I find pro games recommendation interesting. Its a common question, and it could be a good addition to some aspects of Kaya. Adding it to the server could be trivial, or we could simply provide an archives API.
One of the many server supported features that didn't get to the client include something of this: instead of an analytical measure of style, we thought of prompting the user you just played, and make an average of your last 20 "opinions". Our plan was so you could tailor either you who you are going to player, or know beforehand a little bit more about your opponent. We could eventually add this to a players profile: https://beta.kaya.gs/users/conanbatt Send me a private email if you are interested in collaboration. On Sat, Apr 13, 2013 at 6:07 AM, Josef Moudrik <[email protected]> wrote: > > Hello, > > as a part of my thesis, I've just launched an initial version of a webapp, > which tries to estimate strength and style of a player (based on a sample > of games he submits) and give some teaching reccomendations: > > http://gostyle.j2m.cz/webapp.html > > > It works by analyzing patterns and some other stuff from the games, and > applying machine-learning tools to determine the style and strength. The > methodology is somewhat detailed in a article we've written with Petr > Baudis, see gostyle.j2m.cz for more. I plan to add more info soon. > > As this is a sort of coming-out, I would be glad for any feedback :-) > Do you like it? > > > Thanks & regards, > Joe Moudrik > > _______________________________________________ > Computer-go mailing list > [email protected] > http://dvandva.org/cgi-bin/mailman/listinfo/computer-go >
_______________________________________________ Computer-go mailing list [email protected] http://dvandva.org/cgi-bin/mailman/listinfo/computer-go
