Thursday, August 30, 2018

Pawn structures part 2

All of my students are talented but unfortunately I still need to meet the first one, willing to work hard at home. Only a few players in Belgium study at chess besides playing. Meanwhile I got used to the fact that none of my students make a serious analysis of their games. Last year I gave my students some tasks see holidays part 3 but I won't repeat this anymore. That is not because some people are thinking that I am too strict. No I don't care about such comments. I simply don't want to spend anymore efforts at those useless discussions. Besides I already decided anyway to continue teaching for another year so irrespective of the efforts made by the students.

This school-year my son Hugo starts in step 4. That is the final step he can do without me in the club of KMSK. If Hugo will not start to work independently during this step at chess then I will stop. After you achieved step 4, you are mature to enjoy competitive chess for the rest of your career. Steps 5,6 and anything higher are made for the more ambitious students. Of course anybody has ambitions but those higher courses are only interesting when they are combined with a lot of study at home (my student Sterre is likely an exception). I have connected my teaching to Hugo following courses. Once I see that Hugo barely learns something from his courses, then I will stop mine and his classes. I still can give Hugo advice after his games of which he can/ will still learn a lot. That is something I can do very easily, contrarily to the courses at which I often spend several hours of preparation per course.

I have shown at this blog many times examples of how useful it is to study chess. However I can't stress this enough so in this article I show some new positions which popped up in a recent tournament I played. Last year I wrote that it is also useful to replay games outside ones repertoire as several ideas can be used in different openings see pawn structures part 1. In this article I want to elaborate on this topic by showing some connections even inside my repertoire.

First I start at a micro-scale. Within a variation of an opening we discover that the same concept can sometimes appear several times. The first position comes from a game-preparation which I made for the Belgian interclub. In the round against Wetteren, there existed the possibility that the Belgian expert Galeh Khonghaloos would be my opponent. Therefore I studied his games from the database. One of those games is shown below.
[Event "BEL-chT 0910"] [Site "Belgium"] [Date "2010.02.28"] [Round "9"] [White "Decoster, Frederic"] [Black "Khonghaloos, Galeh"] [Result "0-1"] [ECO "C18"] [WhiteElo "2280"] [BlackElo "2217"] [PlyCount "90"] [EventDate "2009.09.27"] [Eventtype "team-tourn"] [Eventrounds "11"] [Eventcountry "BEL"] [Sourcetitle "EXT 2011"] [Source "ChessBase"] [Sourcedate "2010.11.26"] [Sourceversion "1"] [Sourceversiondate "2010.11.26"] [Sourcequality "1"] [CurrentPosition "rnbqkbnr/pppppppp/8/8/8/8/PPPPPPPP/RNBQKBNR w KQkq - 0 1"] 1.e4 e6 2.d4 d5 3.Nc3 Bb4 4.e5 c5 5.a3 Bxc3+ 6.bxc3 Ne7 7.Qg4 O-O 8.Bd3 Nbc6 9.Nf3 Ng6 10.Qh5 c4 ( 10...Qc7 11.Be3 c4 12.Ng5 h6 13.Nxf7 cxd3 14.Nxh6+ gxh6 15.Qxg6+ Qg7 $15 ) 11.Bxg6 { (Earlier this year during a preparation of a game for the Belgian interclub I had noticed that Ng5 is winning in this position. I played that round at board 1 so there existed the possibility that Galeh would be my opponent which eventually didn't happen.) } ( 11.Ng5 { (If black's queen is not at c7 then this move is very strong.) } 11...h6 12.Nxf7 Rxf7 ( 12...Qa5 13.Nxh6+ Kh8 14.Bxg6 Qxc3+ 15.Kf1 Rxf2+ 16.Kxf2 Qxd4+ 17.Be3 Qc3 18.Nf7+ Kg8 19.Qh8# ) 13.Bxg6 $18 ) 11...fxg6 12.Qg4 Qa5 13.Bd2 Bd7 14.h4 Rf5 15.h5 gxh5 16.Rxh5 Rxh5 17.Qxh5 Rf8 18.Qh3 Nd8 19.Ng5 h6 20.Nf3 Qa4 21.Kd1 Be8 22.Qg3 Kh7 23.Nh4 g5 24.Nf3 Bg6 25.Ne1 Bf5 26.Qh2 Kg7 27.f3 Nc6 28.g4 Bg6 29.f4 h5 30.f5 exf5 31.Bxg5 fxg4 32.Bf6+ Kh7 33.Qd2 Nd8 34.Ng2 Ne6 35.Ne3 Qb5 36.Rc1 Qa5 37.Qe1 g3 38.Rb1 b6 39.Rb4 Qxa3 40.Kd2 g2 41.Rb5 Nf4 42.Qf2 Qa2 43.Bg5 Qa1 44.Nxg2 Nxg2 45.Qxg2 Rf1 0-1
I got to play somebody else but in the second round of the last Open Gent I still was able to employ the learned concept. The position is almost identical (compare position after move 10 of the previous game with the position after move 11 of the next game) but a small difference of the details causes a chain-reaction of changes.
