Ol' Reliable versus Something New--I'm puzzled

Williams 728
Homebrew Talk - Beer, Wine, Mead, & Cider Brewing Discussion Forum

Help Support Homebrew Talk - Beer, Wine, Mead, & Cider Brewing Discussion Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.
To be honest, I'm not convinced that you don't want to argue. You have a couple of favorite beers that yo will always order if they're available. But if they aren't, you'll try a flight to find one that you like. That's great. Lots of other people do pretty much the same thing. But then you claim that there's something inexplicable (i.e., wrong) about people who enjoy the second part of your own approach. You're a smart guy; I bet you understand it just fine. But you set up a "two kinds of peple in the world" dichotomy that must be resolved. IOW, you started an argument (in the academic sense).

It's already been said, but beers can change over time. So can people. You yourself may someday find that you've lost your taste for something. Not Peeper's of course, but something. In fact, there's a pretty good chance that you already have. You may also find that you've acquired a taste for something else. Exploring new things can help one's tastes evolve (or help one realize that one's tastes have changed, for that matter). This is actually a lot less complicated than college stats.

I like single malt scotch. Always have. Used to be a real snob about it - all other liquor was inferior and not worth my time. Fortunately, I got over that eventually and now I have a whole world of other spirits that I enjoy. Maybe you're right and you will never taste anything you like better than Peeper's, but there's still a whole lot of joy to be had in trying to find out.
 
mongoose33 said:
Forgive me, but I'm not employing this in an incorrect fashion. It's simple distributional probability. Period.

Or by another guise, regression to the mean.

I'm pretty up on this stuff. You don't have to take my word for that, but after having taught college statistics for 30+ years, I have a fair amount of confidence that I'm parsing this correctly.
Click to expand...
You are not drawing from the same distribution over time. Let's just say, over the course of your life, you have become a more "sophisticated" beer drinker. That could mean that you sampled from a changing population of beers (beers you would drink). Over time, the quality of the beers you have sampled from has also changed, brewers have improved and expanded styles.

You are also assuming that your choice of beers is random, which it's not.

Back to being sophisticated. That would truncate your beer score distribution by lopping off the left side of the distribution where the beer scores are less. You most likely aren't going to the Anheuser-Busch brewery in Newark, NJ looking for the best beer of your life. Those low scoring beers dropping out will shift your sample mean higher. Your chances of finding a best beer have now improved because the one you thought of as being the "best" is now closer to the mean. (The probabilities change and inflate because they have to sum to 1.) Now if your sophistication stabilizes, you might be back to sampling as you think you were but there are also external changes to the population you are sampling from as well. I wouldn't necessarily argue that your probability isn't very small that you will find a best. Given your statement of trying thousands, that would place your "best" out beyond 3 sigma from the mean. Two sigma from the mean and higher is roughly 95% of the population. Would you be happy with a beer that's in the top 5%? Now there's a 1/20 chance for that, 1/5 for a 4 beer flight, by random chance. Also, if you just had the best beer ever from a craft brewer, the next beer you try from them isn't an independent observation.

As a statistician, you may be evaluating many of your beers on one observation. Do I need to say anything else about that?

I'm happy to have a good beer in a style that I like or sometimes a style that I am not particularly fond of if the beer is well made. At first introduction to a brewery, I am willing to sample a new beer from each style that I like, I'll grab a flight usually. Palette-fatigue is not the reason I might switch, it's usually a desire to explore the other offerings. I'll mix it up if there is a couple of good ones later and I am still there. I might stick with one if it is fantastic. I don't walk into every bar though and order my exact favorite. Really hard to narrow it down to just one anyway.

As a statistician, you might find a statistically significant difference between two treatments but the difference may be of no practical value whatsoever.
 
mongoose33 said:
Forgive me, but I'm not employing this in an incorrect fashion. It's simple distributional probability. Period.

Or by another guise, regression to the mean.

I'm pretty up on this stuff. You don't have to take my word for that, but after having taught college statistics for 30+ years, I have a fair amount of confidence that I'm parsing this correctly.
Click to expand...
You’re not though. I don’t need to give my pedigree to know that what you’re presenting is a logical fallacy. You can’t have “the best beer you’ve ever tasted” everywhere you go. Period. You don’t want to try new stuff because you found something you like and don’t want to venture to new stuff, that’s a fine choice. But don’t try to cover it in some sort of mathematical guise that is built on a foundation of sand and present it as bricks.

