On August 27th I reported that I had been experimenting with a tweaked version of Stockfish, Stockfish 250816 64, and it had performed well against the regular version. Earlier in July I had decided to test it out in some rapid games (10 basic plus 1 day per move, no vacations) on Lechenicher SchachServer.
Of course you have no idea what engines your opponent is using, what hardware they are running or how much time and effort they put into their analysis, but it's pretty safe to assume that in these rapids most of them are using home laptops and one of the usual engines: Houdini, Stockfish or Komodo and they are not conducting a lot of lengthy analysis because they are "rapid" events.
In the section I entered to test this engine, I lost one game fairly quickly, but that was in no way the engine's fault.
As black in the King's Gambit I failed to check the opening book and omitted a move with the result that white came out of the opening with a clear positional advantage. By move 18, I had a weak c-Pawn, a lack of space and discombobulated pieces and by move 26 the position favored white by about 1.5 Ps. After some fairly long analysis sessions and several Shootouts using different engines almost all the games were lost...no point in continuing.
As white, a Belgrade Gambit in the Four Knights Game fizzled out to 0.00 by move 13 so we agreed to a draw.
As black in the Chebanenko Variation of the Queen's Gambit Declined (1.Nf3 d5 2.d4 c6 3.c4 Nf6 4.e3 a6) black has a solid structure and we reached a dead even ending after 26 moves and agreed to a draw.
As white I varied in the Sicilian just a tad. After 1. e4 c5 2. Nc3 d6 I played 3.Nc3 and got surprised by 3...e5. The resulting closed position fizzled out to a 0.00 evaluation and a 30-move draw.
In the other finished game, in the early middlegame I did not go with Stockfish's top choice, but tried a move that looked promising and within a few moves my advantage was clear...about one Pawn. Black tried to hang on, but it soon became apparent that after we traded down to the ending that black was lost, so he resigned.
There's one game unfinished, but my opponent is a slow mover with 49 running games and we aren't even out of the opening yet.
So, in this brief test in actual play it doesn't appear that Stockfish 250816 64 has given me any marked advantage.
Next on the agenda is a test of SugaR Pro.
No comments:
Post a Comment