My initial take is that they’re not bondage-haters, but rather shoe-haters. (“No heels, dangit!”). Another factor, I suspect, is the division between those who find “hot girl-girl action!” to be, well, hot, and those who don’t.

So for some people, the sets are like a bowl of ice cream topped with anchovys and ketchup:

“Thumbs down. I rate this a ‘1’”

“What? You don’t like ice cream?”

HI All,

The whole rating/voting idea is something I’ve fretted about for years.

On the plus side, it is about the best way we’ve found to get meaningful feedback on sets. It is simple to understand and easy to do, requiring just an extra click or two.

On the minus side, as Andrew says, it is clear that only a small minority of members vote. Furthermore, that group is self-selecting, so one must take the results with a hefty pinch of salt.

We have no way of knowing whether a majority of members like nude in metal bondage sets, or whether the people who like nude in metal bondage sets are just more likely to vote. Any attempt to get more data (e.g. by a more detailed voting schedule or inviting comments) is only going to make it harder to vote, which will further reduce the participation.

There’s also the unfortunate side effect of it causing stress for the creators. Cast and crew can see the vote, which leads to worry. Why did a set we loved shooting go down so badly? Whose fault was it?

For models in particular there’s a BIG risk of them taking it personally, when the real reason for a low rating may be because the small subset of members who vote prefer nude sets to historical outfits- the reason a set is low rated is hard to determine and may well be nothing to do with the girl in the photos!

Greater detail in the voting/comments might help remove that ambiguity, and inviting comments for each set certainly would, but that only compounds the problem of people taking it personally. Just take a look at the bile spewed on public blogs everywhere to see that. It would take considerable policing to remove nasty personal attacks, and I don’t want any person who has worked hard to create a lovely set getting upset by stupid drongo comments from the interwebs.

I did some more in depth statistical analysis a while ago, made possible because I have access to the whole database and can cross-reference all sets with a given model or keyword tag with the scores of sets with that tag or featuring that model. What I don’t have is any explanation for the observed trends, and given that not all members vote and not all voters even vote on every set, I think it is dangerous to read too much into them. There’s too many variables, and too many inherent biases in the data collection method to be worth losing any sleep over.

There’s also the fact that people’s tastes are different. I’d much rather see a set get a stack of five star ratings and a stack of one star ratings than get a broad “I don’t really care” spread around a “meh. 3 stars” rating. I reckon if a set is just perfect for someone with a particular taste, people who don’t share that exact kink might well HATE it. I think the purpose of art is to generate an emotional response in the viewer, and love/hate is much better than indifference.

Finally, I take the ratings with a pinch of salt because ultimately, it is me who has to create the NEXT shoot plan for RE, and RE will always be my personal art project.

I’m happy to adjust the mix to suit what people like, we shoot enough ideas that that’s never going to be a problem. Shooting more things like what people liked in the past seems like a lovely way to say thank you to everyone for supporting us. You are paying us to create more RE, so of course we’ll try to factor in what people enjoy.

But I’m also going to keep shooting what I the hell I like some of the time, too 🙂 🙂 🙂 I need to keep my enthusiasm for what we do or we’ll end up soulless and chasing the market. In my opinion there’s no faster way to generate all one star ratings than losing the passion for what we do.

One cannot cater to every taste anyway, so at least let’s do the best job we can in the small niche in which we work.

To tickle your interest, with all the above caveats, here are the major trends I’ve noted in the voting data, and a few of my thoughts about them.

1) New sets do better than old sets. That’s fine, it would be depressing if it were the other way around. We’ve demonstrably got better technically so it might be as simple as pic resolution or people not really caring for sets they’ve seen before so much.

2) With any given projection there’s a wide range of average rating for galleries fitting that description. No single factor (model, clothes, bondage, gag, etc.) guarantees a high rating or condemns the set to low rating purgatory.

3) A model should not worry if one of her sets is low rated. She’ll very likely have very high rated ones as well. Any model who has worked with us over a longer period is guaranteed to have a few low rated sets. It isn’t her fault, it’s likely that some other factor is at work.

