Statistical Analysis of Beer Drinking

I've been doing nearly the same thing for the last ten years now. I have a database of every beer I've rated in the last ten years and in the past four years the database includes every unique beer I've tried, even if I couldn't rate it at the time.
Countries:
1. United States 79.3%
2. Belgium 3.6%
3. United Kingdom 3.1%
States:
1. Missouri 27%
2. Colorado 11.3%
3. California 9.1%
Styles:
1. American IPA 8.3%
2. American Pale Ale 5.3%
3. American Adjunct Lager 3.6%
I also track, dates that I have obtained, alcohol percentage, and location drank.

 

More than a quarter of the beer I drink is made right here in Central Oregon, and more than half comes from my own state. Half the beers I drink are APAs, IPAs, or IIPAs.

3258 Total Checkins
1855 Unique Beers
Beers checked in more than once: 32
States: 39 + DC
Countries: 23
Average ABV: 7.2%
Average Rating: 3.54(all beers), 3.75 (>9% ABV), 4.0 (Barrel Aged Beers)

Top Countries
US 93.3% of beers, 95.2% of checkins
Belgium 2.1% of beers, 1.5% of checkins
England 0.8% of beers, 0.5% of checkins

Top States
(Central OR) 23.5% of beers, 25.6% of checkins
OR 45.8% of beers, 57.0% of checkins
CA 15.4% of beers, 14.1% of checkins
WA 10.6% of beers, 9.6% of checkins
CO 4.6% of beers, 3.4% of checkins
TX 2.4% of beers, 1.4% of checkins

Top Styles
American IPA 17.1% of beers, 32.1 of checkins
Imperial IPA 8.4% of beers, 9.1% of checkins
American Pale Ale 7.2% of beers, 8.4% of checkins
Imperial Stout 3.1% of beers, 2.4% of checkins
Saison/Farmhouse 2.7% of beers, 1.7% of checkins

Most Checked In Beers
Laurelwood Workhorse IPA (327)
Worthy Prefunk Pale (40)
Deschutes Fresh Squeezed IPA (27)
Boneyard RPM (23)
Hop Valley Alpha Centauri IIPA (22)

 

I admire your 2000 mark.I too will be there someday.

 

I keep a spreadsheet for reference. I copy and paste my data from BA, reformat, and add the state and country information. I also keep separate worksheets for states, countries, and styles.

I haven't done any detailed analysis on my data, except to calculate that my average rDev is +4.4% and that around 75% of my ratings are on the positive side. (And that, friends, is why you should take these BA ratings with a huge chunk of salt. )

 

I would love the ability to input additional data when rating or reviewing beers such as method of consumption (i.e. draft, bomber, can, 12 oz. bottle), acquisition method (i.e. trade, bottle shop purchase, restaurant/bar) and age in days of beer consumed (if available). I would love to see graph that shows average rating for a specific beer charted out against age in days. I assume it would not be too difficult to establish. Has this been discussed in a forum before?

 

Having the ratings on the positive side is pretty much to be expected since the distribution would be slid upwards before you even get to it.

1) Many breweries will not release a beer they consider a failure and prefer to "drain pour" it rather than packaging and shipping it, realizing that in the long run they could be wasting more money by releasing a beer that just won't sell because nobody wants to drink it. This will make sure the ratings distribution gets slid upwards by removing most of the worst. (i.e., it isn't going to be a normal distribution centered at the mid-point of the rating scale)

2) If you make any use at all of the ratings and reviews on this site as a factor in your decision making (or any word of mouth reviews, or what ever) that will introduce a selection bias in your purchases such that you don't even try beers below a certain quality level, eliminating them from your personal distribution of scores, there by moving the center of your distibution above the mid-point of the rating scale as well.

So what we really have to take with a grain of salt is the assumption that the mid-point of the rating scale represents the average that should be found with group or individual rating scores.

 

Yeah, similar such ideas have been discussed before, but we're not likely to see them implemented if only because large numbers of site users don't know, can't find, or won't take the time to provide such information (e.g., time between bottling and consumption). We all get to enter new beers, but there are some folks who can't or don't even take the trouble to be sure they aren't creating a duplicate entry using a slightly different verson of the beer's name. Method of consumption used to be possible (e.g., draft, bottle, so on) but was removed and it seems at least one reason was that there was some sort of abuses going on, etc.

 

Congrats on 2,000! Truly impressive.

I also have a spreadsheet (as a former accountant, I could not help myself) long before I joined BA so I have about three times more reviews on my spreadsheet than beers I have reviewed on BA. One category I included in my spreadsheet is whether I would buy the beer again. This helps when I am beer shopping.

I like some of the stats you used. I wish I had used some of them.

 

If you seek professional help, see if you can get me a discount-- I also have the same data need affliction. My wife says that as a former accountant, I am probably brain damaged.

 

My statistical analysis is the store has more beer than I have money.

 

Haha, yeah, "I'm an engineer", that's always my excuse.

 

I tried a lot of different excuses but none of them seemed to work.

 

Your mistake is that you're giving them a chance to reply at all. You've gotta fling that excuse out there like it's the word of god, and then grandly stalk away.

 

You need to go back and enter all the hop variety data into the IPA and APA category to see which hop variety makes the best hoppy beers.

 

You must know a different sort of woman than the kind I've been around.

 

That is a good idea. I am not that familiar with the various hops but have noted that the IBU rating does not always indicate how bitter the beer may be. Some high IBUs may not be that bitter, while so low ones may. I am assuming it is the type of hop used.

 

That, but especially WHEN the hops are added. Early = bitter; late = aroma (says the online "expert" who has never brewed beer in his life! )

 

The IBU calculation factors in boiling time and alpha acids. The reason that IBU is not the best indicator of bitterness is that residual sugars reduce perception of bitterness. A more useful statistic is IBU/OG, where you divide the IBU by the last two digits of the OG. For example a beer with 50 IBU and an OG of 1.050 would have an IBU/OG of 1.

 

Thanks for the info. Now I do not have to sound too ignorant when I talk with my beer geek friends.

 

I use Microsoft Excel to track my stats. Still working with pivot tables to get some of the information. I track pretty much the same stuff you do, plus a bunch of other non-numeric fields. I started in 1994 so some of my beers aren't reviewable on BA, or even listed. Unfortunately. I'm still going to rate them for my own data.