| Jul-03-20 | | jith1207: 48...Qd8 would've saved the game for black, making it a Q vs R+B, rather than the Queen trying to promote a pawn against those combined forces. Black seems to have overestimated the chances at that point and missed a good opportunity to get to a favorably drawn game. Love the numerous snares Deisy set up on the way to grab all the pawns, by the end of the game that queen must have felt like a lone warrior, except that there was no fight left in her. |
|
| Jul-03-20 | | Brenin: "Up-a-daisy: an exclamation addressed to a child that has fallen, or when raising it in the arms" (Concise Oxford English Dictionary). Chambers's 20th Century Dictionary has "up(s)-a-daisy". In fact I have never heard anyone say "up-a-daisy", it's always "ups-a-daisy" where I live, or "oops-a-daisy" after a fall or in Yorkshire. Whatever, well played Deysi! |
|
Jul-03-20
 | | OhioChessFan: Oops-a-daisy in these parts. |
|
Jul-03-20
| | OhioChessFan: Or occasionally, upsy-daisy. |
|
| Jul-03-20 | | Honey Blend: At first I thought 27. ... ♖fd3 was the mistake, but turns out Black's got some crazy stuff going on if she tried to stop White from queening by 27. ... ♖e8 28. ♘d6 ♕d7 29. ♘xe8
click for larger viewIf 29. ... ♖xb3 30. ♘f6+ gxf6 31. ♗xf6
If 29. ... ♕xe8 30. ♖ad1 ♕xe7 31. gxf3
If 29. ... ♕xe7 30. ♖fe1 |
|
| Jul-03-20 | | jith1207: Whoopsy-Daisy, and the modern day whoops or oops derived from there. Yes, we don't hear it nowadays because we have evolved versions more commonly used and no longer use the original variant. |
|
| Jul-03-20 | | Ironmanth: Great game! Thanks, chessgames. All stay safe, please! |
|
Jul-03-20
 | | AylerKupp: <Saving the game for Black> (part 1 of 3) <<jith1207> 48...Qd8 would've saved the game for black, making it a Q vs. R+B> I don't know that Black's game needs saving at this point. Stockfish 11 at d=60 (it went very deep, very fast, in les than 25 mins, maybe because a lot of the moves were forced and/or the positions were simplified) evaluates 48...Qd8 and the move actually played, 48...Kh8, as equal, [0.00]. Losing the knight does not seem to tilt the game in Black's favor, and the two lines transpose into each other after 48...Qd8 49.Rxg7+ Kh8 50.bxc4 Qxe8 51.Rb7 and 48...Kh8 49.Rxg7 Qe6 50.bxc4 Qxe8 51.Rb7
click for larger viewStockfish 11's analysis continued for both lines: 51...Qa4 52.c5 Qc2 53.Rb6 Qxc5 54.Rxa6 Kh7 55.Ra8 Kg6 56.a6 b3 57.Rb8 Qa5 58.Rxb3 Qxa6 (Now, with R+B+P vs. Q and all pawns on the same side of the board, the chances seem pretty even to me)
click for larger viewStockfish's analysis further continued for both lines: 59.Rb8 Qa7 60.Rd8 Qc5 61.f3 Kf5 62.Rh8 Ke6 63.Rh7 Qb5 64.Ra7 Qd5 65.Ra6+ Kf5 66.Ra7
click for larger viewIt still seems to me that both sides have equal chances. White's king, though passive, seems quite secure and Black's king, though active, much less so. But now for some unknown reason Stockfish decided to diverge in both lines, although it still evaluated both of the resulting positions at [0.00]: 48...Qd8 ... 66...Qd4 67.Rc7 Ke6 68.Rh7 Qc5 69.Rb7 Qd5 70.Rb6+ Kf5 71.Rb4 Kg6 72.Rb6 Qd4 73.Rb5 Qc4 74.Rb6 Kf7 75.Rb7+ Kg8 76.Rb6
click for larger view48...Kh5 ... 66...Kg6 67.Ra6 and here for another unknown reason Stockfish stopped listing moves. But the position hasn't changed much.
click for larger view |
|
Jul-03-20
| | AylerKupp: <Saving the game for Black> (part 2 of 3) Thinking that perhaps Stockfish got "distracted" early due to its aggressive search tree pruning, I tried Komodo 12.3. It is much slower than Stockfish to reach deep search depths but after about 28 mins of calculation at d=29 it also evaluated both 48...Qd8 and 48...Kh8 to be effectively equal, with a slight edge to 48...Qd8, and the lines did not transpose: 1. [-0.66]: 48...Qd8 49.Rxg7+ Kh8 50.Rb7 Qxe8 51.bxc4 Qf8 52.Rb6 Qc5 53.Rxa6 Kg7 54.f3 Kg6 55.Rb6 Qxa5 56.Rb5 Qa4 57.Rb6 b3 58.c5 Qd4 59.Rxb3 Qxc5 (Now we have a similar position as in Stockfish's 48...Qd8 and 48...Kh8 lines, R+B+P vs. Q and all pawns on the same side of the board).
