To watch an opponent blitz moves early on in a game while you’re forced to sit, hand to the temple, and scrupulously calculate every probability as the clock races madly, is perhaps among the most unsettling scenarios for a chess player. World No.3 and tournament favourite Hikaru Nakamura lived through this unenviable predicament with White, in Round 2 of the Candidates in Toronto. Opposite him, gently rocking in his chair, eyes glowing like fireflies, was a self-assured Vidit Gujrathi.
In a Ruy Lopez Berlin opening, Vidit unleashed a novelty on move 8 with c6 – calling on White to win a pawn. A novelty is essentially a new move in a position that’s been played earlier. Vidit sacrificed two pawns and a bishop in a row. The American tried to comprehend the trappings and nuances of the position, his chaotic pattern shirt mirroring the turmoil of his mind. Hikaru had already used up over 40 minutes on his clock while Vidit appeared to be well within his preparation and had taken up just five minutes. The Bishop sacrifice that followed on move 11 by Vidit was dynamite and the beginning of the end for Hikaru in the game. Vidit credited his team of trainers for coming up with the winning ideas.
The evolution of preparation in chess has led up to a point today where computers show a greater number of acceptable ways to play as opposed to earlier, especially as Black. Finding new ideas in the opening also comes down to being able to go in new directions, which isn’t the computer’s obvious choice. “The human element in preparation is a lot more today than say 2015-2016 when I started working on these things,” said GM Srinath Narayanan, who previously worked as Arjun Eragaisi’s trainer. “There was a lot more of looking at what the computer said and organising that information back then. Now a lot of conscious attention is required to be aware of what the computer says but disregard it and choose your path. This requires the human touch.”
“You can generate a lot of work in a month or two. Typically for Candidates, at least three months of preparation goes into generating ideas. We saw it happen even post the pandemic lockdown. Players had accumulated months of training and knowledge and when everyone started playing over the board again, there were a lot of new ideas.”
To find these ideas, you try to get your hands on any information available. “You look at the correspondence chess, games played by computers and different tests that the computers run and the Computer Chess Championship – which pits the strongest chess engines in the world against each other – for instance. But again, everyone has access to these things.” The whole effectiveness of an opening idea or a novelty is after all based on how least likely it is to be looked at by someone else or at least getting to use it first.
A trainer/player could suggest moves that aren’t the engine’s top moves and look at how the corresponding positions are evaluated. “Also, sometimes it’s about suggesting moves, which the engine doesn’t seem too enthusiastic about at the beginning,” said Narayanan. “But you also sense that it’s not so easy for a human to find and evaluate. It’s clear now that the world of Stockfish and the world of humans is so different. So, a certain kind of human evaluation, as I see it, has become prevalent today, where, regardless of the evaluation or what the engine shows, there is an aspect of how difficult it is for a human to figure things out over the board.”
Vidit crushed Hikaru in 29 moves, a game that the American professional streamer went on to describe as “easily the worst chess” he’s played in the last couple of years, on his YouTube recap soon after.
“Hikaru has generally been able to smartly navigate his relatively less work on preparation. But that can work only at a limited probability, I suppose. If he plays regularly, I think the person who spends more time on chess, like Vidit, will be able to outplay him in openings.”
