Shankland on His Rise from GM to Top Hundred: Part I

I am American born and raised, there is no chess culture in my family, I played my first tournament at an age when most top talents are already masters, I never had more than one hour of coaching per week, I went through the full American K-12 education system in the traditional timetable, and I completed a four year degree from a highly ranked university. Yet in spite of all this, I have been able to consistently improve my chess, and I have recently reached a new milestone- a spot comfortably within the top hundred players in the world at 2661 FIDE. All throughout my career, many players of all levels have asked me how I am able to keep improving, and this article will serve as an overview of one leg of the journey. While the title and general theme is a ripoff of an excellent book that I thoroughly enjoyed, From GM to Top Ten by GM Judit Polgar, I decided to write the article because every chess player should have fond memories of reaching new levels, and it brings me great joy to share part of my journey. As all aspiring chessplayers know, improvement is highly nonlinear. A player can work very hard on their chess and not improve at all for a long time, and then all of a sudden have a huge growth spurt seemingly out of nowhere and reach much higher levels. Players will get stuck at different places. After getting my GM title, I made it to 2600 without much trouble, first crossing the barrier in September 2012. But, my quest from there to the top 100 took just over two years. It's those two years I wish to write about here. Just as Judit did, I could easily write a book on this timeframe because so much happened, but I'll save that aspiration for a later date and stick with some of the most important factors that lead to my improvement. I started and ended 2013 with exactly the same live rating: 2602. The whole time I had bounced around between 2590 and 2610, and I was getting very frustrated with my inability to reach higher levels. A close examination of all of my games, regardless of the result, showed several weaknesses that I needed to address, and lessons that I needed to learn. While it took most of 2013 to identify these problems and this may be why I did not make any progress, once I was able to effectively address them in 2014, I shot to 2660 like it was nothing. I'd like to share 10 of these lessons here (in no particular order), with some examples from my own play. 1. One must challenge one's own success and to learn as much from victories as losses It's very important to look for mistakes in all of one's games, not just ones that ended unfavorably. The most glaring example from the frustrating 2013 calendar year can be seen here:

[pgn][Event "Dresden ZMDI Open"]
[Site "?"]
[Date "2013.08.18"]
[Round "?"]
[White "Shankland, Sam"]
[Black "Vorobiov, Evgeny"]
[Result "1-0"]
[WhiteElo "2591"]
[BlackElo "2591"]
[SetUp "1"]
[FEN "r2q1rk1/4p1bp/1p1p2P1/p1n5/2P5/2N1Bb2/PP1Q2P1/3RK2R w K - 0 19"]
[PlyCount "27"]

19. Rxh7 {It pains me to award this move a question mark, but it is definitely
a clear example of how critical it is to challenge one's own success. This
shot led to a swashbuckling final round victory over a 2600 GM to clinch first
place in a strong European Open, but had I faced stiffer resistance, things
would have worked out very differently.} (19. gxh7+ {Something like this was
called for. White looks a little better following:} Kh8 20. gxf3 Rxf3 21. Ke2 {
With Rdg1 and Bd4 coming, I would prefer to be white in a very double edged
position.}) 19... Bxd1 20. Bh6 Rf1+ $2 {Throwing away the win, although black
is still not worse. Cold calculation shows holding all the material to be
correct:} (20... Be5 $1 21. Bxf8 Qxf8 22. Rf7 Bh5 (22... Bg4 23. Nd5) 23. Qg5 {
I had gotten this far in my calculation and stopped- it seemed to me the h5
bishop was trapped and I would win the queen as well, with a winning material
advantage. However, one should always end their calculation with a search for
candidate moves, as we can see here:} Nd3+ $1 24. Kf1 Nf4 {Black saves the
bishop, and he will have far too much material for the queen. White is
completely lost.}) 21. Kxf1 Qf8+ 22. Kg1 Bxh6 23. Rxh6 {Now white is only one
piece down and his attack is roaring ahead. It takes very precise play to
maintain the balance, and black was unable to cope.} Bc2 $1 (23... Bg4 24. Rh7
$18) 24. g7 Kxg7 (24... Qxg7 $2 25. Qd5+ {Snares the rook.}) 25. Rh3 {With a
double threat of Qh6+ and Qxc2.} Qf6 26. Nd5 Ne4 $2 {Finally black cracks.} (
26... Qe5 $1 {Cold calculation proves this very suspicious looking move to be
correct. After:} 27. Qh6+ Kf7 28. Rf3+ Kg8 {White has nothing better than a
draw.}) 27. Rh7+ $1 {A very nice tactic.} Kf8 28. Qe3 $18 {Black's queen is
overworked and unable to defend both h6 and h8.} Qe5 29. Qh6+ Ke8 30. Rh8+ Kd7
31. Qh3+ Kc6 32. Rc8+ 1-0[/pgn]

