The number one pain in the neck for chess fans following top-level events online is having fun, interesting, thought-provoking, instructive commentary by knowledgeable GMs obliterated by as...les flinging handfuls of meaningless computer evaluations all over the place: 0.00 (“dead draw!”), +2.35 (“blunder! what a prick!! I always knew this guy was overrated…”)
The normal step to take to keep enjoying the games themselves is to not use a computer to check variations. Although it does make sense, why should we deprive ourselves of such a useful tool? Anyway, it’s not working unless you refrain from looking at and/or taking part in the chat… in which case you are missing out, one has to admit – not just on the opportunity to hurl witty comments (“That’s not a draw, you stupid! My comp says +0.21”) in the face of those idiots, but also because the damned machine does unearth incredible and often beautiful (not to mention interesting, thought-provoking, instructive) variations.
However, there is a way out. And, counter-intuitive though it may be, the solution is not less computer input, but more. As anyone interested in sports – especially US sports – knows, fans want stats. Fans relish stats. Fans abuse themselves over stats. Fans can have endless, heated discussions over the compared batting averages of players who died at a time when their own father – the guy who got them into it in the first place – hadn’t even been born. Actually, if there’s alcohol involved, it’s highly likely the discussion will end up in a brawl. Point being, the more figures fans are fed, the happier they get. And why am I using the third person here… oh, the hypocrite. Give ME numbers!! I don’t understand chess anyway. Obviously. How could I? Chess is so deep (Kramnik).
OK, so what figures are we talking about here? Well, commentators keep reminding us that 0.00 just means that the computer has found at least one line which ends up in a balanced position, possibly – but not necessarily – a draw. Similarly, +2.35 means one of the players has a winning advantage, but as a matter of fact, this might be an only move leading to a 30-ply variation… and all the others are losing! That’s called walking the tightrope.
What we need is at least one more figure telling us how difficult the variation is from a human point of view – this can take into account the depth of the variation, the number of only moves in it (both are numbers), the counter-intuitiveness of the idea (can be quantified somehow, surely), etc. We’re talking about a number that gets Magnus to say “I’m mightily impressed here”. For those who were not there, that is something he said when Jakovenko won his first rapid game against MVL recently at the 2019 World Cup, finding the computer’s first move five or six times in a row when MC and his fellow commentators were focusing on other moves. Say, the higher the number, the less likely you are to find it. That would help put things in perspective. But that’s not all. I’m sure today’s AI would be able to assign a relative value to each piece, if only for the purpose of discussion. Actually, the commentators can do that: basically, when you willingly give up an exchange, that’s because you believe a Rook and Knight, instead of being worth an arbitrary 5 and 3 respectively, were closer to 4 each, for instance. Likewise, king safety, space advantage… this needs to be quantified for the amateur player – even very weak – to get a better grasp of it, and one feels today’s software can do that already with some tweaks: picture a nice, technological-looking dashboard with numbers and graphs and some green here and a bit of red there and lights flashing all over the place and everything… wouldn’t it be nice?
There are many advantages to this approach. Not only would it – hopefully – make chat discussions during a live event more interesting, since instead of focusing on one number, we would have a bunch of them to toy with, but it would also give commentators something to discuss during quieter phases of the games. Most importantly, it would give all of us a better appreciation of how otherworldly strong top level chess actually is. And who knows, we might even improve our chess in the process.
We respect your privacy and data protection guidelines. Some components of our site require cookies or local storage that handles personal information.