[Event "Open Gent 2de ronde"] [Site "?"] [Date "2018.??.??"] [Round "?"] [White "Brabo"] [Black "Deveque, G."] [Result "1-0"] [ECO "C18"] [WhiteElo "2310"] [BlackElo "2020"] [PlyCount "71"] [CurrentPosition "rnbqkbnr/pppppppp/8/8/8/8/PPPPPPPP/RNBQKBNR w KQkq - 0 1"] 1.e4 e6 2.d4 d5 3.Nc3 Bb4 4.e5 c5 5.a3 Bxc3+ 6.bxc3 Ne7 7.Qg4 O-O 8.Bd3 Nbc6 9.Qh5 Ng6 10.Nf3 Qa5 { (This move doesn't fit in this system. I already noticed during the preparation of the game that my opponent knows very little theory.) } 11.Bd2!? { (More challenging is 0-0.) } ( 11.O-O Nce7 12.Bg5!? Qc7 13.dxc5!? b6! 14.Bxe7 Qxe7 15.c6 Qc5 16.Nd4!? Qxc3 17.Bxg6 fxg6 18.Qg4 Re8 19.a4!? $14 ) 11...c4 12.Ng5 { (I knew Pg5 is strong after 10....c4 but that is also here the case. However there are much bigger complications now.) } 12...h6 13.Nxf7 cxd3? { (In the game I mainly looked at Nf4 but I assume my opponent never considered it as he played very fast. Stockfish recommends an interesting novelty.) } ( 13...Ngxe5!? 14.Nxe5 cxd3!? { (Or first taking at e5 and next at d3.) } 15.cxd3 Nxe5 16.dxe5 Bd7 $14 ) ( 13...Nf4!? 14.Nxh6+ gxh6 15.Qg4+ Kh8 16.Bf1 Ne7 17.g3! Nfg6 18.Qh5 Rf7 19.Qxh6+ $14 ) 14.Nxh6+ gxh6 15.Qxg6+ Kh8 16.Qxh6+ Kg8 17.Qg6+ Kh8 18.h4 Qc7 19.Bg5?? { (Played after thinking for more than 20 minutes as I didn't manage to discover the win. It is even worse as after this move I lose all advantage.) } ( 19.Qh6+! Kg8 20.Rh3 Rf7 21.Qg6+ { (The win is not simple which was proven by the Qatari Mohammed El Sayed in 2006, playing the weaker Rxd3. Today he is a grandmaster and at that time was already rated 2486 so definitely no wood-pusher.) } 21...Rg7 22.Qe8+ Kh7 23.Rf3 $18 { (I had calculated this line till this point but couldn't discover black has no good defense against Rf6 followed up by mate at h6. ) } ) 19...Nxe5 20.Bf6+ Rxf6 21.Qxf6+ Kh7 22.Kd2 { (The planned 0-0 fails spectacularly due to Ng4. Here I realized the game is out of control.) } 22...Nc4+ 23.Kxd3 Nd6 24.f3 Nb5 25.Ke2 b6? { (This is too slow. After Bd7 white has nothing better to force a perpetual.) } 26.Kf2?! { (I evacuate my king out of the danger-zone but the engines consider h5 as slightly more accurate.) } ( 26.h5! Ba6 27.Qg6+ Kh8 28.Qh6+ Qh7 29.Qf6+ Qg7 30.Qxg7+ Kxg7 31.Kf2 $16 ) 26...Ba6? { (Too optimistic as black underestimates the dangers around his own king.) } ( 26...Qg7! 27.Qxg7+ Kxg7 28.a4 Nd6 29.a5 b5 $14 ) 27.Rae1 Qxc3 28.Qf7+ Kh8 29.Qf6+ Kh7 30.Qf7+ Kh8 31.Qh5+ Kg8 32.Qg6+ Kh8 33.Qh6+ Kg8 34.Qxe6+ Kh7 35.Qf7+ Kh8 36.Qh5+ 1-0
Another example is also from the past Open Gent, my game from round 6 against the German expert Taylan Guelsen. This time we look at a macro-scale. There is even a connection between 3 openings. I start with my game against the Dutch FM Frank Wuts, played in Open of Avoine at the year 2000. A Philidor was chosen in which I tested an interesting idea for the first time.
