Changes made:
- Changed "number of active eHarmony members" and source of information. May provide a more accurate number of dates that you'll get when you first subscribe. (Or, it may not.)
- Factored in the increased likelihood that women with attractive pictures will get more dates. (Why didn't I think of that before?)
4 comments:
My comments:
2. 16,000 comes from dividing the 3 million gain in membership between February and August 2007 (180 days). I didn't make it up. :) It's already four months -- I still await the next news release with an updated membership figure.
4. Buckwalter says, "Our statistical criteria for making a match is that there’s an 80% statistical probability that a match would be in the top 25% of relationship satisfaction." You know Statistics better than I, how much percentage does that figure out? 1/4? 5/16? Do you mean that 3/4 of test takers are rejected?
16-19. You should also acknowledge the desire to raise offspring: What do sources show: how many are Yes, Maybe, or No? This factor reduces matches a lot, I think.
22-24. For aesthetic reasons, I suggest you move this up to between 15 and 16.
26. All your other criteria have their counterpart (e.g., 22 and 24). Number 26 doesn't.
26. Label this figure as "How many matches will I expect?", because the subsequent factors depend on how the communication (or non-) pans out.
31. You might want to mention that this one factors out non-paying matches.
Cool tool. Quite entertaining.
Thanks for your comments! (And I hope you appreciated the well-deserved plug in #2.)
I'll probably be incorporating some of these ideas in the next version of the "Date Predictor". But a few things came to mind...
No, 3/4 of test takers aren't rejected, per se. They're just not considered "compatible matches" to you, and you'll never see 'em on your list. (Other "more compatible" people may see those people.) To be in the top 25%, that means that you don't see the bottom 75%. The "80% statistical probability" means that there's a 20% false positive rate, but, unless eHarmony discloses it's false negative rate, I can't really incorporate this in my model. One of many weaknesses caused by a lack of information.
"Lack of information" is also what led me to leave out the desire to have offspring. If you can find data on this, I'd appreciate it! Until that happens, I figure that users can just incorporate it into the "miscellaneous reasons to reject" section at the end.
I request permission to repost today's RelyID article in its entirety, duly credited. Please let me know.
Permission granted... Somehow, I though you'd ask. :-)
Post a Comment