No! Not the Swiss System again!

Every year this debate rages within the English pétanque community. It’s started again with the decision of the EPA to use a 5-round Swiss System (‘Swiss’) format in the forthcoming World Singles Championship Qualifier. What is the debate all about?

Swiss, as its many detractors say, is not exactly used by many sports (Chess, Bridge and Badminton are the most common), yet seems to be ingrained in international level pétanque competitions.

The system itself was first used in a chess tournament in Zurich back in 1895, which is how it earned its name.

Swiss – the basics

In a Swiss tournament, players are never eliminated. Instead, teams or players are paired in each round against an opponent that has the same (or a very similar) number of wins as them so far in the tournament.

After the first round, all participants proceed to the next round in which winners are pitted against winners, losers are pitted against losers, and so on. In subsequent rounds, each competitor faces an opponent with the same, or almost the same, cumulative score. No player or team is paired up with the same opponent twice, one of the major benefits of the system.

Depending on the number of entries (the system can cope with byes, although this is far from ideal as it can impact on a tie break – see below), it is possible to use Swiss to arrive at an outright winner without the need to go to knockout.

This will work where the binary logarithm applies – the same as any traditional knockout tournament. Three rounds can handle eight players, four rounds can handle sixteen players and five rounds can handle 32. If fewer than this minimum number of rounds are played, it can happen that two or more players finish the tournament with a perfect score, having won all their games but never having faced each other.

In the case of the forthcoming English World Singles Qualifier, there is a maximum entry of 64 players. This actually implies that there should be six rounds of Swiss to find an outright winner, although the EPA has decided that there will be 5 rounds to try and find 4 semi-finalists for a knockout. The drawback with this approach is that 5 rounds of Swiss will find 2 players with a perfect score of 5 wins, but then a further 10 players with 4 wins. The tie-break (see below) will have to be applied to find the best two players out of these 10.

Typically in a Swiss pétanque tournament in England, the first round is randomly drawn, so it can be the case that the stronger teams can play each other early on and effectively determine the outcome of the tournament in the very first game.

However, Swiss is used most frequently in English pétanque to arrive at the top 8 rankings (in the case of both the International Qualifiers and the Home Nations Qualifier) to then either move to a barrage/knockout format and or select the squad for the tournament.

At international level, the Swiss system is always used to determine who plays in the ‘A’ competition and who plays in the ‘B’ competition (known respectively as the Championship and the Nations Cup at the World Championships). It is always set up to work on the first day of the tournament with the second day being a seeded knockout based on Swiss rankings.

There is always an element of uncertainty in who you might draw in the following rounds even if you win as it appears that the computer programme generally used by the FIPJP and its affiliated organisations makes a random draw amongst the players or teams on similar wins.

More sophisticated approaches involve seeding where the draw is based on some predetermined ranking (like performance in the previous competition or some ongoing rankings table). Here, the top half is paired with the bottom half. For instance, if there are eight players or teams in a score group, number 1 is paired with number 5, number 2 is paired with number 6 and so on.

Without over complicating things, it is important to understand that there are in fact a number of different variants of Swiss. Believe it or not, there is the Danish System – where the restriction that the same players cannot meet twice is dropped. Here, the draw strictly follows the rankings after each round. But let’s not confuse things further!

Tie-break

When people complain about the Swiss System, what they are mostly unhappy about is the operation of the tiebreak that is often applied. As mentioned above, if the binary logarithm applies, then the system will always find an outright winner if you play the correct number of rounds.

However, when you are trying to find the top eight rankings in a qualifier, or (as is the case with the intended format of the English World Singles Qualifier) you do not play the required rounds to find the outright winner, then its time for the use of the dreaded Buchholz tie break.

This is what causes the most confusion and unhappiness in the pétanque community, largely because it is complicated to calculate and unless you fully understand what is going on, your natural instinct is to feel that there’s been some weird fix to the outcome. It is definitely quite disempowering as a player to await the results that usually spew out of a computer and are impossible to work out unless you have been noting down every result yourself and working things out as you go along.

So what is this Buchholz? Bruno Buccholz developed this tiebreak system for Chess in 1932. What it does in a situation of equal wins is to add up the number of wins of a player’s or team’s opponents. This effectively shows how comparatively strong the opponents were. Put simply, if the players or teams you have played have performed well in the competition then you must have had a tougher draw than other players and teams whose opponents did not perform so well.

