New Rating System Feedback

Well he also said the ratings before wern't making an impact. It wont change anything for you all. I'm still going to rate how I want, now it just takes 5x longer. You're still not going to look at my rating and jedge a beer based on what I have to say, so again, why does it matter so much? I don't see how this change makes my rating anymore useful to you.

 

That's pretty much it right there in bold

 

I may not be an expert in the field of mathematics, but five times pretty much nothing still leaves you with pretty much no time spent. Sure, it's longer, relatively speaking, but still pretty damn negligible.

 

Where do you get five times nothing? I'm not sure how iTunesUpdates does it, but as I've stated I typically input ratings for anywhere to 40-100 beers at a time. You will never convince me that having to input five numbers instead of one is a negligible difference. Surely you see that.

 

I meant per rating, but even in the aggregate I don't anticipate all that much extra time. I mean, it's not like many words are gonna be typed or anything...

 

OK, fine. You can't admit that 5x the clicks equals 5x more work/time/thinking, then I guess I'll just have to argue "BeerAdvocate's new scoring system is leading to carpal tunnel inflamation!" @#$%!

 

It's not just the clicks, either. When I enjoy a new beer, I pay attention to the nose, the color, the lacing, and of course, the taste. No way in hell, though, do I sit there and think about how each attribute rates on a 5-point scale. I don't care THAT MUCH about that stuff! I like drinking beer, not studying it. I know other people do, and that's fine! I'm not asking them to do anything different than they've ever done before. This change, though, IS requiring people to change their beer-enjoying habits, and that sucks.

 

The only way for rating additional attributes to not require any significant increase in time is to ignore the attributes and just rate all values the same. Which defeats the purpose of the change. If you are going to consider all factors, that is going to take more time.

My point is, I think there is an argument to be made that adding the attributes will give more justification for the numbers because it requires more thought and therefore gives more meaning to the aggregate ratings. But I don't think that can be followed with and it takes no additional time.

I'm personally adjusting to the attribute ratings. I am not complaining about the change or advocating to reverse it. But it is taking me more time, not to tick the ratings, but to decide what those ratings will be.

In the end, I think it is more productive not to try convince people who are spending more time with their ratings that they are not, but rather that the time is well spent.

 

I'm happy to be proven wrong when the intent of the change shines through. My whole argument regarding the time spent is that, if you're only clicking one field before and now it's five, it's barely any difference. Trust me, if you're really dead-set against the change, you're just gonna be clicking quickly through things anyway, if for no other reason than to (silently) make a point. If you're actually slowing down and considering things, putting some thought into it, then, yeah, there's gonna be more time spent, and it's certainly time well spent, at least in terms of the overall community. This is what we pro-change/pro-review people have been advocating for some time now.

 

I've done a lot of reviewing under the original core requirements which required a full review, including scoring all 5 attributes and more than 200 words as a review. I've done rating under the changes that were made to allow use of a single rating score only. I've tried out some of the possible variants in between. What I have learned is that when I review I do it because I'm cheap, lazy, selfish, hate wasting time and want to enjoy good beer.

At a personal level, numbers can tell me that I liked a beer but a full review can help me remember what I liked about the beer and why I liked it or what disappointed me about it. That way I don't have to spend money making bad choices more than once. That way I don't have to waste time searching for good beers to share with my friends. That way I can keep on improving my enjoyment of a new good beer when out with friends or in some other situation when I'm not reviewing. Its all about increasing my enjoyment of what is for me a is a good beer.

So to maximize the return on my investment, I’m moving back to investing the personal time in doing the attribute numbers and even full reviewing for no more than one beer an evening when I have a new beer. In doing so I'm including both my level of enjoyment of the different dimensions of the experience and some reflections on why there is that enjoyment in the experience since I am cheap, lazy, selfish, hate wasting time and want to maximize my enjoyment of good beer.

Do I drink fewer beers because of reviewing? No, not really. Do I try fewer new beers because of reviewing? Probably, but that's OK because, for example, in a few weeks Troeg's Nugget Nectar will be out on the market and I'm already getting impatient for the case I'll buy and for the experience of once again enjoying the flavors of that combination of hops, the backbone provided by the malt, and the changes that will unfold over the month it takes me to finish the case. Plus when one of my favorite bars offers it on cask, I'll be able to enjoy once again the differences, both subtle and not so subtle, between the bottled, the kegged, and the cask versions of the beer.

