My sincerest apologies to post members; it never occurred to me that the manufacturer of Houdini 3 and the Pro would program the computer to play junk in play mode.
Houdini works well, within its inherent limitations.
I could not expect the author of Houdini to program YOU, or pretty much address every case of the PEBKAC syndrome effectively. He probably thinks that the usage manual should suffice, but alas, it does not seem being the case here...
I didn't get a manual and Houdart won't return my e-mail (well, I did ask him several off topic questions )
I feel your pain, bro. To find the online manual, I actually had to Google "Houdini chess engine manual"!
Man, am I ever exhausted! I'm going to have to take a sabbatical leave now.
P.S. Thanks pfren, I learned a new acronymn! PEBKAC!
For the moment it seems that Houdini 3 and Houdini 3 Pro have been created for analysis to get the very best possible move in any analyzed position, but that it was not considered needed to make the play mode take systematically the first line of Houdini 3's analysis as its move each time. As everybody seems to point out, when you play against Houdini 3 at max level and then a few minutes later analyze the game, you can see it did not play the best moves it itself recommends. That's obvious from the very first game played.
That's strange to me and so I have asked why that is, and how hard it would be to make the maximum playing mode "simply" be the first line of the analysis of each move. Of course some kind of opening has to be included to save analysis time, maybe that's the problem. Maybe it's not as simple as it sounds.
I'm assuming that this is communication to you from a Chess King representative? The information is a bit odd, as it seems to imply that the problem originates with the Houdini engine and not the Chess King GUI. However, this might just be due to the rep's ignorance of the issues, or even willful misdirection (not wanting to 'fess up - let's just keep obfuscating the facts, etc.). I seriously doubt that the problem comes from the engine itself, unless Houdart went to the trouble of making a special Houdini version just for Chess King. (I would be surprised if he did that.)
Anyway, I would be interested in knowing the results of the investigation and how Chess King decides to resolve the problem.
Yes---this is from the Chess King representative, Diego. He also indicated surprise that when he input the Two Knights' Defense, his Houdini also played the blunder---4.Ng5 d5 5.exd5 Nxd5 6.Nxf7 Kxf7 7.Qf3+ Ke8? so the problem does not appear restricted to my computer.