What is often hard for players to accept is that the tiebreak has nothing to do whatsoever with the scores in the games that you played or your opponents played.

Taking the example of the World Singles Qualifier and those 10 teams that would be in a tiebreak on 4 wins out of 5 after five round (assuming 64 teams), you would not get any credit whatsoever if you had won four of those games 13-0 and lost the 5th 12-13. The determining factor is how many games your opponents won in their five rounds.

It gets more complicated as the Buchholz scores (or BHN as you will see it in the EPA competition results table) can be tied as well. That is when the Fine Buchholz score comes in (fBHN). Instead of the sum of your opponents’ wins, this is the sum of your opponents Buchholz scores. It is actually the sum of the wins of the opponents of your opponents.

We mentioned above that things could get complicated further when there are an uneven number of teams. This happened in this years English Men’s International Qualifiers where byes had to come in. The problem with byes in a Buchholz tiebreak is that the Buchholz score for a bye is zero. A bye ‘team’ cannot win a game. That means that if you are a loser in a round and unfortunate enough to draw a bye (it is randomly drawn not given to the worst performing team in the round – you can lose 0-13 in the first round and avoid it), then you are definitely placed at a disadvantage.

If you want to be an anorak, you can look further and investigate further tie break systems that can be used such as the Harkness System, where the wins of the worst and best performing opponents are discarded. Let’s not go there now!

Swiss v Leagues

The most often heard complaint from pétanque players who don’t like Swiss (although as we have considered above, it’s probably the tiebreak they don’t like), is that it’s all so much easier to understand if you play leagues (pools) and then knockout. Everybody knows how that works and it’s a familiar format in most sports.

Simplicity is not always the best though and there are some specific and real advantages to the Swiss System such as:

  • Each player/team is guaranteed to play in all the rounds;
  • No player/team plays the same opponent twice;
  • Many players/teams can be handled without an impractical number of rounds;
  • Competitive pressure is at its greatest as there is an imperative to win every game – Buchholz is not a factor if you win every game;
  • There is a certain tension gained in the tournament as you never know who your next opponents will be until the conclusion of each round;
  • There will always be a final ranking that shows the respective performance of all players/teams.

Aside from the complexity of the system and its tiebreak- enough for many to dismiss Swiss alone – there are disadvantages:

  • You have to await the conclusion of each round, meaning that the pace of the competition is dictated by the slowest game; and
  • You have to play every game and if you lose two games (where you are trying to find an outright winner rather than say a top 8) you will not be able to win and you may give up or lose motivation; in more extreme circumstances, you might just leave the competition all together (this has happened at an English Home Nations Qualifier in the past).

So where might a league and knockout system be better? It probably comes down to this:

  • It’s simple to understand – you know who you will be playing and there are no surprises with computers using complicated formulas;
  • You can lose a game in a league of 4 and still get through;
  • You can get a nice draw in a league if there is no seeding (but you can get a bad one!);
  • You can play games at the pace of the length of the game in the individual league, rather than the pace of the whole tournament;
  • The knockout that follows a league is straightforward and has greater tension as it is a ‘do or die’ match; and
  • You are not forced to play all the games at a tournament – if you get knocked out you can go home.

On the downside, leagues plus knockout can:

  • Lead to a perverse outcome where teams/players that perform worse in comparison with players/teams in another league can still get through;
  • Reward failure – you can afford to lose a game or even two in a league depending on the format;
  • Eliminate strong teams early on and make it easier for weaker teams because of a random draw pitching the best teams against each other (this has certainly happened at the EPA National Titles Weekend);
  • Lead to a situation where you play the same team twice;
  • Shortchange players/teams who get knocked out in the first round after the leagues – sometimes it’s a long way to travel for just four games;
  • Give no indication of the respective finishing positions of players/teams.

Conclusion

At the end of the day, if you want simplicity, don’t mind playing the same player/team twice, don’t care too much about the randomness of leagues or the respective rankings of players/teams in tournament then, sure, the league format plus knockout has got to be for you. It’s straightforward.

The berating of Swiss by many though fails to recognise the advantages it offers. For the top players they should be playing more of it, because that is exactly the system they will face if they achieve the highest levels in the sport.

