keekerdc.com

essays and pithy thoughts

Disqualification for rational behavior

in down is the new up

And now, some stuff that will seem horribly outdated about the Olympics that I just didn’t have the time to write about at the time, but in actuality is a lesson that all tournament organizers should heed.

The story dominating any coverage which badminton received at the London games was focused on the efforts of several teams to throw their final match of the group stage. These several teams were summarily ejected from the competition, and one pair even retired from the sport over the mess. Most of the musings on this topic, both from members of the press and the bottom half of the internet, served the purpose of lambasting these teams for their lack of sportsmanship and a failure to uphold the Olympic spirit, as well as cheating the paying spectators out of the sporting spectacle they paid for.

This commentary suggests a very large misunderstanding of the nature of spectator sports, and just how large and important a role the schedule itself, as well as the group of people who writes it, takes in keeping a tournament ‘on the rails.’

Defining victory

Yes, it’s generally expected that two badminton teams entering the court would both be looking to achieve victory. This is where most commenters’ analysis on the subject ends, stating that when they pay to see a sporting event, they expect to see two combatants go all-out for their entertainment. Such analysis is wrong, in that entertainment of crowds is not actually the main purpose of sports, but is actually a side effect of some sports which contributes to the development of professional scenes. Such analysis is also incomplete as it ignores the larger storyline of a team or player as they make their way through a competition.

There’s always a larger context to consider. Sometimes the context makes this goal of winning every match less absolute. Not always will two teams be completely bound to the goal of winning every indivisible match they play. Matches played at a serious level are almost never played in isolation as one-offs; there’s a bigger goal that each match is only an incremental part of.

Take baseball’s regular season, for instance. Teams play 162 games every season. To think that every team is going to engage in an all-out last-ditch effort to win every game is folly. Teams would be toast halfway through the season engaging in such behavior. Sometimes you need to rest players, or players get injured; sometimes as a manager you need to make lineup changes to send a message to a slacking player, or just shake things up a little; sometimes you’re down by a handful of runs after three innings, you’re already hitting your bullpen, and there’s little chance of you pulling out a win, so you sit your better players; sometimes you move pitchers in your starting rotation, putting your weakest starter against their strongest, trading a more likely loss in one game of a series for more likely wins in others. There’s situations in which a manager may have priorities ahead of winning the specific game unfolding in front of them. This is all reasonable and acceptable because the real goal is to win the season, not every individual game.

In Olympic competition, that context is defined by the medal podium. If you’re a competitor, not only are you trying to get there yourself, but you’re part of a larger effort on behalf of your country to get on that podium as many times as you can, across many sports. With all the heavy breathing over medal counts, as well as the very real benefits some countries attach to medal-winning performances at the games, this is an important context to understand. It’s especially important when trying to decide who is truly to blame for the mess in women’s badminton.

Disqualification for rational behavior

Four teams were disqualified from the tournament – an entire quarter of the field for this event. Tossed out were both teams from South Korea, one of two teams from China, and the team from Indonesia.

In typical fashion for these sorts of tournaments, things started with a group stage, where groups of four teams were created, and all teams in a group would play each other, resulting in three matches for each team, and the top half of each group would advance to the next stage. In groups A and C, we found a situation that should only happen under the longest of odds: these groups were already decided after the first two matches, and the third match scheduled were between the two teams we already knew were advancing.

You might ask, “why is this a problem?” It’s a problem because it directly encourages the very behavior that led to the disqualifications. Typically the purpose of a tournament structure is to make it obvious that winning is the more advantageous outcome in any given match, thus giving teams a direct incentive to win, thus creating more engaging and exciting matches. What happens here is the opposite.

These teams, both having won their first two matches, were already advancing to the next stage, making the third match in which they played each other utterly meaningless. But since performance in the group stage determined the seedings – where each team in placed – in the final elimination bracket, and since the players in question are thinking human beings and not just badminton playing automatons, these teams did the math on what a win or a loss in these matches would mean for not only themselves, but also their sister teams from the same country.

The Korean and Chinese teams – arguably the strongest teams in the tournament – found the following: winning their matches would get them matched up against their sister team earlier in the bracket than the final, meaning that one team would be knocking the other from medal contention, thus making a loss more desirable; and that losing would increase the odds that the above situation would happen for their rivals. The team from Indonesia could also help put the Koreans in that situation, letting them knock a stronger team out for them, by losing their third group stage match against one of the Korean pairs.

For these teams, losing the third match was the more desireable outcome. So, they went about trying to achieve it. This is why laying blame for this situation at the feet of these competitors is wrong: they were taking the obvious course of action which maximized their country’s chances of landing on the podium; it just so happened that losing this third group stage match was part of it.

This situation was made possible by a colossal blunder in the scheduling of the tournament.

So, you want to be a tournament director

The first rule of scheduling a group stage which wasn’t followed is that, given (x) groups in your group stage, the best (x) teams are each placed in a separate group. This didn’t appear to happen, as the Chinese and Korean teams were likely favorites coming into the competition, and one team from each of those countries were placed into Group A. This allows room for weaker teams to stage upsets, and presents the best opportunity for observably better teams to make it further into the tournament, both of which are good for everyone.

The second rule which wasn’t followed is that, given that you may have multiple teams from the same country or geographical area in a tournament, said teams are placed in opposite ends of the tournament structure so that they can’t possibly be placed in a position where they are eliminating each other until which point it is unavoidable. The World Cup follows this paradigm closely, where the host nation and the seven best teams in the world are in a pool for the first slot in each group, and the pools for the remaining three slots are divided by geographical regions.

The third rule which wasn’t followed is that the two teams perceived to be the strongest in each group should not play against each other in the last match of the group stage. This ensures several incentives that will keep all teams fighting throughout the group stage: it means the two strongest teams are more likely to put on a square fight, not wanting to count on victories in matches yet to be played in order to push them through to the next stage; and it means that the last group stage fixtures will include one match where a non-favorite has a chance to defeat a favorite and scrap their way into the finals. (That, incidentally, is what made the USA-Algeria match in the 2010 World Cup so important, and why that moment is so permanently struck into the memories of American soccer fans.)

Thus, the optimal way to schedule a group stage is to first rank all teams, put the top four in their own group, ensure that teams from the same country are on opposite sides of the tournament structure, and pit the top two teams in each group in the first fixture they play, or the middle one if absolutely necessary.

The organizers of this Olympic tournament did none of these things, and they’re ultimately to blame for the disqualifications they deemed necessary to restore order to the tournament they so spectaularly botched.