Rankings (based on "mean" rating data)

Guess everyone has "plans" tonight or just don't want to bite
This is... kind of... interesting. It may take a long time to do this with the data available but confidence intervals or histograms would be better for making the types of conclusions you seem to want to make based on your second post.

 

Only real issue I see with this is that it's not weighted appropriately. Wyoming shouldn't be #2 with a 3.95 average on just one qualifying beer ahead of say, Illinois for instance, that has a 3.89 average on 20 different qualifying beers. The same applies with Scandinavia being placed ahead of the U.S.

If we re-did the states portion of the list based on a reasonable minimum, the top five would be 1.) Vermont, 2.) Illinois, 3.) Massachusetts, 4.) Florida, 5.) Ohio. In my humble opinion, at least. That's what makes sense to me, statistically.

 

Well, there are a lot of lagers brewed in WI and Germany so...

 

OR, just like all of New England given your numbers above, thinks every beer they taste is the second coming of Christ.

Seems to me what you measured essentially confirms that certain places rate beers higher than others given that Colorado is a beer mecca yet apparently isn't, according to ratings from BA, the place to find excellent beer.

Could just be that some places hype better than others and some others buy into the hype more than others. I can speak for New Mexico, which scores a lot better on this than on other rankings like poverty and child hunger, that scoring 29 is a great victory. However, I also know that all the New Mexicans I know treat are beer as mediocre or good while treating every other flavor-of-the-month as the beverage for the Last Supper.

Either way, appreciate your analysis.

 

Interesting, and I definitely see your point. I think the order is based off of the overall rating averages throughout a state. So beers who brew consistently great beers rank above states that provide some top 250 beers, but don't deliver across the board.

Shame about Missouri, but I definitely see it overall.

 

Thanks OP for putting the analysis together.

A few comments:

1 This data alone does not prove whether one region or country is better than others in terms of beer quality.
2 It does give some clues about that though
3 There are other valid hypotheses that may explain the data as well, such as people in some regions being easy graders or buying into hype more easily.

The main problem with drawing scientific conclusions here is that the people rating in new england are not the same as the ones rating in colorado (what is called an "endogenous problem", one of the main culprits of "bad science", and misinterpretation of statistical results), although there is overlap.

 

If you want to get really scientific with it, scientists never use the word prove. Papers generally state their data "indicates", or use other vague words like "suggests a trend" or "seems to verify the hypothesis".

 

But to touch on the OP, I would highly suggest adding in standard errors man.
Doing so give a better indication of the distribution of data points. If a state/region has a couple of 5/5's and a bunch of crap beers, but the mean is pulled up by the 5's, that will be reflected in the SEM's.

Also I don't know if you have access to the correct software but can you do any sort of P-value calculation on this data? Anova is probably best. But just ask because the differences between most states are very small, in the range of 0.01, so knowing whether or not there is a certain P-value would lend this data much more credibility.

But to the others stating their state is better or whining about why people might be hyping their state's beer or others, keep in mind regions like New England suffer from massive influxes of people from many different places in the country (Jersey I'm looking at you...). So there is definately some credibility to these numbers based on that fact.

 

I'd be curious to see MO, CO and WI if Bud, Miller and Coors products were removed from the averages.

 

Looking at BeerAdvocate notes all the time, I feel like some countries are overestimated. The most overestimated is obviously Belgium. There are a a few trappiste beers out there that are overestimated by Americans rating them, the best two examples being Chimay or Duvel. They're good beers, but they don't deserve to be so close in ratings to Cantillon or Rochefort. There's a bit of that for Germany too. It's really strange to find Weihenstephaner Hefeweissbier in first place. I'm not an expert in German beers, but I've tasted the top 10, and there surely must be a less widely distributed beer that can beat them.

As for other countries, some simply aren't exported enough for the world to be able to judge them fairly. In France, a couple of breweries in Brittany or the Alps do great stuff, but they're unpasteurized and some of the bottles aren't good, due to poor bottling. Australia is also pretty forgotten, as well as New Zealand.

There is also a rating which seems to be style dependent: Pale Ales are a lot less generously rated than IPAs. There's barely 10/15 pale ales in the world that have a 4+ rating, and 5 having 4.5+ while for IPAs it's more like 50 and 15. Why is that ? On the other end of the scale, almost any Imperial Stout seems to attract a 4+ rating average...

As for your results for New England, I'm not surprised, it's really up to Trillium, Tree House, Alchemist and Maine.

As for countries, I would like to see Scandinavia split and not considered a "country". It would be nice to see who would win. Iceland would be last for sure, I think Denmark would be first, with Sweden and Norway close together (Finland isn't a part of Scandinavia for those who wondered).

 

Interesting stuff. I was surprised, like others, about WY being high or MI being low. But fun stuff to read through.

 

It is true that belgian styles are better for keeping. I'm also wondering if there isn't this mystical aura about Belgium in the US, like anything from there must be good, especially with the pricetags you're getting there. I'm not sure how much a Duvel costs in the US, but here in France it's like 2.5 €. A Rochefort 10 is like 3.5 €. When I lived in Melbourne, a Rochefort was the equivalent of 12 US dollars for a 33 cl bottle. That helps put things in perspective when it's cheap. You don't automatically think "this is so expensive for a reason, I must work hard to find all the aromas in there, it can't be bad for the price I've paid". On the other hand, Weihenstephan beers are cheap pretty much all over the world. And they're top of the list. I love them, but they don't have the complexity and craftmanship of a Cantillon or Rochefort, the unforgettable punch of the best American IPAs or the unforgettable coffee, chocolate and oak aromas of the best stouts out there, European or American.

 

I am assuming WY is so high b/c they only have 10 beers listed in their "Top Beers" selection. Since the sample size is much lower than the others, its score is a bit inflated.

 

For me, this is really interesting data, but when you look at the "top" ratings there are so many beers that have like 100 reviews or less that I simply can't consider them a top beer.

Their Galaxy Pale has 10 reviews, still shows up, I'm sure it's great, but why is a beer with 10 reviews counting towards the top 100? . Again they make great beer, so maybe it's deserving, but # of ratings is not something taken into consideration enough. Way to much potential for bias in there. I'm curious where the statistical significance comes in.

I love Vermont Beer, there's some great stuff coming out of there, but the top 100 is basically 3-4 breweries. I counted at least 70 Hills Farmstead (if you include their Grassroots line) beers in the top 100. Very impressive on their part, no doubt they are all pretty nice beers, but a bit silly too. Still, interesting info so thanks for compiling. I sure indicates where some of the best beers come from.