Every year, the American Checkers Federation (ACF) hosts a series of tournaments that culminate in a Championship Match. About 56 people compete in the Nationals each year. This last year featured the incumbent champion, Cleveland native Alex Moiseyev, defeating Michele Borghetti of Livorno, Italy with a score of 7 wins, 6 losses, and 27 draws.
Some might find it strange that checkers tournaments still exist, as checkers is in fact a solved game. Checkers is vastly more complicated than, say, tic-tac-toe or Connect-4, but all three games are similar in that ‘perfect play’ sequences are now known. It happened for checkers in 2007. Computers had been working on the checkers problem since the 1980s, meticulously working out each sequence of moves from 39 trillion endgame positions, then verifying that any beginning position could wind up as one of those endgame positions given perfect play. There now exist computer algorithms that literally cannot lose to human players. (It turns out, incidentally, that perfect play by both sides forces a draw in checkers.)
So where do we go from here? And, also, why are there still checkers tournaments? Hasn’t humanity, in some sense, finally beaten the game for good? What’s even the point in playing, when no matter what clever and brilliant stratagem one unleashes, the counter is already known and recorded in some hard drive somewhere?
For the first question, one response is to move to a new game with more complicated play that still hasn’t been entirely solved. If checkers is finished, move to, for example, chess, a game whose complete solving by computers is expected to occur at around the same time as the heat death of the universe . ‘Regular’ chess (8x8 board, 64 squares, FIDE rules and regulations) still hasn’t actually been solved yet, even if some would argue that it’s getting close. There are still new openings and tactics being developed. People can’t even agree if either side has an advantage or if the game starts on an equal footing, or if perfect play would yield a white win, a black win, or a draw; all of which indicates a fair degree of uncertainty about what ‘perfect play’ would look like . Grandmasters still do win and lose against each other. So games in which there is still uncertainty about the outcome -- still room for a clever player to add something new to the books -- do exist. Then why do people play checkers?
After all, while playing checkers can be fun, and can help maintain friendships, and can keep the mind sharp, et cetera, the ostensible purpose of playing is to win. To beat other people on a level playing field, your wits against theirs. Essentially, to be better than them at the game. And for those who want to be ‘good’ at checkers, their objective is to be better than lots of people. Is there any metric in the game that doesn't, at some level, require comparing one's own ability against others? For aspiring champions, they must want, by definition, to be the best. But now anyone can only hope, at best, to tie the number one champion, a network of computers that ran for eighteen years. Solitary possession of the number one rank is, quite literally, impossible; and any victory against a human opponent may well pale in comparison to the knowledge of what would happen were one to challenge the perfect computer player. Alice may be better than Bob, but both are far inferior to Multivac.
A response might then be that, while computers know all four billion move sequences, people don’t. Most human beings aren’t capable of that sort of analysis and memory. Nor do most people play Deep Blue-esque checker supercomputers, but rather other people – and then it’s a game that can be won. Not everyone wants to be the best; some only want to test their mettle against another
‘But,’ the naysayer might add, ‘Can someone really take up a game, play it routinely, and not at some level want to be the best? And if you really don’t care about how well you play overall, why do you care if you win or lose anyway? And if that doesn’t matter to you even the least bit, why are you even playing? What satisfaction does it give you that just socializing wouldn’t?’.
Or it might be that, while perfection is impossible, maybe the pursuit of it is still worthwhile and engaging. One player may never beat a computer, but if she lasts forty moves in one game, and forty-five in another, that’s still progress – and quite good progress to boot, considering the opponent. ‘But if the ending’s always the same, how could that not get soul-crushing after a while? The futility of it all?’
Or perhaps it’s human pride. For the same reason that a bicyclist might not see a motorist’s time in crossing a course as something to compete against, a player that doesn’t use tools or computers might simply not care what a computer can do. That’s another division; and in the human division, there’s still improvement to be made.
