Touhou statistics

[Akane]

Hello everyone!
A friend of mine is doing a scientific research on statistics in his university, and he decieded to choose Touhou for a subject. So, the guy basically wants to make a tool that would measure the player's efficiency by measuring several key features of their success in games. Ideally he wants to make a rating system that's not only based on score but on various factors like the average difficulty, level progression and others. The guy is all into statistics, and I like his idea.
For now, however, the project is in the trial stage, and is only focused on one game - IN (which, I believe, has the most data in score.dat). And in order to see if this system is working, we need a lot of different score.dat files from players of various skill levels. If you wish to participate - just send your score.dat to his mail: slokinson@yandex.ru
Your cooperation will help this project a lot. And of course, your comments and suggestions are welcome.

[RegalStar]

There are replay archives online though. Unless you need more information than what they keep?

[Mikuru]

My IN statistics are a mess now thanks to spending a while on high-fps attempts so at the moment it would look like I'm worse at Easy than at Normal.

[Akane]

There are replay archives online though. Unless you need more information than what they keep?

Yeah, unfortunately, replays are not enough. Otherwise we'd already invaded the archives. Basically we need all the info from "player score" tab in the menu. And I think you can only get this info from score.dat.

My IN statistics are a mess now thanks to spending a while on high-fps attempts so at the moment it would look like I'm worse at Easy than at Normal.

Oh, I didn't think about that x) yeah, things like high-fps and ultra patches kinda interfer.

[Drake]

Why do people not make separate score files for challenges and stuff

I guess because you don't need to care about it but yeah

As for OP and friend, I'm curious as to what you can glean from score files that have existed for long periods of time. If a player significantly improves over the course of years, where does that factor into this? Without dates on the information, which is explicit in replays at least, how will you discriminate / will you? Will you just take info from e.g. the highest difficulty they have some decent history in? This seems like a difficult problem to find trends in overall.

[Akane]

As for OP and friend, I'm curious as to what you can glean from score files that have existed for long periods of time. If a player significantly improves over the course of years, where does that factor into this? Without dates on the information, which is explicit in replays at least, how will you discriminate / will you? Will you just take info from e.g. the highest difficulty they have some decent history in? This seems like a difficult problem to find trends in overall.

Laaaaaaaaaaate answer, but better late than never I guess.
It would be great to consider dynamics, yeah, but for now we'll go without that and will only try to get something out of data in recorded score. We'll see how it goes, for now we only need to do a lot of tests, and primary requierment is a decent amount of scores.