent with bottled and Soma being indistinguishable. Which is the opposite conclusion.

The suspicious inability to distinguish bottled water from Soma agrees with this. He wouldn't be able to cheat on that test if he actually couldn't tell the difference, but could cheat on tap vs Soma like you said about water and orange juice.

If he didn't even know which two types he was comparing, that would protect from that kind of bias. However, from the setup, it looks like he does know.


If you think he's lying he could have made up the entire blog post too... :P

But more seriously, how would you suggest designing an experiment that worked even with liars determined to thwart it?


You don't need biased subjects to have ambiguous results. Suppose for instance all people from group A can taste chlorine at 20ppm, and below that threshold, those individuals have no perceived chlorine taste. Individuals from group B can taste chlorine only at 80ppm and above. If the tap water has chlorine starting at 100ppm and the filter removed half (as stated in the article) then we would have water at 100ppm and 50ppm. Some observers would perceive a difference and some would not. If the Pur filter removed 90% of chlorine, then individuals in both groups would perceive a difference in taste.

To point, this experiment depends on your taste preferences, as the author has stated that there is a physical difference in the makeup of the water samples after filtration.


He would have to convince his wife who participated in it, so he can't lie about everything. To her, he can pretend two water types taste the same when really they don't, but he can't pretend the opposite.

He can also lie to himself with subconscious bias. The point of blinding is that you can suffer from this even if you're trying to be honest.


Why was Soma so hard to distinguish from Bottled? Article seems to gloss over this point?


This bothered me a lot too.

If he can tell the difference between 'tap v bottled' but not 'tap vs soma'... then 'soma vs bottled' would intuitively have a different result if his conclusion is 100% correct.

I suspect there's more to it, but the author was focusing on winning an argument so... I guess that killed the curiosity required to dig further for a more thorough conclusion.


It wasn't and I think that's the point. Essentially soma was hard to distinguish from tap water, which OP had said tasted chlorinated, given that soma was supposed to filter and improve tap water it was failing to meet its stated objective.


gp is right,

there was 8/23 wrong classifications of "Bottled vs Soma" vs. 10/23 wrong classifications of "Tap vs Soma"

In Figure 6 , the author shows there's no statistically significant difference between "Bottled vs Soma"

So like he wrote :

> Soma filtered water performed the worst, having a taste statistically indistinguishable from tap water

I could write using the same data:

Soma filtered water performed the best, having a taste statistically indistinguishable from bottled water


He says he can differentiate tap v bottled.

He also says he can _not_ differentiate tap v soma.

So the difference between soma and bottled should be obvious (or there is some other factor in play), but it's clearly not.

Obviously taste is not commutative, but there's something fishy here regardless.


For further reading: The Difference Between “Significant” and “Not Significant” is not Itself Statistically Significant (Gelman and Sterns, 2006): http://www.stat.columbia.edu/~gelman/research/published/sign...


I recently did a similar blind taste-test for mineral waters vs my tap water: https://www.gwern.net/Tea#water-experiment

I used a lazy susan to implement randomization/blinding, and an interesting Bayesian variant on best-arm finding to try to optimize sample selection. I found my tap water is weakly in the middle and perhaps 1 or 2 mineral waters were better, but not strongly so.

The results are not quite comparable to OP because I designed it in terms of deciding which in a pair tasted better, not whether I could guess which was which correctly.


Eh, I get the author is being sarcastic, but for someone who buys water filters for their looks, his Soma jug looks just like a slightly plump version of Brita. (What's wrong with Brita, anyway?)


in stereotypical gender role parlance, it does not pass the "wife test".

also filters are overly expensive and useless. issues that the fancy filter only aggravates, so...



If there are two and only two roommates and they disagree, and they settle an argument with evidence-based results, what do you call that?

If there are two and only two roommates and they disagree, and they fuck but are not married, and they settle an argument with evidence-based results, what do you call that?

If there are two and only two roommates and they disagree, and they fuck but are not married, and they differ in gender, and they settle an argument with evidence-based results, what do you call that?

But if you have a married couple, and while adhering to traditional gender roles, they happen to disagree about anything, uh oh! Everybody sulks about the patriarchy.


It may be worth noting that if your water is only chlorinated with chlorine (and not chloramine, which is quite common now as well), leaving it in the fridge overnight can allow some of the chlorine to evaporate, making your water taste better even without a filter.

In this case he allowed both his tap water and filtered water to cool overnight, which would tend to understate the taste difference between freshly pulled tap water and freshly filtered water (if that filter effectively removes chlorine).


Fun little article, but I find the overall premise a bit weird. I always thought of water filters as something that you use to remove harmful things in the water, not improve the taste. Filtered water may taste different, possibly even better, but that's not the goal when I use one.