4) That said, there are some models whose average rating over all sets are higher than others. This may be because she’s good, or because she’s new and exciting, or that we shot a lot of sets which happen to hit the spot on the first shoot, or that she’s brought a fan base with her to the site (whoo! we like that when it happens!)

5) You can get some idea of who our favourite models are: count the number of updates. You really have to factor in how long ago we first worked with them and a few other things. There’s still many reasons why we might rebook one model and not another- some GREAT models retired soon after we worked with them, some good models we’ve just not got around to rescheduling, etc..

6) Broadly speaking, the models whose sets are most highly rated on average are also the ones who we’ve worked with most often. Which is nice. At least our opinions and the voting members aren’t completely divergent.

7) The spread in average rating from model to model isn’t huge, except for a few outliers (I’ll name no names). There’s a distorting effect from the “newer sets do better” factor which has to be accounted for, though.

8 ) On the keywords, the variation is big from set to set (so there will be some very low rated sets even in a theme which broadly does very well, and vice versa).

9) Nude in metal bondage sets go down well- average rating 3.92. Historical outfits tend to do less well (average 3.23), but again a huge spread, with some of the most popular sets ever being historical ones.

10) Reverse prayer, hogtie, frogtie and handcuffs do well (around 3.8 ) with suspensions being less highly rated in general.

11) Gags don’t seem to affect the ratings. I’m sure some people love a particular sort of gag, others prefer ungagged- they all end up with the same average rating. So for every ballgag voter there’s maybe a no-gag voter to balance it out?

12) Foot torture and bastinado don’t do so well on average, but tend to be very polarised, so if you like them, you REALLY like them and will give them all 5 stars, or you’ll tend to give them all 1 stars. Reason enough to keep doing them even if it wasn’t my personal kink (which it is).

13) Sets with uncommon themes do less well. Bit of a chicken and egg situation here, clearly, as if your interest was purely in a theme we don’t often do, you’ll not have joined the site in the first place.

14) At the time I did the big analysis, stills sets were rated higher than videos. This is no longer the case, so maybe the 3-year loan to acquire the RED wasn’t such a bad idea.

After all that, you might be getting a hint of the headache that rears up every time the voting/rating system is discussed 🙂 🙂 🙂

So in the end I have come to regard it with interest, factor the voting results in to my thoughts at a background “be aware of it” level, but try not to get stressed or lose sleep over it. I try to make the models do the same- what do you think?

Cheers, Hywel.

Speaking of the voting system, we just implemented an under-the-hood revision to the system which should greatly speed up a few pages (e.g. top ten list, advanced search). There shouldn’t be any effect on the user side except that things might feel a bit speedier. If you notice any oddities, let me know.

For the curious, the issue was that we now have over 6 years of voting data. Compiling the statistics from individual votes had become impractical, so we’ve archived the older votes and pull the numbers from a summary table. Hence, a big speed-up.

Cheers, Hywel.

For the Katarina Nikita set, I suspect the low ratings came in large part because of the long-sleeve blouse, the long pants, and (most of all) the shoes. Put the model in “daisy duke” shorts and a halter top, and take her shoes away, and I’d bet that the ratings would have been much higher, even with everything else unchanged. Or even if the set didn’t have the bare breasts show up. For that matter, simply having Katarina lose her shoes, after the first dozen photos, might have boosted the set’s ratings.

Hywel, I know you’ve said that nude sets rate higher, but have you done any analyses on how different kinds or levels of clothing rate? In particular, analyses on how “clothed, but the arms/legs/breasts/navel is exposed” sets rate? If it turned out that any sort of clothing dragged down the average rating vs nude sets, that would be interesting to know. But it would also be interesting if it turned out that (e.g.) long sleeves were what mostly dragged down the ratings of clothed sets, with “clothed but sleeveless” sets turning out to have nearly the same average rating as nude sets.

We do that where practical, and where it wouldn’t do violence to the story.

The main trouble is that if the model is in lots of bondage, getting her naked part way through the set requires either clothes cut off, or undoing some of the bondage. Fine if that’s just wrists, not so fine if it took 45 minutes to rig and covers her head to toe 🙂

But yes, our aim is to shoot a lot more sets with both clothed and nude elements where the bondage and story supports it.

Cheers, Hywel.