Like many magic players all over the world, I spend most of my weekends travelling to play in competitive tournaments both small and large. Whatever level of tournament I end up playing in, whether it’s a 20 player Grand Prix Trial at a local store or 3,000 man Grand Prix, the structure of the tournament is pretty much the same. Almost every tournament is run using a modified swiss pairing system followed by a top cut. Using the same system for every level of play has huge advantages for the players. No matter how far we travel we arrive understanding how the system works and what sort of record we will need for success. Experienced players become adept at reading the between round standings and calculating whether or not they can safely intentionally draw or whether their  count back is likely to hold up.

The level of familiarity we all have with system make it comfortable and predictable. However, lately I have been questioning whether or not the status quo is really good enough. Is the current system fit for purpose? Whilst many of the players I have discussed this with have taken the view that “if it ain’t broke, don’t fix it”, several felt that it was at least worth exploring other possibilities or finding ways to tweak the system to improve it.

How It Works

Before digging in to possible changes it is worth looking at exactly how the current system works. The majority of Magic tournaments are run using a modified Swiss pairing system. This system is based upon one originally devised for Chess tournaments as far back as the 1890’s. The main purpose of Swiss pairing is to allow for large multi-round tournaments that enable players to continue playing rather than being knocked out after 1 or 2 rounds. The size of these tournaments makes it infeasible for each player to play every possible opponent, the Swiss system instead seeks to pair each competitor against progressively more challenging opponents until the best players have been identified.

In each round of a Magic tournament (except the last round of Swiss) the system pairs each player against a random opponent from amongst those who have the same number of points. Three points are awarded for a win and one for draw. As such your opponent is entirely random in the first round, whilst the second round will pair each of the first round’s winners against each other. If there are an odd number of players on any given point total, one of the players will randomly be paired down against a player on the next highest point total. If there is an odd number of players in the tournament one player from amongst the players with the lowest point total is awarded a bye, and automatically wins the round 2-0. Where possible no player can be awarded a second bye this way.

No player may be paired against the same opponent twice in the tournament, so if the only players on any points total have already played one another they will be paired against a random opponent from those with the next highest point total.

In the final round of Swiss pairings each player is paired according to their current standings. The player in first position plays the player in second place, and so on. If this would pair a player against somebody they have already played they will instead be paired against the next highest placed player they have not already faced. After this round most tournaments will then have a top cut, to either the top 4 or top 8 players.

Many Magic tournaments are quite large, and even in smaller tournaments it is not unusual for several players to be tied on points at the end of the Swiss rounds. When this is the case various tiebreakers are used to decide where players finish. The main tiebreaker is Opponents Match Win Percentage. This looks at all of your opponents during the tournament and calculates the percentage of matches they won during the entirety. To prevent very poor performances from unduly damaging a player’s tiebreakers, the lowest any one players match win percentage can fall for this calculation is 33.3%. The concept behind this tiebreaker also comes from chess, where it is known as Strength of Schedule. The idea is that the player who has faced the more challenging set of opponents during the tournament should finish in a higher position.

The other tiebreakers are your own Game Win Percentage, rewarding a player for winning 2-0 and losing 2-1, followed by your Opponents Game Win Percentage. Both of these statistics never fall below 33.3%.

So, What Is Wrong With It

Magic is a very different game to Chess, and whilst the system has been adapted to take account of the differences between the two, these changes are not quite enough. The concept of Strength of Schedule works very well in Chess, the game is entirely skill based and it is far easier to measure relative player ability. We only have to look at the number of words dedicated to debates about who is the best Magic player in the world, or who deserves election to the Hall of Fame to know that it is very hard to measure player ability in our favourite game. Whilst a chess player can be expected to beat a weaker player in the vast majority of games, this is not always the case in magic. As such, Strength of Schedule does not translate particularly well to Magic.

Whilst every match we play is subject to a degree of variance, we, the player are able to play our best in an attempt to overcome this. However, we have no control upon the performance of our various opponents in their other matches during the tournament. It can be quite depressing going in to the last round reliant on our previous opponent’s performance to ensure we manage to make the top cut. This issue is further exacerbated in tournaments where an odd number of players or the number of games drawn cause a player to be “paired down” against an opponent on a lower point total. As the player on the lower point total has a lower match win percentage our tiebreakers become worse every time we are paired down.

This becomes worse during the final round as if we have been paired down we are likely to have lower tiebreakers than those on the same points and thus a lower position in the standings, this increases the chance that we will be paired down again further hurting our tiebreakers. This can feel particularly unfair when the initial pair down was random. This effect becomes more pronounced in smaller tournaments as there are fewer rounds and each opponent has a greater impact upon our tiebreakers.

So How Do We Improve it

I have three suggestions which I believe would improve the Swiss system in Magic

1) The primary tiebreaker should be our own Game Win Percentage.

This would allow the player more control over their own position in the final standings, and reward players for their performance in each game of each match. Whilst seemingly a small change this would have a significant impact upon the standings at the end of many tournaments.

2) The pairing computer should assign which player goes first in each match. The secondary tiebreaker should be Wins Against The Play. Whilst I have argued that Magic is a very different game to Chess, the latter has had many years to develop and perfect its pairing system. Whilst being the player assigned White in Chess is a far greater advantage than going first in Magic, it is generally recognised that the advantage in magic is tangible. In Chess the majority of Swiss systems used assign a player to a colour, and then attempt to pair the players in such a fashion that they rotate between going first and second from round to round. It prevents any player from going first for three consecutive rounds. A similar system in magic would lower the variance caused by relying on dice rolls or other random methods of choosing the player who goes first, it also gives us another tiebreaker that is both measurable and also rewards a player for their successes.

3) Rather than award 3 points for a win and 1 point for a draw, we award points only for wins.

Drawn games award 0 points to each player. This may seem like an extreme suggestion, however it has several positive impacts upon the tournament structure. Firstly, it limits the number of times a player will play against an opponent on a different point total. It will also lower the number of games that end as draws, as it effectively eliminates intentional draws and also will likely result in more players conceding games that they cannot win. It takes away an incentive for players to play slowly in game 3 when they begin to believe that they cannot win, which should make matches more pleasant and make judges lives a little bit easier. Furthermore, the lack of reward for intentional draws will cause more players at the top of the standings to play their final rounds, which will make for a better spectacle in live streamed tournaments. Viewer experience is becoming more and more important, and having the best players and decks on camera in the last round of Swiss can only improve this.

The changes I have suggested range from the minor to the radical, and by no means are they the only options available. If you have any thoughts about my suggestions, disagree entirely, or have some great alternatives to propose, please comment below or tweet me @TTMalfico