World Draughts Forum

I don't understand your paragraph, but I can comment on the "play to the end" part. It is customary in other games, and chess in particular, to always play the game to the end. Tournament directors use special rules to shorten boring endgames, like king+rook vs king+rook (similar to king vs king in...

The major draughts engines, such as Scan, Kingsrow, Damage, Dragon, et cetera, all seem to be closed source. Are there any open source draughts engines out there? Scan and Moby Dam are open source: https://hjetten.home.xs4all.nl/ Thanks, I found the source code for the engines in Rein Halbersma's g...

Na enige tijd zal hij wellicht TurboDambase, KingsRow of Flits willen aanschaffen. Kingsrow is sinds enige tijd helemaal gratis en het beschikt over een heel makkelijke analyse-functie. Je kunt bijv. een partij uit Toernooibase copy-pasten en dan gewoon klikken op Partij/Analyseer Partij. Als je di...

Hierbij nog een artikel uit het Engelse taalgebied: Bobby's Binds Ook dit stukje geeft aan dat het hier gaat om de onderlinge binding van twee witte pionnen (hier f5 en g5). Het heeft dus niet zozeer met een penning door een vijandelijk stuk te maken, maar vooral op de onderlinge rugdekking die de ...

Uit de schaakwereld, in het bijzonder de Duitse schaakwereld: "Eine Bindung besteht, wenn ein Stein (A) einen anderen (B) decken muß, ansonsten dieser verloren ginge". Bekend is ook de zogenaamde "Maróckzy-bind", een pionnenstructuur met c4 en d4. In de damwereld staat binding meer algemeen voor "s...

Thanks Fabien for this update on the NN front. The huge resource requirements are not very helpful to do quick experiments with fast feedback. But there is some progress on the Go front here: https://blog.janestreet.com/accelerating-self-play-learning-in-go/ In poker people also used huge (multi-Tb ...

- think hard about "original work"; I don't consider the current draughts engines "original" enough for healthy competition (hopefully one day), but your rules would make them so, IMO by construction I agree that Scan, Kingsrow, Maximus, Damage etc. who are now using the Scan-like patterns + MTD-f ...

So a few more remarks on this topic. First of all, I would like to thank Fabien for a more than 3 year long discussion (on and off, every time either one of us would get back to it, as we developed new insights and expanded into new territory (patterns, Frisian, Stratego) that demanded slightly more...

In my new Tabula library (https://github.com/rhalbersma/tabula) that I am developing (not yet ready for primetime), I have been experimenting with new board layouts / square numberings. Here's the classic layout for a 10x10 draughts board by using 1 ghost column (also called "separating squares" or ...

I'm guessing 1 or 2 elo at most, but who knows? My hunch is that it is indeed very little Elo-wise. And if it does gain a program without a prior book some Elo it is probably not that a Kingsrow book would suggest radically different moves than a program would make on its own, but mainly the fact t...

That's exactly what I was referring to when I said a book could be copied with a little bit of programming. If I'm correctly interpreting the replies so far, Bert, JJ, and I are opposed to it, Rein doesn't see a problem. I could see this leading to some kind of book wars, where at each iteration on...

In any case, very specific feature requirements and perhaps a lack of familiarity with high-quality numerical software might be the main reasons for lack of use. Another reason might be the cognitive load from learning new tools and also the tight coupling of different parts of a software program t...

The gradient descent code itself is probably best done using a professional library. Rein, I'm sure you know much more about this than I do, but from my point of view it was easier to write my own. I looked at a few of the big packages, and I was intimidated by what looked like a non-trivial exerci...

In my opinion, giving one person access to such a powerful tool and not making it publicly available makes for an unfair competition. It is unfair to the people who don't have access to an optimization program and in a way also to the people who put in the work to make their own. This is the point ...

I think Bert has it exactly right. The data generation process is much more important than the optimization routine. IIRC, in AlphaGo they use >95% of resources on the data generation (half a billion games) and very little time on the actual learning. So tinkering with the search depth, features and...