Beer Reviews Biased

I'm CANming late to this discussion, but as a long-time reviewer & rater, this was a hot-button issue for MANY years. The CANsensus was rate as much as possible to style & NOT to taste, with the exception of the "Overall" score, which is your ChANce to be as subjective as you like. That said, it does help if you qualify, in your review, why the others scores may have been decent, but the "Overall" score is seemingly out-of-whack.

I read some others saying this, but this site's updated "Beer Styles" guide is MUCH better than it had been - more user-friendly, more CANcise, more detailed. If it is too much to toggle, then I CAN also reCANmend the BJCP Style guidelines. They will allow you to drill down deeply, perhaps more so than you intended, but more info is always better than less. The nuances between some styles CAN be tricky, but that is why I feel that a full-blown review is better than a simple rating - we/I CAN see the thought processes behind your numbers.

As to this last - my ocCANsional partner-in-crime, @NeroFiddled , does a great job with his "Stream of CANsciousness" reviews. Thoughtful, humorous & insightful. Worth taking a gander for potential inspiration.

Gotta go! The CANQuest (tm) awaits!

 

I would say giving a review based on what you're experiencing as you drink a beer and judging how you feel it stacks up to the expected characteristics for a style is the worst way to make a usable consumer beer recommendation. Except for every other way there is to do it.

 

If folks were truly rating to style, every style would have a beer rated 5 or at least in the hugh 4's and that isnt the case with the exception of a few styles, such as stouts, iPAs and sours

 

When I try a style that I really just don't like, I don't rate it because it's not the style, it's me.

 

I haven't reviewed many beers, but years ago I rated Jolly Pumpkin Madrugara Obscura very low, and justified it (at the time) because it claimed to be a dark, roasty stout with no indication anywhere on its label that it was a sour beer. I made it clear in my review that I was giving it a low score mostly on its merits as an imperial stout, and a little bit because I (at the time at least) did not like sours. Since then they've changed the label to call it a "sour stout", which would have altered my review.

 

I don't see why we can't strike a chord between pure like/dislike and style guidelines. Obviously, we should try to avoid reviews where appropriate style characteristics are used to downrate a beer (things like "malt-bomb" for styles that should have those characteristics, etc.) But at the same time, aren't we in most cases liking a beer because it accurately represents the style?

To the point, would it be possible to have a dropdown menu when rating that didn't just have a number but had some language keyed to the style guidelines about what that number represents? Like a 3 in feel for imperial stout might include "too thin" among other things? I know this could be inferred from the Beer 101 page, but having it drop-down would help. Clearly, just numbers attached to traits is pretty abstract and I doubt many people really know why they chose a 4 instead of a 3.75 when rating said beer.

As a teacher, this would represent the kind of rubric I give back to students so they understand what traits actually led to the score they got.

 

Well it has been awhile since this old topic was hashed out so I will reiterate some basic things. Yes there is a style bias, but you can mitigate this by going to the style lists and look for top beers in the style. Also consider this, the sheer number of reviews for any one beer are nearly as informative as any score. Take for example, Heady Topper, a small brewery with no real distribution has 14, 500 reviews yet Bud light has 1,500 . What do you think would happen if every Bud Light drinker in the US joined BA and rated their favorite beer? Bud light would easily be the #1 beer. But more importantly, it shows that the vast majority of BA reviewers / raters only bother to review or rate beers they LIKE.

 

An easy solution - and, when I say 'easy,' I mean 'easy to me,' because I have no intention of doing the programming - would be to to allow the ratings to be filtered by style and then apply a relativity to each beer based on it's rating within that particular style such that, for example, the highest rated beer in any given style, regardless of it's actual rating within the unified database, would display as 5 when filtered by style and the lowest rated would be a 1. When the filter is removed, the display reverts to the actual rating. See? Easy-peasy!... for whoever actually has to do it (@Todd).

 

While I like having the ratings, I struggle with rating beers as bad 1 and 2 range. I guess a drain pour is a one and something really off would be in the 2 range. Are there other reasons for rating a 1 or 2?

 

I try my best to rate based on style. If a beer doesn't represent a style I give it lower ratings. However, often I get a complex, mixed-profile beer that purposely deviates from style, and is absolutely delicious. In that case I will give it a good rating. I'm not sure how to balance the difference between accuracy and preference. Some beers are absolutely delicious, but completely inaccurate of the style. How can you not give it a good rating? Is creativity not appreciated? If a brewery claims to represent a style and does not I give it lower ratings, however if it acknowledges the beer is unique and has a special twist then I am more lenient. I don't know if there is a right answer.

 

I use noticeable brewer's errors when rating anything below a 3. Here's a link to common off-flavors from John Palmer's book, How To Brew: http://howtobrew.com/book/section-4/is-my-beer-ruined/common-off-flavors These descriptions can clue you in on what to notice.

Not really included in these descriptions is the over-the-top tart taste characteristic of an infected beer caused by improper sanitation conditions inside the brewery. If it burns your tongue like an over-carbed soft drink, and likely is a gusher when you open it, it's a bad beer. Sometimes sour fits here as a characteristic taste of infection too, and some beers with a sour character from an infection can still taste acceptable (especially if you like sours), but if the beer isn't supposed to be a sour, it's a brewer's error and deserves to be marked down. Infected beers should fall within your term drain pour, because sanitation is inexcusable.

 

just bring the top 250 back please... I still want to venture and try all of the top beers rated by enthusiasts like us. Only suggestion would be to stop adding a beer with less than 50 reviews onto the top list

 

From another thread:

 

Right, that's why using personal preference to rate, vs rating to style, makes ratings useless. I might enjoy the the shitty Brut more than the Murk, but, I would likely give the Murk a higher score.

 

I have favorite styles, for sure, so I try like heck to rate based on the style, but at the end of the day, I'm not going to beat up a brew that is "offbeat" for the style if it is a tasty one, especially if it is unique and daring.

 

I completely agree here, right thru the last sentence. But the details and environment could perhaps make that Gansett a 4.5 experience. Your on the beach it’s 95 degrees and your sweating your balls off, ya pop that ice cold Gansett and tell me if that’s a 4.5 beer experience. Or you could sip on a 9% ipa, or 14% BA Whale of a stout. Conditions could impact ratings well beyond the ability of the beer in and by itself to give pleasure. Flip that to a coolish Oct evening and the beers reverse themselves.

 

Rating shouldn't be a summary of what styles you like most to least.

 

Perhaps... people can and should be able to rate how they'd like. If it were coming from experts who get paid, sure. Everyone is learning. A few will never change, but that's on them

 

Kind of agree. But I just had this thought. What if your BA profile included a list of styles you favor. I know we could kind of get this by looking at "styles rated", but that's just an inference.
Obviously the most useful ratings are rated "to style". It would be good if we could see which reviews really are that.