The Tennis Draft
Imagine picking or influencing your position in a draw. Would you want the easiest first round match, just to get a win? A “winnable” first round match with a potentially easier second round? Do you just want to NOT face an opponent you’ve played too many times? Or maybe there is someone you’ve never had the chance to play? How would you decide?
The only place a draft is currently used in tennis, apart from fantasy sports platforms, is the “ITA Kickoff Weekend Draft”. NCAA Division I teams, if they are unseeded, get to select their position in a draw — with the actual matches taking place almost a half-year later. The order in which teams get to make their selections is determined by team rankings. This draft happens only once a year, team rosters change year to year, and the location where first round matches will be played is often a big factor in the decision making process, so directly applying this idea to singles/doubles tournaments doesn’t necessarily make a lot of sense. But there is something interesting here…
In many tournament management systems there is support for “avoidance rules”, meaning that the generation of the draw will attempt to separate, by at least one round, participants who are doubles partners or from the same country or state or city or zip code. There is sound reasoning here, though these capabilities are not often employed. None of these factors give participants any agency, however. The only agency players, parents, or coaches have in tournament planning ends with the registration process.
I’m sure that anyone who has been around tennis for more than a season has tales of draws which just don’t seem right. I’ve seen players who travel half way around the world only to play their doubles partner or someone from their own city in two tournaments back to back; players who face the first seed in the first round in sequential tournaments… I know parents who suspect tournament directors of manipulating the draws against them, or of there being known flaws in the software which no one is bothering to fix.
How could the idea of a draft be fairly applied in tournaments which happen every week? A strict rank-order selection process, as in the ITA draft, seems unfair to lower ranked or rated players who would get last pick week after week and mostly end up with un-winnable matches. The “luck of the draw” (when it is good luck) at least holds out the possibility of a first round match where it is possible to put some points on the board.
The Power of Preferences
Being a seed means something, is an achievement. The first principle in what I’m about to propose is that seeded participants don’t get to choose, they don’t get to specify their preferences, their positions are determined as they have always been determined. Every other participant, however, gets to study the draw and nominate, say, three positions (this is configurable) where they would like to play, in the order in which they would like their preferences to be considered.
Obviously if all remaining participants select the same positions the end result would be no different from a normal, mostly randomized draw. Only one participant can occupy a single position, after all. Preferences can only influence positioning to the extent that they are unique.
As a thought experiment, let’s take a draw of 32 singles participants. After the 8 seeds have been placed there are 24 open positions. Preferences could be solicited from all 24 remaining participants, or they could be grouped by ranking and preferences ‘resolved’ one group at a time. For this example, imagine creating 3 groups of 8. The first group to nominate their preferences can place themselves as far from seeds as possible, selecting amongst the 16 positions which will not face a seed in the first round, or they can elect to play a harder first round match (perhaps they know a seeded participant’s ranking is based on chasing points). The second group of 8 has fewer choices, but more information, and the final group of 8 has the most information and the fewest choices, but still choices, and even possibly very good choices.
Now imagine what this looks like in practice. Are preferences made online in specified blocks of time? Does the draft occur in public the night before the tournament (when people are often waiting hours for the draw to be published). Do participants in the same preferences group get to share information about their choices?
Think it over. There are a number of possible variations on this theme. What are the pros and cons? How would you decide? If you’re inspired, please let me know your thoughts: email@example.com.
An algorithm to implement this is available as part of the Competition Factory. I will follow up with a more technical article, if there is interest, which details how the preferences are selected. I’m even thinking of an online tool that makes the selection process 100% transparent. If you know a tennis player or fan who likes code, send them here!