[Event "Open Avoine 6de ronde"] [Site "?"] [Date "2000.??.??"] [Round "?"] [White "Brabo"] [Black "Wuts, T."] [Result "0-1"] [ECO "C41"] [WhiteElo "2245"] [BlackElo "2335"] [PlyCount "44"] [CurrentPosition "rnbqkbnr/pppppppp/8/8/8/8/PPPPPPPP/RNBQKBNR w KQkq - 0 1"] 1.e4 d6 2.d4 Nf6 3.Nc3 Nbd7 4.Nf3 e5 5.Bc4 Be7 6.Ng5 O-O 7.Bxf7+ Rxf7 8.Ne6 Qe8 9.Nxc7 Qd8 10.Nxa8 b6 { (I never played this line before even not in a blitz-game. However it was Eddy Verledens, a strong correspondence-chess-player, whom showed me this idea. I thought it was interesting so decided to give it a try. An important alternative of b6 is b5 which has been played by Miguel Najdorf. Below I tried to summarize the many lines but things are too complex to get a final view. My conclusion is that b5 leads to an interesting battle with chances for both sides.) } ( 10...b5 11.dxe5 ( 11.Nd5 Nxd5 12.dxe5 ( 12.exd5 Bb7 13.dxe5 Nxe5 14.Nb6 axb6 15.f4 Bh4+ 16.g3 Nc4 17.O-O $11 ) 12...Nxe5 13.Qxd5 Qd7 ( 13...Bd7 14.O-O Bc6 15.Qe6 Bxa8 16.Re1 Bh4 17.Re2 $14 ) 14.a4 Bb7 15.Qxb5 Qg4 16.Be3 Qxg2 17.O-O-O Bf8 18.Rhe1 Bxa8 $13 ) ( 11.f3 b4 ( 11...Bb7 12.Nxb5 ( 12.dxe5 dxe5 13.Nc7 Qxc7 14.Nxb5 Qa5+ 15.Nc3 Bb4 16.Bd2 Qb6 17.Qe2 $11 ) 12...Bxa8 13.d5 Nh5 14.Nc3 Bb7 15.O-O Ba6 16.Rf2 Nf4 17.Rb1 $44 ) 12.dxe5 ( 12.Nb5 Ne8 13.dxe5 ( 13.Nxa7 Bb7 14.d5 Qxa8 15.Be3 Bh4+ 16.g3 Bd8 17.a3 Bb6 18.Bxb6 Nxb6 19.axb4 Bxd5 20.Nb5 Qc6 21.Nc3 Bc4 $44 ) 13...Nxe5 14.f4 Nc6 15.a3 Ba6 16.c4 bxc3 17.Nxc3 Qxa8 $44 ) 12...Nxe5 13.Nb5 a6 14.Nbc7 Bb7 15.O-O d5 16.Be3 Bd6 17.Bb6 $44 ) ( 11.Qd3 b4 12.Nb5 a6 13.Nbc7 { (Nac7) } 13...Bb7 14.f3 exd4 15.Nxa6 Nc5 16.Nxc5 dxc5 17.O-O { (e5) } 17...Qxa8 $44 ) 11...dxe5 ( 11...Nxe5 12.Nxb5 ( 12.Bf4 Ng6 13.Bg3 Bb7 14.f3 Bxa8 15.Qd4 d5 16.e5 Nh5 17.O-O-O $44 ) ( 12.O-O Bb7 13.Nxb5 { (Rd1) } ( 13.f4 Ned7 14.Nxb5 Bxa8 15.Be3 Qb8 16.c4 a6 17.Nc3 $44 ) 13...Bxa8 14.f3 Qb6+ { (d5) } 15.Nd4 d5 16.exd5 Nxd5 17.c3 $13 ) 12...Qa5+ 13.Nc3 Nxe4 14.Qd5 Nc5 15.Qd2 Bb7 16.O-O Bxa8 17.Rd1 $44 ) 12.Nd5 ( 12.Nxb5 Qa5+ 13.Nc3 Nxe4 14.Qd5 { (Bd2) } 14...Bb4 15.Qxa5 Bxa5 16.f3 Nxc3 17.Bd2 Bb7 18.Bxc3 