When I first reviewed this game, I brushed off the computer's disdain for my speculation as unimportant. This was a very big mistake. Rxh7 is simply a bad move and the result of poor calculation, but I was so blinded by the success of the tournament overall that I was unwilling to accept silicon criticism. It did not take long for my coach to rain on my parade by listing several players who he believed would have disposed of me in 25 moves for my misplaced bravado. Looking at the position now, I entirely agree- I won the game because I got lucky, and it would not have been that hard for a strong black player to defend and win. While calculation can always be improved and is something I spend a great deal of time on, the big lesson I learned from this game is that I was not doing myself any favors by not looking critically at my victories. 2. Finding candidate moves and atypical or counterintuitive ideas is very difficult When we first learn to play chess, every legal move is a candidate move. As we improve, the list of moves rapidly shrinks as our understanding prevents us from looking at moves we can dismiss with a glance as poor. In my case, this process went too far and I had started to miss candidate moves I should have been noticing. A particularly painful example can be seen here:

[pgn][Event "Cuitat de Barcelona"]
[Site "?"]
[Date "2013.11.11"]
[Round "?"]
[White "Narciso, Marc"]
[Black "Shankland, Sam"]
[Result "1-0"]
[WhiteElo "2509"]
[BlackElo "2509"]
[SetUp "1"]
[FEN "8/3k4/3p2p1/4p1Pp/2P4P/2P1KB2/8/1b6 b - - 0 43"]
[PlyCount "16"]

{In the final round of a GM round robin, I needed a win to tie for first.
Although I managed to get a modestly better pawn structure, I was very
frustrated that I was never able to create winning chances, which might
explain the following move:} 43... Ke6 $4 {Black has some weak pawns on the
light squares, he cannot zugwang the king away from e3 because white has
unlimited bishop moves, and a protected passed pawn should be decisive in a
pawn ending. This led me to not consider white's response, which was
immediately decisive.} (43... Bf5 {Pretty much any other move should lead to
an equal position.}) 44. Be4 $1 $18 {All of a sudden black is just lost- after
Bxe4 Kxe4 he finds himself on the wrong side of a reciprocal zugzwang.} Ba2 (
44... Bxe4 45. Kxe4 {And white to move would lose, but as is black's king has
to give ground. White will then penetrate with Kd5 and break up the pawn mass
with c5.} Ke7 46. Kd5 Kd7 47. c5 dxc5 48. Kxe5) 45. c5 $1 {The most precise.}
dxc5 46. Bxg6 Bd5 47. Bxh5 Kf5 48. Be8 Bb7 49. Bd7+ Kg6 50. Bg4 Bc6 51. Be6 1-0[/pgn]

If there's one good thing about a game like this, it's that the pain it causes is so great that we are forced to address the problem immediately to avoid it in the future. My coach had me do a lot of exercises, and even just a few months later, the difference was staggering. After two good results I immediately jumped to 2630. My favorite example can be seen here:

[pgn][Event "Fagernes International"]
[Site "?"]
[Date "2014.04.18"]
[Round "?"]
[White "Shankland, Sam"]
[Black "Turov, Maxim"]
[Result "1-0"]
[ECO "B12"]
[WhiteElo "2616"]
[BlackElo "2593"]
[SetUp "1"]
[FEN "4r2r/1k4pp/2pb4/3pPb2/B2B4/8/PP2K1PP/R6R w - - 0 20"]
[PlyCount "21"]
[EventDate "2014.??.??"]
[Source "ChessBase"]
[SourceDate "2006.11.23"]

