Swiss Champs 2024

Hey Quentin,

First the facts:

  • 14 teams are registered
  • I calculate with 18min games and 4-5 mins break (average)
  • the double elim on sundays will take about 10 hours with 14 teams (26-27 games) —> this fills the whole sunday, start probably 8:30
  • 4 rounds on saturday should be feasible if we start around 9

Second, my opinion:
I think that the final ranking will be more balanced if there’s one single group, especially because groups of 7 are very small (with 4 rounds it’s almost a round robin within the group). Also, it’s more fun if everybody’s at the court at the same time and not just for the duration of their group games.
However, average waiting time with 14 teams in one group will be about 2.5 hours, which is very long. But in my opinion the two other points have more value than the last one.

With 5 rounds we could also think about a mixed solution (e.g. 2 groups with 3 swiss rounds each, then 2 swiss rounds between the 14 teams). But with only 4 rounds I think this is not a good solution.

We could also think about a split up on sunday (bottom 6 [or 4] “challenger” and top 8 [or 10] “slayer” —> 24-26 games). Each group plays a double elimination. This will reduce the waiting time on sunday. However, this is less of an issue during the elimination anyways…

So much on my thoughts on this. What do you guys think?

1 Like

@Mamar @coolquentin

did all the teams find a flinta?

about the format:

i think that for a national turney there are 2 priorities:

  1. ranking before elimination : as fair as possible
  2. games : as many as possible for each team (during elimination, if elimination is planned), not only for the top ones.

thus, i agree with Mario that for saturday ranking :

1 GROUP
( on saturday + perhaps continued on sunday morning)

offers clearly more precise ranking, especially compared to the option with 2 groups having 7 teams, which means that at every round 1 team will get a bye = a score results that represents nothing. so bringing the whole thing far away from fair ranking. also the separation on 2 groups allows not the face off between teams who should normally play one against the other if there was only 1 group.

but, 1 group of 14 teams and only 4 rounds is not enough,
because: 4 rounds is the lower limit, that only allows to find out the very top and very bottom of the group.
image

here the formula for the minimal number of rounds in function of the number of teams.

and we all know that the most important is to seed as finely as possible the mid fraction of the group before the elimination. so it has to be at least:

image
image

1 GROUP WITH 5 ROUNDS

(5 rounds x 7 games/round ( no byes, byes are shit ) x (18min games + 5 min gap))/60 =~13,5h

in general, more rounds → better ranking, but we have no time to add the 6th round.

  1. elimination…

maybe, as Mario said, the split bracket is an option,
because : since the early rounds the games are competitive ( less possible to predict the winner, not like seed 1 against seed 14 in classic double elimination)

BUT: if the number of rounds during seeding is not sufficient, the team that is ranked lower than it should be ( with 1 more round), has their chance to join top bracket is confiscated.
i d be ok with split bracket if in seeding there are minimum 6 rounds…

single elim: i m totally agaist it as always, unless during seeding there was a round robin. which is not so realistic with 14 teams and 18min/game. concretely single elim here would mean that 7 teams would only play 1 single game after seeding. NEIN for an annual nationnal meeting that also should target new players encouragement.

double elim: YES, we will be very tight on planning, yes. like always yes :D. if the switches between the games are fast, no reason we cannot finish

27games x (18 + 5 min) =~10,5h max.

image

sat 8 to 20 + sun 8 to 20h perfection.

of course, this is in case that games are not stopping at 5 goal difference and if everybody needs 5 min to get ready to play…
if just 1 min is spared in between each couple of games, 1h of total time is already saved, for ex.

for seeding, there s also the order of tiebreakers which is important for fair ranking cause the number of rounds is low, even with 5 rounds we risk to get fucked up situations for some mid teams.
to be discussed later when the global format is set i guess.

P.S: you can ask chaton gpt about the number of rounds in function of number of teams, or check wiki.
the simplest version is here below :slight_smile:

[[[[ moi:
salut chaton :). dans un groupe de 14 participants, combien de rounds suisses sont nécessaires au minimum? tu peux aussi me donner la formule pour le calcul du nombre de rounds en fonction du nombre des participants, stp?