Whatever you think, enjoy the sport and remember – if you win every game it makes not the blindest bit of difference what competition format you play!

 

9 replies
  1. Rager
    Rager says:

    Many years ago I used to play backgammo n competitively in the UK and Swiss was the norm for all comps. Everybody knew and understood the system. I never heard any complaints – in fact, I think everybody liked the fact that ”winners play winners” and the strongest players are always paired together. Each player had a national ranking, based on tournament results and you improved your ranking more if you beat a stronger player – the ranking depending on the score and the difference in the rankings of the 2 players.

    The difference, however, is that backgammon is an individual game, whereas petanque is normally a team game, so rankings are harder to impliment.

    However, one question about the description above: is it correct that scores don’t count, only the results of the other teams? This doesn’t seem right. I would have thought the results should depend on: games won, then points difference – the results of the other teams only coming into play if there is a tie after games won + points difference?

    Having now played a lot of petanque as well as backgammon, I have to confess, my gut instinct for smaller petanque events is for Leagues + KO – despite the ‘theoretical’ flaws. I think Swiss is a good system for larger events where rankings are required and is a good choice for an initial ranking comp followed by a KO.

    Basically, ”horses for courses” – every system has it’s pros and cons.

  2. martin
    martin says:

    No, the scores achieved in each round do not count for the purposes of a Buchholz tie break. BHN and fBHN only recognise games won and not any points achieved in those games.

    • Rager
      Rager says:

      Don’t you think Points Difference should come before the relative results of other teams? Otherwise, you could even apply the same ‘logic’ to games won: supposing Team A won 13-0 and Team B lost 12-13 (against different opponents). But Team A are World Champions and had an easy win against a novice team. Wheras Team B is also a novice team but were narrowly beaten by a strong team. You could argue that Team B performed better than Team A, despite the actual scores.

      • martin
        martin says:

        Ray, I always struggle a bit with these hypothetical examples. To try and answer a partial scenario, it is very unlikely that Team A would end up in a tie break with Team B as if they are World Champions they will probably win enough games anyway (I am assuming that the Swiss is being used to find say a top 8). If Team B end up in a tie break situation then if the team they played performed well, then they will get the credit for their opponent’s performance throughout the competition and that will be reflected in their Buchholz score. Usually the Swiss is set up with sufficient rounds to limit the need for the tie break (which is why I do not personally agree with the EPA’s apparent decision to play only 5 rounds of Swiss in a projected 64 player World Singles Qualifier next month as it will lead to two teams on 5 wins and 10 on four wins, with Buchholz required to find the top two out of the 10) and it is there to decide between teams that have lost games. For example, in a 5 round Swiss at Home Nations in England where there are 40 odd teams, the tie break is usually required to split those teams that have lost two games out of 5. In this year’s International Qualifiers it was to split teams that had lost 5 games out of the 10 played. This debate is not so much about Swiss, but the tiebreak system used in Swiss. Remember that the pure application of Swiss will always find an outright winner without a tiebreak if sufficient rounds are played.

          • martin
            martin says:

            It is a complicated system, Ray. People do not generally like things that they do not understand are which they think are complex for the sake of it. Certainly it’s way easier to have leagues and knockout or barrage and knockout. I do not advocate it for general use, but I do think it offers distinct advantages to top level competitions, especially where there are a large number of teams and you need to establish rankings as a precursor to knockout. The article I wrote also sets out a number of other advantages to the system.

  3. Rager
    Rager says:

    I’m still pondering whether Points Difference should be the first – or the last – factor in ranking teams on equal wins or whether the relative performances of the opponents is the best way. I would love to know the origins on the Swiss system for pétanque and how this was thought through.

    Martin, your homework is to produce an example that makes this clear 🙂

      • Rager
        Rager says:

        Thanks, Martin. How about a simple case, where Team has won their 1st game 13-0 and Team B has won their 1st game 13-12. Who did better? Does it depend on points difference or the relative results of the opponents?

        Is it a question of ‘logic’? My gut instinct is that the 13-0 team did better than the 13-12 team. Or do you argue: the 13-0 win was against a weak team, so isn’t worth as much, the 13-12 win was against a strong team, therfore counts more?

        Is it a mathematical proof or a question of judegement as to who performed better?
        I await your results.

Leave a Reply

Want to join the discussion?
Feel free to contribute!

Leave a Reply