@Jinn86 - the way you and Z22 use the system is a perfectly valid way to do so, Some of use use a very different method of assigning the ratings - the effect of the clashes between the two methods is that the system is broken.

I claimed that Z22 was deliberatly trying to break it because he said so himself - or rather he said words to that effect when he said he was "gaming the system".

The rating system is useless because it's been made useless by many of those doing the rating.

I think the rating system is fine once enough people rate the card. Also, people need to be aware that earlier cards in a series will often have lower rating because people are "trying" out the girl. The lowest rated cards generally have technical issues or have steered clear of what totem normally offers. So really low quality, non-full nude, more pricey generally piss people off...and this is o.k as it clearly shows to totem that they need to release high quality cards with the most options for nudity, and they need not raise their prices just because of a costume. The gap between a 4.6 and a 4.8 is generally just preference and those qualities can be gleaned from a demo. The rating system is fine but the "preview" system could use a dedicated textual previewer. A person like zzmaskers that gives a mostly objective run down of what is in the card. Demo, pictures give about 50% of a preview.

@orclover

The rating is broken because it is not at all clear what the scale of 0 to 5 means. In particular people provide ratings based on very different ideas as to what the extremes represent. As a result we have ended up with something where the vast bulk of the cards have a rating somewhere between 4.0 and 5.0.

Those of us that want to submit ratings expressing our opinion as the quality of the card that is somewhat above average can only choose between offering 4.5 or 5.0 stars which in effect gives us a choice between "a rather average card" and "an exceptionally good card". However, if we think a card is below of average quality we seem to have much more choice but if we are trying to indicate what the general quality of the card is compared to others in the collection then the real choice is between 4.5, 4.0 or 3.5 or effectively a choice "rather average", "somewhat below average" and "much worse than the bulk of available cards". For those that are trying to be objective in this way a rating of 3.0 or below would represent an unbelievably bad card

On the other hand there are those that use 0.0 to indicate that they personally don't like the card - often for some very ***** perceived flaw such as not taking her shoes off. Their are also those that award 5.0 to any card that they like while yet others think that 5.0 should represent an absolutely perfect card and that a ***** flaw in one clip should be enough to disqualify a card from a 5 star rating. This subthread has been triggered by one user who was simultaneous claiming that a single flaw in a single clip should disqualify a card from being 5 stars while submitting a large number of 0.5 star ratings because he personally did not like a card.

As a result what we see for a rating is a mixture of at least two very different rating systems and we have no idea what the mixture means. A card can be rated as a "average" card with 4.5 stars either because everybody has agreed that it is "average" or because many people have rated it as 5 stars and a few (roughly one tenth as many) have rated it as 0 stars. The ratings cannot distinguish between a rather ordinary card and one that most find exceptional but a few don't like for ***** personal reasons.

Despite these problems the ratings manage to produce useful results, but ones that are not nearly as useful as they would have been if there was only one way for people to interpret what was meant by each level when submitting a rating (and everybody abided by this when submitting ratings) .The system is very definitely broken, though not to the point that the result is completely useless.

You can't even sort the cards by rating in the store or collection, it's only "real" use is the preview tab and that is broken as everyone has different tastes (many of us don't like the "curvy" girls, so won't buy the card anyway, but those that do seem to rate 5 stars regardless of performance perhaps thinking totem will hire more girls of that body type). I have loads of cards rated by others that have a total rating less than 4.0.

This subthread has "triggered" YOU TheEMU posting wall after wall of text. Your insistance that i rate cards your way is out of order. However, this thread may go someway to the team understanding why the rating system is broken and why the preview tab is useless after a while. So at least there is that...

As for my single fault = -0.5.... have you ever watched gymnastics or ice skating? Rhetorical question.

@orclover

The rating is broken because it is not at all clear what the scale of 0 to 5 means. In particular people provide ratings based on very different ideas as to what the extremes represent. As a result we have ended up with something where the vast bulk of the cards have a rating somewhere between 4.0 and 5.0.

Those of us that want to submit ratings expressing our opinion as the quality of the card that is somewhat above average can only choose between offering 4.5 or 5.0 stars which in effect gives us a choice between "a rather average card" and "an exceptionally good card". However, if we think a card is below of average quality we seem to have much more choice but if we are trying to indicate what the general quality of the card is compared to others in the collection then the real choice is between 4.5, 4.0 or 3.5 or effectively a choice "rather average", "somewhat below average" and "much worse than the bulk of available cards". For those that are trying to be objective in this way a rating of 3.0 or below would represent an unbelievably bad card

On the other hand there are those that use 0.0 to indicate that they personally don't like the card - often for some very ***** perceived flaw such as not taking her shoes off. Their are also those that award 5.0 to any card that they like while yet others think that 5.0 should represent an absolutely perfect card and that a ***** flaw in one clip should be enough to disqualify a card from a 5 star rating. This subthread has been triggered by one user who was simultaneous claiming that a single flaw in a single clip should disqualify a card from being 5 stars while submitting a large number of 0.5 star ratings because he personally did not like a card.

As a result what we see for a rating is a mixture of at least two very different rating systems and we have no idea what the mixture means. A card can be rated as a "average" card with 4.5 stars either because everybody has agreed that it is "average" or because many people have rated it as 5 stars and a few (roughly one tenth as many) have rated it as 0 stars. The ratings cannot distinguish between a rather ordinary card and one that most find exceptional but a few don't like for ***** personal reasons.

Despite these problems the ratings manage to produce useful results, but ones that are not nearly as useful as they would have been if there was only one way for people to interpret what was meant by each level when submitting a rating (and everybody abided by this when submitting ratings) .The system is very definitely broken, though not to the point that the result is completely useless.

I think we are ***** review and preview. A review should tell us the technical quality of the product. How is the sound, image quality, duration of the card. How many true levels of nudity does the card have, does it have have a diversity of outfit pairings and how is the makeup, hair, and lighting all interact. You know, the things that totem hires people for...A preview kind of tells us the flavor of the card...is it western style dancing, ballet, slow tease, raunchy...I think this sort of thing is best left to an animated gif of some nice shots from the card...maybe 10 seconds long...Previews should merely be visual teases of what is in the card...and a review should tell us that the card is technically solid before we pay money. The proof that the current review system works is that I think it would be a very bad thing if it was removed. There are many cards that have technical issues, or weird things going on in the card that are explained by purveyors of the card. This system provides a layer of accountibility that is really important. If a card pisses of buyers it will have a lower rating...if it is a normal card it will be between 4.5-5...that is enough for most of us.

LMAO, todays release went from n/a to 5 stars before any customer could have watched the whole thing. Yeh, carefully considered rating...

Edit: This was before Razielle commented and rated 5 stars on the card.

You can't even sort the cards by rating in the store or collection...

How is that so? You can do that without problem, I am doing it all the time.

You can't even sort the cards by rating in the store or collection...
How is that so? You can do that without problem, I am doing it all the time.

Oh yeh... oops. lol

Re: Subjective ratings

All rating systems have this problem. Even with published definitions of the levels there will be people who don't read them, disregard them or don't care. Although it seems counter intuitive, one solution is to reduce the number of choices. At present there are 11 choices. This is far too many. Three to five choices is probably best. I'd recommend three.

I think you are thinking too hard about the rating system. Every rating will always be subjective no matter what. But the results are relatively objective. Objective is what majority thinks about it, it doen't reflect what minority thinks of some card. That is automatically achieved when there is enough people in the system.

You can always read what problems with some card have people who are rating it too low (if there's not the message: game the system) and then you can evaluate if it matters to you or not at all (like some shoe lovers giving 0 for bare feet).

About the rating system, 95% of voters are either giving a 5 or a 0. There is hardly any vote in between. I guess unless you have a strong opinion about a model, you don't bother to vote.

@Orclover - no I am not ***** review and preview - both of these are different to "rating" and I am talking only about the rating system, not the review system even though they are closely tied together.

@Rex - thank you very much for the information regarding the statistics. This reinforces the view expressed in my recent comment that if I was to replace the rating system with something new it would be with something like a simple binary up-vote/down vote system that would not ***** from the current system's ambiguities. If for the most part people are only using two of the 11 possible choices then just offering those choices seems to be a sensible course - however I see that there are a significant number of those of us that bother to comment who do award 4.5 stars so a three choice system might be preferable (even though the effect of voting for "average" logically should have little or no effect). Part of the ambiguity arises because although we can get an idea about the vote distibution for those that provide comments we have no idea about the distribution of ambiguous votes - I had hoped that the two distributions were similar but from what you say I would guess that the annoymous votes are even more biased towards the extremes than the visible votes.

@Dorsai6 - the rating system would not ***** from the problem if we had access to the raw data, or to a resonable summary of the distibution of awarded ratings and could make our own judgements rather than trying to express a complex situation with a single number - however most users presumably would prefer a single number no matter how meaningless it is when you try to analyse it. I fully agree that a three level system would be much better than the current 11 levels.

If 95% of voters are either giving a 5 or a 0 then a three level system would have almost the same results like it is now, only the average number would be 0.8 (in system from -1 to 1) or 1.8 (in system from 0 to 2) instead of 4.5.

Yeh, a like/meh/dislike might work out better. Adding a popup to rate after watching all clips once may be an idea too rather than being able to rate/comment without even downloading the card nevermind watching all the way through. That may be a bit much though as you might decide you don't like a card in 5 mins. So maybe after one clip has played it unlocks the rate/comment button.

Yeh, a like/meh/dislike might work out better. Adding a popup to rate after watching all clips once may be an idea too rather than being able to rate/comment without even downloading the card nevermind watching all the way through. That may be a bit much though as you might decide you don't like a card in 5 mins. So maybe after one clip has played it unlocks the rate/comment button.

It still doesn't make the rating more valid or accurate...I might rate the card a zero because the girl didn't twerk, although she was a master at striptease, the card was high quality, had a long duration, and a great outfit. If my desires aren't met I would be ***** to give a 0, because it obviously isn't a 5. With the current system I can acknowledge that the card is high quality in a lot of ways but I find it boring by giving it a 3.5 or a 4. I have rated many cards between 3 and 5....and feel my rating counts in such a way. Otherwise I am ***** to downvote the cards that aren't my favorite, which isn't fair either.

@Z22: In fact it will be worse, because 95% of votes is already delivered - you cannot anticipate 5 voters will give a negative or meh rating in new system and vice versa all your 0.5 will not get meh or higher rating from you. On the other hand I would change almost all my rating to positive and only very few cards would go to meh from me, never a single card negative, because I simply don't buy such cards or don't think any card deserves such rating.

The advantage of reducing the choices to just two or three levels would be that it would remove the false impression that the current system gives that it provides a more nuanced view. I would much prefer to see separate numbers for up-votes and down-votes than trying to express their meaning by a simple number.

Like Rex said 95% of users is already using upvote/downvote system - 0 or 5. I am one of those who is not using it. If you change it, everyone will be ***** to use it, but the results will be almost the same like it is now.

I don't like the idea of changing the system too much because the vote before and after the change would have a different scale or meaning no matter what we do. I would be more inclined to display stats about the votes distribution on roll-over on the note for ex.

@Rex - I agree that it would not be sensible to change the system now as there would be no way to make any new system consistant with the old. Changing to a two or three level system would only make sense if it accompanied other major changes to the extent that the whole program was different - much more radical changes than from VG or VGHD to iStripper. I would however welcome any extra information that would help me understand what the distribution of awarded ratings actually was, but also you do not want to overwhelm us with too much data which is why I have on a couple of occasions asked if we could optionally see the results as a histogram.

Yeh, if you changed the system you would need to wipe the current scores really and i guess that may annoy a few people. A histogram of the rating spread in a rollover might be usefull for those that buy based on ratings and interesting anyway. I guess the team is quite busy with the VR (yay!) and 3D istripper apps atm though.

I would ask that the team take a look at the way previews are chosen to be pushed as it eventually becomes useless (full of unwanted "high" rated cards) Perhaps allowing ~25% of the cards to be from the full pool regardless of rating would help.

I don't like the idea of changing the system too much because the vote before and after the change would have a different scale or meaning no matter what we do. I would be more inclined to display stats about the votes distribution on roll-over on the note for ex.

Rex, maybe you could add a "trending" feature based on sales figures. You guys obviously have data that tells you if a model is popular or not...this could be communicated using a positive trend arrow on the card. No need to use a negative trend arrow. So if cards increase or stay similar in sales as you move through the set they could said to be trending..I feel like this would communicate that people are continuing to buy a set even when they have 3,5,10,30 cards of the same girl. The ***** side of this is card ratings often go up even as sales decrease because the people rating are the only people still buying the card. Ie...the new Sofie Lilith line.

Converting from the current scale to a 1,2,3 scale is easy and no more ambiguous than the current results:

<= 1 in current system -> 1 in new system
> 4 in current system -> 3 in new system
everything else -> 2 in new system

Since Totem has the raw votes this can be done. Then recalculate the average score for each card.

I would label the choices:
Above average (3) - I feel this card is much better than most cards I won
Average (2)
Below Average (1) - I was disappointed by this card

@Dorsai - the conversion would be far more complicated than that because some of us in effect decide between 4, 4.5 and 5 while others decide to either award only 0 or 5. There were at least two "systems" of voting used and it is this that causes the current system to be ambiguous. I suspect that even with a choice of three levels as you describe there would be many who would still only ever use the two extreme ends of the scale while some would try to use it as you suggest so even a three level system would end up with some ambiguity. The problem is not one of defining a sensible system but is one of getting the users to all use it consistantly. There is also ambiguity in what constitutes "disappointment" - if the purpose of the rating is to be of use to others in general then it would be "wrong" to down-vote just because, for example, shoes were kept on until the last few seconds no matter how much that is important to the person providing the rating, some will ignore things like this in order to try to be objective while some will just go with their own likes and dislikes. There is nothing wrong with either approach, but they are not compatible with each other.

Even if Totem could convert from the old system to the new as you say it is likely that the distribution of new votes would be different with a three level system - it certainly would be if everybody did decide that 2 was "average" as you, and I, would like and so the scores for new cards could be expected to be rather different to that for old cards. I do not think that there could be a sensible way to maintain continuity.

Re: the conversion would be far more complicated

Given the ambiguity of the current system, I don't see the need for anything more complicated. I do data analysis and data cleansing for a living. Given that Totem knows everyone's votes, it would be possible to profile each member and adjust the conversion results based on their individual voting profile. Normally, I do a lot of data analysis before I suggest any data conversion or cleansing. In this case I don't have access to the data and from what Rex said I don't think that would be worth the effort.

Certainly the voting going forward with a 3-choice system is very likely to show a different pattern than what the converted votes would show. Again, I don't think it would matter that much.

All subjective scoring systems have this problem. If it were really important, I'd implement a quota system: No more than 20% of your votes can be 3 and no more than 20% can be 1. This could be enforced, but it would take programming effort and might upset some members. I've done this in a few cases where the voters were all employees and they were being paid for their time. In some cases the system prevent too many high or low votes. In other cases management was given reports of people who were exceeding the limits and the worst offenders were told to revise their vote.

@Dorsai6 - you say

Certainly the voting going forward with a 3-choice system is very likely to show a different pattern than what the converted votes would show. Again, I don't think it would matter that much.

On the contrary the fact that there would be a discontinuity in the meanings between old and new ratings would mean that you could not directly compare the ratings for old cards with new cards. That would be OK for those like you who do "data analysis and data cleansing for a living" or those like me who have a scientific background. But the majority of iStripper could not be expected even to know that they would need to take account of the difference, nor should they be expected to know. This, for me, would rule out making a change of this sort.

In any case all of this is rather academic as it is extemely unlikely that any such change would ever be made to iStripper. Where it would make sense is in a new product such as the recently announced VR Strip Club.

Rather than just a star rating it could be a series of questions about the performance/ look/ clothing choice/ dancing ect ect, with 3 to 5 possible selections once a watch threshold has been passed. You know the kind. strongly disagree/disagree/ meh/ agree/ strongly agree. People tend to respond better to questions than a plain rating system.

@Z22 - are you sure that "People tend to respond better to questions than a plain rating system". In my admittedly very superficial exposure to such things the main effect of being presented with a a more complex set of questions (unless cornered by an interviewer) is a tendency to ignore it. Admittedly the results of those few who do respond to a more complex set of questions is probably more useful than those that would respond to a simple rating question.

I guess the spread may just end up the same whatever. As there will be people that just click the max/min values to get rid of the popup. Just the addition of a histogram will help more in the end, for anyone that buys based on rating that is.