ChatGPT
Salut ! :)
Dans un tournoi suisse, le nombre minimal de rounds nécessaires
pour s’assurer qu’il n’y ait qu’un seul gagnant invaincu
peut être déterminé par une formule mathématique. Cette formule est généralement basée sur le nombre de participants 𝑛

Formule pour le nombre de rounds en tournoi suisse
Le nombre minimum de rounds 𝑟
r pour un tournoi suisse peut être approximativement calculé par la formule suivante :

𝑟 ≈⌈log 2(𝑛)⌉
Où ⌈x⌉ représente la fonction plafond, qui arrondit x au nombre entier supérieur le plus proche.

Pour un groupe de 14 participants :
Calcule le logarithme en base 2 de 14 :log2(14)≈3.807
log 2 (14)≈3.807
Applique la fonction plafond pour obtenir le nombre entier supérieur le plus proche :
⌈3.807⌉=4
Ainsi, pour 14 participants, le nombre minimum de rounds nécessaires dans un tournoi suisse est de 4 rounds.

Explications supplémentaires
Pourquoi cette formule ? Le principe du tournoi suisse est d’éliminer progressivement les participants en ne les faisant jouer qu’une seule fois contre chaque adversaire possible. Le logarithme en base 2 donne une idée du nombre de divisions successives nécessaires pour réduire le groupe de participants à 1, sachant qu’à chaque round, le nombre de participants invaincus est approximativement divisé par deux.
Résultats précis : La formule
⌈log2(𝑛)⌉ donne une estimation minimale.
Parfois, selon le déroulement exact du tournoi et les tiebreakers nécessaires, un round supplémentaire peut être ajouté pour garantir un classement définitif.
J’espère que cela répond à ta question ! Si tu as besoin d’autres précisions, n’hésite pas.

moi:
comment on sait que 4 rounds ont été suffisants pour classer dans l ordre juste tous les 14 participants et pas seulement le gagnant?

ChatGPT
Excellente question ! En effet, la formule que je t’ai donnée ⌈log2(𝑛)⌉
est principalement utilisée pour déterminer le nombre minimal de rounds nécessaires pour garantir un seul gagnant invaincu dans un tournoi suisse. Toutefois, classer de manière précise tous les participants en fonction de leur performance peut nécessiter un nombre de rounds supplémentaires. Voici quelques points à considérer :

Classification complète des participants
Garantie d’un unique gagnant invaincu :
Avec ⌈log2(14)⌉=4 rounds, il est garanti qu’il y aura au moins un joueur invaincu, mais cela ne signifie pas nécessairement que tous les participants seront correctement classés dans l’ordre exact de leur performance relative.

Classement précis :

Pour un classement plus précis de tous les participants, le nombre de rounds doit être suffisant pour différencier les performances de manière significative. Une règle générale souvent utilisée est d’ajouter des rounds supplémentaires pour augmenter la précision du classement.
Une formule empirique courante pour un classement plus complet dans un tournoi suisse est :
𝑟≈log2(𝑛)+1
Cela compense les situations où des joueurs peuvent avoir des performances très similaires.
Application pour 14 participants
Nombre de rounds minimaux pour un classement plus précis :
En utilisant la formule améliorée
𝑟≈log2(𝑛)+1 r≈log⁡2(14) + 1≈4.807
​En arrondissant au nombre entier supérieur, on obtient 5 rounds.

                                                        !!!!!!!!!!!

Conclusion
Pour un classement plus précis de tous les 14 participants et non pas uniquement pour déterminer un gagnant invaincu, il serait recommandé de jouer au moins 5 rounds dans un tournoi suisse.

                                                        !!!!!!!!!!!!

Ainsi, bien que 4 rounds puissent suffire à identifier un gagnant invaincu, 5 rounds seraient plus appropriés pour classer les 14 participants de manière plus précise et fiable.

moi:
*et qu est ce qui se passe si on joue plus de rounds que la formule empirique *
log 2 (n)+1 suggère? si par exemple on joue 6 rounds pour 14 participants?

