TalkChess.com

I vaguely recall Tord Romastad's post on fishcooking list where he explained that overvaluing drawn positions is actually good for modern selective search. Getting "good drawn position" might indicate that there is some conversion to superior ending along the way, but it has been pruned one way or t...

Tested that, probably with some depth limit, but failed narrowly (which does not mean anything for Your engine, the rating gap implies I must have failed on several techniques that work for You). IIRC Andscacs might be using it. The real fun comes when you print all the positions when quiescence sea...

Michael, I will be glad if you prove me wrong. But even then I would search for ways to reduce game length. You want to feed transposition table with additional information, trusting that it will help to shape the final search. The idea looks good, as long as this information has a chance to be acce...

I'm afraid this algorithm would search too many moves irrelevent for the second search to be competitive. But when I'm analysing positions under Arena, I often do the following: short search, taking back, longer search. Some tactics are found faster that way. I'd suggest the following: try that appr...

A nameless entity, still weaker than Rodent (I'll release only if it becomes clearly stronger), got this amusing draw after a ridiculous pawn sack in the early middlegame: [Event "Impersontor2019-01(2)"] [Site "LAPTOP"] [Date "2019.01.09"] [Round "18"] [White "Rodentiii"] [Black "Impersonator"] [Res...

I add noise in eval, pseudocode follows:
This way random component of evaluation is always the same for the position, centered around zero, different in each game and does not pollute search code.

Regarding history, one possible enhancement is to update history for a hash move if hash score exceeds beta. At least it helped for me a tiny bit.

Old books say: this is not too good compared to 1.e4 c5 2.Nf3 e6 3.b3. The reason is that now Black can try to block long iagonal with ...e5, immediately or on move 3. So let's try: 1.e4 c5 2.b3 e5 3.Bb2 d6 White's best try is probably 4.f4, but this looks worse than King's gambit, because Bb2 looks...

It seems that TCEC uses automatically generated number of AppVeyor build rather than correct version number 0.275.

A good intermediate goal would be to code an engine whose elo is greater than its number of lines (not chars, like in these little marvels). You can go pretty far with that. I have a 2400 line program that plays at 2700 Elo, CCRL scale.