Bxc3+ 19.bxc3 Bxa8 $44 ) ( 12.Bg5 Bb7 13.Bxf6 Bxf6 14.Nxb5 Qa5+ 15.Nc3 Nc5 16.f3 Bxa8 17.Rb1 $13 ) 12...Bd6 ( 12...Nxd5 13.Qxd5 Nf6 ( 13...Ba6 14.Be3 Qc8 15.O-O-O Nf6 16.Qxe5 Ng4 17.Qg3 $16 ) 14.Qxd8+ Bxd8 15.Be3 Bb7 { (Be3) } 16.O-O-O Ba5 17.b4 Bxb4 18.Rd8+ Rf8 19.Rxf8+ Bxf8 20.Nc7 Nxe4 21.Nxb5 $16 ) 13.Bg5 Bb7 14.Qd2 ( 14.Nxf6+ Nxf6 { (Qe1) } 15.f3 Bxa8 16.O-O a6 { (Rd7) } 17.Qe2 $13 ) ( 14.O-O Bxa8 { (Qc4) } 15.Qd3 Bxd5 { (h6) } 16.exd5 e4 17.Qxe4 $44 ) 14...Qf8 { (0-0-0) } 15.Nxf6+ Nxf6 16.f3 Bxa8 17.O-O-O Rd7 18.Qd3 $13 ) 11.Be3 Ba6 12.f3 $146 { (There was a lot of noise during the game by spectators which let me burn a ridiculous amount of time of my clock. The played move is an interesting novelty. In the past has been played dxe5 leading again to a complex middle-game.) } ( 12.dxe5!? dxe5 ( 12...Nxe5 13.Nxb6 axb6 14.h3 Nc4 15.Qd4 Nxe3 { (Nd7) } 16.fxe3 { (Qxe3) } 16...Nd7 17.O-O-O $13 ) 13.Nxb6 axb6 14.a3 Qa8 15.f3 Bc4 16.Qd2 Bb4 17.Rb1 { (Qxd5) } 17...Nxe4 18.fxe4 $13 ) 12...Qxa8 13.Qd2? { (I have to admit that I totally underestimated black's next move. Stronger was d5 and black has surely sufficient compensation.) } 13...d5! 14.dxe5 Nxe5 15.Nxd5 Nxd5 16.Qxd5 Bh4+! 17.Bf2?! { (Kd1 was likely stronger but also after that move white has issues.) } ( 17.Kd1 Qxd5+ 18.exd5 Nc4! 19.Bc1 Rd7 20.c3 Rxd5+ 21.Kc2 Re5! $40 ) 17...Nxf3+! 18.gxf3 Qxd5 19.exd5 Re7+ 20.Kd2 Bxf2 21.Kc3!? { (Rad1 is likely stronger but again black is much better after e.g. Re2+. ) } 21...Re3+ 22.Kb4 Rxf3 { (I resigned as I considered my position hopeless especially as I was very low on time. My opponent however was surprised as he told me that his endgame isn't very good and the win still didn't look easy. Anyway the win is straight-forward by first locking out the rooks by Be2 and next to conquer the d-pawn. ) } 0-1
I still remembered this debacle 18 years later. The same idea can be used in the Spanish Breyer-variation but obviously I wasn't very excited about it. The computer also demonstrates that black has good counter-play despite the small material defecit.