{At the moment, black seems to have an absolutely fine position. He will soon
recollect his lost pawn and his pieces have found good squares with a strong
central presence. It took a thorough quest to look for less than obvious ideas
to find a way to apply the maximum amount of pressure, which led to a quick
collapse} 20. Rac1 $1 Bd7 21. Rhf1 $1 Bxe5 22. Rf7 $1 {The point. White allows
a bishop to be captured with discovered check (and he won't even threaten Kxd4
due to Re4+)- a completely ludicrous idea on the surface- but once Rf7 enters
the mind as a candidate move, the rest is not too hard to sort out.} Bxd4+ (
22... Bf4+ 23. Kd3 Bxc1 24. Rxd7+ Kc8 25. Ra7 $16 {White will win several
pawns in compensation for the exchange and is much better}) 23. Kd3 {And we
see white's point- it's extremely difficult to deal with the coming threats of
Rxd7+ and/or Bxc6+} Kb6 $1 {The only defensive try} 24. Rxd7 Re3+ $1 25. Kxd4
Re4+ 26. Kd3 Rxa4 27. a3 {At this point the position is about equal, but black
is under concrete pressure and has to find a way to meet white's threats. The
computer finds one way to maintain the balance, but especially right after one
has lost control of the game, it is quite difficult to defend precisely} Rg4 $2
{Now black loses a pawn by force} (27... Re8 $1 {This highly computer-y
defense was the only way to maintain a tenable position. The point is Rd6 can
now be met with Rc4! when the weakness of the e2 square can be felt, and Rxg7
is met with c5! with strong counterplay in the center} 28. Rd6 (28. Rxg7 c5 $1
{Rd4+ and Re2 coming}) 28... Rc4 $1 29. Rxc4 dxc4+ 30. Kxc4 Re2 {And black is
fine}) 28. g3 h5 {Missing white's threat, but already the position was very
difficult to defend} (28... Rg6 29. Rf1 {Black will not be stopping Rff7}) 29.
Rd6 $1 Rc8 30. Rxd5 $1 $16 {And I went on to win} 1-0[/pgn]

3. Be very wary of "automatic" moves There is a great story about Alexander Morozevich. He was playing a game and another GM was watching, shocked that Moro was thinking so long on a move where he clearly only had one choice. The GM kept walking and saw some other games, and then he realized there was actually a second possible move that looked very interesting. He then went back to Moro's board only to see that Moro had played a third move. Playing "automatic" moves without thinking is a recipe for disaster. In a rather subtle case I was quite proud of avoiding temptation:

[pgn][Event "Eastern Class Championship"]
[Site "?"]
[Date "2014.03.15"]
[Round "?"]
[White "Shankland, Sam"]
[Black "Lenderman, Alex"]
[Result "1-0"]
[ECO "C08"]
[WhiteElo "2611"]
[BlackElo "2580"]
[SetUp "1"]
[FEN "3rr1k1/1pq1bppp/p4n2/3p4/3Bn3/3QNN2/PPP2PPP/R3R1K1 w - - 0 16"]
[PlyCount "21"]
[EventDate "2014.??.??"]
[Source "ChessBase"]
[SourceDate "2006.11.23"]

