Unfortunately this is an issue with your old OBS version, so you may be out of luck until you upgrade – Matchplay Next uses some modern JS features that the Chromium in older OBS versions doesn’t support.
3 Likes
Question on balanced pairing in a group match play tournament.
We recently hosted a 6 round group match play tournament with 21 players. Balanced player pairings were used. Based on the description of pairing players with previously unfaced opponents I would assume that each player would play all others at least once with these numbers. But one player brought up to me that he didn’t play 7 other players at all, which is one third of the field.
Is there other criteria the balanced player pairing looks at, or does the order in which players are assigned to a game affect the match making?
Just looking for an answer for one of my players. Thanks for your help!
It’s always nice to see the actual tournament data (so drop a link), but with 21 players you don’t have very many groups each round so a number of repeats is to be expected. If you want a better distribution you can play 12 rounds of head-to-head – that’ll give the balancing algorithm more flexibility to do it’s thing.
The balancing algorithm for group match play looks at previous opponents and number of three-player groups. More weight is put on balancing the number of three-player groups than the number of repeat opponents (so 21 players is a though starting point due to the number of three-player groups).
2 Likes
Here’s the tournament:
https://next.matchplay.events/tournaments/120412
The player who brought it to my attention is Tom Bentley.
Balancing three player groups first makes sense and I know with a smaller # of players exacerbates that issue. With that in mind I looked at the number of three player games each player had and realized they would only face max 16 opponents in a six round tournament of 21 players. So no matter what he wasn’t going to play at least 5 people.
Thank you the explanation!
Looking at the number of repeat opponents the distribution here looks pretty good. Go to the player stats and scroll down to the last table: Match Play Events
Tom was the “worst” in the sense that he had a couple of repeat opponents, but the differences here are really small and he’s definitely no outlier.
Similarly, on the topic of Balanced Grouping, if I set a 16 player, 5 round tournament, there are arrangements in which you get a true round robin in the sense of each player grouped with every other player a single time. (16 player 5 rounds is just one player-round setup that allows this, there are various others). Would this set up on a group matchplay tournament on Match Play Events correctly end up at the round robin result? or, should I test it myself to find out how consistent it might be with this?
And, if it won’t calculate those groups on its own, would you consider hard-coding it if the player-round formula fits:
4 player groups ( (players-1)/3=rounds) -
28 players, 9 rounds
40 players, 13 rounds
52 players, 17 rounds
64 players, 21 rounds
3 player groups ( (players-1)/2=rounds) -
9 players, 4 rounds
15 players, 7 rounds
21 players, 10 rounds
27 players, 13 rounds
There’s no guarantee that you’ll end up with a perfect round robin. My gut feeling is that you’re more likely to not. Hard coding a special treatment for this specific schedule is not something that’ll happen.
If you’re looking to play a round robin I recommend the Max Match Play tournament format. As a bonus it’ll play faster than a group tournament.
Thanks for the response.
That’s true, but with a small number of machines, 4 player groups reduce the perception of time spent not playing.
This “perception of time spent not playing” idea confuses me.
Is this something we should care about and/or spend effort mitigating?
Our group plays at a barcade on a weeknight, 8 machines, 30 players. I just think people would think the night was longer if it was all two player games.
I have a feature request. It’d be really cool to follow a location/set of locations, and see everything happening there within a single filter. I could add the half dozen or so comp spots near me and easily see what’s going on at those places when I’m not in a tournament there at any given time, versus searching for the venue or tournament name, or using the tournaments near me meature (for example, when I’m traveling). Low priority request, but it’d be cool to have all of my “favorite” locations present in a given spot.
@arenapoutine I’m working on something right now that I think will scratch this itch for you. Keep an eye on the “next generation” topic for updates: Match Play Events: The Next Generation
1 Like
I was trying to get a arena preferred for my tournament this Saturday its a launch party but it doesn’t seem like i can plz help.
The preferred arena feature is only available for certain tournament formats. I’m guessing you’re planning to run a format where it’s not supported?
Running a classics tournament this weekend, and there are a few areas in my arcade that have a high concentration of classic games. What I want to do is spread out the randomly selected games each round so that there aren’t 7 four player matches going on in one small room.
I thought I could do this with categories, but it doesn’t seem like it.
Games that are in areas with enough space around them could be in every round, but there are other spots where half of the games could be used in one round, and then the other half could be used in the next round.
So basically, at the beginning of the round, I want to choose which “bank” is being used, and the arena draws will come from that bank. Something like:
Bank 1
Paragon
Black Knight
Space Time
Playboy
Wizard
Doodle Bug
Bank 2
Paragon
Black Knight
Space Time
Captain Fantastic
Prospector
Frontier
Is there a way to do this? I swear I’ve been in tournaments that do something like: this round is EMs, the next round is SS, the next round is modern, repeat.
Why won’t categories do this? Call one bank A and the other B. Draw from A round one, then B round 2 and so on.
Okay, yes…they do.
I tried this the last time I ran this format, and was getting an error message (that I can’t remember now) so it was unusable. But…I ran a test tournament just now and it worked.
Under Category Banks in the matchplay documentation, it says: “Category banks only really make sense when the tournament is configured to have more than one game per round” and I guess that statement seemed to support it not working for me previously. So, I would respectfully disagree with the documentation!
Thanks Erik!
1 Like
That’s apples and oranges. Category banks is different from drawing exclusively from one category every other round.
Category banks will use arenas from different categories for each match within a group. So group 1, game 1 will come from category A and group 1, game 2 will come from category B and so on. If you only have one match per round then all groups will get an arena from category A. That’s not really what anyone wants hence the warning.
When you start a round you can limit the arenas used to a single category but that’s a separate filter that’s applied before the arena assignment algorithm starts doing its work
Delayed reply, but I have used labels to try to address this.
You have three sets of machines
Set 1
Playboy
Wizard
Doodle Bug
Set 2
Captain Fantastic
Prospector
Frontier
Set 1 and 2
Paragon
Black Knight
Space Time
In the Labels section of the Area, I add Set 1 to the six machines and Set 2 to the three others plus the three that are also in Set 1.
Then, when I do the draw I draw against the label.
Since only the first label shows up for display, when I did this I had 5 New, 5 Mid, and 5 Old machines. I made the Mid labels all ‘Mid, Old, New’.
When I drew rounds, I either did Old or New and then I had 10 machines available, 5 of which were available every round.