Guessing the Evaluation: Learning and Fun in Chess Analysis

🔵 My Chessable Courses: chessable.com/drcan
♟️ Find me on Chess.com: canka19
♟️ Find me on Lichess: cantosh
🏆 2022 Chessable Community Author of the Year! www.chessable.com/blog/announ...
00:00 Introduction
03:13: Position 1
09:30 Position 2
17:00 Position 3
23:52 Conclusions
I am excited to have started this series on guessing the evaluation. Thank you to those of you who have sent me interesting positions. My assistant picked 3 positions you had sent me and I gave my judgment after around 3 minutes.
This is a multi-purpose series: You will be witnessing how a candidate master thinks out loud and thinks through a position. This is great for learning, especially for novice players (modeling example). Moreover, this guessing element makes it fun to watch, especially if I fail miserably according to the engine. Finally, we compare our human judgment with the engine evaluations and take lessons accordingly. Both education and entertainment!
Please send me more positions by writing a comment (FEN number of the position) on this video for my future recordings of this series! Also please give me feedback on this first episode so I can improve the series accordingly.

[@eschiedler]

Very good coaching tool. Blind evaluation of near equal positions and the differences between tactics and space and compensation.

[@Dr.CansClinic]

Thank you!

[@roberthansen5727]

SwedishWalrus is a friend of mine, the position around 9:30 is from one of my games! I managed to find the best move and ultimately win the game. It started out as a weird Queen's Gambit Accepted with an early ...g6, which is how I got in the h4-h5-h6. I felt like my position was overwhelming enough to justify sacrificing a piece to enter this endgame where Black's position is falling apart and his activity is miserable. Blunders were made on both sides and there were cleaner ways to convert the advantage than this line, but SwedishWalrus will be able to contribute a comment about my habit of playing silly lines rather than the best lines.

[@Dr.CansClinic]

Wow, so nice to hear! Thanks for sharing it. So you found Bc3 in the game? This became a think out loud de Groot protocol rather than evaluating a quiet position :) But it was fun to analyse!

[@roberthansen5727]

@@Dr.CansClinic I hunted down the game and it turns out I misremembered slightly -- I found the Bc3 idea, but played the wrong move order and went into a 0.0 position. However, we were both short on time and I ended up regaining the advantage. There was one moment I was lost, but my opponent was down to 30 seconds and I'd started playing against his clock. If you want to analyze the full game on your channel or just review it yourself, the link after the regular Lichess address is NHbFTC1T

[@jimmccann3856]

Postscript: The later positions are also fascinating, and you provide the seminal comment (serially): "Perhaps the position is not settled yet.." Ya, in the middlegame, it often (usually) not settled yet, and any time spent Evaluating While Assuming Quiescence may well be better spent Proving Quiescence. Rhetorical question: How relevant is Evaluation anyway, if the game is usually won by the person who can trace lines better? I am definitely looking forward to the next video in this series!

[@Dr.CansClinic]

That is true - proving quiescence can take a long time. And you do not need to know the evaluation to find the best moves as you said. But you usually end your calculations somewhere, after you properly 'evaluate' the resulting position, right? You judge it to be better for you, thus you choose that line.

[@brigidwell]

Thanks, these were really fun. I paused before you solved and did pretty decent. My guesses were: 1. -0.7 (all French players know the first move) 2. +2.0 3. -0.2 I really liked the computer’s …Qd7 in the last position.

[@Dr.CansClinic]

Thank you so much for your participation! You were pretty good with your guesses!

[@jimmccann3856]

Thank-you for (mis)using my (very rich) position as your first example. (I deliberately used a basic tactic to require a Ply 3 evaluation, rather than a Ply 1 evaluation. You correctly required quiescence before evaluating. Well done.) Initially, you refused to read the comments I attached to the position, stating that they might color your evaluation. Fair enough. But please consider them now: The computer is WRONG! My opponent was almost certainly surprised by Nxe5, and as the old chess adage goes: The man surprised is half-beaten! I evaluated -1.00, as you did, and was astonished to see Stockfish later giving only -.40 or so. My actual chances of winning against a human opponent are very much better than a -.40 score might suggest. The real point here, that both you and Stockfish missed, is that the Scandinavian Advance has a rather poor reputation, and an oblivious opponent, who then drops a center pawn quickly on top, is highly unlikely to make computer-like use of compensation to recover. I play 1...d5 exclusively versus 1 e4, see the Advance dozens of times per year, have sprung this trap perhaps 20 times over the years, and to my recollection, have never seen White win. That ain"t -.40! Pablo Picasso: "Computers are useless. They can only provide answers..." Therefore, Doctor, be very careful about the questions you ask of the machine. As they say in the computer biz: Garbage In, Garbage Out!

[@Dr.CansClinic]

Haha great insights. Once you put it on the board, you see some activity compensation for White, especially around the knight on h6 after fxe5. But if Black survives the immediate storm, Black should achieve a winning position with an extra pawn, the bishop pair, as well as the nice pawn duo on d5 and c5. Yes, advanced Scandi is quite poor.

[@ibiwisi]

A fun visit inside the head of DrCan! As you said in the video, you provide a wonderful model for improving the thought process of us lesser mortals. Thank you, and I look forward to more videos in this series.

[@Dr.CansClinic]

Glad you enjoyed it! I will speak more in the future episodes :)

[@pavelalg1163]

A few positions from my games. Alekhine defence (1.e4 Nf6 2. e5 Nd5. 3.d4 d6 4. Nf3 dxe5. 5. Nxe5 c6 6. c4 Nb4 7.a3) Black to move FEN: rnbqkb1r/ppp2ppp/4p3/4N3/1nPP4/P7/1P3PPP/RNBQKB1R b KQkq - 0 7 English opening (1. c4 Nf6 2. Nc3 b6 3.e4 Bb7 4.e5 Ne4 5. Qf3 Nxc3. 6.Qxb7) - Black to move FEN: rn1qkb1r/pQpppppp/1p6/4P3/2P5/2n5/PP1P1PPP/R1B1KBNR b KQkq - 0 6

[@Dr.CansClinic]

Thanks for sending those!

[@southernrun9048]

Very neat idea for a video. Was fun following along

[@Dr.CansClinic]

Thanks so much!

[@masonparkman5567]

This was fun! Thanks Can!

[@Dr.CansClinic]

Thanks for sending the position Mason! That was the least chaotic one for sure!

[@hooptron9]

excellent video!!!

[@Dr.CansClinic]

Thank you so much!

[@eschiedler]

The most difficult positions to evaluate for an amateur are those with equal material, queens still on the board, multiple scattered pawns, but active play only for white and the computer says Plus 4.50!!!

[@Dr.CansClinic]

Yes, piece activity/coordination/quality can be difficult to evaluate precisely, whereas everyone can count the material.