Swiss system: match distribution problem!

[Sakhura]

I think the same. I played 4 games in a tournement against the same players. It's less funny ( sorry for m'y Bad english )

[Janelle2]

All was said. Same constatation.

[Keldeofus]

Yes I agree with this, it's less funny vs the same player

[Al1fr78]

same here !

[Pinovitch]

Encore un exemple : https://boardgamearena.com/tournament?id=112487

Match 20 je rencontre un adversaire déjà rencontré en match 3
Match 18 je rencontre un adversaire déjà rencontré au match 15
Match 17 je rencontre un adversaire déjà rencontré au match 7
Match 16 je rencontre un adversaire déjà rencontré au match 14
Match 13 je rencontre un adversaire déjà rencontré au match 1
Match 11 et 12 sont les mêmes

Impressionnant comme l'option "Try to prevent same players to be opponents twice" ne fonctionne pas !

[NCFCRulz]

Unfortuantly they don't care about improving tournaments so this is unlikely to be fixed

[SwHawk]

Unfortuantly they don't care about improving tournaments so this is unlikely to be fixed

Before trying to improve the tournaments, perhaps you could blame the player who created the tournament, instead of the system, which actually does what's stated : TRYING to create games with different opponents.

That being said:

Encore un exemple : https://boardgamearena.com/tournament?id=112487

For this tournament, there were 23 players present, so 12 or 13 games each round, and there were 20 round, statistically, you were bound to spend some games with opponents you've already seen (and keep in mind that those were the parameters the tournament creator chose)... Also, you're supposed to being matched with played of the same level which reduces the pool of available opponents...

I think the same. I played 4 games in a tournement against the same players. It's less funny ( sorry for m'y Bad english )

If you refer to this tournament : https://boardgamearena.com/tournament?id=111788, same explanation as above, 20 rounds with 24 players, you were bound to have rematches.

All was said. Same constatation.

This tournament https://boardgamearena.com/tournament?id=112208 ? Same as above...

If I look at random posters in this thread, you all seem to play at the same tournaments, swiss system, 20ish players, 20 rounds... With those seetings, it's obvious you're going to be rematched with previous opponents for the reasons mentioned above. As I said, instead of blaming the system, blame the tournament creator, and the way they chose the tournament settings.

Going back to OP:

1v1 Championships : the players meet several times while it is possible to do otherwise.
Example : https://boardgamearena.com/tournament?id=113200 : I meet the same player in stage 15,19 and 20
Example : https://boardgamearena.com/tournament?id=112039 : the same, I meet the same player at stages 15,19 and 20

You created the tournament with 20 steps... There were 30 players around the table. The swiss system will match players of the same level together... Obviously there are going to be rematches. After the first step, the player pool will be divided in 2 or 3 groups depending on the game results (Winners, Loosers, Draws) then the system will try to match winners against each others, losers against each others, and draws against each other. And so on and so forth for each step. You want all players to be matched against everyone, then use a round robin tournament. You want a specific number of games, then swiss system is for you, but you should keep the number of games within reason with regards to the number of participants. With 20 to 30 players, I wouldn't advise more than 10 steps to avoid rematches, and even so, some are bound to happen. Basically, if you want to avoid rematches in a swiss system, don't use more steps than the maximum number of available games for each step. For example, if games are 4 players, and there are 24 players, that means that each step will produce 6 games, so I wouldn't advise more than 6 steps.

TL;DR The system is working correctly, but if the tournament creator/administrator can chose settings that will create rematches. Remember, the option states TRY to avoid, which is what it does, it tries. But if the player pool is too small, it doesn't really have any choice but to create rematches...

---
[EDIT]
To further elaborate, let's take one situation here:
A swiss tournament with 20 rounds, 20 players, and 2 players per game. Having 20 players play 2-player games means there are 190 possible games if all players went to play with everyone. With those tournament settings, there would be 200 games in all the tournament. At that point it's already obvious there are going to be rematches. But that would be the case if everyone could be matched with everyone, which is not how the swiss system works. It matches you against players of the same level (no taking ELO into account actually, only performance during the tournament), so for each following step, the available player pool dwindles... It's actually 20 for the first round, but then it starts to diminish to 10 for the second round (assuming you don't get matched if the score difference is 1 or more), and then actually 5, and so on... Widening the score difference only delays the matter, but after 5 or 6 rounds, the player pool is so small that there are going to be rematches, especially for the players who perform the best/the worst...

