<Bridgeburner:> PART I

Schlechter vs Lasker, 1910 is an extraordinary game played by two superlative grandmasters at an extraordinarily high standard.

Quantitative mapping of game 7 of the 1910 title match between these players is below. Figures in brackets are the corrected engine evaluations generated on the return slide. The reverse slide smoothed out many, but not all fluctuation in the engine’s evaluations. The complexity of some variations was very likely too great to enable a fuller reconciliation from the forward slide. General methods used are described in the bio.

The values in the opening are of no great moment, but of interest is that they come at the end of a full forward slide to the end of the game and a full return slide from the last move to the first.

<1.e4> ( 0.28) <1…c5> ( 0.28)

<2.Nf3> ( 0.28) <2…Nc6> ( 0.28)

<3.d4> ( 0.28) <3…cxd4> ( 0.28)

<4.Nxd4> ( 0.28) <4…Nf6> ( 0.28)

<5.Nc3> ( 0.28) <5…g6> ( 0.51)

<6.Bc4> ( 0.35) <6…d6> ( 0.35)

<7.Nxc6> ( -0.35) <7…bxc6> ( 0.35)

<8.e5> ( 0.35) <8…Ng4> ( 0.35)

<9.e6> (=-0.15)

Capablanca thought this advance was “at the least very risky”.

<9…f5> (=-0.17)

<10.0-0> (=/ 0.32) <10…Bg7> (=/ 0.30)

<11.Bf4> ( 0.54) <11…Qb6> (=/ 0.36)

<12.Bb3> ( -0.73) <12…Ba6> ( -0.73)

<13.Na4> ( -0.73) <13…Qd4> ( -0.73)

<14.Qxd4> ( -0.74) <14…Bxd4> ( -0.74)

<15.c4> ( 0.74) <15…0-0> (=/ 0.57)

<16.Rad1> (=/ 0.57) <16…Bf6> (=/ 0.34)

<17.Rfe1> (=/ 0.34)

<17…g5> (=-0.25)

<18.Bxd6> ( -0.72) Capablanca gave this an ! , commenting that the advanced e-pawn now becomes very strong. The engine is less enthusiastic about this assessment, preferring the conservative <18.Bc1> or <18.Bd2> with a corrected evaluation of =-0.25 and =- 0.35 respectively, compared to -0.72 for the move actually played – or about a third to half a pawn...

However, either of these alternative continuations would have been much less interesting than that which transpired in the game. It’s unlikely Schlechter didn’t see these moves, so perhaps the “drawing master” (sic) was being adventurous against one of the greatest players of the era.

<18…exd6> ( -0.72)

<19.Rxd6> ( 0.72) <19…Be5> (=/ 0.46)

<20.c5> (=/ 0.46) <20…Rfe8> (=0.00)

<21.g3> ( 0.45) <21…Bf6> ( 0.45)

Better was <22…Bb5> with the better game, as noted by Capablanca in the annotations. This is a confirmed evaluation shift of 0.94, equal to a bad move. Weighting = 1.

<23.Rc7> ( 0.49) <23…Be4> ( 0.49)

<24. Nc3> ( 0.49) <24…Bxd3> ( 0.49)

<25. bxc3> ( 0.49)) <25…Ne5> ( 0.49)

<26. Rd1> ( 0.49) <26…Nf3+> ( 0.49)

<27. Kf1> ( 0.49) <27…Nxh2+> ( 0.49)

<28. Ke1> ( 0.49) <28…Nf3+> ( 0.49)

<29. Ke2> ( 0.49) <29…Ne5> ( 0.69)

<30. Rdd7> ( 0.69) <30…f4> ( 0.69)

<31. Rg7+> ( 0.69) <31…Kh8> ( 0.69)

<32. Rxg5> ( 0.69) <32…Bd3+> ( 0.91)

<33. Kd1> ( 0.91) <33…fxg3> ( 0.91)

<34. fxg3> ( 0.91) <34…Ng6> ( 0.91)

<35. Rd5> (=0.24)

<35…Be4> (=0.24)

<36. Rd6> (=0.24) <36…Bf5> ( 0.55)

<37. Bd5> (=0.07) <37…Rab8> (=0.20)

<38. c6> (=0.20) <38…Nf8> (=0.20)

<39. Rb7> (=0.08) <39…Rbc8> (=0.08)

<40. e7> (=0.08) <40…Ng6> (=0.08)

<41. Bf7> (=0.08) <41…Rxe7> (=.0.08)

<42. Bg6> (=0.08) <42…Bg4+> (=0.08)

<43. Kc1 > (=0.00) <43…Re1+> (=0.00)

<44. Kb2> (=0.00) <44…hxg6 (=0.00)

<45. Rxg6> (=0.00) <45…Bf5> (=0.10)

<46. Rf6> (=0.10) <46…Be4> (=0.10)

<47. Rxa7> (=0.13) <48…Rb1+> (=0.19)

<48. Ka3 > (=0.19) <48…Bxc6> (=0.19)

<ANALYSIS>

<Note> The fluctuations generated in the relatively low (16 minimum) ply forward slide were smoothed out in the equivalent return slide. The corrected evaluations extracted from the return slide are used in this analysis, as they are considered more reliable than the raw evaluations generated on the initial forward slide. All moves, including the starting position have been evaluated on forward and return slide for completeness.

<Evaluation range>:

- Between <-0.73/-0.74> applying to the move group <12.Bb3 Ba6 13.Na4 Qd4 14.Qxd4 Bxd4 15.c4>, representing a significant but not necessarily a winning advantage for Black and

<0.91> in respect of the move group<32…Bd3+ 33.Kd1 fxg3 34.fxg3 Ng6> representing a significant but not necessarily winning advantage for White.