Regression to the mean is far more brewery dependent than beer dependent. A good brewery that makes a beer you like is far more likely to brew another one you like than a bad brewery is make any you like. The mean is among all beers in all breweries, not just the one beer you happen to try. So while you’re confident in your correctness, I think you’re just confidently incorrect
 
acrowe said:
Well I guess I fall in the middle on this sort of thing. I can relate this to things other than beer, for vacation times, different people have different preferences. My partner always wants to fly somewhere new and do new things, experience new places. She rarely wants to go back to the same spot. Myself if I had my way I'd stay on Vancouver Island (where I live in the summer) and do the same biking, hiking, fishing and beer drinking we always do. With beer, I enjoy trying new things, and going new places but I also really like to have stuff I know I am going to like.
The original poster suggested this was down to understanding the probabilities. I think there is something to this, particularly for gambling, I am not a statistician (though I do work with stats as a social science prof), and I know that long run you can't win at casinos, or VLTs, its a negative sum game (for players) where the odds are against you. I have never gambled in a casino - I'd rather buy a beer and drink it.
For other aspects I am guessing there are some aspects of neurodiversity (simply different people's brains functioning slightly differently). Some people are going to get a lot of enjoyment out of trying something new, and that rather than simply the quality of the actual beer experienced, is very valuable to them. Other's simply by the way their brain works are going to be more concerned than excited about the risk of getting something they won't really enjoy, and are happy to enjoy the familiar.
There are other aspects to this as well such as age, we get more conservative generally as we age(not everyone but many people) and I don't mean politics (though for many that too). The OP said they'd tried thousands of beers, and if that's true they are probably not really young (or are an amazing drinker). I would guess many people are more open to trying new things when they are young, and as they age they tend to stick with things they know and like. Consider music, in general people experience a bunch of different music when young, develop preferences, then don't shift much after their mid 20s.
Anyway I think its an interesting question, and my not definitive opinion is that some of whether you are always looking for the new favourite or stick with the known and loved comes down to exactly how much pleasure you get out of the excitement of trying something new, and we generally get more of that when we are younger (though not everyone).
Click to expand...
Breaking all beer drinkers into Just Two Types not a very nuanced starting point. I find myself on both sides of the spectrum at the same time. Like me Lagers as standard fair, but seek out new flavors and styles as much as possible. Seems like the emphases of the original author of the thread was more or less an opportunity to crow about his grasp of statistics (a regression to the me) which to be fair, he seemed pretty expert.
 
aceluby said:
You’re not though. I don’t need to give my pedigree to know that what you’re presenting is a logical fallacy. You can’t have “the best beer you’ve ever tasted” everywhere you go. Period. You don’t want to try new stuff because you found something you like and don’t want to venture to new stuff, that’s a fine choice. But don’t try to cover it in some sort of mathematical guise that is built on a foundation of sand and present it as bricks.
Click to expand...

That's the whole point. You CAN'T have the best beer everywhere you go. And so the odds of the next beer you try being as good or better are virtually nil.

And pedigree or no, you're not parsing this correctly. I won't continue to try to teach statistics and probability here; what everyone arguing about this are emotionally tied to is the idea that it ISN'T this way.

Because they don't want it to be. I'm a scientist by training and inclination, and good scientists, which I am, don't let their emotions overrule their reason.
 
Deadalus said:
You are not drawing from the same distribution over time. Let's just say, over the course of your life, you have become a more "sophisticated" beer drinker. That could mean that you sampled from a changing population of beers (beers you would drink). Over time, the quality of the beers you have sampled from has also changed, brewers have improved and expanded styles.

You are also assuming that your choice of beers is random, which it's not.

Back to being sophisticated. That would truncate your beer score distribution by lopping off the left side of the distribution where the beer scores are less. You most likely aren't going to the Anheuser-Busch brewery in Newark, NJ looking for the best beer of your life. Those low scoring beers dropping out will shift your sample mean higher. Your chances of finding a best beer have now improved because the one you thought of as being the "best" is now closer to the mean. (The probabilities change and inflate because they have to sum to 1.) Now if your sophistication stabilizes, you might be back to sampling as you think you were but there are also external changes to the population you are sampling from as well. I wouldn't necessarily argue that your probability isn't very small that you will find a best. Given your statement of trying thousands, that would place your "best" out beyond 3 sigma from the mean. Two sigma from the mean and higher is roughly 95% of the population. Would you be happy with a beer that's in the top 5%? Now there's a 1/20 chance for that, 1/5 for a 4 beer flight, by random chance. Also, if you just had the best beer ever from a craft brewer, the next beer you try from them isn't an independent observation.

As a statistician, you may be evaluating many of your beers on one observation. Do I need to say anything else about that?