click for larger viewKomodo's analysis continued, still similar to Stockfish's: 60.Rb8 Qc4 61.Rh8 Qd5 62.Rc8 Qb5 63.Rg8+ Kf7 64.Rc8 Kg7 65.Rc7+ Kg6 66.Ra7 Qc4 67.Rb7 Qd5 68.Rb6 and again nothing much has changed:
click for larger view2. [-0.42]: 48... Kh8 49.bxc4 Qc6 50.c5 b3 51.Re3 b2 52.Rb3 Qxe8 53.Rxb2 (A different approach but the result is the same, White still gives up its knight for Black's advanced b-pawn) 53...Qc6 54.Rb8+ Kh7 55.Rb6 Qxc5 56.Rxa6 Kg6 57.f3 Kf5 58.Ra8 Qc6 59.Ra7 Kg6 60.Rc7 Qb5 61.Ra7 Qf5 62.a6 Qe6 63.Bf2 Qd5 64.Kg1 Qd1+ 65.Kh2 Qc2 66.Bg3 Qc4 67.Bf2 Qe6 68.Bg3 Qb6 69.Ra8 Qb5 70.Ra7 Kf5 71.Bf2
click for larger viewKomodo tries a somewhat different approach. After 71.Rxg7 Qxa6 we have a similar situation as in Stockfish's lines and Komodo's 48...Qd8 line, R+B+P vs. Q and all pawns on the same side of the board. After 71...Qc6 the position (and the result?) is slightly different, White still has R+B+P vs. Q but White's a-pawn will require constant watching. So, I restarted the analysis from this position to see if a different result was likely or not. It wasn't. At d=38 (a good search depth for Komodo in order to have reasonably good confidence in its evaluations) and about 9 mins of calculation it evaluated the position at [-0.31] after either 72.Be3, 72.Kg1, or Kg3; effectively equal chances for both sides. I won't bother posting the lines, this sequence of posts is long enough already. |
|
Jul-03-20
| | AylerKupp: <Saving the game for Black> (part 3 of 3) Undaunted, I decided to try Houdini 6. It, like Komodo, does not reach nearly as deep search depths as Stockfish but like Komodo it also evaluated 48...Qd8 and 48...Kh8 as effectively equal at d=29 and about 36 mins of calculation, with 48...Qd8 again having a slight but not significant edge over 48...Kh8 as played:: 1. [-0.18]: 48...Qd8 49.Rxg7+ Kh8 50.Rb7 Qxe8 51.bxc4 Qe4 52.Rxb4 Kg7 53.f3 Qc6 54.Bf4 Qc5 55.Rb6 Qxc4 56.Bg3 Kg6 57.Bf2 Qc2 58.Be1 Qc8 59.Bg3 Kf7 60.Bd6 Qc4 61.Bg3 Kg7 62.Rb7+ Kg6 63.Rb6 Kf5 64.Bf2 Qc1 65.Bg3 Qf1
click for larger viewHoudini tries to take a more active role with its bishop but the results are similar, R+B+P vs. Q with 2 pawns on the q-side. But preserving its Pa6 requires Black's queen to stay on the a6-f1 diagonal or move to c8, limiting its mobility. So I doubt that either White or Black can make progress in that case. 2. [-0.03]: 48...Kh8 49.bxc4 Qxc4 50.Rxg7 Qe2 51.Ra7 (Somewhat different, Houdini goes after Black's a-pawn instead of the advanced passed b-pawn. Seems risky to me, in a race between pawns Black's b-pawn seems to have an edge since it's more easily defended than White's a-pawn.) 51...Qxe8 52.Rxa6 Qb5 53.Ra8+ Kg7 54.a6 b3 55.Ra7+ Kg6 56.Rb7 Qxa6 57.Rxb3 (But the result is no different, R+B+P vs. Q and all pawns on the same side of the board) 57...Qa2 58.Rb7 Qa5 59.Rb8 Qd5 60.Rb6 Qc4 61.f3 Qa4 62.Rd6 Qb5 63.Rd2 Qa5 64.Rd6 Qa7 and the position is not much different than others we've seen before.
click for larger viewSo these 3 engines reinforce the impression that, if Black was going to save the game, whether he plays 48...Qd8 or 48...Kh8 it makes no difference in the final result. In order to lose Black must have gone wrong later. It would be interesting to see when this happened, but I'm very busy at the moment and I don't have time to look for it. Maybe later I will do so. Oh, and I forgot to mention that all 3 engines were supported by 5-piece Syzygy tablebases, perhaps the reason for all 3 engines to reach deeper than normal depths in the relatively small amount of time that I let them analyze. |
|
| Jul-03-20 | | LivBlockade: 
click for larger view
Why not 54...Qxf6? |
|
| Jul-03-20 | | jith1207: Black, it seems based on few resident Stockfish checks, should've played 53...Kg6, 54..Qxa5 followed by 55..b2 that seems to force white to go for Repetition to avoid two Queens on the stage while also capturing White's a pawn and guarding its own while not allowing Black's rook and knight to freely torture its king. Your research is very useful info, When I suggested 48...Qd8, I didn't mean that's the only way or turning point of the game, but it was a good point of time for black to simplify into the pieces I noted, from where it is psychologically easier for OTB play with much simpler intuition than trying to play perfectly computer-like moves with only a queen against three pieces combo that was on the verge of forming a mating net after the capture of g7 and f6 pawns as it turned out in the end, especially for players not of grandmaster strength, far more from being replicators of silicon outputs. |
|
| Jul-03-20 | | jith1207: <LivBlockade>: I had the same question, but resident Stockfish seems to have comfortable winning game taking the Queen, letting the pawn promote to a new queen and still come out on top on R+B+P vs Q ending. |
|
|
|
|
Kibitzing
Annotations by 
sac: 26.exf6