ChatGPT
Si tu décides de jouer plus de rounds que le nombre suggéré par la formule empirique
log2(𝑛)+1
comme 6 rounds pour 14 participants, plusieurs effets peuvent se produire :

  1. Précision accrue du classement
    Différenciation des joueurs : En jouant plus de rounds, il devient possible de distinguer plus finement les compétences et les performances des joueurs. Les scores finaux seront plus variés, permettant une hiérarchisation plus précise des participants.
    Réduction des égalités : Avec plus de rounds, il est moins probable que de nombreux joueurs se retrouvent avec le même score, ce qui facilite le classement sans avoir à recourir à des tiebreakers complexes.
  2. Avantages pour le tournoi
    Fair-play : Plus de rounds offrent à chaque joueur plus d’opportunités pour compenser une mauvaise performance ou une mauvaise paire dans les premiers rounds.
    Stabilité du classement : Des performances exceptionnelles ou des erreurs peuvent être mieux équilibrées sur un plus grand nombre de rounds, menant à un classement final plus stable et juste.
  3. Considérations pratiques
    Temps et endurance : Chaque round supplémentaire demande du temps et de l’énergie, tant pour les joueurs que pour les organisateurs. Trop de rounds peuvent fatiguer les participants et affecter leurs performances.
    Logistique : Plus de rounds impliquent une organisation plus complexe, nécessitant plus de temps pour planifier et gérer le tournoi.
    Simulation des 6 rounds pour 14 participants
    Pour illustrer, imaginons un tournoi suisse de 14 participants avec 6 rounds :

Premiers rounds : Les premiers rounds permettront de séparer rapidement les joueurs les plus forts des moins performants.
Mid rounds : Les rounds intermédiaires affineront le classement, éliminant les joueurs invaincus et permettant aux joueurs ayant subi une défaite de se rattraper.
Rounds finaux : Les derniers rounds serviront à départager les joueurs ayant des scores similaires et à déterminer un classement final précis.
Conclusion
En jouant 6 rounds pour 14 participants :

Le classement sera probablement très précis, avec une bonne différenciation entre les niveaux de performance des joueurs.
Les égalités seront réduites, et il sera plus facile de déterminer des positions claires pour chacun.
Cependant, il faut s’assurer que les joueurs sont prêts pour la durée et l’intensité supplémentaires que cela implique.
En résumé, ajouter des rounds au-delà de la formule
log2(𝑛)+1 peut améliorer la précision du classement mais nécessite de prendre en compte la logistique et la capacité des participants à maintenir leur performance sur une durée plus longue. ]]]

Thanks for the input both of you.

So far my cold reaction :

  • 2:30 between games = dealbreaker
  • 8:00 to 20:00 on both Saturday and Sunday = dealbreaker and a bit unrealistic

One group : For me there’s more than just having the maximum games played, the quality of the day is important. 2:30 between game is really not ideal in my opinion (it’s awful).
I know having 1 bye is shitty and make you wait twice between two games but it happen 1 time during your group.
I don’t see how it’s unfair since every time will have this “free win” ?
Edit : well my bad, we need to play 7 rounds to all have a bye.

Double elim, mmm it’s more a 11h long thing since there’s not tie and so golden goal are long…

I’ll think a bit and come with a proposal too, we need to keep it realistic.

Edit 2: what was the format last year @Migcha ? 2 groups RR + single elim ?

Edit 3 :
So yes, maybe one group is the most fair solution after all + having enough rounds to keep it single elim.
Or two groups (one of 6 and one of 8 teams) like we have done one year, with the risk of having uneven group if one team drop.

Editedit : sorry for the brain vomit next time i’ll think more before writing…

@coolquentin I Will recheck what we did and come back to you. I could share all the documents directly to @Mamar if this is any help for him.

1 Like

Last year there were less teams… The two Lausanne teams fuck up everything :wink::smiling_face_with_three_hearts:

I think 5 rounds and double elim is quite unrealistic for the timing, even though I agree with you Elena that 4 rounds is very minimal!
Finishing very late on sunday is not very cool for people coming from far away… We could optimize the schedule mainly by starting a bit earlier on sunday morning. Then, as you guys said, optimising time between games is crucial, but it is hard to plan.

MY PROPOSALS:

  1. Plan 4 rounds and a double elim. IF we manage very fast changes between games and everyone’s down for it, let’s add a 5th round on saturday evening. But we cannot count on it beforehand.