I'm happy to have a good beer in a style that I like or sometimes a style that I am not particularly fond of if the beer is well made. At first introduction to a brewery, I am willing to sample a new beer from each style that I like, I'll grab a flight usually. Palette-fatigue is not the reason I might switch, it's usually a desire to explore the other offerings. I'll mix it up if there is a couple of good ones later and I am still there. I might stick with one if it is fantastic. I don't walk into every bar though and order my exact favorite. Really hard to narrow it down to just one anyway.

As a statistician, you might find a statistically significant difference between two treatments but the difference may be of no practical value whatsoever.
Click to expand...

You're trying to make this about significance, which it is not.

Funny how this thread moved away from what I wanted it to be--which is my trying to understand why others don't approach this the same way I do--to people wanting it to be about statistics.

It's regression to the mean. Pure and simple. Use whatever measure of "quality" you like, and it's still....regression to the mean.

I used to teach this to my students in this way: suppose you get a 100 on an exam. What are the chances the next exam will be a 100? Answer: unless it's trivial in nature, the next exam will be lower. The same goes for earning a "12" on the exam, out of 100. What are the chances the next exam will score higher? Answer: pretty good.

[The joke was that I told them to be sad when they got a 100, because it's only down from there, and be happy if they got a 12, because it was likely to improve. They didn't buy it. :) ]

Regression to the mean is everywhere, and those who know to look for it and use it usually make good choices. It's ALL over the stock market, for example. It's in the golf scores people shoot, in the trap shooting scores they shoot, and in the distribution of beers they drink.
 
mongoose33 said:
Funny how this thread moved away from what I wanted it to be--which is my trying to understand why others don't approach this the same way I do--to people wanting it to be about statistics.
Click to expand...
No, you're refusing to understand why people don't approach this the way you do and insisting that it's all about statistics. You've refused to even acknowledge the validity of others' personal experiences, which is frankly pretty damned obnoxious.

But I'll try one more time - by your own logic, I should never try the beer you think is the best in the world, because I've already had the best beer in the world and the statistical probability that your favorite is better is virtually nil.
 
mongoose33 said:
You're trying to make this about significance, which it is not.

Funny how this thread moved away from what I wanted it to be--which is my trying to understand why others don't approach this the same way I do--to people wanting it to be about statistics.

It's regression to the mean. Pure and simple. Use whatever measure of "quality" you like, and it's still....regression to the mean.

I used to teach this to my students in this way: suppose you get a 100 on an exam. What are the chances the next exam will be a 100? Answer: unless it's trivial in nature, the next exam will be lower. The same goes for earning a "12" on the exam, out of 100. What are the chances the next exam will score higher? Answer: pretty good.

[The joke was that I told them to be sad when they got a 100, because it's only down from there, and be happy if they got a 12, because it was likely to improve. They didn't buy it. :) ]

Regression to the mean is everywhere, and those who know to look for it and use it usually make good choices. It's ALL over the stock market, for example. It's in the golf scores people shoot, in the trap shooting scores they shoot, and in the distribution of beers they drink.
Click to expand...
Here's the problem I pointed out to you that just totally glossed over with your exam example. You are not drawing from the same distribution, which is what you are saying if the next exam is trivial. My example-a person grows up in a small town with a limited beer variety yet travels around the state they live in and tries lots of beers but unfortunately the state is ranked at the bottom for beer quality too. They take a trip to a beer capital of the world. Now in your mind Professor, you should be thinking of two nearly disjoint distributions, think two widely separated normal distributions if you like. In that case it's extremely likely the first beer is going to be better than the best beer they ever had, even if it's the worst beer in the capital. Regression to the mean doesn't work on that first beer, because the assumptions were violated. It's not a random draw from the initial distribution. It's a completely different random variable.

You said "As far as approaching everything this way: my understanding of statistics and distributions and probability DO inform most of what I do. It's impossible for it to not: when I know what the probabilities are, I have to choose the one with the greatest likelihood of success." Which I can empathize with but it doesn't work when you violate the assumptions Professor. And your statement plus the several times you've mentioned you are a professor is why I felt compelled to discuss the statistical assumptions. A different way if putting it is, it's a great big world, don't assume you know everything about it, you might miss something extraordinary.
 
mongoose33 said:
That's the whole point. You CAN'T have the best beer everywhere you go. And so the odds of the next beer you try being as good or better are virtually nil.

And pedigree or no, you're not parsing this correctly. I won't continue to try to teach statistics and probability here; what everyone arguing about this are emotionally tied to is the idea that it ISN'T this way.

Because they don't want it to be. I'm a scientist by training and inclination, and good scientists, which I am, don't let their emotions overrule their reason.
Click to expand...
So since you've had the best beer you've ever tasted already, you don't try new beers ever, correct?
 
mongoose33 said:
You're trying to make this about significance, which it is not.

