
Image from commons.wikimedia.org
Training Game #22 using Turbo Eval
I'm back! It's been a minute since I last uploaded, so I have some things to cover.In my last post, I stated that I would have a competition between both Turbo Eval and Jeremy Silman's thought process. I played a game against the 1800, 2000, and 2200 chess.com bots in order to get a feel for how each of the thought processes play so that I can decide which would be better for me to use. Below, I will show the 6 games I played in total:
Turbo Eval vs 1800
https://lichess.org/3MaKFyiv#1
I had 99% accuracy in this game.
Turbo Eval vs 2000
https://lichess.org/fm004G13#1
I had 98% accuracy in this game.
Turbo Eval vs 2200
https://lichess.org/K7szIUoI#1
I had 90% accuracy in this game.
Silman's Process vs 1800
https://lichess.org/0e5FZWvm#1
I had 94% accuracy in this game.
Silman's Process vs 2000
https://lichess.org/62y9S6k3#1
I had 89% accuracy in this game.
Silman's Process vs 2200
https://lichess.org/FLVymsgV#1
I had 95% accuracy in this game.
Overall, I performed well with both of the thought processes. In general, I did better with Turbo Eval, though that may be in part because I was more used to using it. However, I did find that some of Silman's steps were certainly useful, namely, the step where you determine which side of the board to play on. It's a small step, but it did help narrow down some of the moves that I would look at. So while I did well with Turbo Eval, I decided to improve it. I used this website as the main inspiration for the changes and format of the thought process. So, without further ado, I present the new and improved Turbo Eval:
Better Turbo Eval
In the opening:
- Play what you know from self study and opening preparation. Once out of what you know, consider yourself to be in the middlegame.
In the middle game:
Every Move:
- Analyze the position for tactical considerations such as loose pieces, possible forks, etc.
- If something seems promising, calculate as accurately as possible.
- If nothing seems promising, then follow your set plan. If you don't have a plan, transition to the planning steps.
Planning Steps:
- Analyze the position for general considerations such as open lines, possible outposts, undeveloped pieces, etc.
- Consider what your opponent might want from the given position based on your analysis.
- Based on analysis, consider what side you should play on.
- Find candidate moves that fulfill some purpose based on your analysis.
- Calculate as accurately as possible.
In the endgame:
- Play carefully and be mindful of pawns.
If you have read the title, you may have wondered where the training game is. Well, now that I have provided a small introduction, below is the rematch against the 2700 chess.com bot:
In the post game analysis, I had an accuracy of 86% with 3 inaccuracies, 1 mistake, and 1 blunder. However, I feel that the 86% accuracy is a bit misleading. Out of the opening, I was able to completely equalize the position and keep it equal for 30 moves. I was able to give up the exchange in order to obtain a decent amount of counterplay, which was the correct thing to do, and in the endgame, I had a brief moment where I was better, but managed to screw up. After I felt my position fall apart from what it was, it wasn't long until I managed to blunder a drawn endgame into a loss. Below are a few highlights:
On move 16, I had obtained an equal game with the black pieces. I had a really insane way to continue the game, which I actually saw, but because it was so very complicated, I chose against it, fearing that I would miss something crucial. The move that can be played is 16... Bf4. I had actually considered this move as it opens up my queen to the pawns, attacks a crucial pawn, and puts a lot of pressure on the white pawns, but because when I attempted to calculate it, it proved very complicated, I decided against it and went with the move c5. I knew that if white took my pawn on c5, I would be winning due to immense pressure, and I had the plan that after white plays d5, I would reroute my knight to e5 via d7.
On move 26, I had the ability to capture a free pawn with 26... Rxe3, however, I saw that after 27. Be5, weird things would happen. After some thought, I came to the conclusion that I could give up my rook for a bishop and two pawns, and have enough counterplay to consider it a good idea. Obviously, this would only be good if I could follow it up correctly, and I was actually able to hold onto my advantage and provide enough pressure to keep at an equal evaluation for another 25 moves. I am very pleased with this.
But all good things must end, as on move 50 and 51, I had a chance to prove that my efforts could be rewarded, but I didn't play the critical moves in the positions given and as such, I brought myself into a completely drawn endgame, where I had two pawns and a bishop for the computers rook. The same amount of material I had exchanged earlier. Finally on move 55, I lost my second pawn and blundered the game. After 55. Rxf5, I had to play 55... Bg3+, but instead, I gave the check with my pawn, which completely blundered the game.
Final Thoughts
As I said earlier, I am pretty proud of my game, despite my loss. With more practice, my improved Turbo Eval should be a very effective weapon for me to utilize. Anyways, that's about all I have for today.
If you are interested in the idea of training to fight against the computer, consider joining The Club of Believers. As I have been posting these games, more people have joined, which is absolutely amazing. The more likely it is that we will uncover new ways to fight back against the computers.
If you find this content interesting, consider giving my post a like so more people can read and enjoy.
Thank you for reading, and enjoy the rest of your day!
More blog posts by HelloImRiley

Training Game #23 with Turbo Eval
This is the 23th training game I have played against the computer with my thought process, Turbo Eva…
Training Game #21 with Turbo Eval
This is the 21th training game I have played against the computer with my thought process, Turbo Eva…
Training Game #20 with Turbo Eval
This is the 20th training game I have played against the computer with my thought process, Turbo Eva…