[Event "Analysevariant 1"] [Site "?"] [Date "2018.??.??"] [Round "?"] [White "Brabo"] [Black "Guelsen, T."] [Result "*"] [ECO "C95"] [WhiteElo "2310"] [BlackElo "2170"] [PlyCount "32"] [CurrentPosition "rnbqkbnr/pppppppp/8/8/8/8/PPPPPPPP/RNBQKBNR w KQkq - 0 1"] 1.e4 e5 2.Nf3 Nc6 3.Bb5 a6 { (My opponent likes to switch a lot in the opening confirm the database. I have not found any older games with this move from him.) } 4.Ba4 Nf6 5.O-O Be7 6.Re1 b5 7.Bb3 d6 8.c3 Nb8 { (This is played in less than 1 out of 1000 games. Personally I think black was mixing up the theory which wouldn't be a surprise as he likes to change openings a lot.) } 9.d4 Nbd7 10.Ng5 { (During the game I used more than 20 minutes at this move. I guess my opponent never considered any sacrifices at f7 as suddenly I saw him becoming nervous when he discovered what I was trying to calculate. In the end I played slightly disappointed the normal developing move Nbd2 as I couldn't find any advantage with the sharper moves.) } 10...O-O 11.Bxf7+ Rxf7 12.Ne6 Qe8 13.Nxc7 Qd8 14.Nxa8 Bb7 { (White has 2 pawns + rook for the 2 pieces. That looks interesting but I still remembered my defeat against the Dutch FM Frank Wuts of 18 years ago in Avoine. Black has excellent counter-play which also the engines confirm.) } 15.d5 Bxa8 16.f3 Nb6 $11 { [%eval 0,34] } *
There is a second link with a position from my game against Stefan Beukema played in the Belgian interclub at the year 2014. At that time I missed the spectucal winning idea Bxf7 see my article achilles. The opening of that game was a Spanish Chigorin-variation.
[Event "Analysevariant 2"] [Site "?"] [Date "2018.??.??"] [Round "?"] [White "Brabo"] [Black "Guelsen, T."] [Result "*"] [ECO "C95"] [WhiteElo "2310"] [BlackElo "2170"] [SetUp "1"] [FEN "r1bqk2r/2pnbppp/p2p1n2/1p2p3/3PP3/1BP2N2/PP3PPP/RNBQR1K1 w kq - 0 10"] [PlyCount "12"] [CurrentPosition "r1bqk2r/2pnbppp/p2p1n2/1p2p3/3PP3/1BP2N2/PP3PPP/RNBQR1K1 w kq - 0 10"] 10.Bxf7+ { (Earlier I wrote that I have looked at Ng5 during the game but most of my thinking was spent at Bxf7. Finally the quiet a4 is here also interesting but that is outside the scope of this article.) } 10...Kxf7 11.Qb3+ ( 11.Ng5+ Kg6 12.f4 exf4 13.Ne6 Qg8 14.Nxf4+ Kf7 15.e5 dxe5 16.dxe5 Nxe5 $11 { [%eval -21,35] } 17.Rxe5 Bg4 18.Qb3+ Kf8 19.Be3 Qxb3 20.axb3 Rd8 21.Na3 Kf7 22.Bc5 $11 { [%eval -18,35] (This line recommended by the engines popped up recently in a game between amateurs see Clarke - Maroroa played in the 4NCL interclub of 2016-2017. Black won in 38 moves.) } ) 11...Ke8 ( 11...Kg6?? 12.Nh4+ { (I remembered this idea from the analysis I made upon my game against Stefan Beukema played in the interclub of 2014.) } 12...Kh5 13.Qf7+ Kxh4 14.Qxg7 h6 15.Re3 Nf8 16.Qxh8 N8h7 17.Qg7 $18 { [%eval 2409,33] } ) 12.Ng5 Nf8 { (I discovered this move while calculating at the board. I didn't see a good continuation for white so abandoned the concept of Bxf7.) } 13.dxe5 Ng4 { (The only move. Despite I can't find any advantage in this line, I do think Bxf7 deserves a practical test. Unfortunately this will probably not happen very soon as this is a very rare line.) } 14.Qf7+ Kd7 15.f4 Ng6 $11 { [%eval 0,42] } *
This time Bxf7 wasn't winning but surely it would've been a good practical choice. So I also want to warn the reader that ideas/ concepts/ plans can't be copied blindly. Analyzing or preparing our games let us learn a lot of useful things which we often can reuse later. Yes some of these things we can find during a game without any foreknowledge but for sure this will cause a loss of time. In other cases we won't discover some hidden possibilities and won't take advantage of some offered chances.