I wasn’t sure at first why this interested me. After all, I don’t play checkers, and I’m a rank amateur at chess—nowhere near the level where I could use solved strategies. But I think that this is applicable to more than games. Automation and computer ability is ever increasing, and machines are increasingly able to do things better than humans. But humans don't always seem to want to exploit these capabilities.
To take one trivial example, a digital scale can be used to measure things much more precisely than the traditional analog balances that used to hang in supermarkets. Yet one still finds people using said analog balances in their homes and workplaces. In another case, I’m an amateur baker – I bake bread as a hobby. I could make better loaves (rounder, fluffier, richer) with a fancy bread machine, but I find myself preferring to make it by hand. I also play piano as a hobby, and I mostly play classical scores (as opposed to improvisational pieces). Player pianos and keyboards can, in some sense, play ‘better’ (or at least more accurately) than I. Based on the research I’ve seen, I think that robots and computers will be able to play as well as humans (including expressivity, emotion, and other features that are commonly thought to be exclusive to humans) within ten years. And yet I keep on playing...
So – questions for the audience. Do you do anything as a hobby or fun pastime that a machine could do ‘better’ in some sense? Why do you still do it?
 Of course, some would argue that chess itself may be slowly approaching ‘solved’ status, at least among the grandmasters (World Champion Bobby Fisher famously worried that chess was going to enter a ‘drawn death,’ where the greatest players would just draw each other all the time because they’d reached the limits of the game). This has in fact happened with some openings; the Poisoned-Pawn Variation defense to the Sicilian opening very often results in a drawn game, simply because the (supposedly) best moves after such a defense are well known to most professional players.
So what then? Fisher chose to make the rules still more complicated. He invented Chess960, chess with a randomized back row, a vastly more complicated game. Another World Champion with similar concerns, Jose Raul Capablanca, designed a 8x10 board and two new pieces that necessitated new strategies. A third World Champion, Emmanual Lasker, opted for a different tactic; he proposed changing the scoring so that draws could be weighted towards one side or another (i.e., if white is winning and stalemates black, white gets a larger fraction of a point than black). This would discourage draws and, hopefully, lead to more innovative play. Other ideas include giving ‘draw’ odds – draws are awarded to black as a win, and white gets a bonus on the timer to compensate – and reducing the amount of time for moves, to force play that isn’t as well thought out.
But, again, that chess is becoming solved is by no means settled upon. Lasker worried that it was… and Lasker’s World Championship was in 1894 (Capablanca's was in 1921, Fisher's in 1972). We’ve gone more than 100 years since then, and people are still winning and losing games, so perhaps these fears are a bit too soon. ↩
 For those who don’t play chess: each player has the same pieces and the same movement options to begin the game, but white always goes first. Statistically, about 56% of grandmaster-level games that aren’t drawn go to white, and 44% for black. Thus, many argue that white has an advantage. But some counter that having an extra move can occasionally be a major disadvantage in chess (this is called zugzwang), and that white has to reveal itself and its strategy before black, which has the option of picking a response rather than striking into the unknown. They argue that the white victory odds are due to psychology – those playing black don’t believe they’ll win, so they don’t – and most people not knowing how to play black (i.e., the great debate over whether black should attempt to equalize position first before attacking, or if should attack first and not worry about white’s ostensible first-move advantage).
Grandmaster Weaver Adams, whose title indicates substantial knowledge of chess, once wrote a book with the premise that white should be able to win with perfect play. This earned him widespread ridicule. Of course, without knowing whether black, white, or neither has an advantage, determining perfect play is not really possible. Grandmaster Andras Adorjan responded that it’s best to play as black, also earning him his share of mockery. For more information, see Wikipedia: First move advantage in chess. ↩
The Slacktiverse is a community blog. Content reflects the individual opinions of the contributors. We welcome disagreement in the comment threads, and invite anyone who wishes to present an alternative interpretation of a situation to write and submit a post.