Also, I think it might be buggy, or am I just not getting your UI? Each of my 10 submissions have one lost game and one won game. The lost games are lost on move #1, while the other game always wins. I first thought that it had to do with the AI spending too much time, but it seems improbable that I'd get the same result 10 times in a row. Username is helgefmi.
The score of a game is very much dependent on luck, though.. I've made an AI bot (https://github.com/helgefmi/c2048) that averages on ~50k score, but its highscore is ~220k. Sometimes it gets 13k. So to get a good score (or your best score) on hackerrank.com, you'd need to resubmit 500 times. At least with the algorithms that I use.
Why not? That would be awsome. There's already a python repository for interfacing with bots, and it works using unix socket. It's here: https://github.com/matslindh/4096
I'm not sure 1vs1 battle matches is the best idea (it's not chess), but it shouldn't be too much hassle to make it into a platform that tracks the highest scores, best averages etc. It could be divided into different time controls (1 sec per move, 3 minutes per game etc.). The point would be to see who could make the best AI.
It could have a simple flask app, or similar, that just shows the high scores and makes it possible to reserve bot names (so other people can't play as you.)
I'm just putting ideas into someones (noones) head. But if people who's currently/who wants to play with 2048 AI responds to this, I might make an effort to get it starting.
I was coincidentally working on the same thing when I saw this yesterday. Pushed the code to https://github.com/helgefmi/c2048 if it's of anyones interest.
It's just a wicked fast board implementation with a simple depth first search, as of now. But the idea of "making up" an opponent to make it possible to do alpha beta pruning is a cool idea. I might try to implementet it myself.
I regularly get scores above 50k with AI_DEPTH=5-6 and NUM_TRIES=20-30. My record so far is a score of 220k :-).
I'm new to this space, but reading about it it seems quite impressive to me. So enlighten me.