I have lived in cities where the tap water, to me, was undrinkably gross tasting. It was probably just fine to drink, but I found myself drinking less water than I should because of the taste. A Brita filter helped a lot.

Edit: and not all cities I've lived in are that way. In my current city the tap water is just fine.


The two times I've made major moves in my life I thought the new water tasted awful. In both cases I acclimated and the water was tasteless within a month or two.


My parents have used a Brita filter for years.

For them it started out because they were fed up of descaling the kettle, as they live in a very hard water area (lots of dissolved minerals) and the kettle would scale up easily. But they also always use the filtered water for drinking, because they prefer the taste now.

My sister doesn't, she drinks it straight from the tap when she's visiting.

But it's not at all about water safety. It's just about flavour and the longevity of the kettle (and the iron).


Not in the US. The vast majority of US tap water is perfectly safe to drink. The only reason to filter it is taste.


For me, it's a hedge against incompetence by the local utility (e.g. Flint), a hedge against problems with my own plumbing, and (for the most part) to shut up that part of my brain that worries too much.


It's definitely my goal -- I assume my water's perfectly healthy, but removing the chlorine makes a world of difference in terms of taste.

But I live in a city. If you're in the country and drink well water, your priorities might be the complete opposite.


Some well water also has a very high iron content which could taste good to some but to others it tastes overwhelmingly metallic. This is the case in a lot of mountain areas where drilling deeper than normal to hit a purer part of the aquifer is more expensive.


These two things are not necessarily uncorrelated.


Right, but neither are they necessarily correlated, and they might even be correlated oppositely.


The part I find interesting is the 100% success rate between bottled and tap, which is something bottled water drinkers insist is true (being able to taste the difference, that is), but are often told is just in their heads.


I live in south Texas, where the water is quite hard, and therefore somewhat bitter. There's no mistaking the difference. I've always described bottled water as being more pure; to me it has no taste (as it should), and filters seem to do a decent job of masking that taste out.


Why do you think water should have no taste? Good water should has minerals and those have lots of flavors.


I think it's because the "bad" taste of water (bitter, sour, or kind of metallic) in the areas where I live is so very strong, that "good" water is tasteless to me by comparison. I'm sure there's something there, but it's really subtle.


It will depend on your tap water.


It also depends on the bottled water. Different brands have different mineral content, and sometimes other things that can affect taste.

Furthermore, I think the packaging may affect it sometimes. At least that is my guess after buying a case of small bottles of one particular brand to see if they would be more convenient than filling my own bottles with filtered tap water, and finding that the damn things had a strong flavor that is what I imagine you would get if you ate an ear of corn on the cob for lunch at a picnic on a hot day, and then left the cob sitting out on the table for a few hours until it gets to that state where it has that "it was once part of food, but now it is garbage" smell, and then stuck it in your water and let it sit there for a few more hours, pulled it out, and then drank the water.

Since I cannot believe that they intended this flavor, I'm assuming it is something that leached out of the cheap plastic they used for the bottles.


I have a Soma water filter... their filters used to work well and quickly, then they "improved" them and it takes 10 minutes to fill the pitcher. Yeah.. i'm probably gonna toss the top soon and just enjoy the glass pitcher


"Initially I thought just to buy a Brita, but Brita filters always seemed like something you’d shove in your dorm room mini-fridge and not display on your kitchen counter."

This is a pretty abrupt context switch after coming from the article on stereotypes, trailer homes and rural folk that's on the front page right now.


> this point I just decided to throw the pitcher into the fridge (cold water masks poor taste) and use it for a few weeks to see if there were any changes.

if chlorine was the actual cause of the bad taste, wouldn't the water "aired out" the chlorine after a hour or so?


> First, I filled all the pitchers the previous night

Chlorine evaporates. If he's worried about chlorine taste this step - allowing the jug to sit overnight - will fix it.


I have a Soma pitcher and PUR pitcher at home (I drink enough water to need both). They taste about the same to me.


Why not a reverse osmosis system? Not terribly difficult to maintain and they are much better than a single carbon filter at removing impurities and improving the taste.


I didn't even know these existed until reading your comment. Finally, a new topic to obsessively research :) The last one was indoor air quality and air filters. It sounds kinda lame but it's a really fascinating topic (with a very scammy industry attached to it).

Do you use a reverse osmosis system? Any recommendations?


> Do you use a reverse osmosis system? Any recommendations?

