One of the standard recurring discussons in the beer comment-o-sphere is about the relevance or otherwise of Ratebeer ratings. My personal view is that they're basically an extremely good way of finding out whether or not Ratebeer users like a beer. This isn't always a particularly useful bit of information, but following on from my "craft cred" post, this might be one of the times when "what do Ratebeer users think" is actually an interesting question to ask. Hence, being a massive nerd, I've collated some information.
This doesn't really answer the question that I asked previously - there are other factors involved in "credibility" than how well regarded your flagship Best Bitter is, and Ratebeer users are just one part of the complicated and heterogeneous mess that gets lumped together as the "craft scene", but it was an interesting exercise anyway. So without further ado, here are the results:
|Beer||ABV||Rating||Number of rates|
|Timothy Taylor's Landlord||4.30%||3.51||353|
|Fullers London Pride||4.10%||3.36||510|
|Harveys Sussex Best||4.00%||3.3||282|
|Hook Norton Old Hooky||4.60%||3.2||144|
|Sam Smiths Old Brewery Bitter||4.00%||3.12||180|
|Shepherd Neame Spitfire||4.20%||3.11||195|
|Adnams Southwold Bitter||3.70%||3.09||238|
|Greene King Abbot Ale||5.00%||3.04||310|
|Mc Mullens AK||4.30%||2.99||90|
|JW Lees Bitter||4.00%||2.94||54|
|Greene King IPA||3.60%||2.48||358|
Note that I've generally tried to pick the most popular brownish beer in the 4% to 5% range for any given brewer to give a decent base for comparison. For Adnams and Greene King I've included two as there wasn't an obvious "mid to high 4's" option. In almost all cases, Ratebeer treats the "cask" and "bottle" versions of the beer seperately, and I went with the cask version.
Any conclusions? Well, Harveys, Adnams and Fullers all do well. Spitfire, Bombardier and Abbot all score surprisingly highly, and Ratebeer agree with me on the Batemans vs (the Banks subbrand formerly known as) Thwaites issue. Bathams Bitter scores highly but apparently isn't particularly sought after. Meanwhile, Taylors' Landlord and Greene King IPA are clear outliers at the top and bottom respectively - I suspect that this is actually a case of credibility influencing the scores! Diehard real ale traditionalists might be pleased to hear that almost all of these soundly beat Tetleys Smoothflow (2.55), John Smiths Extra Smooth (2.51) and Worthington Creamflow (2.28), not to mention Stella (2.48), Carlsberg (2.0) and Fosters (1.76), although they do mostly lose out to the likes of Brewdog's Dead Pony Club (3.51) and Beavertown's Neck Oil (3.53). Interestingly, London Pride is clearly the most commonly drunk, followed by Greene King IPA and Taylor's Landlord. My guess would be that this is because Landlord and Pride are both relatively well thought of, while IPA and Pride are extremely widely available.
 I can't resist elaborating a bit. As far as I can tell, Ratebeer users tend to rate beers in terms of how immediately interesting it is to give them your full attention for a few sips.This is a bit like rating cars based on how much fun they are to nail round a track for a couple of laps - a fairly mediocre roadster is probably still going to seem better than a really great family estate in that context, but I know which one I'd prefer to have to take a family on holiday to Wales. On the other hand, this is arguably a feature - or at least a known limitation - rather than a bug, and if you do want a flashy sports car to take out at the weekend then it's not an unreasonable test. In practice, I tend to find that the ratings for things like US IPAs and Imperial Stouts are often useful for getting a general idea of whether a new-wave craft brewery are much cop, but that they're best ignored for more traditional breweries and not worth paying too much attention to for individual beers.