| |
BGonline.org Forums
Cubeful equities: how to derive from cubeless?
Posted By: Matt Ryder In Response To: Cubeful equities: how to derive from cubeless? (NJ)
Date: Friday, 8 January 2010, at 5:58 a.m.
From personal experience, I can tell you that there is very little financial incentive to improving a bot past the current world class levels.
This is very sad. I just bought your iPhone app for $5, if that helps! :-) It's a classy piece of work.
I would've thought that the 'average backgammon player' isn't really the target market for the author of a new bot (it seems there are already many programs out there with more than enough 'features' for the casual player).
I still think the community of pros will all want the bot that demonstrably provides the best evaluation engine. And most of them would be prepared to pay for it. So I think there's still a largish specialist market to be tapped. (Certainly the 'average backgammon player' wasn't shelling out $300 for Snowie!)
Apple's rumoured upcoming tablet/slate platform might be a lucrative avenue to expand into in this regard. (You'd have a headstart over the open-source community there...)
The "killer feature" I'd like to see in a mobile backgammon app is the ability to enter positions and get an evaluation from a strong neural net such as yours. It would be handy to have an ultra-portable arbiter in those inevitable chouette disputes/wagers about which play is best. At the moment, I take a photo and analyse after the fact at home. Not nearly as much fun.
In fact, I've thought up many ideas on how current bots could be improved, but I don't have any reason to work on them other than out of my own curiosity.
Keep going. Ideas born of curiosity can occasionally turn into gold, quite unexpectedly.
2) Add an opening book, like in chess. After all, every bot has a database of bearoff positions so why not opening positions as well? Stick has already done the rollouts.
The key difference is that the bearoffs are calculated a priori. Opening positions are still being investigated by Nack, Stick et al. Even just as far in as the third move, there are thousands of as yet unexplored variations, especially if you add AtS scenarios into the mix.
How about a bot that (with a player's permission) automatically records early rollouts to a central database over the web, assuming they haven't been undertaken before at a given strength? Then we could all download regular updates and this "opening book" would rapidly improve based on a cumulative global effort.
| |
BGonline.org Forums is maintained by Stick with WebBBS 5.12.