I think it would be interesting to see an average of a individual reviewer's scores by category. Then we could see (as an example) that bono-one generally scores gay dvds higher than straight gonzo.

[Link]

DenverDon <<<<[[[[

[Link]

Anything else we can think of to fill up Drew's "spare time"?

"Vegetarianism for me is about saying ‘yes’ to things - even meat." - Coupling

[Link]

I mean, I can see the interest that reviewers themselves have in these kinds of stats (I'd be curious myself), but I'm not sure I understand how stats like this would help those who READ the reviews.

If a reviewer rates on the high side, there are lots of possible reasons for that. It might mean the reviewer just likes porn in general. It might mean that the reviewer is super picky about the movies he reviews. It might mean that the reviewer has a bias, and that bias may or, in fact, may not be a debilitating factor for any particular review READER. In some cases, it might be an advantage. It just depends.

I would hate to see readers pass a reviewer by because their review averages are really high... or really low. That really has no reflection on the quality of the reviews themselves or even of any possible biases involved. It's simply a stat that might seem to reveal something, but in all honesty... doesn't reveal anything.

I'd be wary of including any statistic on the review pages that discourages people from actually reading the content of the reviews in order to decide which reviews are useful to them, and which aren't. If Drew just posted some stats in the forum, though, I would find that interesting and totally cool.

Anyway, porn is such an issue of personal taste that trying to cull all that down into a complex system of ratings and averages just doesn't seem that it would be effective to me. In fact, I think it would be misleading.

Perhaps that's just me. Am I wrong there?

[Link]

If I get around to doing this it would only be for posting in the forum.

[Link]

There are many things that affect the sales of an item other than the item itself. It may be on sale, it may have better placement than other items, etc... The first thing that we do in order to forecast is to "normalize" each item. So, using historical averages we factor out the ancellary reasons why an item may do better or worse than other items and then we forecast it. Based on how this item will promoted, we will then apply those same averages that were used to "normalize" each item to get a more accurate forecast for each item being promoted.

In looking at a review, I think that knowing the average score for a reviewer is important because it allows the reader to know whether this reviewer rated a particular DVD above average, below average or just average based on the average for that reviewer.

A 4 star review from someone who's average is 2.5 stars is a higher score than a 4 star review from someone who's average review is 4 stars.

I just think that in something subjective, knowing an objective baseline is a good idea.

See why I'm so fucking boring at parties?

[Link]

I think that a lot of people that read reviews tend to find a reviewer that has the same viewpoints of porn as they do, and then they follow 'that' reviewers recommendations and reviews. An average for that reviewer is somewhat meaningless for those that follow that reviewers recommendations, or reviews.

If I read the averages for a reviewer, it wouldn't make any difference for me. It's just a number. What the reviewer says about a DVD, and why he liked it or didn't like it is more important.

DenverDon <<<<[[[[

[Link]

"Vegetarianism for me is about saying ‘yes’ to things - even meat." - Coupling

[Link]

One case observation versus measuring and looking for averages.

[Link]