MINI ESSAY

When You Just Know The Answer

Intuition is real, but not everywhere

Aaryan Harshith

--

Image by Clark Van Der Beken

Within the first ten moves, a game of chess can exist in one of 69,352,859,712,417 permutations. Despite its rigid rules — pawns that shift up, knights that move in L’s — the power of emergence quickly transports chess into a world of immense complexity.

And yet, when the current world champion, a Norwegian player named Magnus Carlsen, was asked the secret to his prowess, he simply replied “Sometimes a move just feels right.” Faced with a similar question, former world champion and grandmaster Vishy Anand offered up an equally head-snapping answer: “Intuition is the first move I think of.”

At first glance, these statements make the world’s best chess players sound like prophets: out of an inhumanly large haystack of options, the best move simply materializes before them. Given how automatic their decisions appear, it becomes difficult not to see these individuals as prodigies.

But neuroscience points in the opposite direction. Intuition, it seems, isn’t a divine force, but a skill that can be learned just as readily as riding a bike. Even beyond chess, it suggests that mastery in general isn’t a matter of divination, but carefully refined judgement.

On Intuition, Recall, and Skill-Acquisition

Consider a landmark study conducted by two researchers, Gobet and Simon, from Carnegie Mellon back in 1996. Here, researchers gathered a control group — randomly sampled from the population — as well as a group of intermediate and advanced chess players.

For 5 seconds, they showed their participants snapshots of boards midway through games, then asked them to recall the positions of the pieces. As expected, the participants with backgrounds in chess showed substantially better recall than those without, and advanced players performed better than their intermediate peers.

An example of how Gobet and Simon turned a real chess position (left) into a contrived one (right) by swapping quadrants of the board. Note the placement of the white king with three pawns for context.

But there was a twist: shortly after their first test, participants were shown boards that were randomly distributed with pieces, in a way that would be impossible to arrive at in a real game. This time, though, the chess players fared no better than the non-chess players. Extra experience didn’t seem to help, either.

I like to think of this study as one of the great democratizers in chess because it finally proved that the competitive advantage of great players doesn’t lie in superior brains, but in the careful study of openings, variations, and counterattacks. Chess isn’t elitism disguised as a board game. It’s pattern recognition.

Skill arises as we learn rules, apply them in the real world, and reflect on our mistakes. Stir in thousands of games (perhaps sprinkle in 10,000 hours of practice), and the best move becomes as self-evident as the fact that two and two are four.

Here’s how chess historian Bill Wall described one of Bobby Fischer’s most iconic series of simultaneous blitz games, played during a trip to Yugoslavia:

He [Fischer] used half his allotted time in 5-minute chess and scored 19 out of 22 against some of the strongest players in the world, including three past world champions Petrosian, Tal, and Smyslov. After the tournament, he rattled off the moves of all 22 games he played without a mistake

…To the experienced player, chess is not a random bunch of moves, but is as meaningful as a literary work, like reciting a poem. The logical sequence of moves links them together in memory as are verses by their cadence.

In the psychology of expertise, certain non-negotiable requirements must be met for skills to become procedural (second nature). At a minimum, they must be predictable, repeatable, and verifiable.

Predictability

Learning must occur in a bounded environment with rules. In chess, this corresponds to the limited ways in which each player can move. For aviation or music, it would be the laws of physics or composition theory. In contrast, you can’t become a master at predicting whether a roulette wheel will end up red or black since the game is founded on randomness.

A glimpse into the science of skill-acquisition, and the environments that allow for mastery (Source)

Repetition

Learning requires many reps. When a toddler first learns to tie their shoes or recite their times’ tables, it isn’t habit. Biologically, this is because we are exposed to countless stimuli that our brains ultimately have to filter and slot into their limited storage. One-off actions live in our short-term memories and fade within minutes. Regular rehearsal signals to our brain that the action is valuable enough to get shuttled to the hippocampus, where it can remain intact for decades.

Verifiability

Learning requires accurate, timely feedback. In other words, your performance needs to be compared to an ideal, and you should be reinforced or penalized accordingly. This is where coaches and mentors (or chess bots) come in. Slowly, they help you recognize the merits and shortfalls of your decision till you can do it yourself.

Intuition is simply skill turned unconscious. It’s a hallmark of expertise because it can be gained through deliberate practice.

But it’s not present everywhere.

Intuitive No Man’s Land

Take the political commentators that are so ubiquitous in today’s news channels. Despite holding numerous degrees in history and international affairs, and extremely specific knowledge of the tensions in Libya or the famines sweeping India, their predictions consistently underperform.

Another study, courtesy of Holly Donaldson and colleagues at Hamilton University, evaluated the accuracy of 472 political predictions from 26 op-ed columnists and anchors — rated on a likelihood scale of 1 to 5 — only to reveal that the majority’s forecasts fared worse than a coin flip.

An excerpt from a 2013 paper by Metaxas et al. highlighting a testimony from a commentator (Source)

The straightforward explanation to this discrepancy is that there simply aren’t thousands of conflicts in the Middle East or famines in Asia to serve as case studies. To boot, each event traces a never-before-seen tree of causes and effects that make it fussy to match with an already small reference class of similar conflicts.

Finally, commentators can sit on the fence (“X or Y might win”) to mask uncertainty. In any case, their predictions have no way of being wrong, and they have little incentive to adjust their theories even if they are. After all, no one’s checking.

Unreliable predictions aren’t exclusive to political commentary, though it is a striking example. Any field that claims to out-strategize random noise — whether it’s stock trading or fortune-telling — is conflating speculation with science. Intuitions without a background of reliable, deliberate practice shouldn’t be trusted.

Where We Stand

Telling the tides of complex global issues and predicting the movement of the markets are fundamentally different from the controlled, discrete environment of chess.

But when it comes to the few skills that can be honed, masters can make their work seem indistinguishable from magic. We idolize them. We even mythologize their feats of athleticism and intellect because they seem so far removed from the bland hubbub of everyday life.

Except, Carlsen was never born pre-loaded with 1001 chess games. The only thing that separated him from the average chess player was the steepness of his learning curve — how effectively he applied the principles of skill-building.

Few people consider that — hiding behind every master’s exquisite intuition — is a trail of bad decisions that gets slightly better with time. What could be more human than that?

We can all get better, but perhaps not in the ways we might first think. And yet, that path is now a science, equally accessible to us all. Waiting for you to use it.

With that, it’s your turn.

--

--