Parallel Universes: the Chess Player’s Fear of the Simul
In this article county match captain, David Jenkins, offers some reflections on simultaneous displays with two illustrative games.
Simultaneous chess, universally referred to as the “simul”, is a particularly entertaining and informative kind of chess, having as its principal benefit that the opposition get the chance to play against and perhaps even momentarily disconcert a master player. It is quite exciting for the headstrong patzer or average club player just to be one-to-one with an exceptionally strong player and see the whites of his eyes as you line him up in the cross hairs. Anyway, that is how it feels, but of course “one to one” is not strictly true: Andrew will be taking on twenty players at the same time. How exceptionally strong players manage this feat will be an early consideration in this article, which aspires, no less, to offer advice to those at the lower end of the chess food chain who fancy having a go at some serious opposition in a simul.
The inexperienced horde player will fear being obliterated in a few moves, but will have fun, be protected by low expectations, and hope for a surprise. The fear for the master is less substantial but just as real, for what is to count as a decent result? A top GM would expect to survive a simul against average club players with barely a single loss or draw to disturb his winning sequence. There is an essential chutzpah of giving a simultaneous display against multiple opponents and the intrepid exponent will be intent on demonstrating the validity of his acceptance of the task.
What are the conditions under which the simul is played, and can knowledge of its quirks assist the weaker player in pulling off a surprise? This takes us into an unexplored backwater of chess psychology, for there are non-trivial differences between simultaneous play and standard chess. Some commentators have argued that the arrival of chess-supporting software is likely to kill the simul, as you can now play a silicon opponent of any putative grade level in the comfort of your own home and on any evening of your choice. But I do not think so. Chess is a deeply inter-personal game and the often arrogant and dismissive pseudo-Germanic banter on voice-over Fritz is no substitute for watching real egos crumble across the board.
One might enquire what supportive literature is available in preparing an article seeking to offer tentative advice on how to handle the unusual features of the simul. Weirdly, according to Chess and Bridge, there is not a single text in existence on how to play in a simul, so we are bravely breaking new ground. In a simultaneous match the visiting master usually claims the privilege of playing White, although interestingly GM Tony Miles was prone to waive that privilege and varied the colours. The “psych-ops” (military jargon for psychological operations) of the simul can begin early. Some GMs giving simuls remain faithful to their usual opening repertoire and showers out a whole flurry of d4s or Nf3s at manic pace, many of which the opponent needs to j’adoube into the centre of the intended square. This is presumably calculated to have two effects on the startled opponents; it signals that the current repertoire of opening theory has been internalized to the extent that it can be played as if by a zombie on automatic pilot, but it also hints that during the opening phases he will be round again before you have even thought of your initial reply.
This psychological tactic exactly matches the conditions under which a simul is played, since as Hooper and Wylde’s The Oxford Companion to Chess puts it, “the master walks around the inside of an area, surrounded by tables and chess sets, with the opponents on the outside. Each participant must move when the master arrives at the board, neither sooner nor later”. This exerts a subtle and novel kind of pressure, forcing opponents to keep a “candidate move” in the head while continuing to muse and/or calculate, as well as simultaneously charting the progress of the master around the tables in their peripheral vision. Add to this mixture that the master giving a simul speeds up as opponents crumble to defeat, and it easy to see that this is chess, but not as it is usually played. This defending of the squared tables, too: does it not remind you of how a single heavily armed Cowboy with rotating barrels might see off the attacking hordes of American Indians?
A parallel psych-op has been practised by GM Jon Speelman in simultaneous displays. Jon, in contrast to the immaculate concierge’s garb of Raymond Keene, will have turned up in a sartorial style somewhere between smart casual and professorial scruff, his luxurious and tousled hair crackling with static electricity. He will appear laid back, even languid, and have offered a courteous, rambling and often amusing account of his expectations (“Don’t try moves on the board before I get around to you, in case we get confused about what the position is”). Then, as if a switch in his brain had been triggered by remote control, he will have been off like a jack rabbit, each successive board subjected to almost as many opening moves as are legally permitted (a3, a4, b3, b4, Na3, Nc3, c3, c4, d3, d4, e3, e4, f3, f4, Nf3, Nh3, g3, g4, h3, h4; they all might come tumbling out, taking the majority of the opponents instantly out of their books. The semiotics of this gesture can plausibly be decoded into two dispiriting messages, that Jon Speelman is indifferent to whatever path his game with you initially takes since he does not think it will impact on his winning chances, and that he suspects that your average club player’s opening theory may operate as no more than emperor’s clothing to keep you warm until the cold blast of thinking time arrives in the God’s own preferred battle ground, the middle game.
The first condition of simultaneous chess that the club player needs to adapt to is the physical absence of the opponent. Like Halley’s Comet, the master’s ice cold brain has taken off on some celestial orbit, but it is not possible, as it is with Halley’s Comet, to estimate accurately the time of return. The subsequent withdrawal and absence (death?) of the God who initiated the big bang at which all the games were initiated cannot fail to carry disturbing theological overtones and must be a subliminal source of ontological insecurity, with your pain at your deteriorating position lacking even the intimacy of torture. Some address the absent and assumed-to-be-uncaring master in a parody of prayer and supplication, or spend their time exchanging furtive asides with any gathering spectators in the language of the lyrics of Road Runner (“What do you think of that, you guys?”).
