Tom Roeder's Dry Yeast vs Liquid Yeast Experiment

[Bobby_M]

Interesting?, definitely. Definitive?, hardly. Repeat the experiment another 99 times, 9 times exactly the same and 90 times changing the malts, hops, temps, length of primary and secondary fermentation, etc and you will start to be able to draw an educated conclusion on the differences.

I was wondering when the voice of reason would finally show up and make sure we all don't go patting ourselves on the back for finally putting this debate to bed once and for all. I'm looking forward to tasting your A/B contribution.

Douchy Comment? Likely.

 

[FatherJack]

Subscribed.

I'm a huge fan of Safale, especially US-05.

 

[brewit2it]

I was wondering when the voice of reason would finally show up and make sure we all don't go patting ourselves on the back for finally putting this debate to bed once and for all. I'm looking forward to tasting your A/B contribution.

Douchy Comment? Likely.

Sensitive much? Sounds like you're the one who could use some massengales or at least some midol and a hot soak. You'll feel better in a couple days.

 

[dantheman13]

Interesting?, definitely. Definitive?, hardly. Repeat the experiment another 99 times, 9 times exactly the same and 90 times changing the malts, hops, temps, length of primary and secondary fermentation, etc and you will start to be able to draw an educated conclusion on the differences.

I don't think anyone said it was a "defined" anything. You are right though, definitely interesting.

 

[remilard]

I was wondering when the voice of reason would finally show up and make sure we all don't go patting ourselves on the back for finally putting this debate to bed once and for all. I'm looking forward to tasting your A/B contribution.

Douchy Comment? Likely.

I'm curious why this would settle anything not settled in previous kitchen science experiments or in commercial sensory analysis experiments using appropriately sized and trained tasting panels.

You have 4? tasters? Unless all 4 correctly pick the odd sample, the probability of type I error is huge (not that anyone will think about such things before forming a conclusion).

You don't settle things but ignoring basic experimental design as is taught in high school statistics courses.

 

[Bobby_M]

Sensitive much? Sounds like you're the one who could use some massengales or at least some midol and a hot soak. You'll feel better in a couple days.

The guy spent hours of time and at least $40 in shipping to make this happen. At least hold back your hater BS until he actually claims some result is conclusive.

 

[Bobby_M]

I'm curious why this would settle anything not settled in previous kitchen science experiments or in commercial sensory analysis experiments using appropriately sized and trained tasting panels.

You have 4? tasters? Unless all 4 correctly pick the odd sample, the probability of type I error is huge (not that anyone will think about such things before forming a conclusion).

You don't settle things but ignoring basic experimental design as is taught in high school statistics courses.

I hope you caught the sarcasm. I don't know who actually thought any of this was going to be even minimally conclusive, but it's not me and I doubt Tom had any such hopes either.

I think the alternative is to never do any casual experimentation in any public setting and never publish any findings. Afterall, it's best to leave science to scientists because they know all the hazards of error and bias. If people are foolish enough to draw definitive conclusions from such things, they would have taken off the cuff advice from someone who didn't even bother experimenting anyway.

If I find that my preferred sample was dry yeast, I'm not going to stop using liquid cultures but it will force me to try the experiment for myself. Curiosity and actually doing something about it is what this is all about.

 

[BioBeing]

n/m -misread...

Interested to see the result though...

 

[doctorRobert]

I'm curious why this would settle anything not settled in previous kitchen science experiments or in commercial sensory analysis experiments using appropriately sized and trained tasting panels.

You have 4? tasters? Unless all 4 correctly pick the odd sample, the probability of type I error is huge (not that anyone will think about such things before forming a conclusion).

You don't settle things but ignoring basic experimental design as is taught in high school statistics courses.

The most likely scenario, is that the results are mixed, meaning one isnt superior than the other, but that they produce different beers and different people have different preferences. If that wasnt the case then we would all prefer the same beer style.

 

[dantheman13]

You have 4? tasters? Unless all 4 correctly pick the odd sample, the probability of type I error is huge (not that anyone will think about such things before forming a conclusion).

Good point. I don't know if you watched the videos or not, but the difference between the two beers so far sounded to me that it was pretty easy to tell the difference between the two samples. Therefore, I predict that each of the 4 testers did in fact describe the same sample as having more body and hops.

 

[remilard]

The most likely scenario, is that the results are mixed, meaning one isnt superior than the other, but that they produce different beers and different people have different preferences. If that wasnt the case then we would all prefer the same beer style.

The most likely conclusion is that the probability of type I error will be something massive, like 35%, but nobody will have the discipline to not try to draw conclusions anyway.

That is most likely because it almost always happens.

ETA: almost always happens because the sample sizes are almost always ridiculously small. Most panels for triangle tests in industry are 20+ people.

 

[RichC999]

The most likely conclusion is that the probability of type I error will be something massive, like 35%, but nobody will have the discipline to not try to draw conclusions anyway.

That is most likely because it almost always happens.

ETA: almost always happens because the sample sizes are almost always ridiculously small. Most panels for triangle tests in industry are 20+ people.

He's just upset he didn't get a beer in the mail Only kidding.

 

[killian]

I watched about 40 min of video and read 40 some posts with a bunch of people getting pissy and NO PAY OFF!!!! DAMN IT which is which, just tell me I wont say anything.

 

[dantheman13]

I watched about 40 min of video and read 40 some posts with a bunch of people getting pissy and NO PAY OFF!!!! DAMN IT which is which, just tell me I wont say anything.

Sorry, lol! We're still waiting on Nige's tasting.

 

[bsdx]

Well, I'm going to patiently wait for the results but I'm also more inclined to duplicate the experiment myself too after seeing this.