[EDIT 2]
I suggest reading through the wikipedia article about Swiss-system tournament, which states the following:

Assuming no drawn games, determining a clear winner (and, incidentally, a clear loser) would require the same number of rounds as that of a knockout tournament, which is the binary logarithm of the number of players rounded up. Thus three rounds can handle up to eight players, four rounds can handle up to sixteen players and so on. If fewer than this minimum number of rounds are played, two or more players could finish the tournament with a perfect score, having won all their games but never having faced each other. Due to the fact that players should meet each other at most once and pairings are chosen dependent on the results, there is a natural upper bound on the number of rounds of a Swiss-system tournament, which is equal to half of the number of players rounded up.[9] Should more than this number of rounds be played, the tournament might run into the situation that there is either no feasible round, or some players have to play each other a second time.

[sprockitz]

sorry swhawk, but that is not what's happening. the top players are paired with the bottom players, this is the opposite of how a swiss should work. the implementation option of preventing players from meeting multiple times is severely flawed.

[SwHawk]

sorry swhawk, but that is not what's happening. the top players are paired with the bottom players, this is the opposite of how a swiss should work. the implementation option of preventing players from meeting multiple times is severely flawed.

When you're saying top players, are you referring to ELO ? If so that's not taken into account by the matchmaking algorithm of the Swiss-system.
I'm currently doing the analysis of this Patchwork tournament which has been reported in this thread. I'm at step 5, and I can't see any inconstitencies in the matchmakings (meaning no top player matched against a bottom player, taking into account the score in the tournament, and not ELO, I haven't studied rematches so far as they shouldn't happen until around step 10). I'll post the results when I'm done.

Then again, with 20ish players, setting 20 round is going to force rematches as stated in the excerpt from the wikipedia article (that particular claim is actually backed by a scientific publication). There should be at most 10 steps if the administrator doesn't want rematches.

[EDIT1]
I have progressed to step 10, and so far I haven't seen inconstencies in matchmaking... Although at step 10, the algorithm is already allowing a score difference of +/- 3 to avoid rematches (maximum score difference is +/- 6 at the start of step 10) . As I said, I haven't studied rematches yet. This is already time consuming doing it by hand... Attached are my preliminary results up to step 10, feel free to look at it. I'll explain my method for filling out this sheet when all is done.
Link to file: https://nextcloud.swhawk.eu/s/5fkdqj8sX2K3cjF

[EDIT2]
I've now reached Step 13, which should be the maximum step of the tournament taking into account the total number of players, and the number of players per game... Two players are nearly impossible to beat (they already have a score difference of +4 with the players below them). Matchmaking algorithm is forced to resort to creating games with a score difference of +/- 5 to avoid rematches, while the maximum score difference at that step is 10. Leaderboard is already very close to the end of the tournament, so I'm wondering what's the point of continuing the tournament after this step... Basically when a game starts with a score difference of 3 or more it is almost certain that the winner will be the player who has the most points. So basically trying to force mixing players only confirms the already present difference.

As I said earlier, so far, the matchmaking algorithm is working as it should regardless of how you feel about it. Poor tournament settings are creating those conditions. Basically if you want each player to meet everyone else, then a roundrobin tournament is what you're looking for. But having those tournaments with those settings will end up in rematches everytime, and as the tournament proceeds further, the TRY to avoid same opponents twice part of the algorithm, it will try to matchmake players with a bigger score difference. So the system in itself is not flawed, but selecting poor settings will obviously create flaws...

[sprockitz]

I’m referring to difference within the Swiss itself nothing to do with Elo.

The problem I see is in multiplayer mode and trying to avoid playing same players twice option. Meeting the criteria strictly quickly becomes impossible, at which point the algorithm throws the bottom players to fill the remaining spots.

So in smaller tournaments this can become impossible by round 2. Take a 6p game with 24 participants. In R2 the 4 winners get paired with 2 last place players because ranks 5-24 all fail the no rematch criteria. So instead of going through the list again to look for best Swiss score match it simply throws the 2 lowest ranks in. By R3 this grows to the bottom 3-4 playing against the tournament leader.