What Exactly Is A 5.0 Beer Rating?

IMO we will never see a normal distribution centered on 2.5 because a lot of raters obviously know what beers are "good" or better and will be reluctant to spend money on crap beers just to build up the frequency of " below average" beers for that portion of the bell curve. I think the best you can hope for is something like a truncated normal distribution or maybe lognormal.

 

Not only a paradigmatic example of the style, but a superlative one. It does what it does better than anyone else, and it works.

Saison Bernice by SARA is a one of the few examples I can cite.

 

Great point. The samples are probably already biased towards "better/good" beers overall, since to say otherwise would imply that consumers intentionally purchase beers they expect to be bad. But still it doesn't sit right to me that the true top end of the scale, 4.5 and above, is basically unused.

 

That and breweries doing due diligence with their QC will get rid of their off batches.

@TootyMcButtface

 

For example, see the above quote where another user is arguing a 3/5 or less should be average on a 1-5 scale (a 60% or failing in nearly every grading scale).
My reason for choosing 3.75 as an average is based on most BeerAdvocate reviews and the scaling system in place. A beer is great maybe a 9/10, but not quite exceptional, so on BeerAdvocate you'll might it a 4/5. If the scaling were out of 10 I'm sure the average would be much higher and closer to 8-9/10
I cut the difference as most reviewers probably do and would imagine your average craft beer is somewhere between 3.5-4.0 in a rating. That's why I base my average at 3.75, because most of the beers I see on this site, 3.75 seems about average.
You can argue 3.75 is a C+ and far lower than average, or you can argue on a scale 1-5, 2.5 should be the average. I see both points.
I'm basing my scale on what I see most reviewers posting on this site. My estimate is a 3.75 is closer toward average of most reviews here.

 

Let's try it another way. Instead of assigning a numerical score, what if we ordered them? Say you have 11 of the same style in front of you. You have to decide which is best (1) and which is worst (11). Those two would probably be easy, but differentiating and ranking 4-9 would probably be harder. And, every time you have another of that style, you've gotta decide where to slot it in at. Is that new IPA actually better than the previous #1, or is it just new?

Deciding the extremes is simple, the middle, where most beers live, is a tougher task to accomplish.

 

Agree that most beers should fall closer to average and ideally it would represent somewhat of a bell curve. Definitely difficult to differentiate more of the middle-class beers. I try to do side-by-side comparisons to learn the difference and better train my palate.
You bring up another point that some styles are more favorable than others so the average for one style may be higher than another style. NEIPAs will always have a higher average than American lagers. I always try to rate within the style comparing to beers of the same style. My hope is that keeps a somewhat standard average among my ratings regardless of style.

 

I’ve had 2 5s in 1200 beers, and honestly after revisiting one, only mornin delight deserved it. It’s hard because between 4.75-5 it’s all incredible beer.

 

This is really well explained. I think the issue is the granularity in the scale, or rather the lack of granularity. If the scale was bigger, I think there would probably be more scores reaching the upper end of it, but from 1-5 just doesn't give enough room. Fair point, all around.

True that. The overall quality of beers out there is definitely getting better. Also it puts a little smile on my face that you had to spell my name

 

I find it hard to rate anything "perfect" as I feel there is always room for improvement.

But when I went to Parish and had their Imperial Reserve (my only 5.0) on tap, it blew me away me so much, that I finally said, "if there is a beer better than this, I probably won't encounter it. Now is a good a time as any."

I still stand by it. It might not be the best beer ever created, but it is the best beer I have ever had.

 

I likewise do not understand the mindset of rating nearly everything a 4.x or higher. Above 4 to me is generally "would seek out again" which is pretty huge for me considering there is so much choice out there and I'm kinda picky as is (why drink mediocre bunk? In 2006 okay, or at a bowling alley sure...)

So many beers have been a one and done. i.e. "meh." I'm sure anyone can relate.

As stated, there was only one "5" for me, but a few have come close. The qualifiers are either ones that blew my socks off during the review, or beers that are just so insanely good that I buy a flat whenever I encounter it. Consistently good experiences bump the initial scores up a bit.

I've seen artificially hyped reviews as well. Once I saw one say "it's not great, but it's still xxx brewery, so I gotta rate it high 4.9/5". ....Disqualified.

I also had someone tell me once that they rate a certain brewer high because of the can art. Consistent 5/5's. Again, disqualified.

