Mikhail Tal once said that…
“There are two types of sacrifices: correct ones, and mine.”
…but was this really true?
Were his sacrifices really incorrect?
Tal’s sacrifices were sometimes unsound, and that’s okay because Tal wasn’t playing against computers…
He was playing against regular humans, who even at GM level, defend really poorly.
In positions with a king who is vulnerable and with enemy pieces lurking and highly active…you will NOT see everything.
That’s a fact.
And probably a reason why Mikhail Tal also said that…
“You must take your opponent into a deep dark forest where 2+2=5, and the path leading out is only wide enough for one.”
It seems to imply that if you present your opponent with unfathomable complications and an impossible (for humans) defensive task, they are very likely to miss the best defense and lose.
Tal made an art of winning in this way.
So if you’re ready to look at a few examples, I’ll show a couple of my favorites now.
Position 1: Tal vs Larsen
Candidates Semi-final, 1965
I’m sure that when Tal played this sacrifice against Larsen, the Danish GM was fully aware of the dangers, but what could he do?
I mean, in 1965 the idea of 3400 Elo engines like Komodo easily finding a defense was science fiction.
And come on…after the (forced) acceptance of the piece sac, would you be confident in defending this position as black?
White has two bishops aiming at the black kingside and the black queen is MILES away from the area, and will not be able to help in defense.
Most of us would lose this against an attacker of Tal’s prowess…just as Larsen did.
So Tal won a CANDIDATES match game…with an “unsound” sacrifice.
Let’s see another…a WORLD CHAMPIONSHIP game!
Position 2: Botvinnik vs Tal
World Championship Match, 1960
Above we find another position where a 3400 engine thinks Tal is clearly worse after his 21…Nf4!? sacrifice.
But again, after accepting the sacrifice, it feels a bit tough to find a good defense.
Would you feel comfortable defending against all of these active pieces?
World Champion Mikhail Botvinnik wasn’t and lost.
So now I wanna show you something I have been working on with Pawel Koziol – a TOP SECRET project.
How do you get a chess engine to resemble Tal so much, that it likes even his most unsound sacrifices?
This is a part of our new project, and as you can see below…the results speak for themselves.
A Real Digital Tal?
Digital Tal finds and likes Nd5 in just 2 seconds – this is something I couldn’t even manage when tuning OpenTal to emulate the Latvian genius!
And Digital Tal takes 10 seconds to like …Nf4! as the best move.
Pawel and I are excited about the possibilities of replicating a whole host of world chess champions, so that their exact styles, prejiduces and move choices are almost a mirror with the real GM.
So stay tuned for news on that project.
How I Copied Tal’s Style and Went for UNSOUND Sacrifices in MY Games
Whilst working on Digital Tal and researching his games in detail, I have found his style rubbing off on my own play and this was obviously a welcome surprise.
I’ll finish this post by showing you a game I played where I played an unsound knight sacrifice which created so many problems for my opponent that he was unable to find the right defense.