{I was in the bathroom when black made his last move (a6). Not so long ago,
it's very possible I would have come back to the board, played Rad1, hit the
clock, and wrote the move pair down.} 16. Re2 $5 {This is a very interesting
idea. I thought the d5 pawn was well enough protected that I should be putting
my effort more on the efile. In my opinion black is slightly yet steadily
worse if white can achieve this, so he has to try to complicate matters with
something like Nc5 Qf5 g6 Qh3, after which I lost some activity but
compromised the black kingside. Even now I'm not sure that Re2 is a stronger
move than Rad1, but it shows that oftentimes even the most automatic of moves
can have very reasonable alternatives.} Bf8 {Black doesn't do anything to stop
white from grabbing the efile, and he is under pressure. Soon enough I was
able to pick up the d5 pawn with some tactics} 17. Rae1 Re6 18. c3 g6 19. g3
Bg7 (19... Bh6 {I would have preferred this, being more active, though of
course white still has some edge}) 20. Kg2 Qa5 21. a3 Qc7 {Not so impressive
but black has no plan} 22. h4 {Probing for weaknesses and inviting errors by
threatening Ng5. h6 is more or less forced} Ree8 $2 {Just losing a pawn. h6
was needed} (22... h6 23. h5 {I don't have to rush with this but I'm not sure
how else to improve my position} (23. Nh2) 23... Nxh5 24. Nxd5 Rxd5 25. Rxe4 {
I have some pressure but black is almost equal- he definitely should have gone
for this}) (22... Bh6 $2 {This also fails:} 23. Ng5 $1 Nxg5 24. Bxf6 $1 Rxf6
25. Ng4 $1 {I noticed this resource during the game}) 23. Bxf6 $1 Nxf6 24. Nxd5
$1 Rxe2 25. Nxf6+ (25. Nxc7 Rxf2+) 25... Bxf6 26. Qxe2 {And I went on to win
with my extra pawn} 1-0[/pgn]

I was really happy with myself for playing Re2, and after the game I thought that I had fixed this problem for good- obviously a foolish assumption to make after one success. It was then a rude shock when I went back to my old ways the very next round!

[pgn][Event "Eastern Class Championship"]
[Site "?"]
[Date "2014.03.16"]
[Round "?"]
[White "Kamsky, Gata"]
[Black "Shankland, Sam"]
[Result "1-0"]
[ECO "D02"]
[WhiteElo "2709"]
[BlackElo "2611"]
[SetUp "1"]
[FEN "r1b2rk1/pp1Nqppp/2nbp3/2pp4/3P4/2PBP1B1/PP1N1PPP/R2QK2R b KQ - 0 10"]
[PlyCount "42"]
[EventDate "2014.??.??"]
[Source "ChessBase"]
[SourceDate "2006.11.23"]

10... Bxd7 $2 {I played this extremely natural move in about 5 seconds- it
never occurred to me that it might lose the game immediately. If I had
followed my own strategy from the previous round and double checked even very
obvious looking moves, I could have spared myself a crushing defeat} (10...
Qxd7 $1 {Black is a little uncoordinated now but I think he should equalize
without too much trouble. The point is the tactic played in the game does not
work} 11. Bxd6 (11. Nf3 b6) (11. O-O b6 (11... Bxg3 12. hxg3)) 11... Qxd6 12.
dxc5 Qxc5 13. Bxh7+ Kxh7 14. Qh5+ Kg8 15. Ne4 g6 $1 {And here there is no
hanging bishop on d7 after Nxc5, and white is reduced to a perpetual check.})
11. Bxd6 Qxd6 12. dxc5 {Here I realized what was coming, but I chose to take
my chances in a bad but complicated position instead of just playing without a
pawn.} Qxc5 (12... Qe7 13. b4 {Pawn for more or less nothing}) 13. Bxh7+ Kxh7
14. Qh5+ Kg8 15. Ne4 Qc4 16. Ng5 Rfd8 17. Qxf7+ Kh8 18. Qh5+ Kg8 19. Rd1 $1 e5
20. Qf7+ $1 Kh8 21. e4 $1 Ne7 22. Qxe7 Bb5 23. Rd2 Qxa2 24. Qf7 Qb1+ 25. Rd1
Qxb2 26. Qh5+ Kg8 27. Qh7+ Kf8 28. Qh8+ Ke7 29. Qxg7+ Kd6 30. Rxd5+ Kc6 31.
Qf6+ 1-0[/pgn]

