Tourney Possible Heads Up Tournament Setup - Seeking Feedback (1 Viewer)

[Moxie Mike]

I'm thinking of hosting a heads up tournament. Here's where I'm at so far:

Entry: $125 includes green chip bounty. $1600 prize pool.

Participation: Capped at 16 players. Invite only.

Blind Structure: TBD but we'll be targeting a 5-6 hr run time.

Seating: Players will be randomly placed for each round. No seedings.

Event Structure:

Round 1: 8 heads up matches. Winner of each match captures a bounty and gets to keep their own. Bounties only apply in the first round. Winners move on to Round 2. Losers move on to an 8-player SnG (pool 2) with the winner advancing to the final.

Round 2 (pool 1): 4 heads up matches. Winners move on to round 3; losers are eliminated from the tournament and receive $50.

Round 3: 2 heads up matches. Winners move on to round 4; losers are eliminated from the tournament and receive $100.

Round 4: 1 heads up match. Winner moves on to the final to face the winner from Pool 2; loser is eliminated from the tournament and receives $200.

Final:

Heads up match between winners of Pool 1 and Pool 2. Format: Best of 3 unless the winner from Pool 1 wins the first match, in which case they will be declared the winner of the tournament. If the winner from Pool 2 wins the first match, it becomes a best of 3 match.

$650 for the winner; $350 for the runner up.

Dealers: Games will be self-dealt until other matches end, at which point players from completed matches will be required to deal in yet-to-be-decided games. This will expedite the remainder of the round. I may try to find a dealer for the Pool 2 game.

That's as far as I've gotten.

 

[TheOffalo]

Having players who won Round 1 essentially go into single-elimination for the rest of the rounds (though with small consolation prize), while those who lost Round 1 having a second chance to take it all down, seems a little strange to me. Why not a more straight-forward double-elimination bracket?

Can still use bounties, players get to keep the ones they won but their own is still up for grabs through the whole tournament? (People who win matches in the lower bracket won't get any more bounties for winning, since all players in the lower brackets would have lost their own bounties.)

Everyone's guaranteed to play at least two heads-up matches. Loser in the upper bracket final face off against the winner of the lower bracket final. Winner of that plays the winner of the upper bracket final. If the upper bracket final winner wins that then it's done, but if s/he loses then there's a rematch. So it's true double-elimination all the way until the end.

 

[BGinGA]

Ater much experimentaion, I found that using a best-2-of-3 format for all rounds worked best for us.

Each round was designed to take roughly 60 minutes to 90 minutes max (rare), so completing a 6-round progressive-feed double-elimination format with consolation bracket* for 16 players in 6 hours or so was achievable, and every player got to play at least two rounds (minimum four heads-up matches).

* Main Flight = 16 > 8 > 4 > 2 > 1st Main
Consolation: = 8 > 4+4 > 4+2 > 3+1 > 2 > 1st Cons

Payouts (takes 4 round wins to cash):
1st Main = 40% (4-0)
1st Cons = 35% (5-1)
2nd Cons = 25% (4-2)

Can also pay 45%-35%-20% if desired.

 

[Moxie Mike]

Having players who won Round 1 essentially go into single-elimination for the rest of the rounds (though with small consolation prize), while those who lost Round 1 having a second chance to take it all down, seems a little strange to me. Why not a more straight-forward double-elimination bracket?

Thanks for chiming in, Peter.

The reason is time and ease of management. It seemed like an 8 player SnG would be pretty easy to engineer so that it ends roughly the same time as the last heads up match. It also seemed like a more appropriate consolation prize than another series of heads up matches and would be easier to deal with if we wound up with an odd number of players.

Can still use bounties, players get to keep the ones they won but their own is still up for grabs through the whole tournament? (People who win matches in the lower bracket won't get any more bounties for winning, since all players in the lower brackets would have lost their own bounties.)

No the bounties are only applicable to the first round. By earning a bounty and keeping their own in the process, that's basically a $50 prize for winning the first round. Even if they're eliminated in the second round, they receive another $50 - the idea being that even though they only lost once, they're still only out $25.

Everyone's guaranteed to play at least two heads-up matches. Loser in the upper bracket final face off against the winner of the lower bracket final. Winner of that plays the winner of the upper bracket final. If the upper bracket final winner wins that then it's done, but if s/he loses then there's a rematch. So it's true double-elimination all the way until the end.