  2. Just play swiss rounds for the whole weekend, probably like 8 to 9 rounds. This will give a quite precise ranking but I’d definitely miss the big adrenaline and tension which comes from elimination games

PS for your question Elena: I’m only waiting for news from Pololove if they got in touch with Dev. All the other teams are complete now.

From the Challonge i found here it was :
RR 10 teams
then SE 4 teams

Do you know if there’s lights and if we can play after 20:00 ? Like what is the maximum playtime on the court ?
@Joe

So maybe 7 (to 8, according to the light/schedule question) swiss rounds followed by SE top4?
We played 49 games last year. with 7 swiss rounds we’d have 49 games plus the 4 SE games (4 if there’s a game for the third place), which seems feasible to me

1 Like

Really one big group will be a pain in my opinion, let me digg what w’ve done during the past editions cause we did like 2 groups (like a 6 and 8 teams groups) then SE or DE.

I’ll do some maths !

A quick draft / proposal :

Saturday :
2 groups SR, 4 rounds (Group A 6 teams, Group B 8 teams)
start at 8:30 finish at 20:00

Sunday :

  • 2 groups SR, 2 rounds (TOP 6, BOTTOM 8)
  • 14 Teams SE
    start at 8:00 finish at 19:00

It’s not perfect but each teams plays min 4 games on SAT and 3 games on SUN.
The ranking within the TOP and BOTTOM group is not perfect but sufficient to seed the SE 14 teams. Could be tweaked / discussed more.

image

Yes please 2 groups !
Waiting 2,5 hours in between games will be awful and just gave contre performance :(
Also 8:00-20:00 is hard too

2 Likes

on this link i’ve tested other variants too :

This one could be quite realistic :
2 groups plays 5 SR rounds
4 rounds the SAT
1 Round the SUN
Then 14 SE
Finish before 17:00 (on paper…)
It’s VARIANTE 2 :

image

2 Likes

for me, its like Mario said: if enough rounds, then ok for single elim. 7 ( or 8) round+ SE with 4 teams. not bad.

this is the version, among all others proposed here, that seems to me the best compromise:

  • gives better ranking, than any variation with 2 groups
  • gives enough games for everybody regardless the level, the difference between 14th and 1st is just 2 games. very fair.
  • is very realistic, was already done with the same number of games like the past year.

other turneys make 1 group every year since years and the waiting time well spent doesnt impact the “performance”. by performance is concretly ment max 15 min of play per player at 25 km/h 7 times during the whole day, please… sorry, but if we pretend that polo is a sport and not a game, at one point we should be coherent with it. this “maximal” effort is not over-human.
the ones at pressure are the organisers only. they have no rest between games and for them, maybe its better to have more time between their games. i was an organiser of swiss champs and i would have preferrend 2,5h between my games, rather than 30min.

ex: brussel mixed 16 teams, 12 min games 1 group, since ages.

group of 6 + 8 and 4 rounds is not really better for the ranking than 7 + 7 + 5 rounds + 1 bye per round. its still 4 games per team for a total of 14 teams. also there s a risk that in the smaller group of 6, the team matching could be less even than in the group of 8.

If you give organisers half a day to organise and then they can focus on playing instead of 2hours shifts inbetween games, it seems to be way better for them in my opinion.

And « playing » 15min every 2:30 hours doesn’t seem like a fun game if you don’t want to call polo a sport…

Two groups is less heavy for everyone, schedule will already be rough, no need to force everyone to be at the court over 12 hours straight for 2 consecutive days.

2 Likes

Same opinion

3 Likes

Thanks for all your inputs, opinions and calculations.
Regarding the orga stuff, we’ll be fine with 30mins or 2.5h between games, as the orga team plays in different teams and thus there should always be someone who’s not playing.

It’s more important to have fair (legitimate ranking) and fun (reasonable waiting time) swiss champs.