Friday, August 17, 2018

Tactics part 3

On my blog several articles were published about how much efforts I make to allow my son to play chess. This means little free time remains for myself to play standard games. It is an investment in the future as probably within 2-3 years things will improve. Besides even today I already receive some dividends. Hugo starts to understand more and more about chess so our conversations become more interesting. He shows genuine interest in my games and even gives comments or shares ideas about certain positions.

Last he disagreed about my chosen strategy in a critical position which popped up in my game of round 2 from Open Gent. He considered me a coward by not playing the winning move as I wasn't able to calculate everything correctly to the end. His reasoning is that a couple of moves down the line things would become clear. This is not silly as if we look to the variation below which I calculated in the game then the win is very easy to detect in the final position.
[Event "Open Gent 2de ronde"] [Site "?"] [Date "2018.??.??"] [Round "?"] [White "Brabo"] [Black "Deveque, G."] [Result "*"] [ECO "C18"] [WhiteElo "2310"] [BlackElo "2020"] [SetUp "1"] [FEN "r1b2r1k/ppq5/2n1p1Q1/3pP3/3P3P/P1Pp4/2PB1PP1/R3K2R w KQ - 0 19"] [PlyCount "9"] [EventDate "2018.??.??"] [CurrentPosition "r1b2r1k/ppq5/2n1p1Q1/3pP3/3P3P/P1Pp4/2PB1PP1/R3K2R w KQ - 0 19"] 19.Qh6+! { (I chose after almost 20 minutes thinking for the much weaker Bg5.) } 19...Kg8 20.Rh3 Rf7 21.Qg6+ Rg7 22.Qe8+ Kh7 23.Rf3 $18 { (Till here I had calculated the line at the board. Now it is not difficult anymore to see that black has no good answer against Rf6 followed up with mate on h6. Therefore the engine sacrifices desperately a piece at e5 to prolong the game.) } *
Hugo definitely gets support from other players with this view see comments from some strong players on my article tactics part 1. They also believe you shouldn't be able to calculate everything to play a certain move. Some calculations and common sense should be sufficient to decide. However I am not convinced as this sounds more like I knew it. Hindsight it is always easier to tell that the move is better. Besides in above example I can prove that playing a winning move doesn't guarantee finding the win. It is a pure coincidence but I found a game in the mega-database with exactly the same position. In that game the Qatari Mohammed Al-Sayed played with white the strongest move but still deviated 2 moves later from the mainline.