At the heart of the dilemma of the simul is the gulf between real time and psychological time, which is a recurring theme in literature as well as chess. The seventeenth century metaphysical poet John Donne’s The Sonne Rising despite the punctuality of the light-bearing “busy old fool” declares that love:
“…no season knows, nor clime
Nor hours, days, months, which are the rags of time.”
Even more evocative of chess is the last soliloquy of Dr Faustus, the patron sinner of time trouble:
“O lente, lente currite, noctis equi!”
[“O run more slowly ye horses of the night!”]
The horses, unheeding, continue to pull the sun across the heavens at their customary steady gallop; the devil comes and the flag falls.
Psychological time in simultaneous chess is equally a product of the position; the worse your chess dilemma, the faster it goes. Yet in a simul you are denied one of the standard resources of normal play, your right to give a disproportionate amount of your time to a difficult move.
Another seemingly reasonable line of inquiry that might be contemplated by club players wishing to enter the world of the simul is to ask how expert chess players are able to do it in the first place. What is known about the problems of playing multiple opponents and can this knowledge offer any prospect of payoff by using it as a basis for inferring plausible strategies? For example, should one get into tactical complexities on the grounds that the master will be short of calculating time, or follow the advice that used to given to human beings about to take on the emerging silicon monsters of the chess board (David Levy’s now discredited “do nothing but do it well”)? The chess literature is hugely and depressingly silent on these matters, although The Oxford Companion to Chess somewhat misleadingly suggests that in a simul as a general rule, “the master will… play simply, seizing on errors but confident that his superior endgame technique, if needed, will see him through”. This was true of neither of the featured games below.
Interestingly, the reticence of the chess literature on precisely how in practical terms master players are able to cope with the multiple simultaneous processing of the simul is not replicated in the literature on experimental psychology. The processes involved in playing chess have, for obvious reasons, attracted the attention of two further groups of people whose interest in chess is instrumental to other purposes. These two are the academic specialists operating in the fields of artificial intelligence and cognitive psychology. I am minded to take a look at artificial intelligence and chess in a future article (after all, British chess champion Sir Stuart Milner-Barry still remains one of the unsung heroes of Bletchley Park) but for the moment the task is to look at what cognitive psychology has to say about the simul.
Fernand Gobet, the principal investigator for Studies in Chess Expertise at the University of Nottingham’s ESRC Centre for Research in Development, Instruction and Teaching some time ago developed and tested theoretical models for chess expertise and sought experimental and empirical support for them. The focus was on how memory, perception and attention processes meld in highly skilled performance. Although the literature on high level chess performance is divided on whether GM level is defined by superior calculating ability, the current orthodoxy is the canonical conclusion of Groningen University’s Adriaan de Groot that “it is unequivocal that depth of calculation cannot be the prime distinguishing characteristic between the grandmaster and the expert player”, although Gobet’s attempt to replicate this result suggested that across the full spectrum, the stronger the player, the greater the depth of analysis.
Towards the top of the chess food chain, it is more commonly asserted, outstanding results are more to do with what Chase and Simon in “The Mind’s Eye in Chess” (W. G. Chase (1973) Visual Information Processing (1973)call “chunking theory”, in effect a form of pattern recognition. Those interested in seeing how Gobet has extended the model and how it might be related indirectly to skill at playing simuls, might take a look at Gobet and Simon ‘s Templates for chess memory: a mechanism for recalling several boards” (1996) in Cognitive Psychology 31.
Enough of this serious stuff. What are its implications? The first is that the master in a simul may not be instantly calculating “candidate moves” in a speeded up parody of Alexander Kotov’s Think Like a Grandmaster so much as using memory and judgment. Although chunking theory was developed to explain how strong players can recall a number of boards following a brief scan (answer: they remember chunks as in “the king side position after a bishop fianchetto and castling”, but with better players perceiving more complex chunks) it probably covers how a GM or IM in a simul will re-recall and re-invent a position on returning to it. There are two potential points of weakness in the GM’s instant practical judgment in response to your hot-from-the-press move. You have a slight hope if the import of the move crosses the boundaries between two or more chunks, and another sliver of opportunity if your move has a covert secondary purpose at some distance from the current action. Expert chess players do not often make mistakes in simuls, but if they do they are more likely to be of this kind.
Another reassuring thought, now the performance of master players in simultaneous games can be put through the rigours of subsequent analysis by computers, is that at least it has become possible to pinpoint inaccuracies and dent their invincibility. I trawled the two games reported here though Fritz 7 and neither of the GMs escaped Fritz’s wagging finger. Indeed, both Mikhail Tal and Raymond Keene managed to earn themselves the dread double question mark in our featured games, although for different reasons.
The two games featured here, both from simuls, are annotated with some help from Fritz 7. The first game is my own rather fortunate effort against GM Raymond Keene in a simul in support of St George’s Hospital and the second Colin Searle (Stratford and Warwickshire) impudently essaying the Latvian Gambit against the most famous Latvian of all, Mikhail Tal. This game should be of especial interest to Falmouth’s Grant Healey, the only other county player (apart from me) who to my knowledge plays the Latvian seriously.
Ray Keene v David Jenkins
Mikhael Tal v Colin Searle
David R Jenkins
18th September, 2014