I suggested a long time ago that they normalize the opening round to ensure that the second round of elimination tournaments has 2^n players, but they haven’t wanted to do that either, so I doubt anything more robust is coming. To this day tournaments with (2^n) + 1 players to begin with will have a bye every round up until the second-to-last round, so silly.
I’m curious - who is this “they” that you are talking about?
I think I’m probably part of the “they”. It feels so weird being referred to as “they” in a thread that I started 
I’m not sure if you properly understood the approach I’m exploring, but the gist of it is to open up the implementation so anyone can be “they” if they want to.
1 Like
In most tournaments I participate in there is a player or are more players who time out. So even a (2^n) tournament can still lead to a (2^n) + 1 tournament in later rounds.
@GreenAsJade : compliment for the subtle reply.
1 Like
First i’m asking myself which system. Automatic one? Or the ones one can create with (/without?) groups?
Should add
-joining players (before pairing) by proposing a schedule, rules and prize.
-give reward(s) (after over)
1 Like
Maybe all of the above? I guess it depends how it goes.
Yes these are factors, but they are “around” the actual tournament itself. They are important, but at the moment I’m focussed on “what has to be allowed for, to be flexible for all sorts of rules”
2 Likes
It’s crucial, more as around to me. We can try to be exhaustive on the different ways to run a tournament, but the “around” will interact strongly with the system itself.
I need to decouple them, so that I can make the “rules” able to be “plugged in” by anyone who wants to implement a different set of rules.
I mean for example Correspondance and Live. I’m not sure that trying to list what is common for both will help much, as this is quite different (disponibilty of players at first)
Those are differences in the settings for the games being played in the tournament.
They are independent from the rules of the tournament. In the way that I’m looking at a tournament, the rules that I am talking about are specifically “the rules that determine who plays who, and who wins the tournament”.
For example, if I say “Double Elimination” and “Arena” you understand that there are different rules applying to the tournament and who wins.
It makes no difference to the tournament rules whether the games themselves are correspondence or live. That’s about the ‘game settings’ not about the ‘tournament rules’.
I am aware that you could say “I agreed to have a Correspondence Double Elimination Tournament with 19x19 games that have 10 komi” - these are the “Tournament Rules”. I could not argue with you about that, if you define the “Tournament Rules” that way.
However, I’m not interested in the “game settings” part of this - we already have great ways of managing game settings.
I’m interested in “what kinds of rule-sets, like “Double Elimination” or “Arena” need to be supported?”
Because that set of rules is the thing that people want to be able to define for themselves.
2 Likes
Well then an arena Correspondance Tournament could become the novelty on OGS 
Let’s go further, put it in rengo format 
1 Like
I guess the ability to drop out of a tournament or sit out a round might be useful.
It does come up in IRL tournaments, by giving a bye to someone in McMahon say, but other times people opt to be that person or already have something to do and then can return later.
So possibly some flexibility to drop in/out depending on the mode - not elimination of course 
3 Likes
I hope you’ll allow individual games to be customized by the tournament director instead of restricting them to the limited options currently available in tournaments. Handicap - 2, reverse komi, 7x7, etc.
2 Likes
One minor thing which seems popular in european/finnish mcmahon tournaments is to have handicaps based as “MMS-1”.
Meaning that handicaps are calculated based on mcmahon score instead the rank difference (for the first round these will be the same thing but for the following rounds the results of previous rounds matter), and games between players who have mcmahon score differ by 1 will be playing even game, if mms’s differ by 2 they play with 0.5 komi, if by 3 they have 2 hc stones etc.
Currently on ogs, if you have handicap mcmahon tournament, the handicaps are based purely on rank difference at the moment when the pairing is done, so you can have games where player with more mms’s is getting black because the opponent has gained ranks since the tournament started.
Tho i’d assume that changing that part of the code would be pain in the butt without much benefits for most users ://
1 Like
For a moment I thought you meant “individual games, on a per game basis”. That’d be a hassle.
But I think you mean “as a tourney director, I want to specify all parameters of the game setup that will be used in my tourney”. In that case: yes for sure.
Uh-oh this does seem like the thin end of a wedge to let each game be different. I’ll have to think about that. If there’s a formula (such as MMS-1) that the rule-engine can compute based on the player information, then its OK.
2 Likes
That is in fact what I mean. Why not let the tournament director control all parameters of each game?
Who knows what weird formulae a TD might want to use? Even just the ones that people have actually done in the past are so varied, that implementing them all yourself seems like way more of a hassle than leaving the API open so that they can be defined later.
I do imagine that OGS could provide some standard pre-defined tournament types like what we have now, as services that would use the API, but it wouldn’t be a monopoly and the available types could grow in the future.
1 Like
OK - good call. I will look at accomodating a rule-engine that computes what all game parameters should be and sets them up for each game.
5 Likes
If it feels weird being referred to as “they” then the reasonable assumption would be that I’m talking about a particular discussion I had with someone other than you.
I think it’s clear via my statement regarding the intransigence of OGS wrt the tournament system that I’m in favor of your change, as it would allow people to make (one of many possible) common sense changes to a system in which a single player often gets multiple scheduled byes in a row in a single-elimination tournament.
It’s ok that comments that followed seemed to misunderstand how setting up a single elimination properly still benefits players even in the case of uncertainty caused by people leaving tournaments in progress. I just wish that in the world we currently live in, one in which recent events have led to “trust the experts” in the context of medical professionals, also applied “trust the experts” to mathematicians who have studied tournament formats and other applications of combinatorics for a significant part of their life.
“They” were some people who I don’t believe to still be involved with OGS, but their rudeness when I pointed out a clearly undesirable interaction in their tournament formats (people who play in only the first and last round and finish 2nd, or in the most extreme case, who play in no rounds except the final round) disillusioned me of the idea that such a robust common sense idea such as yours would come into play. I wish you the best though, I’d certainly play and organize tournaments on OGS if they weren’t so poorly implemented, and I think your idea might well sidestep the issues I have with OGS tournaments.
The first thing to acknowledge is that personally I am renowned for taking insult too easily when it’s not aimed at me 
I think that’s why the next post complimented me for not flipping out this time 
That said, the reason for responding in the way that I did is that OGS for me is “we” not “they”.
We in the forum, we developers, we moderators, we all together trying to make this a better place.
When you frame OGS as “they” it is dissonance for me.
Maybe that’s just me, I dunno 
1 Like
A suggestion inspired from another thread
Let organizers invite bots.
How about multi phase tournaments, for example group phase followed by a knock out phase?
2 Likes