166.137.88.42

Unregistered editors using this IP address received messages on this talk page years ago. Since users of the IP address have likely changed, these messages have been removed. They can be viewed in the page history.

Halloween Gambit

One of the commentators on this gambit has bragged here of finding refutations or improvements to outdated, published analysis using the chess engine Houdini 3, using a very impressive 16 processors. I would like to point out that that version of Houdini was very likely unable to utilize nearly so many processors, for as Wikipedia notes elsewhere, version 5 of that program could not even do so. More to the point, Houdini 3 is not exactly state-of-the-art these days, any more than say, Eric Schiller's published analysis from the last century.

The go-to engine these days is Stockfish 16.  And yes, it considers the Halloween Gambit to be unsound, with best play.  And FYI, the retreat ...Ng6 is preferred to the alternate line with ...Nc6.  The eval is about two pawns advantage for Black-- a winning (if not overwhelming) edge.  Former world chess champion Max Euwe's line, which has Black jettisoning his d-pawn before playing ...c6, is improved upon by omitting ...d5 and going straight to ...c6.  All this can be found using a moderate number of processors, on an ordinary computer, and the strongest free downloadable chess engine.
 In the last century, there was stiff resistance from the old guard to the idea of computers --or ordinary chessplayers using them-- correcting the published analysis of the supposed greats, as exemplified by, for example, GM Larry Evans in Chess Life magazine.  He seemed to regard these efforts as the equivalent of cheating, since such players ought not to be correcting their betters, and probably wouldn't be so darned good at it, were it not for computer aid!
 Claims that computers might one day match or even surpass experts, masters, grandmasters, and finally, world champions, were summarily dismissed in print by esteemed commentators posing as experts on the subject.  All were eventually proven wrong, as well as no real "experts" on the matter at all.  The requirement here that nothing can be included without it having already been published reminds us of those not-so good old days.  To wit: Eric Schiller's chess analysis has certainly been published (a long time ago), whereas Stockfish's vastly superior --and easily verifiable-- current analysis might not have been.  Bad ideas may eventually die, but the process is agonizingly slow for us mere mortals to observe.
 For some reason, I am reminded of a story about a medieval university class sitting around discussing what Aristotle wrote on the matter of how many teeth a horse has.  One unfortunate student suggested they step outside and count for themselves the actual number of teeth in a horse's mouth... and was promptly beaten and thrown out.  You see, that number had not yet been published anywhere.  And, you guessed it, Aristotle got it wrong. 172.87.18.203 (talk) 10:21, 8 April 2024 (UTC)

---