Funny how this thread moved away from what I wanted it to be--which is my trying to understand why others don't approach this the same way I do--to people wanting it to be about statistics.
Click to expand...
It's about psychology and people, not statistics. Some people are "explorers" and others are "stay-at-homes", and it does seem to be a pretty fundamental part of what makes people tick, it's not just a beer thing. For instance I had an ex who would happily watch a handful of old movies like Singin in the Rain over and over again, whereas I can't bear to watch the same film twice. Our brains just tick a different way and yes, that difference extends to beer.

And the world is a better place for having that division. Most people are better off just staying at home and tending their fields, but society moves forward (for good or bad) by having explorers like Columbus or Cortez who take risks that come off. That's not to ignore the fact that things don't go well for most explorers, they end up like the Franklin expedition. Or in beer terms you need the Sierra Nevadas and Bostons to drive the industry forward rather than have people just accept the status quo, even if that means the likes of New Albion flourish and then fail. But overall society benefits from having a proportion of "explorers", which is why they stick around in the gene pool.

You can also justify it in the terms that you're trying to think of, that people don't find the greatest utility in drinking the "best" beer at all times, but they can find utility in exploring the diversity of beer. Certainly as a scientist and homebrewer I find there's great utility in trying beers with different hops and yeast in them, not only in their own right but for how they help me understand the ingredients I already know.

On the other hand, part of it may just be an age/experience thing, as related by Boak and Bailey (but that's definitely not the whole story, part of it is definitely biological) :
https://boakandbailey.com/2017/09/the-seven-ages-of-beer-geek/
 
I like beer.
Do It Fight GIF by Sam Omo
 
Since the OP is a stats geek, and so am I, I will express myself in statistical terms: I see variance -- i.e., novelty -- as an end in itself. I'm not solely interested in (re)experiencing the best beer I have ever drunk, the best food I have ever eaten, or the best place I have ever visited on vacation. I want variety in all of those areas. Sameness -- even excellent sameness -- is boring to me.

Obviously, I prefer to experience the positive side of the quality distribution curve. But not everything needs to be +3 sigma or greater. Even when a venue carries my favorite beer, I won't have more than one glass of it, but will try something else for my second glass. Beyond a general desire for variety, diminishing marginal utility comes into play here. That second pint of favorite beer will never be as good as the first, so why not try something different?

Finally, there's the issue of different circumstances and different moods. If you ask me, "What do you feel like drinking now?", the answer changes from day to day and hour to hour. Am I thirsty because I earlier ran 6 miles, meaning I want something light? Or do I want something rich and heavy that I can slowly savor? Has it been a long time since I had a stout such that I'm really craving one? And so on. This is why at any given time, I have 10+ varieties of homebrew available, plus a few different commercial bottles and cans.
 
Last edited:
hamachi said:
"What do you feel like drinking now?"
Click to expand...



whoa now! feel? normally that combines me thinking psycology AND statistics...what marketing is playing a bigger role in my psyche, and statistically how much they are trying drain my wallet for.....
 
SaisonMan said:
This reminds me of a debate my wife and I would regularly have about camping. She wanted to pack up the tent everyday and go to another place to see something new. If I loved the original camp site and area, I wanted to stay.
Click to expand...
I married a woman who had always been poor. We had to travel to be with each other while she got her green card. When I planned trips, she always wanted to see as much as possible, while I, as an experienced traveler, wanted to spend a long time in each location.

I call her the Japanese Tourist, although she's not Japanese. It's a reference to something a friend from New Zealand told me. He was on a year-long trip, which is something people from New Zealand like to do, since they don't expect to see much of the world during their lives. He said he saw a busload of Japanese tourists stop in front of a famous site in Europe. They filed out quickly and made a line in front of the monument. The person at one end ran to the bus, took a picture of the others, and got back in line. Then the next person did the same thing. This continued until everyone had a picture, and then they got in the bus without going to the site. They must have seen a huge number of things, but they might as well have been looking at photos in books. People who spend time in places they visit have much more rewarding experiences.

Now my wife says I was right. Which I already knew.

That being said, I can never stop experimenting with recipes, whether it's food or beer.
 

Similar threads

BugAC
Stouts/Porters/Dark Lagers Perceived Off Flavor
Replies
3
Views
689
rmr9
rmr9
R
Something I find very interesting
Replies
14
Views
956
seatazzz
seatazzz
Velnerj
I've gone pro!
Replies
10
Views
1K
50calshooter
50calshooter
beersk
Domestic Ingredients
2
Replies
54
Views
2K
beersk
beersk
refect
Measuring Kettle PH
Replies
25
Views
1K
Indian_villager
I
Brewer's Friend 300

Latest posts

Back
Top