Columns | July 16, 2012 17:05

The relativity of confidence

I’m normally a bit skeptical when chess players give lectures – cobbler, stick to thy last! – but Vishy Anand’s recent Accenture lecture (posted on ChessVibes) is an absolute must-see for every chess enthusiast.

Vishy Anand during his Accenture lecture in Madrid | Photo Accenture España

The lecture, on the ambitious topic of decision-making, is generally funny, interesting and highly entertaining, but there was one fragment which really had me at the edge of my seat.  This is when Anand discusses the trust grandmasters put in chess engines. I couldn’t help thinking Anand had hit upon some profound truth in chess which I have rarely seen mentioned elsewhere. It’s worth quoting in full:

It’s almost scary to realize that generally we tend to trust the computer like we do a calculator. When I do analysis with the computer most of its conclusions I tend to trust blindly at any given moment in time. However, it turns out that as the hardware improves and the program gets more and more sophisticated it will often change its opinion on many positions and it’s scary to realize that in 2006 I went to some games fully confident that the computer’s verdict that Black was much better in that position was correct. I went to the board and played on that basis. And let’s say in 2008, with the same confidence and conviction, I went to the board thinking that White was better. And clearly both can’t be right. But as a competitive skill it doesn’t matter. It’s much better to be deluded and confident than to have the right information but not know what to do.

Clearly, the notion that confidence is all-important in competitive events is not what makes this quote so memorable, but the implicit recognition that the “absolute truth”, even in extremely complex positions, is not what even the world’s best grandmasters are after– not even with the help of computers - is instructive indeed.

Anand’s example of a computer’s changing evaluations of the same position over time reminded me of Isaac Asimov’s concept of the “relativity of wrong”. This concept occurred to Asimov when he received a letter from a specialist in English literature who put it to him that

in every century people have thought they understood the universe at last, and in every century they were proved to be wrong. It follows that the one thing we can say about our modern ‘knowledge’ is that it is wrong.

Asimov’s answer has now become a classic in scientific literature:

John, when people thought the earth was flat, they were wrong. When people thought the earth was spherical, they were wrong. But if you think that thinking the earth is spherical is just as wrong as thinking the earth is flat, then your view is wronger than both of them put together.

Asimov points out that right and wrong are “fuzzy concepts” and that

what actually happens is that once scientists get hold of a good concept they gradually refine and extend it with greater and greater subtlety as their instruments of measurement improve. Theories are not so much wrong as incomplete.

We all know, of course, that this is exactly how chess theory has evolved over the past centuries. Right and wrong have always been fuzzy concepts. It’s hard to imagine now, but the Sicilian Defense was once considered to be a dubious opening. Siegbert Tarrasch considered 3…Nf6 in the Queen’s Gambit Declined to be a mistake.

But Anand is talking about just a few years difference! According to the World Champion, it’s perfectly possible to be confident about some evaluation of a position or variation while at the same time acknowledging the evaluation may be totally different in 2 years time.

From an absolutist point of view this may sound astonishing (or even despicable, depending on how strongly you feel about the issue) but from a practical point of view it makes, of course, perfect sense. It’s just not workable to focus on the “philosophical” side of chess alone – or perhaps it is, but you probably won’t score many rating points with it.

Indeed, Anand concludes the fragment by saying that

in the end what you’re looking for is clarity at the board or clarity in action. You want to be able to play a position. You want to be able to enter it if that plays to your strengths, and that’s all that really matters. So even some false confidence is fine.

It’s easy to relate this to other well-known “pragmatic” aspects of Anand’s game: his speedy play, his often deceptively simplistic-looking game commentary, his easy-going and relaxed appearance in interviews. But does all this make Anand a true “chess relativist” and should all chess players adopt his perspective in order to make the most of it?

Is his a fundamentally different perspective that that of, say, Robert Huebner, who is known for his extensive game analysis with hundreds of branches and sub-branches of variations in order not to miss a single detail? Would Kasparov, whose insatiable quest for absolute truth in chess variations has become proverbial, agree with Anand? How would Bobby Fischer have reacted to the idea of “false confidence”? Isn’t Anand’s perspective a bit too pragmatic?

Not so fast. Anand himself seems keenly aware of the theoretical downsides of this approach to chess. He calls it “scary” not only to trust a computer so completely, but also to realize how confident he was in positions that would have been evaluated completely differently just two years later – implying that in some abstract sense he’s not really happy with it. Anand’s false confidence is relative, at best.

Isaac Asimov also emphasized the conflict between these two sides of the same coin – pragmatism and the demand for absolute truth. The ancient Greeks made excellent maps of the Mediterranean without taking sphericity into account; the Sumerians accurately predicted the movement of the planets while thinking the earth was in the centre of the universe. And to this we can add that one can become the undisputed World Chess Champion by relying, at least sometimes, on false confidence in computers.

Arne Moll's picture
Author: Arne Moll


noyb's picture

It's been awhile since I gave props to Arne, but I have to admit that he nailed this one.

VIktor's picture

Vishy speaks both English and Spanish too.

Chess Fan's picture

Very intelligent and nice article.

VK's picture

Good one Arne!

Fireblade's picture

Awesome article !

Bert de Bruut's picture

Back to the area where you really shine Arne, feed us some more!

Anonymous's picture

I picked up on playing to your strengths is a key point to the article. To me that means do what you know how to do best. Interesting comments by the champ.

jo's picture

So many say that Anand is past his prime...

IF that is true then maybe its a function of when he realized that what was blind confidence is now wise (damn I cant think of the right me out here)

Your comment

By posting a comment you are agreeing to abide our Terms & Conditions