I agree also about the stranded Bb6 - that Thorsson line looks good otherwise, but the bishop's lack of mobility just feels wrong.

It's hard to compare the two lines directly, though. 22.Qg3 sprouts subvariations and can get quite complex, while 22.Nxc6 is mostly a single line leading to a (debatable) late-middlegame/ending advantage.

It's clear that we have a big debate going on between 22.Nxc6 and 22.Qg3 - a crucial choice which may determine the course of the game.

While keeping the existing structure in place, I'd like to find some way of helping out <jepflast> and <GufeldStudent>, who are hosting the two critical lines.

There was a complaint on the main page that the forums weren't really offering digestible answers for people who didn't have much time to go searching. I answered that everything was still in flux, but that we aimed to have summaries ready well before tomorrow's vote deadline.

So I'll make an appeal to anyone with time to spare. Try to help out <GufeldStudent> and <jepflast> in any way you can. Maybe by copying earlier posts from the main page to the relative forum, or maybe by collating the various lines that have emerged.

Between us, we should aim to have user-friendly summaries ready to post early tomorrow - ideally a preliminary report on each move about 4 hours before the deadline, and maybe a final report with 1 hour to go.

I'll help out where I can. It's too early, I think, to attempt an overview yet - but I'll start to collect material for it.

I will try to scour the main forum. It's the other forums where I will need help gathering all the Nxc6 stuff.

I'm now thinking of writing a single compare-contrast piece on our two main moves, and posting it on the main page with about 4 hours to go (with copies to the forums). It'll be based on whatever material I can find in forums, main page, plus my own brain and my own engines.

This is not because I don't trust you two to do excellent jobs. I just feel that in this case - close race, vital decision, haven't decided yet myself - it might be better to bring a single point of view to bear.

Basically, I want to create something that will help to inform the late voter. I think quite a few people are waiting this time to see what we come up with, and I'd like to assist them.

In the meantime, the more (good) stuff that gets re-posted to the forums, the better.

Of course if either or both of you wish to post your own summaries on Thursday, feel free to do so.

One point I'd like you to include in yours is that with 22.Nxc6, the next few moves are absolutely forced, which gives us lots of time to try to find a win after 25. Rf3. We can use good organization to bring down Arno.

On one hand, Arno might be hoping for Nxc6, confident of holding a draw there and worried about a loss in the Qg3 line. Or it could be t'other way round - he'll outplay us in the forests after Qg3. Who knows?

Hiya,

I've been up since 03.00 hrs GMT and it's now 08.30 hrs.

(Had to get up to finish making yoghurt. Don't ask!)

I posted a good line after Qg3 that I came up with when I saw all the indecision. Hope I'm right.

Anyways, I need some shut-eye. Maybe you can push it a little more when you get here. If you think it's worthwhile.

TFN.

And now for about 24 hours of sleep....

Critical move, close vote. Analysis of these two moves has been going on for a long time now, since around our 18th move. It’s easy to get lost in the thickets of repetition, and even the material posted on the forums doesn’t always help to simplify things. Maybe there is no easy answer.

Both moves are playable. If I was annotating this game, whether 22.Nxc6 or 22.Qg3 was played, I’d want to explore the other as a possible alternative.

It’s partly a matter of style. 22.Nxc6 leads down a single, solid line to move 25 (this is 22.Nxc6 bxc6 23.d4 Nd3 24.Qd2 Nb4 25.Rf3, btw, henceforth abbreviated as ‘the Thorsson mainline’) and leads to a small but tenuous advantage for us in later variations. Some people have gone into this very deeply, and it generally stays good for White. In contrast, 22.Qg3 begins branching at once – does Black reply …d5 or not? – and some lines are very complex (although we have RV/Rybka’s assurance that there are no sudden wins or losses lurking in the near future).

In an OTB game, I suspect I’d happily play 22.Qg3 and hope to extract an advantage from the ensuing complications. Or I might spend 30 minutes looking at Qg3, decide it was just too complex, and play Nxc6 instead.

In a real CC game, this would be the time to spend a week getting to the bottom of Qg3. Unfortunately, we’re not allowed to do this. Therefore 22.Nxc6 – with its deeply analysed and fairly straightforward plan – looks like the best way to go.

If I tried to recap all the analysis here, it’d go on for pages, and I’d still be hammering away at the keyboard when the deadline passes. Instead I’ll pick out some examples.

As recently as yesterday, one of the main ideas after 22.Qg3 involved getting a Knight to e6. The main line went like this: 22.Qg3 d5 23.Rf4 Bd6 24.Raf1 f6 25.Qh3 g5 26.R4f2 Qf7 27.Ne6 h5 This isn’t pretty for White, and after some thought the engines agreed. Neither Shredder’s 28.Bc5 nor Thorsson’s idea 28.Re2 lead anywhere for us. In fact, the line was thoroughly busted after analysis by <twinlark> and others, to the point where its main supporters agreed it was going nowhere.

