archive by month
Skip to content

new BASIL bot ladder

Bytekeeper has let me know about the new BASIL Ladder, which is similar in idea to the defunct SAIL: Bots on SSCAIT are copied over automatically unless the author requests otherwise, and matches run continuously. Elo ratings and replays are available. Theoretically, it should be a better test of strength than SSCAIT, since there is no voting to distort results by messing with pairings. Also, without streaming, more matches take place.

It looks brand new, and not many games are recorded. Games are played at LF 3, aka Normal speed. Some of the Elo ratings look suspect (why is AILien so high? why is Krasi0 so low?) and might indicate bugs, but whether ladder bugs or bot bugs I don’t know. It could also be that not enough games have been played for ratings to stabilize.

A peculiarity is that games run with a real time limit of 20 minutes instead of a game time limit, and games which run over are dropped. Authors have incentive to finish off the opponent when winning, and to try to drag things out indefinitely when losing (“I’ll hide that command center in the corner”).

It is “beta”, which in this case means not everything is set up nicely yet. For example, there is a directory full of game results where each result is a JSON file, and the files are sorted by hash rather than by, say, date and time, or bot name. If you want to look over game results, apparently the most convenient way for now is to write a program. I expect that awkward points will be smoothed in time.

A good addition, thanks!

Trackbacks

No Trackbacks

Comments

Barcode on :

Thanks bytekeeper!

(on a sidenote, Steamhammer really doing good on sscait ladder, top7 and top zerg!)

Dilyan on :

100 bots, average 50 60 games each bot have played. That means not every bot has player every other bot.. And they have to play at least 10 games each to show full picture.

MicroDK on :

Bytekeeper is running 3 games in parallel using sc-docker on a 8 -core machine so a lot of games can be played in a short time.

Dilyan on :

But in normal speed? Why not 2x or faster?

Jay Scott on :

It affects latency, not the rate at which the server plays games.

Bytekeeper on :

Thanks for the coverage! You're completly right, I have a lot of work to do :)
Krasi0 is very low because for a long time I didn't notice that BWTA2 caches weren't setup properly by sc_docker. I only noticed when starting one game in headful mode and it told me that I should use supported maps :)
Now it's playing nad climbing the ladder.
Also, while the ELO rating should stabilize, I will reset it soon again. MicroDK made me aware that I use a completly different ELO setup as SSCAIT does.
So I'll restart it. I'm currently rewriting a part of the server software since it's too "hacky" - which wil improve the way I can provide results.

CUNYBot on :

Thanks for the server! It's really nice to see and feels more gratifying than personal testing.

MarcoDBAA on :

The ladder could use (in part) new maps? So that something is fresh and different?

MarcoDBAA on :

Want to report a bug:

Choosing maps randomly doesn´t work.
At the moment (nearly?) all games are played on Benzene.

Marian on :

Great job!
What functionality I would really like to is to filter replays(player name, date from-to, win/loss...) and then download them in a zip.

Add Comment

E-Mail addresses will not be displayed and will only be used for E-Mail notifications.

To prevent automated Bots from commentspamming, please enter the string you see in the image below in the appropriate input box. Your comment will only be submitted if the strings match. Please ensure that your browser supports and accepts cookies, or your comment cannot be verified correctly.
CAPTCHA

Form options

Submitted comments will be subject to moderation before being displayed.