While I've made great progress here, there is still work to be done... 4. One should always work to identify shortcomings in general chess understanding This is a very tough one because everyone has different weaknesses. It took me some time to realize one of mine, but when I finally did, the results really showed. I found that I was often underestimating the bishop pair as a long term asset. I would look at roughly symmetrical positions with no real pawn weaknesses or open lines and one side possessing the bishop pair, and I'd dismiss them as equal- I even have evidence of this in some of my opening notes from 2013. In the following game, I was amazed by how devastating the bishop pair became in a position where they initially lacked sensible targets or good diagonals:

[pgn][Event "Bay Area International 2014"]
[Site "Santa Clara USA"]
[Date "2014.01.05"]
[White "Shankland, Sam"]
[Black "Macieja, Bartlomiej"]
[Result "1-0"]
[ECO "B12"]
[WhiteElo "2602"]
[BlackElo "2583"]
[PlyCount "115"]
[EventDate "2014.01.02"]
[SourceDate "2012.05.23"]

1. e4 c6 2. d4 d5 3. e5 Bf5 4. Nf3 e6 5. Be2 c5 6. Be3 cxd4 7. Nxd4 Ne7 8. Nd2
Nbc6 9. N2f3 Be4 10. O-O a6 11. Rc1 Qd7 12. Qd2 Ng6 13. Ng5 Ngxe5 14. Nxe4 dxe4
15. Rfd1 Rc8 16. c3 Nxd4 17. Qxd4 Qxd4 18. cxd4 Nc6 19. d5 exd5 20. Rxd5 Rd8
21. Rxd8+ Nxd8 22. Rc4 Bd6 23. Rxe4+ Kd7 24. Bg4+ Kc7 25. Bd4 Rg8 26. g3 Nc6
27. Bc3 f6 28. Bf5 h6 29. f4 Bc5+ 30. Kf1 Rd8 31. b4 Bd4 32. Be1 Ba7 33. a4 b5
34. a5 Rd1 35. Re6 Rc1 36. Kg2 Bd4 37. Kh3 Bc3 38. Bf2 Nd4 39. Re7+ Kb8 40.
Re8+ Kb7 41. Bc8+ Kc6 42. Bxa6 Kd5 43. Rd8+ Ke4 44. Bb7+ Kd3 45. a6 Ra1 46. Rd7
Ra2 47. Bc6 Rxf2 48. Bxb5+ Ke3 49. Re7+ Kf3 50. a7 Ra2 51. Bc4 Ra3 52. Bd5+ Kf2
53. Rxg7 Bxb4 54. a8=Q Rxa8 55. Bxa8 f5 56. Rf7 Ba5 57. Kh4 Kg1 58. g4 1-0[/pgn]

This would not be the only time last year that a newfound appreciation for the bishop pair would earn me a victory. Twice more I won big games, each one against a stronger opponent than the last, and in all of the three games, the originally quiet bishop pair at some moment showed it’s explosive potential and lead to immediate material gains.

[pgn][Event "ch-USA 2014"]
[Site "Saint Louis USA"]
[Date "2014.05.17"]
[White "Shankland, Sam"]
[Black "Akobian, Varuzhan"]
[Result "1-0"]
[ECO "B12"]
[WhiteElo "2634"]
[BlackElo "2643"]
[PlyCount "87"]
[EventDate "2014.05.08"]
[SourceDate "2012.05.23"]

1. e4 c6 2. d4 d5 3. e5 Bf5 4. Nf3 e6 5. Be2 c5 6. O-O Nc6 7. c3 cxd4 8. Nxd4
Nge7 9. Nxf5 Nxf5 10. Bd3 Bc5 11. Nd2 g6 12. Nf3 Nh4 13. Re1 Nxf3+ 14. Qxf3 Rc8
15. a3 Be7 16. Qg3 a6 17. Bh6 Bf8 18. Bf4 Bg7 19. h4 Bf8 20. h5 Be7 21. Rad1
Na5 22. Qg4 Nc4 23. Re2 Qb6 24. Bc1 Qc7 25. Qg3 b5 26. Bc2 Qc6 27. Rd3 Rd8 28.
Rf3 g5 29. h6 d4 30. Qg4 dxc3 31. Rxc3 Qb6 32. b3 Qa5 33. Qf3 g4 34. Qc6+ Kf8
35. b4 Bxb4 36. Rxc4 bxc4 37. axb4 Qa1 38. Qc5+ Ke8 39. Re1 g3 40. fxg3 Qd4+
41. Qxd4 Rxd4 42. Bg5 Rg8 43. Ba4+ Kf8 44. Be3 1-0[/pgn]