I got a watts premiere from costco years ago and it works ok. incoming water is around 500ppm TDS(total dissolved solids) in an area that has very clean tap water and comes out about 20ppm(bottled RO can get to around 0-2ppm with the industrial filters) with my water pressure. you need to change the filters once a year or so and I get them on amazon for like 20 bucks or so and a cup or so of hydrogen peroxide to sanitize it. you need to change the membrane every 3-5 years or so depending on water quality and that's about 40 bucks on amazon right now. Whichever way you go, you will want a multi stage ro system with a sediment, carbon (I can't recall which of the sediment or the carbon is doubled up but one is) and polishing carbon filter post membrane.

I used to work on ultra high purity water systems for pharma so I probably care more about water quality than most consumers. If you have more questions ask away.


Cheers. Do you know if these kinds of filters will remove fluoride from water? I ask more from the 'dental health is good' perspective rather than the 'mass mind control' one :)


I do believe RO takes out fluoride and pretty much everything. Bottled water doesn't have fluoride either to my knowledge.


Did you settle on an air filter? The obviously scammy industry always turned me off every time I looked in to them.


I did. Ended up going with a 'Rabbit Air Minus A2', which I think mostly uses components from a popular Korean brand (can't remember the name). Good customer service too; they've been pretty accommodating given I'm in Australia.

With air filters, although there's a lot of 'magic beans' out there, it really just comes down to having good quality physical filters (usually multiple ones in sequence, like particle, hepa and carbon) and a decent 'clean air delivery rate' (CADR) given the size of the room.

Heck, you can make a reasonable homemade air filter (that has measurable effects on air quality) by strapping a large HEPA filter to a box fan (although be aware that this could have safety issues re: fire hazard). Steer clear of anything that puts O2 into the air that's anywhere near EPA approved 'safe' levels (which are probably still too high).

Some folks use O2 generators for mould remediation when the house is unoccupied and subsequently ventilated (known as 'ozone shock treatment'). However, I think even this is a potentially dangerous practice as O2 will react with other things in the room as well (e.g. furniture varnish), potentially releasing volatile organic chemicals (VOCs) into the air.

When I get back from work I'll try and dig up a really good review site I stumbled across. It's one of those gems where the author is clearly just really really into air quality, and who has a very good nose for BS 'science'.

EDIT: Oh, also, another good option to deal with some VOCs (like benzene) are certain kinds of houseplants. NASA studied this so there's a fairly credible list of plants to choose from: https://en.wikipedia.org/wiki/NASA_Clean_Air_Study

EDIT2: Also, assuming you don't live near a major highway or something, one of the best things you can do to improve your indoor air quality is open a few windows and blinds. EPA studies suggest indoor air quality is usually far worse than outdoor air quality. And opening blinds lets more sunlight in, which will retard mould growth. And it's all free. Free!


Not OP but I've also researched this topic and settled on an iqair filter back when I lived in China. I did borrow a particle monitor and the drop in the number of particle was flagrant when turning it on.

Subjectively it also did a lot to improve the air in the house... I'm not sure that something like an iqair is needed in a place that isn't as polluted as China (we sold our air filter when we moved out)


The big Honeywell cylinders w/ charcoal filters.

https://www.amazon.com/Honeywell-50250-S-True-HEPA-Purifier/...

At one point a couple of these handled 6 cats and 5 ferrets. The scammmy stuff made no dent. Commercial use hotel + mold ozonaters worked, but unsafe.


When does Soma sue this guy? This seems like a risky article to me. Whether or not they have a case, doesn't matter. They can make life miserable.


You aren't going to be married long.


This was an All in the Family episode. For whatever reason, Gloria can;t tell the difference in taste between Coke, Pepsi, and RC but Meathead can.


"HOW I USED R TO WIN AN ARGUMENT WITH MY WIFE"


I prefer a Berkey myself.


But do you have the evidence


Yes. I tasted the water and I liked it.


Sample size of 1 works for me. In all seriousness though in issues of personal taste as long as you are happy with your decision then it's the correct one.


"...a practice pioneered in the 90s by “8 CDs For A Penny!” Columbia House"

LOL. Try 1950s. I love how Millennials think everything was "pioneered" in the 90s and 2000s.


I wonder how much efficiency and productivity hours are lost to stupid preconceptions like "Brita isnt something you put on a kitchen counter".


Aesthetics matter and make us happier. It also signals things about yourself regardless of if you mean to signal it or not.


Soma sent out a really self-congratulatory marketing email on Black Friday a few years back, telling people, "Don't buy a Soma today." gushing about living in the moment and enjoying time with friends and family instead of shopping.

I wrote back and asked if they were so serious about that message, why didn't they take the Buy button off their website?

Shockingly, no response.


I wrote back and asked if they were so serious about that message, why didn't they take the Buy button off their website?

If you can't readily answer that for yourself, I wouldn't have replied to you, either.


Send a stupid email get a stupid reply.


Hahah that's so pretentious what they did


Freedom of choice.




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

Search: