You're welcome.
Thanks for your suggestions, the ones on this thread and the ones last year when I was setting this up.
Most of the process is automated. I started this before E-I implemented it's one review per reviewer/escort rule, so I wrote the programs in a way which would use only the most recent review for an escort/reviewer pair. If I were rewriting them completely then I would probably just use all the active reviews, as you say. Practically I think it is unlikely to make much difference. Cases where a review changes from positive to negative or vice versa are rare.This is already done, also by the review team. I suppose you should only have a quick view to see if maybe some duplicates have not yet been deleted. This should work only for reviews of the same kind, but if the positive changed to negative or vice versa, I think you should take both into consideration because it affects the “reliability” factor. This might be done as well by the review team by I think I have also seen situations where a negative review has been covered by a subsequent negative or vice versa.
The goal when I chose the number 7 was to eliminate most of the fakes. If an escort writes a review of herself, that review won't be counted unless she writes reviews of six other escorts, which seems unlikely. I'm not sure if this is really as effective as I had hoped. Many, perhaps most, fake reviews are not written by escorts. Agencies write reviews of their escorts, and they could easily have 7 or more from the same account. But it seems unlikely that an agency escort would stay around long enough to get on the list. A bigger problem is professional reviewers. There are people who offer to write reviews in exchange for money or discounts. These people may have dozens of reviews from the same account. Unless they are lazy these fakes can be very hard to detect.- Only reviews from reviewers with at least 7 reviews count:
Up to you. 7 is a decent enough number. If you make it too high, you’ll leave out lots of excellent escorts a little unlucky when it comes to getting reviews. But keep in mind as well that the more times it will pass, the more escorts you will have for this list. Because of this, I would personally suggest each year to add a few more to this figure, to count for time passing. A figure of 3 – 4 extra this year should be fine. Most active reviewers will have at least this extra figure in one year time.
I think I will do something like this. It would be better to check whether the profile has been active, but I don't have access to that information. I still haven't decided how long to wait before removing people. Probably a year.- Retirement
You can safely list escorts that have active reviews within the last 1 year or maybe 1.5 years. If you ever get an exception to this, you can amend it manually, like you did in my case.
One thing that might be worth pointing out is that I am counting only visible reviews. If an escort opts out of the review system she will stop appearing on the list, although not immediately. Normally an escort with excellent reviews is not going to want to opt out, but some escorts do this when they retire, as a precaution against people later finding out about their previous job. For example, Diana M disappeared from the list when she opted out after retirement. I would recommend this to any escort who is retiring, since you can always opt back in if you decide to start working again.
I'm not sure if I understand this correctly. Currently I am averaging the "overall satisfaction" ratings, and then sorting them. So if I changed from 1, 2, 3, 4, 5 to -2, -1, 0, 1, 2 the effect would be to shift everyone's average down by 3, but it wouldn't change the order, so the list would be the same. But I'm not sure if that's what you're suggesting.- In particular, there are a lot of very good escorts who don't make the cut-off of 50. If you think about the ratings as a star system, where "Very Happy" is five stars and "Very Unhappy" is one star, there are escorts who fail to make the cut because they have "only" 4.8 stars on average
I think here is your big problem, why the list needs to be extended at the moment . I think here should be given a negative score in certain situations (very unhappy, fairly unhappy). “Very unhappy” and “reliability” don’t go exactly hand in hand, and such reviews should be deducted, same way as the Escort of the Month does.
Let’s just take as example 2 escorts with the same number of reviews, with a score right below the cut of point to make it to the list. One of them is touring, one of them is not (like me for example). Assuming that touring escorts has a chance (in theory) to get more reviews, then the touring escort can make it to the list even by getting a negative review.
So maybe you can make something like -2, -1, 0, 1, 2 (whatever numbers you wish preven they are in sync with the satisfaction factor).
The question of whether the list discriminates against ladies who don't tour is an interesting one. I think there is a bias, but it is complicated. For example, Dublin has more than enough punters for an escort to make the list quickly without touring. In Cork or Galway or Limerick it is harder. In a smaller town like Kildare it is very difficult. There you would typically be seeing the same fairly small community of punters repeatedly, and many of those wouldn't have reviewed enough other escorts to qualify as "established". Similarly, for touring escorts the list will be biased towards ladies who tour the major cities rather than the small towns. So travelling escorts do better than settled escorts in a list like this, but it also matters a lot where you are.