I can beat you all on this condition

I see no reason not to take c6. 

12. Qxc6

Right - a pawn is a pawn. Even Houdini can't bluff forever!!

It can mess around with its evaluations as much as it wants to, but when you finally eat all of his pieces, then we'll see who laughs last!!

For now, it is still keeping its cool though. It rather appreciated your move - in fact, it was (again!!) its first choice. You receive for it a healthy +0.49 evaluation. A masochistic streak maybe? Certainly not the Houdini we all knew and loved...

But being two pawns down now, it must consider that it has a hell of a compensation... so better be careful - it's coming right after you!

What I like about Houdini though, is that even with two pawns under, it doesn't lose its cool and doesn't feel like it has to create something immediate. 

As it plays 13...h6 at 17 ply, it would seem that it's quietly lighting a cigarette and lightly sucking from it, then looking at you, its eyes silently asking - "what now?" as it slowly releases the smoke in the general direction of your face...

Psychological play from Houdini - Lasker, watch and learn!

I put my bets on Houdini

It probably thinks it's got good comp because my 2 pawns are split (k-side and q-side), he has a lead in development, more activity, and a target on e4.

14. Qa4. I need to get my queen out before it gets into trouble. This way it can also drop back to c2 to let my DSB get active. let's see how long it takes me to lose. 

Good thinking!

Houdini would have played the same :-)

Its next move (you're giving it a tough game, aren't you?!) is - 14...Rc8.

It gives you +0.58 - and it seems that it matters little where it puts its rook... all three options are its three leading moves, evaluated within a mere four centipawns of the chosen move at 17 ply.  

15.Qc2, to defend b2 and thus free the DSB. 

For once, you only played the second best move. 

Houdini had a very slight preference to 15. h3 here. 15. Rd1 took third place, with the same +0.52 evaluation as your move. 

Houdini here, still two pawns down and with white threatening to develop naturally and keep its material, continues with quiet pawn moves. I would say "tightening the noose" - but frankly, no noose is as yet visible. Gaining space has its merit, though, I guess. 

Houdini answers your 15th move by 15...a5. As it calculated its answer, its outlook on the position became a hair more optimistic, as it adjusts your advantage to +0.49 at 17 ply. 

Finally, I can develop my last minor piece.

16.Bd2. I crossed the halfway to move 30

If you continue like that you'll probably beat Houdini, I don't know in how many moves. 

For now, your move was considered second best, but with exactly the same +0.48 evaluation as the engine's first choice. 

Houdini is answering you now with 16...Ra7

One thing is sure - if you play like Houdini, you're probably not going to lose to Houdini from an advantageous position. Just keep it up and you're all set. 

For me, if you keep playing top Houdini choices for three or four more moves, I will be satisfied that you're perfectly capable of holding your own against this mighty computer. That's very well done!!

I don't understand why houndi is making these moves.

anyway, 17. h3. I need to keep the knight out of g4 so I can play Rfd1 without interference.

This is (again) Houdini's top move, at +0.46. Houdini is answering you now with 17...Rb7. 

I will keep following this wonderful and eventful game up to and including move twenty. At that point either somebody else should take over - or else, you can just keep playing this against Houdini and report to us the further progress of the game, along with the result... :-)

18.b3, to defend the b-pawn

So, <Ajian>, at this point in the game, Houdini would have gone, strangely enough, Bishop back to c1 in order to hold your pawn on b2. This was what it thought was necessary to keep your advantage at the half-a-pawn area. 

It also had a very slight preference to 18. Qd1 over your move, maybe with a view to 19. Na4 in case of a capture on b2. 

When it was considering your move, its evaluation of it went down as low as +0.22 - which is pretty close to equality. However later, on considering its response (18...Rb4), it reassessed the situation, deciding that it was not yet altogether out of the woods (+0.39 at 17 ply). 

I must say that I'm impressed by the way you still manage to keep a slight edge throughout this whole manoeuvering phase. You definitely don't rush it and don't try anything that a human rated around 1800-1900 blitz here would. Astonishing cool on your part, that only finds Houdini's own cool as a match. 

Fascinating as watching two computers playing each other might be (because you certainly play like one, if not actually with one), I still stick to my position of abandoning this no doubt useful experiment at the twenty-move checkpoint. Best luck to you after that!

I, once again, agree, with Mr. Concert Pianist.

Benz, did you listen to his Rhapsody in Blue? :)

Nope, I haven't. Can you post the link?! :)

I think I have found it: https://www.youtube.com/watch?v=3wQJeSDYXIk

That concert was huge fun - my 40th birthday celebration here in Brussels :-)

Now things are getting very awkward. the c-file line-up sucks for me.

19. Qd3, to hang on to e4, and prevent ...Bd4

Your move is Houdini's third-best at 17 ply, scoring +0.32. Qd1 and Kh1 (in this order) scored nine and five (respectively) centipawns better. 

At move nineteen, Houdini saw very quickly that this is a critical position, in which its path is very narrow. It has one good move - and that's it!

Already Houdini's second-best move, 19...Nh5, is evaluated at +0.98!

I'm sure that you would have absolutely no difficulty to beat the engine, 3139-rated as it is, had Houdini played that - and easily so - with practically no effort on your part.

Houdini's better be careful with you! So it now plays 19...Ne5 - Houdini sees this as its only reasonable move.

By the time it has completed its scan of this move at 17 ply, it evaluates your advantage at +0.29. 

As far as I am concerned, I will answer your next move, which will be number twenty - and after that I will leave this thread for others, who are even more gullible than I am :-) 

In the hope and belief that they exist - but my guess is, they don't.