TD Tools: Writing a Tournament Schedule

This article is part of Leaguevine’s “TD Tuesdays” series and are being re-posted on Skyd following release on Leaguevine.

For the original articles: http://leaguevine.com/blog/tags/td-tuesdays/ or subscribe by email: http://bit.ly/q1bmoT

First things first – this isn’t going to be a format-manual kind of article. That would be very long, and there’s sure to be some particular constraint at your venue or with your number of teams that makes your case different… Instead, I’m just going to talk about some of the things that might help if you’re inexperienced at writing schedules. First, and most important:

The Golden Rule of Scheduling:

You WILL get it wrong

I’ve written hundreds of schedules, for big events and small, and the most crucial advice I can give is to get someone else to check it. Really check it, not just glance at it and assume you’ve got it vaguely right. No matter how careful you are, there’ll be an error in there. I still miss something every single time. It might be that a game is missing or played twice; it might be that the pools are seeded so as to have rematches in the Quarters; it might be that you moved some games around to avoid some other problem and now a team is playing two games at the same time.

It might be something tiny, like the order of games in a pool (usually best to play the most important games last) or the fact that you could rearrange it so that quarter-final opponents could watch each other. There are a million ways to make a complete mess of it, and also a million small improvements that could tidy up an already usable schedule. It’s like writing an essay – there’s always something else you could tweak, right up to the moment you hand it in.

An example from this year, which I particularly enjoyed: I put the women’s matches on the far pitches at a big tournament, and got complaints that it was too far from the toilets – girls can’t go in the woods so easily. There’s always something…

Schedules are complicated, and you cannot keep the whole thing in your head; when you make a change, it’s very hard to go through and check that you didn’t cause another problem, because it’s all so familiar already and checking is boring. Get someone to look at it with fresh eyes.

Games and game-breaks

Different countries accept different schedules. In the UK, we run schedules that mainland Europe wouldn’t consider; USAU run schedules that we wouldn’t touch with a 30-foot pole. The UPA format manual, for example, might have you playing in a pool of seven, then a bracket with as many as 3 games to finish – 9 in a weekend. The player base is perhaps used to that, and will bring huge squads that can cope with the demands of playing maybe 4 times in 5 slots; at UK tournaments, or indeed at fun tournaments with smaller squads, that sort of schedule is not going to be popular – we have an absolute horror of 3-in-a-row at official tournaments. In the UK we always try to play 3 games per day, 2 of them back-to-back (so that you only warm up twice); in much of Europe, playing even 2 in a row would be considered a shockingly bad schedule. So I guess a big thing to think about is your intended audience – squads of 20 or squads of 8? Athletes or drunks?

The type of schedule you run depends on the type of event – at fun tournaments it’s crucial that everyone gets a similar number of games, whereas at a regional event it might be more important to qualify the correct 3 teams, and who cares if the guys who got knocked out on saturday just go home? Again, I can only speak for the UK, and say that any schedule in which the busiest team would play more than 2 more games than the least busy team would be no good to us. Whenever possible, we try to write so that there is no more than one game difference between the team who plays most and the team who plays least. You all pay the same entry fee, so you should get the same number of games (within the constraints of funny numbers of teams or additional qualification games).

Basic advice

You cannot write a fair schedule. All you can do is decide where to compromise. Anyone who’s seen a full round-robin, like for example in English football, will know that teams rest players for certain games – so even the league is not completely fair. The order of matches matters. And it’s clear that any schedule where you don’t play every opponent is open to unfairness in seeding. There is no such thing as a fair schedule. Give up on that idea now. Constraints (such as a maximum number of games without exhausting players, or maximum number of fields or time-slots, or horrible odd numbers of teams) merely add to the unfairness that is already somewhere in there. But here’s a couple of things that might help a little…

First off, unless you’re running something complicated (like an event that qualifies a certain number of teams for another event – and let’s face it, if you are part of a bigger championship there’ll probably be scheduling help available anyway) then the first thing you look at is how many games you want people to play. Set a maximum and a minimum, and then choose your pool sizes, number of crossover rounds, and brackets to meet that number. It’s a non-trivial task to fit a fair schedule to the right number of games, but it’s always a far better idea than wasting your time inventing fabulously fair schedules with multiple crossovers and power-pools and then realizing you’ll need until next Wednesday to play all the games.

Deal with teams in multiples of 8 wherever possible, and multiples of 4 at worst. If you’ve got funny numbers, 99% of the time you’re better off pretending that you’ve got a multiple of 4 and putting byes in the schedule. We’ve tried a whole bunch of times to write schedules with clever bits where pools of 3 go into power-pools of 3 then a modified bracket etc… it almost never turns out well, and I don’t think we’ve ever really used one of those schedules at an actual tournament. They lead to things like rested teams playing unrested teams, people having three games off in a row, fields lying empty, and all sorts of weird stuff. That may be fine for a qualifier where the most important thing is simply to make sure that the best x teams qualify, but it won’t wash at an ordinary event. If you want to finish with neat brackets, start off simple, in 4s and 8s.

If you put in any form of crossovers, triple check what will happen in the next matches. Seeding the pools is non-trivial if you want to avoid rematches later on.

Remember that odd-numbered pools eat up pitches – for example, 2 pools of 4 (8 teams) can be played on 2 pitches in a day (6 time slots); 1 pool of 5 also requires 2 pitches all day (6 time slots again – 5 slots if you’re prepared to make some people play 4 games in a row). It still frustrates me, but that’s just the way it is. Thinking of nice 3, 5 or 7 team pools in your head is no use until you actually sit down and squeeze it onto your pitches – more often than you expect, it won’t fit.

Anyway…

I could go on. I could write about 50,000 words on the intricacies of scheduling – for example, the UPA format manual is a fantastic document that tries to cope with any number of teams, and it’s looong; and even then it doesn’t come close to covering all the possible situations that might apply at an event (e.g. not enough pitches, constraints on back-to-back games, the team from far away can’t start before midday) and doesn’t touch the finesse parts of the schedule itself (as opposed to the format) like making sure that back-to-back games are not at opposite ends of the venue, making sure there’s a decent lunch break for every team, making sure that the girls are near the toilets…

All I can say to you here is keep it simple, only accept entries in 4s where possible, and if it gets complicated, find an expert. Offer to pay the guy who wrote that great schedule at that other tournament you went to, even – you’ll get no thanks for writing a good schedule, but the abuse you get for a bad one means it’s worth doing what you can to get it right. A genuinely well-written schedule is of real value to every player who shows up, even if they only notice it when it’s gone wrong – so if you need to pay someone to make sure it’s done well, I’d say this isn’t a place to be scared of spending a few dollars. A good schedule might not make people come back next year, but a bad one might well put them off.

Benji