<The largest evaluation shift>:

- for White was <0.77> between <34…Ng6> (< 0.91>) and <35.Rd5> (<=-0.24>), coming extremely close to the evaluation shift representing a bad move. <35.Rf7> was the engine’s preference for preserving White’s advantage.

- for Black was <0.93> between <22.Rxc6> and <22…Bb7>, inside the range of an evaluation shift representing a bad move. <22…Bb5> was the engine’s and Capablanca’s preference for preserving Black’s advantage.

<Computer statistics>:

• 97.9% of the ply in this game (94/96) coincided with engine preferences 1, 2 or 3

• The two ply that didn’t coincide with engine recommendations 1, 2 or 3 were <9.e6> and <11.Bf4>

• 81.25% of the ply in the game (78/96) coincided with engine preferences 1 or 2

• 75% of the ply in the game (72/96) coincided with the engine’s first preference

• 95.8% of Schlechter moves coincided with the engine preferences 1, 2 and 3

• 100% of Lasker’s moves coincided engine preferences 1, 2 and 3

• 83.3% of Schlechter’s moves (40/48) coincided with engine preferences 1 and 2

• 79.2% of Lasker’s moves (38/48) coincided with engine preferences 1 and 2

• 75% or 36/48 of Schlechter’s moves coincided with the engine’s first preference

• 75% or 36/48 of Lasker’s moves coincided with the engines first preference.

Perhaps some serious questions should be asked about how these players had played such a high percentage of computer recommended moves!

<The engine evaluation of the final position>:

was <=0.19>, in a position agreed drawn.

<CONCLUSION>

This was an exciting game played with extraordinary precision by both players. The slight “inaccuracies” defined the volatility of the game…perhaps they could be simply interpreted as strategic decisions.

Neither player blundered and only Lasker made a move that met the project definition of a <bad move>, namely <22…Bb7>.

Schlechter played one move <35.Rd5>, that was very close to being a defined bad move.

<GAME RATING>:

<Within the parameters of the project, the game is weighted at 1, representing 1 bad move and 0 blunders by Lasker, and 0 bad moves and 0 blunders by Schlechter.>

<Endnote>

I have been following up on the World Championship matches from the era of the Kasparov - Karpov epics. Yet the first time I looked closely at this game here in CG.com, I was flat-out stunned by it. Gawd, these must be two demi-gods playing chess!

Amidst all the double-edged complicated ups and downs, twists and turns, only one bad move and no blunders? Lesser mortals would no doubt have interchanged many mistakes given the game's complexity in order to arrive at a draw.

Your <Computer statistics>:

• 97.9% of the ply in this game (94/96) coincided with engine preferences 1, 2 or 3

• The two ply that didn’t coincide with engine recommendations 1, 2 or 3 were <9.e6> and <11.Bf4>

• 81.25% of the ply in the game (78/96) coincided with engine preferences 1 or 2

• 75% of the ply in the game (72/96) coincided with the engine’s first preference

• 95.8% of Schlechter moves coincided with the engine preferences 1, 2 and 3

• 100% of Lasker’s moves coincided engine preferences 1, 2 and 3

• 83.3% of Schlechter’s moves (40/48) coincided with engine preferences 1 and 2

• 79.2% of Lasker’s moves (38/48) coincided with engine preferences 1 and 2

• 75% or 36/48 of Schlechter’s moves coincided with the engine’s first preference

• 75% or 36/48 of Lasker’s moves coincided with the engines first preference.

is even more astonishing. Schlechter and Lasker swam through the flood of sharp knives of double-edged complications like two gold medal Olympic athletes without a scratch, making virtually the best moves almost all the time!

Given the difficult complexity and high drama of this game, as you say:

<Perhaps some serious questions should be asked about how these players had played such a high percentage of computer recommended moves!>

---

Perhaps after you have analyzed through the rest of the games of this match, you could add a merit system for especially complex games (but not for tame and <Grandmaster> draws).

Say if one player makes the computer's first three preferred moves at least 95% of the time (thus practically always managing to find the best moves), you could remove a 1 from the final weight on the grounds that nearly always finding the best moves neutralizes one mistake.

<these must be two demi-gods playing chess!>

As I was gazing at this game over the last three days, this was exactly my feeling. Beyond awesome. Beyond astonishing.

<Perhaps...you could add a merit system for especially complex games>

I'm open to all ideas but let's talk about that sort of enhancement later. The engine I'm using is a middle of the road kind of beast, and is not in the league of some of the silicon giants striding the landscape, so the efficacy of that sort of weighting refinement may be questionable.

Also, one bad move doesn't necessarily result in loss, it may just make a position difficult to defend. Alternatively, it may result in loss if it has been preceded by a series of "dubious moves" that have gradually eroded a position without any of them actually meeting the criterion of a bad or even dubious (nearly bad) moves, and it pushes the game over the edge.

Other considerations are that the 3rd or even 2nd preferred move may be a blunder. The placement of moves in the placement order doesn't say anything about the evaluation gap between them. To take a simple example, there may be only two options available in a check, one wins and one loses, say to a mate. There are much subtler versions of this phenomenon that occur all the time.

I think there are other more obvious issues to tackle and quantify, like the question of <dubious moves>, <dubious sequences> and even <dubious strategies>.

My thinking is that a game can't be lost or won without at least one mistake...but that mistake may not necessarily be contained in one move. More on that later once I've finished this match. I'll take notes as I proceed and share these thoughts here.

The game leaves you in awe.

If only it succeeds.