[pgn][Event "41st Olympiad Open 2014"]
[Site "Tromso NOR"]
[Date "2014.08.10"]
[White "Shankland, Sam"]
[Black "Polgar, Judit"]
[Result "1-0"]
[ECO "E11"]
[WhiteElo "2624"]
[BlackElo "2676"]
[PlyCount "119"]
[EventDate "2014.08.02"]
[SourceDate "2012.05.23"]
[WhiteTeam "United States of America"]
[BlackTeam "Hungary"]
[BlackTeamCountry "HUN"]

1. d4 Nf6 2. c4 e6 3. Nf3 Bb4+ 4. Nbd2 b6 5. a3 Bxd2+ 6. Qxd2 Bb7 7. e3 a5 8.
b3 d6 9. Be2 Nbd7 10. Bb2 O-O 11. O-O Qe7 12. Qc2 c5 13. Rfd1 h6 14. Rac1 Rfc8
15. a4 cxd4 16. Nxd4 Nc5 17. Ba3 Nfe4 18. Nb5 d5 19. f3 Nf6 20. Qb2 e5 21. cxd5
Nxd5 22. e4 Nf4 23. Bc4 Rd8 24. Kh1 Rxd1+ 25. Rxd1 Rd8 26. Qc2 Rxd1+ 27. Qxd1
Qg5 28. Qd2 Nxe4 29. fxe4 Bxe4 30. Bf1 Bc6 31. Bc1 Qg4 32. Nc3 g5 33. Qc2 Kg7
34. Be3 Qe6 35. Kg1 Bb7 36. Nb5 Nd5 37. Bf2 Nf4 38. Nc7 Qg4 39. Bxb6 Nh3+ 40.
Kh1 Qf4 41. Qc4 Qd2 42. Qe2 Qc1 43. Ne8+ Kf8 44. Nd6 Bd5 45. Be3 Qb1 46. Qd3
Qe1 47. Qe2 Qb1 48. Nc4 Qxb3 49. Qd3 Qxd3 50. Bxd3 Nf4 51. Bxf4 exf4 52. Nxa5
f3 53. gxf3 Bxf3+ 54. Kg1 Bd1 55. Bb5 Ke7 56. Nc4 Bf3 57. a5 f6 58. a6 Kd8 59.
Kf2 Bh1 60. Ne3 1-0[/pgn]

Everyone will have different gaps in their understanding, which makes identifying these gaps very difficult, and one of the many reasons having a great coach is very necessary. Luckily, once a gap is identified, it's usually already fixed. 5. Come to the game ready to fight and play well in the first moves out of preparation There is very little more frustrating than spoiling great pregame work with over the board ineptitude. There were countless cases in the past couple years where I played very poorly on the first few moves out of preparation, but none as ghastly as this:

[pgn][Event "ch-USA 2014"]
[Site "Saint Louis USA"]
[Date "2014.05.09"]
[White "Shankland, Sam"]
[Black "Robson, Ray"]
[Result "1/2-1/2"]
[ECO "C42"]
[WhiteElo "2634"]
[BlackElo "2631"]
[SetUp "1"]
[FEN "r3r1k1/1pp2ppp/p2q4/n3Nb2/3P4/b1P2B2/1R3PPP/2BQR1K1 w - - 0 19"]
[PlyCount "24"]
[EventDate "2014.05.08"]
[Source "ChessBase"]
[SourceDate "2006.11.23"]

