tag:blogger.com,1999:blog-1392866426745021699.post2349724069207645532..comments2024-03-28T07:27:16.088+01:00Comments on The Wine Gourd: Quality scores changed the wine industry, and created confusionDavid Morrisonhttp://www.blogger.com/profile/11578729952036086391noreply@blogger.comBlogger7125tag:blogger.com,1999:blog-1392866426745021699.post-45236836319655748662021-12-02T15:06:59.688+01:002021-12-02T15:06:59.688+01:00All we know is that Luca Maroni never goes below 9...All we know is that Luca Maroni never goes below 98 - the world's most generous wine reviewer ;-)Testhttps://www.blogger.com/profile/01587795839599225977noreply@blogger.comtag:blogger.com,1999:blog-1392866426745021699.post-16051253973614960742020-01-13T20:57:15.954+01:002020-01-13T20:57:15.954+01:00My point is not that there is a problem with inter...My point is not that there is a problem with interpretation, but that there is a problem with comparing and combining scores as though score differences matter. My post on "Laube versus Suckling — their scores differ, but what does that mean for us?" discusses scores from different observers.David Morrisonhttps://www.blogger.com/profile/00276520192744208262noreply@blogger.comtag:blogger.com,1999:blog-1392866426745021699.post-69191761403045464052020-01-13T20:51:43.871+01:002020-01-13T20:51:43.871+01:00Indeed so. I will see whether I have sufficient co...Indeed so. I will see whether I have sufficient consistent data to do this for any region.David Morrisonhttps://www.blogger.com/profile/00276520192744208262noreply@blogger.comtag:blogger.com,1999:blog-1392866426745021699.post-87615386964349280412020-01-13T18:27:17.943+01:002020-01-13T18:27:17.943+01:00It would be interesting to see if there are any ty...It would be interesting to see if there are any types, styles, or regions of wine with reasonably close correlation between price and average score.Austin Beemanhttp://www.austinbeeman.comnoreply@blogger.comtag:blogger.com,1999:blog-1392866426745021699.post-39720088703767040132020-01-13T18:04:38.947+01:002020-01-13T18:04:38.947+01:00Scoring is contextual, even if it's not suppos...Scoring is contextual, even if it's not supposed to be. Otherwise, everyone should prefer a 5 star burrito over a 4 star Kobe steak.<br /><br />I don't know that there's a huge problem with the interpretation of ratings. a 92 is the same as a 93 and almost as good as a 94 or 95. A more significant dimension to the problem is that the sparsity of reviews means that consumers are exposed to ratings by different reviewers. So now we have to figure out why a wine has a Vinous 90, a Suckling 93, a Decanter 96 and a Luca Maroni 99 point score. Since there are no consensus scoring models that compensate for these biases, we don't even have a starting point!<br />Anonymoushttps://www.blogger.com/profile/08612589529043467922noreply@blogger.comtag:blogger.com,1999:blog-1392866426745021699.post-55321246036285971882020-01-13T07:08:37.117+01:002020-01-13T07:08:37.117+01:00I agree that numerical scores of all sorts have th...I agree that numerical scores of all sorts have the same issues. Amazon scores are averaged by the company's algorithm, but this seems to mean nothing mathematically, especially as one cannot give a score of 0.<br /><br />A single score has no practical use, and so the scores need to be compared. The problem is trying to compare their mathematical properties, since these are more complex than people seem to realize.<br /><br />If price does not correlate with score, then it is difficult to identify something as "a bargain". So, I think that such a correlation would be useful, in the practical sense. Indeed, it might be the most useful thing a score could do for us.<br /><br />Otherwise, a score is just an adjective, like all the other adjectives (style, sensory characteristics, etc). As such, it should be treated as such, as you say.David Morrisonhttps://www.blogger.com/profile/00276520192744208262noreply@blogger.comtag:blogger.com,1999:blog-1392866426745021699.post-58648124051464494202020-01-13T01:31:57.643+01:002020-01-13T01:31:57.643+01:00None of this is specific to wine. Every product ca...None of this is specific to wine. Every product category has the same issues. e.g., a 1% difference in rating can push a Yelp review to 3.5 stars (not going there) or up to 4.0 stars (seems like a good place, let's go). <br /><br />What's the difference between a 3.9 and 4.0 star running shoe on Amazon? Collectively billions of dollars in sales... even though something like a running shoe is at least as subjective as wine. <br /><br />And why do price and rating need to be highly correlated? They're not for non-wine products. <br /><br />It's not reasonable to look at a number as the singularity. It's just another data point that consumers can combine with all the other evaluation criteria (price, discount, region, wine style, sensory characteristics, backstory, appropriateness for consumption context, etc., etc.) to make purchase decisions.<br /><br />Michael Brillnoreply@blogger.com