13 comments

[ 5.4 ms ] story [ 38.0 ms ] thread
Since people seemed to like this non-Flash Go game ( http://news.ycombinator.com/item?id=678145 ) I thought I would try and see if I could get any reviews for my side project Mancala Time. I'll admit I submitted this (in a very early form) about 3 months ago but no one made any comments. Thanks.
I thought the implementation looked very nice, although I hadn't played by those rules before. Well done.
Thanks for checking it out. Yeah, mancala is a genre of games. This is actually kalah, but most americans know this version as "mancala." If the site catches on I plan on adding Oware...which is my guess of the version that you are used to?

http://en.wikipedia.org/wiki/Oware http://en.wikipedia.org/wiki/Kalah

Kalah is a simple and somewhat broken game (the first player is heavily favored).

I would very much like to see Oware or Bao supported, these are much deeper mancala games. I haven't seen many implementations of Oware (outside of GNU Aware). The flash/java implementations of Bao also leave a little to be desired: http://www.baogame.com/ http://en.wikipedia.org/wiki/Bao_(mancala_game)

True story on Kalah being broken. This is the main reason why I added in the ability for users to be 2nd player. I will start to work on Oware sometime soon now that it seems like there is at real interest in it. Thanks.
I've played a few different variants (some with 3 seeds, some with 4, some with captures, some without) but all of them involved "relay sowing".
I like it... maybe you can improve the difficulty of your AIs some?
Thanks for checking it out. I do plan on improving them by using the list of moves that I record...I am thinking of making a Robot API or something too...but it's hard to be motivated to improve them too much when a large majority of people lose to Rex. This also makes the idea of finding the best move out of the moves that have occurred flawed (for now anyway) too.
Rex pretty much prioritizes getting a 'second move'. I've yet to see him protect against a capture.
Yeah, all he is capable of doing is finding an extra turn then it's basically pick one at random. Although, there's a bit of randomness mixed in. There's about a 1 in 6 chance that he will basically become a higher level robot on a given turn.
Gotta love these CS 101 C++ projects becoming weekend web apps.