Preserving a true 'double elimination' format isn't a priority. The biggest priority is the competition.

Ater much experimentaion, I found that using a best-2-of-3 format for all rounds worked best for us.

Each round was designed to take roughly 60 minutes to 90 minutes max (rare), so completing a 6-round progressive-feed double-elimination format with consolation bracket* for 16 players in 6 hours or so was achievable, and every player got to play at least two rounds (minimum four heads-up matches).

* Main Flight = 16 > 8 > 4 > 2 > 1st Main
Consolation: = 8 > 4+4 > 4+2 > 3+1 > 2 > 1st Cons

Payouts (takes 4 round wins to cash):
1st Main = 40% (4-0)
1st Cons = 35% (5-1)
2nd Cons = 25% (4-2)

Can also pay 45%-35%-20% if desired.

Thanks for the detailed response, Dave. I'd thought about something like this but I couldn't visualize how it'd work out. Have you ever tried it with the SnG format I outlined? If so how did it work out?

As to the payouts - it's interesting that you pay prize money to multiple consolation prize winners - my opinion is that longevity in Pool 1 should be rewarded as players advance.

Also, how do you deal with an odd number of players?

 

[Nine_high]

Having players who won Round 1 essentially go into single-elimination for the rest of the rounds (though with small consolation prize), while those who lost Round 1 having a second chance to take it all down, seems a little strange to me. Why not a more straight-forward double-elimination bracket?

Can still use bounties, players get to keep the ones they won but their own is still up for grabs through the whole tournament? (People who win matches in the lower bracket won't get any more bounties for winning, since all players in the lower brackets would have lost their own bounties.)

Everyone's guaranteed to play at least two heads-up matches. Loser in the upper bracket final face off against the winner of the lower bracket final. Winner of that plays the winner of the upper bracket final. If the upper bracket final winner wins that then it's done, but if s/he loses then there's a rematch. So it's true double-elimination all the way until the end.

I've never done anything like this, but this double elimination structure sounds a lot fairer to me rather than a losers SnG. Adding the bounties gives a perk for progressing through the winners bracket.

We do something similar in our darts blind draw. Best of three games in the winners bracket, with a chance to win high out (bonus $) since one of the games is 301. Losers bracket is single game of cricket only, which has no out shot, so no more opportunities to get high out unless you make the final. This is analogous to the bounties only being winnable in the winners bracket.

Winner of the losers bracket plays winner of the winners bracket, best of 3. If the losers bracket player/team wins, then single game of cricket for first and second place.

Odd number of players is easily handled with first round byes, you can randomly seed based on drawing ranked cards. The seeded double elimination brackets for any number of players are available online and tells you where to pit losers based on which bracket they fell from.

 

[BGinGA]

Have you ever tried it with the SnG format I outlined? If so how did it work out?

As to the payouts - it's interesting that you pay prize money to multiple consolation prize winners - my opinion is that longevity in Pool 1 should be rewarded as players advance.

Also, how do you deal with an odd number of players?

SNG:
Hard personal no on the "1st round losers get a second chance to win but nobody else does" format. Better and more fair options available imo.

Payouts:
All of the main flight losers go to the consolation bracket, so paying 1st in Main and 1st/2nd in Cons does reward longevity across all players, and rewards those with the most wins (and fewest losses).

You can also run it as a true double-elimination bracket (where the losers bracket winner plays the Main bracket winner for 1st/2nd), but this takes at least an extra round to complete (two rounds, if the undefeated player loses the first Final match).

Odd numbers:
Odd-player fields are dealt with in one of two ways: 1) random first round bye(s) are awarded as necessary, (not my preference) , or 2) optional bye purchase(s) into the second round are made by those willing to pay double entry to skip a round. Bye purchases can be either randomly selected from the pool of interested players, or auctioned off to the highest bidder(s). The Bye purchase approach ensures all players start with the same event equity.


The tourney format concepts described above are commonly used in large backgammon tournaments, and are mathematically sound.

 

[Nine_high]

optional bye purchase(s) into the second round are made by those willing to pay double entry to skip a round. Bye purchases can be either randomly selected from the pool of interested players, or auctioned off to the highest bidder(s). The Bye purchase approach ensures all players start with the same event equity.

Interesting idea, definitely seems viable.

Another option could be to run a quick all in or fold tourney (or some other fast game) to determine bracket seeding and thus first round byes, if time permits.

 

[bigdonkey]

Thanks for chiming in, Peter.

