Certain events just profoundly invite to some serious reflection on the practice of strategy, and Magnus Carlsen’s victory over Viswanathan Anand in the world chess championship (the WCC) in Sochi just some days ago is such an event.
Indeed, the analogy between chess and strategic decision-making may appear to be close to perfect, as many writers about business and military affairs have observed throughout history: deciding on a good move in chess is about assessing your current position; searching through your set of allowable moves (or a sub-set thereof); for each such move, searching through your opponent’s set of allowable moves given your move; and then recursively traversing the resulting search three until either a mate or reaching a certain search depth. In the latter case, the position is evaluated through some kind of internal node evaluation function based on heuristics (incl. counting resources and spatial distribution of resources). The process is augmented with pre-programmed chess openings and end game databases. The process is in this context to be seen as essentially the same for a human player and for a chess computer, though it is often argued that human players tend to use pattern recognition while chess computers typically employ a minimax algorithm augmented with alpha-beta pruning.
There are of course some obvious structural differences between chess and strategic decision-making, including for example perfect (in chess) vs. imperfect (in business) information; deterministic vs. stochastic behaviors; two-player vs. multi-player game; zero-sum vs. at least opportunities for win-win; conflict vs. competition; and fixed vs. fluid rules for what constitutes allowable moves.
However, if we accept that chess is at least a reasonably good analogy for business, it is then of interest to explore why there are performance differences between chess players to better understand why there are strategic performance differences between companies. I will argue that chess players lose because of*:
- Ineffective opening preparations: This is about opening theory, and it is generally assessed that this is one of Anand’s strength. In business, this boils down to having enough experience (or an MBA or the assistance of a bunch of McKinsey or BCG consultants), and do enough of planning.
- Incomplete or imperfect traversal of the search three / situational awareness: In chess this is about thinking through just a few selected moves, just a few steps into the future, or failing to go through the search process in a systematic way (in which a blunder may result, see below). In business and politics, it is called lack of creativity (too narrow search), lack of foresight (shallow search depth), lack of discipline (incomplete and unsystematic search), or lack of acumen (focusing on wrong part of search three).
- Emotional fatigue / no will to win (in Norwegian: “mangel på vinnerskalle”) / loss of nerve: Even the best chess players may experience a cracking of nerves, and Anand admitted in a press conference the WCC, after his loss of the championship, that he had done exactly so in game 11.
- Lack of physical stamina: Carlsen is apparently an eager player of basketball and soccer, also during tournaments, and it has been claimed that he is physically on par with many premier league soccer players. Not surprisingly, many successful CEOs also attribute their success to being in good shape physically.
- Making the random blunder: We saw it happen in game 6 in the WCC, actually as a double blunder when Carlsen first made a blunder in move 26, Kd2, and Anand then failed to detect the blunder. Generally such blunders tend to be due to time trouble, overconfidence, carelessness, or fatigue. Such things happen to experienced line managers as well, often for the same reasons.
- Seeking complexity: Anand’s move to b5 in move 23 in game 11 of the WCC was generally considered spectacular; the subsequent follow-up exchange sacrifice on b4 in move 27 was generally considered overly optimistic. In general, there is in chess an optimum on the complexity scale, somewhere in between the trivial position and the spectacular, but un-analyzable position.
- Poor strategic intuition: The term ‘strategic intuition’ is not really used in chess, but it is about how to construct an internal node evaluation function based on heuristics, for use when not being able to search all the way to the leaves of the search three. As indicated above, this is about in the end coming out on top, without having perfect analysis of the future.
- Imperfect end game execution: Carlsen’s strength in the WCC was to some extent linked to his precise and forceful end game play, with no mistakes and no unnecessary moves. An end game in chess resembles for example compiling a winning a bid for a take-over target, or negotiating a final contract with a major customer: There are few pieces on the board, good knowledge, and the solutions to many end games are known; it is just about getting execution right.
*) Disclaimer: I am not a regular chess player. I have reviewed various articles in online chess journals related to the issue of performance differences in chess, both of human players and chess computers, to try to understand the characteristics of under-performing players. I would be happy to receive comments to my bullet list above from experienced chess players.
When I first developed the argumentation in this blog post, I must admit that I said to myself: This is all good, and the above seems like a plausible basis for a list of good practices for strategic decision-making in business, but where are the really novel insights? Further thinking about the analogy between chess and strategic decision-making made me conclude as follows: i) If we accept the analogy, we will stop this useless bunk of strategic planning that many companies go through every year, and instead focus on getting the really big strategic bets right. Not only that, we will learn to appreciate that there is generally no single optimum strategy in business, and we will insert a solid portion of hard-nosed business discipline into our strategic decision-making. ii) On the other hand, there are good arguments why chess is not a good analogy for strategic decision-making (in addition to what has already been indicated above); in particular, the set of allowable moves is much larger in business, and with plenty of scope for creativity, and winning is not about avoiding mistakes, but about making the singular spectacular, but robust strategic move. iii) Despite what the McKinsey Global Institute is saying about the automation of managerial work, specifically that the economic impact of automating managerial work in 2025 will amount to USD 0,8-1,1 trillion, some of which will presumably be in areas that we call strategic decision-making, it is hard to see how to construct a business strategy computer with the same relative performance (relative to its human CEO peers) as a chess computer (relative to grand masters). I may endeavor to get back to this issue in a later blog post.