|
< Earlier Kibitzing · PAGE 3 OF 3 ·
Later Kibitzing> |
| Jan-19-24 | | whiteshark: All ♖ endgames are drawn. |
|
| Jan-19-24 | | goodevans: <The great Giri missed a win.> Yes he did. I doubt he'd have missed it if the opportunity had presented itself just after the time control rather than just before it. |
|
| Jan-19-24 | | stone free or die: Missing 36.Qf6 isn't really a miss - you'd have to have engine vision to see it through. 38.Qf6 Rxg6
click for larger view39.Qh8+ (must find) Kb7
40.Rf7+ (must find) Ka6
41.Qc8+ (must find) Kb5
42.Rf5+ (Qf5 works less well) Kxb4
click for larger view43.a3+ (must find)
So, before you move the queen to f6 you have to see the pawn check on a3. Hard. |
|
| Jan-19-24 | | stone free or die: And even then you're far from being done. |
|
| Jan-19-24 | | goodevans: <... you'd have to have engine vision to see it through.> It's a fairly straightforward mating net and, I would imagine, well within Giri's capabilities given sufficient time. Hence my previous comment. |
|
| Jan-19-24 | | whiteshark: tablebase verdict, anyone? |
|
| Jan-19-24 | | jffun1958: I think, that's a dead draw and I'm expecting, they soon will shake hands. |
|
Jan-19-24
 | | PawnSac: it seems like he has 1 trick left with Rf5 |
|
| Jan-19-24 | | stone free or die: <<goodevans> It's a fairly straightforward mating net and, I would imagine, well within Giri's capabilities given sufficient time. Hence my previous comment.> I think we differ on that, I thought the <must find>'s were tough. Plus, if you don't see the mate to the end, Black wins. Giri usually plays conservative vs. speculative in such situations. . |
|
| Jan-19-24 | | stone free or die: How much time did Giri have at that point? (I tuned in late, sorry). |
|
| Jan-19-24 | | whiteshark: <64. Ke4 1/2-1/2> It's already a draw 1 ply be4 |
|
| Jan-19-24 | | stone free or die: <I thought the "must finds" were tough> that is <tough (for me)> I'm willing to reassess upon review...
Here's a little clip of Anish playing 3min showing him doing calculations in real time. https://www.youtube.com/watch?v=_v1... I started a little early to get this in: <"That made me the man I am today">, referring to h2/h7. Here's the nut:
<Black to move after 1.Rd1-d4> < 
click for larger view> Of course, he wasn't exactly certain the tactic worked when he played his move, but this was just a blitz game after all. https://www.reddit.com/r/chess/comm... |
|
| Jan-19-24 | | dehanne: Why win when you can draw? |
|
| Jan-19-24 | | whiteshark: Quality control is a bi1ch when operating at warp speed. |
|
| Jan-20-24 | | stone free or die: OTOH - Giri has missed even simpler mates than those after 38.Qf6 subvariations... https://www.youtube.com/watch?v=Mh2...
Guess Carlsen missed it too, so it all balances out. |
|
| Jan-20-24 | | Olavi: <stone free or die>: <Missing 36.Qf6 isn't really a miss - you'd have to have engine vision to see it through.
38.Qf6 Rxg6>
39.Qh8+ is not a must find, Qxd6 is not worse than the game... |
|
Jan-20-24
 | | keypusher: <Olavi: <stone free or die>: <Missing 36.Qf6 isn't really a miss - you'd have to have engine vision to see it through. 38.Qf6 Rxg6>
39.Qh8+ is not a must find, Qxd6 is not worse than the game...> Seems like a quibble, if you can’t work out the mating line, there’s no reason to prefer 38.Qf6 over what Giri played. |
|
| Jan-20-24 | | stone free or die: <Olavi> perhaps I should clarify what I mean by <"must find">. It's largely shaped from all my tactical training on <ChessTempo> (<CT>). You get a position and have to find the most winning line from a position. You fail if you don't play a winning move (i.e. keeping the eval at >= 1.75). You get a retry if you hit an alt, i.e. a move which isn't the best but maintains a winning advantage. So, when I say "must find" I mean the combo has only one winning move at that point. I often use the term <"sharp"> as an equivalent. Note - that on <CT> your "opponent" isn't constrained to play the best move. Often the suboptimal moves are more challenging to us humans. This is similar to my picking 38...Rxg6, which I thought gives White the most challenging continuation. I do think <kp> is correct in saying Giri's line, while missing the mating attack, allows White to solve the pawn problem the most directly. In fact, it's 2nd best move according to the engine, (38.Qf4, pinning Black's queen, is 3rd best, and anything else loses). |
|
| Jan-20-24 | | goodevans: Whilst <must finds> are quite important, the thing that makes tactics hardest to calculate are the number of <branches> that need to be considered. If you look at the winning line after 36.Qf6 then almost all such branches for Black can be nipped in the bud (e.g. needlessly giving up the Q). It turns out there are <just three branches> to consider and two of these (39...Kc7 40.Rf7+ Kc6 and 39...Kc7 40.Rf7+ Kc6) amount to the same thing, so actually just two. Easy peezy for a Super GM. I'm coming to the conclusion that Giri simply didn't consider 36.Qf6. |
|
Jan-20-24
 | | perfidious: The reason computers are far superior in playing strength lies in their ability to plough through such branches with tremendous speed and pinpoint accuracy, at a rate far beyond the capability of any player living or dead. |
|
| Jan-21-24 | | stone free or die: <goodevans: Whilst <must finds> are quite important, the thing that makes tactics hardest to calculate are the number of <branches> that need to be considered.> Branches are very important - and definitively complicate some tactics. But generally, tactics are usually by definition forced lines and not so bushy. Again, I'm thinking of my <CT> training. I think it would make an interesting discussion. Could I ask for a couple of examples of "bushy" tactics that you think are worth note? If nothing else, it would give me a gauge on your tastes in tactics. (I love tactics - it's my favorite part of the game. Alas, if only I were better at 'em!) |
|
| Jan-21-24 | | goodevans: <sfod> Sunday puzzles here are usually a good source of "bushy" tactics. Today's is not a bad example: M Bartel vs Nepomniachtchi, 2015 |
|
| Jan-21-24 | | stone free or die: Thanks <goodevans>. I have to admit that I normally don't do the <CG> tactics too much. But on your recommendation I'll have a look at that one. |
|
| Jan-21-24 | | stone free or die: Just curious - but do you ever go over to <CT>? |
|
| Jan-21-24 | | goodevans: I have to confess that I'd never even heard of ChessTempo until seeing your posts! |
|
 |
|
|
< Earlier Kibitzing · PAGE 3 OF 3 ·
Later Kibitzing> |
Kibitzing