The reason is time and ease of management. It seemed like an 8 player SnG would be pretty easy to engineer so that it ends roughly the same time as the last heads up match. It also seemed like a more appropriate consolation prize than another series of heads up matches and would be easier to deal with if we wound up with an odd number of players.


No the bounties are only applicable to the first round. By earning a bounty and keeping their own in the process, that's basically a $50 prize for winning the first round. Even if they're eliminated in the second round, they receive another $50 - the idea being that even though they only lost once, they're still only out $25.


Preserving a true 'double elimination' format isn't a priority. The biggest priority is the competition.


Thanks for the detailed response, Dave. I'd thought about something like this but I couldn't visualize how it'd work out. Have you ever tried it with the SnG format I outlined? If so how did it work out?

As to the payouts - it's interesting that you pay prize money to multiple consolation prize winners - my opinion is that longevity in Pool 1 should be rewarded as players advance.

Also, how do you deal with an odd number of players?

I haven't used this for poker, but I do a weekly cornhole tournament and this is the format we use: standard double-elim bracket but we cut out the traditional 3rd place game and championship game(s). So the last 2 standing in the Winner's Bracket play for 1st and 2nd. Everyone else drops into the Loser's Bracket and plays for 3rd.

 

[JustinInMN]

Odd numbers:
Odd-player fields are dealt with in one of two ways: 1) random first round bye(s) are awarded as necessary, (not my preference) , or 2) optional bye purchase(s) into the second round are made by those willing to pay double entry to skip a round. Bye purchases can be either randomly selected from the pool of interested players, or auctioned off to the highest bidder(s). The Bye purchase approach ensures all players start with the same event equity.

FWIW, the WSOP heads-up tournament does method 1. I share your preference for preserving equity and making players with byes pay double. (An auction system is interesting too.) But frankly for simplicity, it's hard to beat method 1.

But just to contribute something different, I have been studying the match.vision pots system which will accomodate any even number of participants (not just a power of 2.)

How this would work is you just randomly assign a number of "guaranteed" matches, say 3, and make a preliminary schedule for the number of rounds. (This is just a truncated round-robin.)

Round 1: 1 - 10, 2 - 9, 3 - 8, 4 - 7, 5 - 6
Round 2: 1 - 9, 10 - 8, 2 - 7, 3 - 6, 4 - 5
Round 3: 1 - 8, 9 - 7, 10 - 6, 2 - 5, 3 - 4

It's not nearly as time consuming as a full round robin, but it can accommodate any number of participants without having to draw groups.

So then you decide who advances to the playoff round. For this field, top 2 players in W-L go to the semifinals the next 4 players play in the quarterfinals for the remaining two semifinal places.

The advantage of this system is that it accommodates any number of players and everyone plays in every round. The disadvantage is there is a high potential for ties at the bubble positions requiring some sort of tiebreaker. Heads up poker does not lend itself to tiebreakers, but I might suggest "most hands played in defeats" followed by "fewest hands played in victories" to separate players. If it's distasteful to eliminate players on a tiebreaker, you could have an optional tiebreaker round and just use the tiebreaker for seeding purposes only.

 

[JustinInMN]

Interesting idea, definitely seems viable.

Another option could be to run a quick all in or fold tourney (or some other fast game) to determine bracket seeding and thus first round byes, if time permits.

I kind of like this idea too. Play a 60m sit and go to seed the tournament. When time expires, remaining players are seeded by chip count, and then eliminated players are seeded in the reverse order of their eliminations. This seems a fair means of determining byes and you can accommodate literally any number of players, including odd number fields. Also has the advantage of accommodating late arrivals.

The downside is it is a bit time consuming. You could reduce that by terminating this stage early if enough players get eliminated to establish the first round, and the players receiving byes can just randomly draw bracket position. (11 players, and end the game when 6 get eliminated. Those 6 contest the first round and the remaining 5 get byes to the quarterfinals.)

 

[buzzmonkey]

We had one once that we did double elim. Got done in about 6 hours or so with 8 players. The only downside was I finished up the winners bracket and basically sat around for 2 hours while the last games finished. Then I got beat twice to end up in second against a guy I beat in the first round. Was still fun though.