Why do I favor the site move back from allowing just a single number rating and asking for reasons when an overall score is an outlier from the norm for a particualr beer? Because I'm cheap, lazy, selfish, hate wasting time and want to enjoy good beer. The various attributes are mostly separate, separable, and important parts of what is in that glass to be enjoyed about the beer (and this is not just personal opinion, there are solid research data to support the importance of including each). Since I'm cheap I want to know whether to consider investing my money in a beer to maximize the return on my dollar by minimizing my mistakes. I lazy so I don't want to have to search all my local area for a place that might have it on tap and might let me sample it. I'm selfish so I want to drink beer that I enjoy and the experience of others can be a helpful. I hate wasting time and don't want to let chance determine how often I encounter a new beer that will enjoy. In other words, I don't want to have to make all the mistakes on my own if I can learn from the experience of others?

Sure people can just tick off all 5 numbers to be the same as the overall score they want to give anyway. Sure people can just throw out 5 words at random to deal with that outlier requirement. But then they are just wasting their own time by simply creating a bit of noise which the statistics will eventually "cancel" out. But on the other hand some folks may find that interesting questions come up that they want to know and understand about beer, for example, why is the head on an 6.7% IPA so different from the head on an 11.0% stout, or why are some Pilsner style beer more popular than others.

 

In my opinion, the fact that it will take 5x longer is a positive, as it might force some more thought into the rating. If it forces a few people that were about to give a quick "1" or "5" rating to a beer to stop 10 seconds and consider all aspects of the beer. More information, then more useful. And if an outlier, then at least a minimum of additional info is required to back up the score.

 

http://www.beeradvocate.com/community/threads/outlier-detection.239068/

Sorry, but color me unconvinced. I just entered in a couple weeks of notes and every time I gave a beer less than a 3, I had to explain why. This includes beers with 3.7-3.8 rAvgs, which is "average" for most people, apparently. I try to rate to a 3 average. If a beer is less good than average, IMO, I give it less than a 3. If a beer is truly unpleasant, I'll give it a 1-2 rating, and I'll likely have noted down exactly what the problem was. But it's a bit problematic if I beer has neither glaring flaws nor particularly noteworthy elements, but is just "okay" and I'm forced to come up with something to say purely because most reviewers won't call a 3 a 3. Fortunately "this beer is just average" is 5 words, I guess.

 

Another user here that's happy with the reversion. Keeps the consistency amongst all of us and in my own reviews. Thanks for the consideration and making the actual change. Can't please everyone, but this worked out for at least this one user here.

 

I am happy with the current direction as well. I find the "Like" function for reviews obnoxious but that's a different issue.

 

My apologies if this has been asked before, but I don't have time to read everything in this thread.

Is there a way to remove the outlier required review for users that have rated a certain number of beers?

My friends and I encounter a lot of beers that we feel should not be rated as highly as they are, for whatever reason, and it's frustrating when we have to justify many ratings that we believe are appropriate.

 

Having read all or most of the thread I can say that there isn't at this point, and I'm predicting that it won't be added.

There's a bit of an explanation for why its there to be found here:

http://www.beeradvocate.com/communi...s-to-its-roots-rating-system-revamped.238804/

 

It says the point is to discourage fraudulent ratings. My ratings are not fraudulent...

 

And I didn't mean to imply they were fraudulent, but how is some one looking only at your number for a single beer to know that?

Unfortunately the number of ratings done is really enough to solve the problem of fraudulent ratings. But it might be possible to use something like number of ratings, time on BA, number of reviews vs. ratings or some more complex measure than just deviation from the mean. Not sure.

However, I am pretty sure that if you can think of a better strategy that can be implemented in a computer program they'll look at it, test it out, and introduce it if it works. (There was something else a couple of us pointed out awhile back did produce a significant change/improvement in another aspect of the site, so it can happen.)

 

Correction:

First sentence in Para 2 should read, "is not really enough to solve the problem..."