[Event "Bayern-chI Bank Hofmann 10th"] [Site "Bad Wiessee"] [Date "2006.11.06"] [Round "3"] [White "Al Sayed, Mohammed"] [Black "Zill, Christoph"] [Result "1-0"] [ECO "C18"] [WhiteElo "2486"] [BlackElo "2270"] [PlyCount "59"] [EventDate "2006.11.04"] [Eventtype "swiss"] [Eventrounds "9"] [Eventcountry "GER"] [Sourcetitle "EXT 2007"] [Source "ChessBase"] [Sourcedate "2006.11.23"] [Sourceversion "1"] [Sourceversiondate "2006.11.23"] [Sourcequality "1"] [CurrentPosition "rnbqkbnr/pppppppp/8/8/8/8/PPPPPPPP/RNBQKBNR w KQkq - 0 1"] 1.e4 e6 2.d4 d5 3.Nc3 Bb4 4.e5 c5 5.a3 Bxc3+ 6.bxc3 Ne7 7.Qg4 O-O 8.Bd3 Nbc6 9.Nf3 Ng6 10.Ng5 Qa5 11.Bd2 c4 12.Qh5 h6 13.Nxf7 cxd3 14.Nxh6+ gxh6 15.Qxg6+ Kh8 16.Qxh6+ Kg8 17.Qg6+ Kh8 18.h4 Qc7 { (The same critical position as in my game despite the different move-order.) } 19.Qh6+ { (Contrary to my game, white does play the best move.) } 19...Kg8 20.Rh3 Rf7 21.Rxd3 { (Stronger is Qg6+ but also after Rxd3 white should win in the long run.) } 21...Qe7 { (After this move black loses quickly. Much more stubborn is Rg7.) } 22.Rg3+ Rg7 23.Bg5 Qf8 24.Rd1 Qf5 25.Rdd3 Qg6 26.Qxg6 Rxg6 27.h5 Rg7 28.Bf6 Rxg3 29.Rxg3+ Kh7 30.Rg7+ 1-0
At that time he was rated 2486 elo. Today he is a grandmaster so not a fish. I mean if he can't find the optimal line after playing the strongest move then it is very harsh to condemn my chosen continuation as cowardly. I don't think my hesitation was misplaced. Besides nobody manages to calculate everything to the end. I am in good company as a couple of weeks ago Magnus also expressed no regrets when he wasn't able to complete his masterpiece. After the game he stated:  "The position screams for 20.Bg5 but if you don't see mate...".
[Event "51st Biel GM 2018"] [Site "Biel SUI"] [Date "2018.07.24"] [Round "3.1"] [White "Carlsen, Magnus"] [Black "Svidler, Peter"] [Result "1/2-1/2"] [ECO "B94"] [WhiteElo "2842"] [BlackElo "2753"] [SetUp "1"] [FEN "r4r1b/1p1q1p1k/p2pb2p/4nP2/4P2B/2N4P/PPP2Q2/2KR2R1 w - - 0 20"] [PlyCount "14"] [EventDate "2018.07.22"] [CurrentPosition "r4r1b/1p1q1p1k/p2pb2p/4nP2/4P2B/2N4P/PPP2Q2/2KR2R1 w - - 0 20"] 20.Bg3 { (After the game Carlsen stated: 'The position screams for Bg5 but I didn't see the mate'. So he also thinks that you should not play something which is impossible to calculate properly.) } ( 20.Bg5 hxg5 21.Rxg5 Ng6 22.Rdg1 Bg7?? { (Both players didn't know how to continue but the engine demonstrates immediately the win. Anyway Carlsen's intuition wasn't wrong as black can still draw with Rac8.) } ( 22...Rac8! { (The Dutch expert Han Schut miss this saving move in his article at schaaksite about this game.) } 23.Nd5 Bxd5 24.fxg6+ fxg6 25.Qh4+ Kg8 26.Rxg6+ Kf7 27.exd5 Rxc2+ 28.Kd1 Rxb2 { (The engine tells us 0.00 in all lines. There is nothing to say more of course.) } ) 23.Rh5+ Kg8 24.f6 Qd8 25.fxg7 Kxg7 26.Qd4+ f6 27.Rhg5 Bf7 28.h4 $18 ) 20...Rac8 21.Bf4 Qe7 22.fxe6 fxe6 23.Qg3 Rg8 24.Qf2 Rgf8 25.Qg3 Rg8 26.Qf2 Rgf8 1/2-1/2

Today I strongly believe that a decision should be based upon analyzing concrete knowledge. So I am an adept of calculating countless lines during the game which is hard work. Intuition is nice but competitive chess doesn't allow much space to gamble. The best pilots are ashore which is today always the case as they can access the strongest engines to detect in a couple of seconds any blunder.

Of course some exceptions exist when a gamble can be interesting. I am thinking about bad positions without any hope. Sometimes a risky move can improve the odds. Also a half point can be insufficient in some situations. Solid play can lead to lower winning chances compared to chaotic moves. Chess is not a casino-game so gambling should be restricted to that one special occasion.