As I said earlier, I can't guarantee that the savedgames information will continue to be there in the future.  It was a quick hack that I pulled together in about 20 minutes one evening  just to see how well it would work, and it was never intended to be an official FIBS resource.  I don't monitor the process that generates it, and while it's been quite stable thus far I probably wouldn't notice for a while if it went south.

The more I see decisions being made based on misunderstanding of the information, the less likely I am to want to continue providing it.  Just today, I had the joy of witnessing something like this:

(someone) shouts:  Don't play so-and-so, he's a dropper.
(someone else) shouts: He is?
(someone) shouts: Yeah, he has 20 saved games.  He's a dropper.

I was curious, so I went off and reviewed the guy's saved games.  In the vast majority of them he was either significantly in the lead, or the game was about even... far from a classic dropper pattern.

Beyond that, I'm coming to understand that the numbers are a less accurate predictor than I hoped they'd be.  A couple of weeks ago I investigated someone in the dropper's hall of shame, only to find out that most of his matches were actually finished.  He'd played quite a few matches in the hours before the list was calculated, and that was enough to push him into the top 30 list.

In short, I'm not too enthusiastic about providing an improved interface to data that have proven to be unreliable, that are often misinterpreted, and that might cause unintended harm to innocent people.

(Oh, and if I *was* going to provide an official API, it wouldn't be XML.  Morse code would be more likely than XML, or maybe Swahili and BCD... I despise XML.  But that's another topic altogether.)

Actually, spielberg, it wasn't rif_raf, but someone else.  Their appearance on the hall of shame page was a temporary aberration-- they were only there for one day.  I'd hate for people to shun someone just for something like that.