 

[SwampassJ]

I love this site.

Primary vs. Secondary
"I did this at first now I do this and my beer is clearer, cleaner and better tasting."
Someone responds you just became a better brewer overtime and there is no difference.

Compare oxygenation vs airation
"Someone gets personal because you didn't own a lab to test the dissolved oxygen to the hundreth decimal and that oxygenation is useless outside a lager and you should just shake.

Same strain different company (Dry vs Liquid)
Someone is upset that you didn't spend 100x the money and time to rebrew it in 9x the same and 91x different , you didn't send it to the 1000 people required and that you're not doing it with the "the mind of reason".

 

[Jaimez]

What i wana know is were did you get a Sun King Growler if your in Nevada? Lol. I see your using it as your blow off.

 

[Bobby_M]

It's probably not obvious to everyone but the guy running this thing is Tom Roeder and he hasn't posted in this thread yet.

 

[BenS]

I wish he would......
But I am glad to know that now I should consider US-05 and 1056 two different yeasts because no one who has tasted the beer yet has considered them the same beer.

 

[Jaimez]

It's probably not obvious to everyone but the guy running this thing is Tom Roeder and he hasn't posted in this thread yet.

Oh I see, i just realized he's not the OP, Lol.

 

[Bobby_M]

I wish he would......
But I am glad to know that now I should consider US-05 and 1056 two different yeasts because no one who has tasted the beer yet has considered them the same beer.

The beers where not even remotely similar. I almost wish he didn't tell me why he wanted me to taste the two beers and just let me blindly perceive. I guess he set up the yeast A vs yeast B premise so that the tasting would be done in that context. The problem is, I know which beer I think is clearly better but I have no idea which yeast made it. I've made good beers with both.

 

[TheSeether]

I wish he would......
But I am glad to know that now I should consider US-05 and 1056 two different yeasts because no one who has tasted the beer yet has considered them the same beer.

Did you consider them the same before, and if so why?

 

[brewit2it]

It's probably not obvious to everyone but the guy running this thing is Tom Roeder and he hasn't posted in this thread yet.

As I said, I wasn't meaning to be an idiot but just stating the fact that there isn't going to be enough data to draw any definitive conclusions. But that being said it is awesome of him to do this and spend the time and cash to send it for tasting to folks he respects, so I'm quite interested in the results.

One thing I may have missed (or it was left out intentionally to not bias the tasters further) but did Tom mention what the FG's were, or at least if they finished the same?

If I had to make a guess it seems like the most likely outcome is going to be that the dry yeast resulted in what is being judged as the winner which also seems to be the one with more residual hop aroma and flavor. The main difference we have seen is that the US-05 took off slower and never really got as active so presumably it blew off less hop aromatics. It may have even been a degree or 2 cooler during the first 24-36 hours of fermentation. I'd also guess the US-05 finished with a slightly higher FG. Of course I could be totally wrong and it might all be the exact opposite.

 

[JJL]

The most likely conclusion is that the probability of type I error will be something massive, like 35%, but nobody will have the discipline to not try to draw conclusions anyway.

That is most likely because it almost always happens.

ETA: almost always happens because the sample sizes are almost always ridiculously small. Most panels for triangle tests in industry are 20+ people.

I never thought I would hear anyone reference type I & type II errors after I got out of grad school.

 

[jakecpunut]

One thing I may have missed (or it was left out intentionally to not bias the tasters further) but did Tom mention what the FG's were, or at least if they finished the same?

Here's Tom's answer to that on the video where he did his taste test..

Yes, they both finished at exactly 1.012..

 

[bruin_ale]

well.. I'm hooked. I'll admit it, I've only ever used S-05. I hate making starters, I've made great beer with S-05 and enjoy the convenience. I tend to use liquid yeast for specialty stuff like belgians, hefe, lagers, etc. Lately I've thought I should at least try the 1056 and see which I prefer, so this is a timely thread at least to open my eyes that I could see a much more noticeable difference than I may have expected.

 

[biochemedic]

definitely cool....dry yeast is definitely kind of easier to use (and cheaper...)

 

[remilard]

The beers where not even remotely similar. I almost wish he didn't tell me why he wanted me to taste the two beers and just let me blindly perceive. I guess he set up the yeast A vs yeast B premise so that the tasting would be done in that context. The problem is, I know which beer I think is clearly better but I have no idea which yeast made it. I've made good beers with both.

Thats been my experience with 1056 vs US05 so you actually might go 4/4 or whatever but I find these things rarely have reasonable significance because of tiny samples. Sean Terrill's pitching rate thing had a pretty fair sample size given the constraints people who can't hire a tasting panel have.

 

[remilard]

I never thought I would hear anyone reference type I & type II errors after I got out of grad school.

In the right crowd I would just say alpha (or beta but I did not actually reference type two error) which is clear enough given context but I figured "type I error" would give more fruitful google results if someone had no idea what I meant.

 

[BenS]

The beers where not even remotely similar. I almost wish he didn't tell me why he wanted me to taste the two beers and just let me blindly perceive. I guess he set up the yeast A vs yeast B premise so that the tasting would be done in that context. The problem is, I know which beer I think is clearly better but I have no idea which yeast made it. I've made good beers with both.

I did not mean to imply that one yeast makes better beer than the other, only that you will not get the same final beer using either of these yeasts as I had assumed before these videos.

Did you consider them the same before, and if so why?

I did consider US-05 and 1056 to be the same yeast before. I had heard they were both cultured from Sierra Nevada's Pale Ale. I have used them both in my own beers before but never at the same time. Without comparing them side by side, I never thought that they weren't the same.