I thinks that the two-group solution is only appropriate if the two groups mix at some point. Otherwise, in my opinion, there will be a bias due to the randomness of the group distribution (what Elena said: “teams who should play each other won’t”).
That’s why I would strongly recommend the solutions 1 or 3 in the excel sheet, where we’d have 4 SR in two random groups, followed by 2 swiss rounds in two resulting groups (“Top 6 / Bottom 8”).
Regarding the time we can still decide spontaneously if it will be followed by a SE top 14 / top 8 / top 4 / …

As the decision should not just be taken by the orga team, I’d suggest to make a little poll in the clubs to get the most convenient solution for all players. Let’s try to get as many answers as possible until next sunday.


PROPOSAL 1:

  • Two random groups ( 6 and 8 teams) play 4 swiss rounds
  • According to the global ranking, the two groups will be mixed and split up again into a Top6 and a Bottom8 group. These groups will play two swiss rounds again
  • After these 6 rounds, a single elimination will be played (goal: all 14 teams play the single elimination)

(PROs: short waiting time between games, ranking seems good, min 7 games per team, “adrenaline games” from SE;
CONs: little less legitimate ranking than with one big group, randomness of the two groups)


PROPOSAL 2:

  • Two random groups ( 6 and 8 teams) play 4 swiss rounds
  • According to the global ranking, the two groups will be mixed and split up again into a Top8 and a Bottom6 group. These groups will play three swiss rounds again
    Top 4 will play a single elimination

(PROs: short waiting time between games, ranking seems good, min 7 games per team, everyone has +/-2 the same number of games
CONs: little less legitimate ranking than with one big group, randomness of the two groups, less adrenaline because only few will play alimination games)


PROPOSAL 3:

  • All 14 teams play 6 swiss rounds as one big group
  • Followed by a single elimination of the 14 teams (can be split up into Top8 slayer/ Bottom 6 challenger)

(PROs: precise ranking and everyone got to play elimination games;
CONs: very long waiting time between games (2.5h average))


PROPOSAL 4:

  • All 14 teams play 7 swiss rounds as one big group
  • Top4 play a single elimination

(PROs: precise ranking, everyone has +/- 2 the same number of games;
CON: Very long waiting time between games (2.5h average))


Do you agree to vote on these options?
Even though a majority in this discussion already stated out that they prefer the 2-group solution, I still wish to integrate the one-group solution in the poll. This allows all participating players to give their opinion.
I hope my 4 proposals reflect more or less what was discussed above. Feel free to correct me. If it seems good to you, say so and we can start the poll as soon as possible

5 Likes

I like to idea of two groups, but I think they should be seeded as much a we can to get better ranking.

One big group is madness. Tourneys that plays 3v3 with 16 teams are already crazy boring, add 4 minutes to each game and (even with 14 teams) and it become even worst.

2 Likes

@uolmo It’s not + 4 minutes it’s + 6 min per game (12 vs 18)

@Mamar it’s not clear on my sheet but groups should be seeded as much as we can. Past years when we use two groups (or more groups, 2021 was 3 groups if I recall well) we always seeded manually the groups, it’s not perfect but better than pure randomness.

1 Like

Totally agree with you @coolquentin . We need to level the groups and more if we aren’t able to have a double elimination that it’s a way rectify an unlucky results.

3 Likes

I’ve added your proposals to the excel sheet. This way we coud see how the planning is affected by each.
I’ve added a Proposal 5 (which is more or less the Variant 3).

I also made a version of the sheet that you can edit to play with the values :

Optimisation :
18 min per game is great, i think it’s how the euro / world was played.
But 16min (or 17) could work too and you will save 1h per day.
I’d suggest to have (or keep) rule that says that +5 goals difference ends the game.
Having winning team goal reffing etc, etc.

I think it’s important to keep in mind that usually not everything works according to the plan and a tournament aimed to finish around 18:00 will likely finish around 19:00. A proposal that aim to finish at 19:30 could very likely finish around 21:00 (worst case scenario).

It’s why my personal preference is the Proposal 5 with your proposal that consist of being able to add more teams into the Single Elimination if there’s time while the default is SE 8 teams and finish around 17:00
SE 10 teams will finish around 18:00 etc, etc.

If you need help to seed the group i think there’s plenty of knowledge around here to help you.
Of course there’s news teams and mixed ones but making group as much as even as possible is doable and will be better than pure randomness.

BTW : this is great, it shows that maybe we need to plan a two courts tournament to have perfect Swiss champs in the future <3

5 Likes