My nephew beat HOUDINI?!?!

Here are results of a quick analysis of that game, using Houdini 1.5a, 512MB hash, 10s/move, at least 14 plies, analysis starting from move 26 (per pfren's diagnosis of 25 book moves). CPU: 2.66 GHz i7, single processor used.

For those unfamiliar with top-3 methodology for detecting engine use, top-1 is the percentage of positions where the 1st egine choice was played, top-2 is the percentage of positions where either the 1st or 2nd engine choice was played, et.c. 

Adam Top1 match: 24/36 = 66.67 %
Adam Top2 match: 33/36 = 91.67 %
Adam Top3 match: 35/36 = 97.22 %

The 10s/move analysis time was based on the OP's statement of having given Houdini 5 minutes for the entire game. This setting is not really correct, but not much can be done without knowing the exact time that the opponents took for each move. I have only the free version of Houdini, so that's another source of inaccuracy.

Given the inaccuracies and small sample, there's no way to draw firm conclusions, however, the numbers are consistent with what might be expected from analyzing a Houdini-Houdini match.

I've ran stockfish through the entire game for black only. For each move he had 5 seconds to think (thus around 5 minutes total, same as the OP said Houdini had to think). Brief results:

21. Qc8?! (suggested Ne5). Difference: 0.57

31. Bd8?! (suggested Ra4). Difference: 0.81

43. Kf7?! (suggested Bb6). Difference: 0.53

44. Ke7?! (suggested e4). Difference: 1.30

48. Bf4? (suggested d3). Difference: 2.86

54. Ke5? (suggested Kc5). Difference : 1.76

Houdini blundered really hard, especially at the end. 2.86 in particular is just too much. A variation stockfish suggested on the 48th move:

This is a huge difference with the moves that were actually played.

Of course this analysis only took 5 minutes and was only for black, it is very likely that at least one of the following may be assumed:

1. Houdini suffered greatly from limited time (especially at the end).

2. The hardware on which houdini was run lacked.

3. Houdini was not playing on full strength.

My conclusion: I will not speak about your nephew's skill in chess, but implying that Houdini was beated by your nephew is misleading. Houdini was handicapped, in one or more ways.

This is kind of fun when you pretend you have a chance :p
5/0

Wow!

This is a troll thread.

Note: in move 54, Ke5 was played, not Ke7.

Houdini 1.5a's ideas about these moves added above in blue. Very different from Stockfish; the magnitude of differences is quite surprising.

From looking at the full analysis, it is evident that horizon effect is in play, e.g. on move 40, white plays top-1 choice Bc1 (which evaluates 1.19 at 18 plies), and black responds with top-1 choice d5 (which evaluates 1.66 at 18 plies). Note that advancing the analysis by 1 half-move changes the evaluation by almost half a pawn.

Yeah, you can't trust a straight engine evaluation.  You have to give it a lot of time and then explore different moves.  It's often a computer "changes it's mind" a few moves down its preferred line... go back to the decision point and it suddenly likes a different move better.

I'm told pros use multiple engines too to cover for particular weaknesses but I've never tried that.

No need to make top3 analysis, although they are a 100% proof.

Anyone whose playing strength is over 1400 and had played against a strong engine should know that this thread is full of crap...

I played against my Shredder around 40 times and never won once. I drew him 3 times (2 rook endgames and 1 opposit color bishop endgame)... Oh, did I tell you that the time limit was 2 seconds per move for engine and I had no time limit? Yeah... this win seems legit.

Isin't it strange the original poster has left no additional comments to any of the comments posted on the thread? My nephew is autistic, and if he played like this I would be arguing every post made against me . Its suspicious, indeed. I would also jump on the 10K bet.

That was a typo, in analysis Ke5 was used.

And about the differences, the GUI is quite 'strange' on how it handles the differences. The suggested move showed the same difference as in the half-move before. So don't take it too serious. Nevertheless, it was more to show how 'bad' the moves made by black were and how disadvantageous it really was in numbers.

Page 3 #47 is the OP again.  He says he'll write to chessbase to try to get them to publish the story.

About the autism you mention suspicious... the 3 s words that come to my mind are spurious, specious and stereotyped.  Obviously counting on the mystic of autism to try and make it more believable... when in reality neither chess nor autism works that way.  Beat an engine in opposite castle najdorf because you're autistic?  lol no.

Correct. Its a huge stereotype (to make the story believable???). Anyway, I don't know enough about chess to say anything about the game. Its the stroy and the approach to the responces recieved that was the big red flag for me.

Plus...why are all these talented, booked up players never online? Why are they always discovered from the outside, then brought in by someone who has been a member for a day? Get him online and let him play a master or GM.

I would have my Nephew online right now playing.

The point is, if he would really beat a titled GM online, then I simply accuse him of using engines.

didn't consider that..lol

ofc course and there are people in india that havent eaten for like a year. There is also the tooth fairy and the mighty warrior that slayed a bunch of dragons to rescue princess...

I mean all this stories are ofc better then to face the truth that he was just more intelligent, a much better worker and had a much stronger will when it came to persue his dreams. I dont understand if i can accpet that someone is just smarter then me, why is it so impossible to other??

God is the life of most people like a cardhouse that will implode if you just say well, maybe im not that smart and others are smarter

well i ment ofc the autism story not the threadstarter

Stop believing this guys, cmon, there is no way this is true...

kasparov carlsen not can beat H3 !!!!!!!!!!!!!!!!! fake !!!!!!!!! only on low level you can beat it !!!!!!!!!!!!!!

Savants exist. The real life 'Rain Man', mentioned earlier in the thread, was amazing. He could speed read and retained about 95% of everything he read. The documentary I watched on him guessed at 12,000 books. Isn't it possible that this ability could be applied to chess? If he could memorise 95% of all the games, positions, strategies he had seen or read about, would it be so unrealistic for him to beat Houdini? Even just once?

Now if that guy existed, what's to say there are not others with the same ability?

For the record, the OP is probably full of it. If not, we would be watching a documentary about this kid.

The fact remains that people like this exist.