{I was not quite warmed up yet, and it showed. My round 1 game basically
featured 0 moves and was bound for a draw from the first half hour of play,
and this may partially explain my very poor play at the start of round 2.
After one of the nastiest pieces of preparation I have done in recent memory,
I emerged from the opening armed with the knowledge my position should be
close to winning and an 80-1 time advantage. So it was very disappointing when
the game was drawn at the first legal moment.} 19. Rbe2 $6 {This throws away
some of white's advantage} (19. Ra2 $1 {This was the simplest route to a large
advantage, and not particularly difficult to calculate} Bxc1 20. Rxa5 Bf4 21.
Bxb7 $1 Rab8 22. Rxa6 $16 {White has a clean extra pawn}) 19... Bxc1 20. Qxc1
$6 {Another poor move} (20. Nc4 Qf8 21. Nxa5 $14 {Again white should win a
pawn, but here black's bishop pair will provide some practical compensation})
20... Nc6 21. Qb2 Nxe5 22. Rxe5 Rxe5 23. Rxe5 Bd7 24. Qxb7 Re8 25. Rxe8+ Bxe8
26. Qa8 $2 (26. Be2 {This was the last chance} c5 (26... a5 27. Qa8 Qe6 28. Kf1
$16) 27. d5 $1 {And white can still press, although I do believe the position
to be defensible} (27. Qxa6 $2 Qxa6 28. Bxa6 cxd4 29. cxd4 {Black should hold
a draw without much trouble, as I cannot see a convincing way for white to
remove his king from d6 or threaten the kingside pawns})) 26... Kf8 $11 27. Be2
Qa3 28. Qxa6 Qxc3 29. Qc4 Qxc4 30. Bxc4 Ke7 1/2-1/2[/pgn]

There is not much specific training one can do for this, but I decided to impose a minimum time amount I need to spend on my first move out of preparation. It definitely helped out at Tata Steel last week:

[pgn][Event "77th Tata Steel GpB"]
[Site "Wijk aan Zee NED"]
[Date "2015.01.16"]
[White "Shankland, Sam"]
[Black "Michiels, Bart"]
[Result "1-0"]
[ECO "C06"]
[WhiteElo "2652"]
[BlackElo "2563"]
[SetUp "1"]
[FEN "2b1rrk1/1pq3pp/p1nbpn2/3p4/N2P4/3B1N2/PP1B1PPP/2RQR1K1 w - - 0 17"]
[PlyCount "33"]
[EventDate "2015.01.10"]
[Source "ChessBase"]
[SourceDate "2006.11.23"]

{This was a case where I got a very pleasant position from the opening, but I
was also able to play very well right away to maintain a big advantage.} 17.
Ne5 $1 {This was not a wildly difficult move} Qb8 18. Bf4 $1 {This was the big
decision I was very proud of. My first instinct was to be solid with Bc3, but
this is considerably less effective. White gets absolutely everything he could
ever want in a Tarrasch French if he can play Nxc6 followed by Bxd6. This
would leave him strategically winning, which forces black's hand:} (18. Bc3 {
This modest move also secures an edge, but white has no real threats anytime
soon.}) 18... Nxd4 19. Bxh7+ $1 Kxh7 20. Qxd4 {And here we see the point.
Despite ceding the bishop pair and trading a central pawn, white has an
incredible and unbreachable dark square bind. After making triply sure black
could never get e5 in, I went with this line confidently, and indeed white is
pretty much winning.} Ne4 21. Bg3 Rf5 22. Nf3 Nxg3 23. hxg3 Bd7 (23... e5 {
Does not work} 24. Qxd5 e4 25. Qd4 $1 {And the knight is immune due to the pin}
) (23... Rxf3 24. Qh4+ $1 (24. gxf3 Bxg3 {Is also winning but less clear})
24... Kg8 25. gxf3) 24. Nc5 Bxc5 25. Rxc5 Qd6 26. Ne5 g5 27. g4 Rf4 28. Qd3+
Kh8 29. Qg6 Re7 30. Rc3 Rh7 31. Qxg5 Re4 32. Rxe4 dxe4 33. Qd8+ 1-0[/pgn]

Look for the second part of Shankland's instructive series on breaking the top 100 next week. Sam recently went to Corus Wijk aan zee along with another star Sam, GM Sevian, through the support of the Kasparov Chess Foundation. Find out more about Sam and follow his adventures on his official website, http://www.samshankland.com/ and his facebook page.