This is the problem with 22.Qg3 – despite the analysis that has been done, there is no single line where we emerge with a definite advantage. There are several optimistic lines where we come out on top, but we just don’t have the time to go into every last corner of them. The line I just cited stands as a warning: a promising idea in which things gradually and subtly go wrong for us.

Of course there are other plans after 22.Qg3. <Monad>, for example, has suggested this: 22. Qg3 d5 23. Rae1 Bd6 24. fxg6 fxg6 (24... Nxg6 25. Rxe8+ Rxe8 26. Qg5 Qe7 27. Qf5 Rf8 28. g3 1.10) 25. Nf5 Nf3+ 26. Qxf3 Qxf5 27. Qxf5 gxf5 28. Rxe8+ Rxe8 29. Rxf5 Re5 30. Rf6 Be7 31. Rf4 Re1+ 32. Rf1 Rxf1+ 33. Kxf1 Kf7

And <boomie> gave some engine-assisted lines involving his idea of 23.Nde2:

22. Qg3 d5 23. Nde2 Rac8
(23...Qd6 24. Nf4 Bg7 25. Nh5 Bh8 26. Rae1 Rac8 27. Re3 h6 28. fxg6 fxg6 29. d4 Nc4 30. Qxd6 Nxd6 31. Rg3 1.01/13) 24. Rae1 Qd6 25. Nf4 Bg7 26. Nh5 gxh5 27. f6 Nf3+ 28. Qxf3 (28. Rxf3 Rxe1+ 29. Qxe1 Bxf6 30. Qf2 Be5 31. d4 Bg7 32. Rxf7 0.43/14) 28...Rxe1 29. Rxe1 Qxf6 30. Nxd5 Qxf3 31. Ne7+ Kh8 32. gxf3 Re8 33. d4 Bf6 34. Bc5 h6 0.59/15

But it is more than possible that lines such as these contain hidden flaws. Looking at it as objectively as possible (I still haven’t voted), I have to say that 22.Qg3 worries me. We might be winning – but we could very easily lose.

Thorsson’s line after 22.Nxc6 is quite different. It has stayed relatively unchanged in its main outlines for several days. All possible early deviations by Black have been dealt with. We know, for instance, that 23…Nc4 (instead of 23…Nd3) just doesn’t work: it is effectively refuted by 24.d5.

Speaking of refutations... One poster last night suggested that the Thorsson line was ‘refuted’ by 25…Bg7 (instead of 25…Nd5, which had been the focus of recent analysis). This is how Thorsson replied: <Let's remind everyone of the analysis of 25...Bg7 in the Nxc6 main line. Bg7 is an engine favourite, which is no doubt why it has cropped up again and again. <22.Nxc6 bxc6 23.d4 Nd3 24.Qd2 Nb4 25.Rf3 Bg7> 26.Raf1 and now: A. 26...Nd5 27.Nxd5 (we could also hold off this as per Dio's line) 27...cxd5 28.Qg5 We analysed both f6 and h6 here to some depth and found White had a good ending in all lines. B. 26...gxf5 27.Rxf5 Re7 28.h3!? Rae8 29.Kh2 Qe6 30.R1f3 with significant advantage. White will play Qf2 or Qf4 and then start to pressure on g- or h-files (although note that 30...f6 is answered by 31.d5 allowing the B to d4. C. 26...Rac8 27.Qd1 threatening Qb3, shows that Black can't play passively.>

Which pretty much refutes the ‘refutation’.

I’ll just deal briefly with some other reasons why 22.Nxc6 might not appeal to voters. Some simply prefer complications – fine, though it also increases the risk of losing. Some – including me – have worried about letting our Bb6 become hemmed in by its own pawns. In practice, though, the Bishop is fine. We can combine kingside pressure with the b2-b4-b5 break to create a passed a-pawn, for example.

Here’s one line given by Dionyseus:

Dionyseus: 22.Nxc6 bxc6 23.d4 Nd3 24.Qd2 Nb4 25.Rf3 Nd5 26.Raf1 Qe7 27.Nxd5 cxd5 28.fxg6 fxg6 29.b4 Bg7 30.b5 axb5 31.Qb4 Rab8 32.Rf7 :

click for larger view

32.Rf7 Qe6 33.Kh1 Bh8 34.h3 and it's really hard to find any good moves for black here. 34...Bg7 35.Ba7! Rbc8 36.Qxb5 :

In summary – there is actually far more and far deeper analysis to be found, if you still haven’t made up your mind. I think I may have actually been the first person to mention 22.Qg3 (about 120 pages back, when I was still suspicious of Nxc6 on broad positional grounds) – but I’ve now decided to reject it. It’s just too uncertain, and contains too many hidden variables.

22.Nxc6, by contrast, looks like the best move to play – at this point, in this game, given the time and resources that we have. So I’ve persuaded myself, at any rate. 22.Nxc6.

Thanks. If I've left out your favorite line, and it's crucial, please remind us why.

I think collectively we did a good job today, on maybe the most critical move in the game.

I put my emailaddress on the bottom of my profile.

Still embarressed here...