But hey, it's an open section, and there's reviewers of different experience levels too. I remember when I was rating ambers, pilsners, and blondes in my notebook straight 4s. (In my defense, this was also back about a decade when the shelves weren't quite as eclectic.)

Bottom line, we can't control what people do. Only the mods can on this site. I guess if someone sees it fit to rate an average beer a 5, we can't stop 'em. We can only bitch about it.

 

A few thoughts, some already shared but anyways:

1) Rating should be based on style - meaning you shouldn't rate all lagers as 1 and all IPAs as 4's because that's your preferred style. To me it's just like BJCP judging, the rating should be how close it relates to world-class examples and standards, but does require some baseline understanding.

2) Rating based on quality, not quantity - As quality and craft has drastically increased in recent years, there are a LOT of great beers out there. Does that mean because there's a lot of great examples that a majority of them should get a 2.5 (average) vs. a 4.0+ rating? I don't believe so, in fact you just have more 4.0 ratings. The beers should not be discounted simply because we have more great examples.

3) Don't be afraid to use the 5.0 rating:. Ratings will change over time based on the fact that beer will continually be made better and people's tastes change and get more refined. Over time the reviews will reflect that a beer that may have been a 5.0 a few years ago may be a 4.5 in today's world as better examples of the style have come out. If you think a GOAT is a one-time designation, time will show you you're wrong. If you enjoy a beer today, don't be afraid of the 5.0 because you may not like it 'quite' as much later. If it's a world-class example TODAY, then rate it a 5.0 TODAY and don't look back. As time and tastes change, the rating will correct itself.

4) Everyone's voice should count. Most people's criticism with BA/Untappd is people complaining about people not rating beers correctly. If in fact a majority of those who are educated about beer SHARE our thoughts and RATE beers, it puts those "outlying" biased and erroneous reviews in the minority and helps CORRECT the rating simply by law of averages. If you are bothered by someone else's high or low rating, the first question you should ask yourself is, did you review/rate it yet?

It's beer, have fun with it. Just like food, music, etc. there's no wrong answer, it's mostly subjective in how an individual enjoys it. These ratings are public opinions, not the BA/BJCP certified competition, let's not pretend it is or think it should be.

 

Actually the 5.00-scale (BA rounds to 2 decimals) has a finer level of granularity than the older 1-100 scale.

 

Wait, but I thought the ratings were only on a 0.X/5 for each individual category? With only one decimal place. Meaning technically 50 possible scores. I thought the overall score was just the result of some formula taking weighted averages of each individual review, and not actually users choosing to rate a beer 4.83/5.00. (As an example.)

 

The user technically has 85 choices (it's a scale of 1-5 with 0.25 increments x 5 categories), I guess I was taking BA's "weight" to each category into account and the cumulative score.

Then again, do we really need 100 point scale for appearance? Or a 100 point scale for mouthfeel? It's already a 17 point scale for each aroma, appearance, taste, mouthfeel, and overall.

If I did my math right, with 17 choices and 5 categories, there are 6188 possible scores.

 

Technically because of the 6,188 possible scores, my individual review for KBBS for example came out to 4.63 whereas the overall score is 4.90. What's nice is that BA is helping standardize the reviews by "weighing" each category the same, so there's a bit more consistency and likely a LOT more granularity in the result even if you aren't choosing it, it's factored in because of the weight.

4.5 * 0.06 = 0.27 appearance
4.5 * 0.24 = 1.08 smell
4.75 * 0.4 = 1.9 taste
4.75 * 0.1 = 0.475 mouthfeel
4.5 * 0.2 = 0.9 overall
== 4.625 total

source: https://www.beeradvocate.com/community/threads/how-to-review-a-beer.241156/

 

Sorry, so what are the actual ratings you are allowed to select? Is it an every 0.25 interval?

1.0
1.25
1.5
...
4.5
4.75
5.0

?

Because if so, then technically the scale would go from 1-20. See what I'm saying? Even though your review came out to 4.625 as a result of the formula, you were only really picking a score between 1 and 20 for each category. Unless I missed something.

 

My head hurts *drinks beer*

 

I definitely feel like lagers get the shaft.

They're not as flavorful as other styles by nature, but you'll rarely find a decently rated lager. I wish there was a way to change that because there's some really good to-style lagers that need more recognition.

 

Absolutely right, and is extremely subjective on those middle beers.