You can play around with brackets on Challonge if you'd like to simulate some things.
https://challonge.com/

 

[Nine_high]

I guess a losers SnG could maybe work, but I'd want players who lose in later rounds to still have a crack at winning the whole thing. Maybe if you added them as "late entries" to the SnG, with bigger stacks? For instance all round 1 losers get 10k, Rd 2 losers 20k, Rd 3 losers 40k, etc. Not sure how the timing of all that would work out though, probably not great. It'd probably need to be a pretty fast structure so the winner isn't waiting around too long.

 

[BGinGA]

FWIW, the WSOP heads-up tournament does method 1. I share your preference for preserving equity and making players with byes pay double. (An auction system is interesting too.) But frankly for simplicity, it's hard to beat method 1.

But just to contribute something different, I have been studying the match.vision pots system which will accomodate any even number of participants (not just a power of 2.)

How this would work is you just randomly assign a number of "guaranteed" matches, say 3, and make a preliminary schedule for the number of rounds. (This is just a truncated round-robin.)

Round 1: 1 - 10, 2 - 9, 3 - 8, 4 - 7, 5 - 6
Round 2: 1 - 9, 10 - 8, 2 - 7, 3 - 6, 4 - 5
Round 3: 1 - 8, 9 - 7, 10 - 6, 2 - 5, 3 - 4

It's not nearly as time consuming as a full round robin, but it can accommodate any number of participants without having to draw groups.

So then you decide who advances to the playoff round. For this field, top 2 players in W-L go to the semifinals the next 4 players play in the quarterfinals for the remaining two semifinal places.

The advantage of this system is that it accommodates any number of players and everyone plays in every round. The disadvantage is there is a high potential for ties at the bubble positions requiring some sort of tiebreaker. Heads up poker does not lend itself to tiebreakers, but I might suggest "most hands played in defeats" followed by "fewest hands played in victories" to separate players. If it's distasteful to eliminate players on a tiebreaker, you could have an optional tiebreaker round and just use the tiebreaker for seeding purposes only.

There is also the Modified Swiss movement matched-record format, where players play subsequent rounds vs a player with a matching/similar W/L record, while those players with 3 (or optionally 2) losses are eliminated. The random Byes offer no advantage, as they don't apply to a player's W/L record. All rounds are a new random draw vs similar records but with no repeated opponents.

Example:
Rd 1 -- random draw
Rd 2 -- 1-0 vs 1-0 players and 0-1 vs 0-1 players
Rd 3 -- 2-0 vs 2-0, 1-1 vs 1-1, and 0-2 vs 0-2
Rd 4 -- 3-0 vs 3-0, 2-1 vs 2-1, and 1-2 vs 1-2
Rd 5 -- 4-0 vs 4-0, 3-1 vs 3-1, and 2-2 vs 2-2
etc.

Eventually you get down to one undefeated (the winner), several players with one loss (who play off down to a 2nd/3rd tie), and players with 2 losses (who play off for 4th/5th tie). Tiebreaks are previous heads-up winner (if applicable) or otherwise chop the $$.

Imo, it's much better than RR, but requires a lot more management to track and implement. It was a popular format on the backgammon circuit.

 

[bigdonkey]

There is also the Modified Swiss movement matched-record format, where players play subsequent rounds vs a player with a matching/similar W/L record, while those players with 3 (or optionally 2) losses are eliminated. The random Byes offer no advantage, as they don't apply to a player's W/L record. All rounds are a new random draw vs similar records but with no repeated opponents.

Example:
Rd 1 -- random draw
Rd 2 -- 1-0 vs 1-0 players and 0-1 vs 0-1 players
Rd 3 -- 2-0 vs 2-0, 1-1 vs 1-1, and 0-2 vs 0-2
Rd 4 -- 3-0 vs 3-0, 2-1 vs 2-1, and 1-2 vs 1-2
Rd 5 -- 4-0 vs 4-0, 3-1 vs 3-1, and 2-2 vs 2-2
etc.

Eventually you get down to one undefeated (the winner), several players with one loss (who play off down to a 2nd/3rd tie), and players with 2 losses (who play off for 4th/5th tie). Tiebreaks are previous heads-up winner (if applicable) or otherwise chop the $$.

Imo, it's much better than RR, but requires a lot more management to track and implement. It was a popular format on the backgammon circuit.

There are a number of programs for running a Swiss format, some free. I don't have any real experience with any of them, however, beyond a cursory test.

 

[JustinInMN]

There is also the Modified Swiss movement matched-record format, where players play subsequent rounds vs a player with a matching/similar W/L record, while those players with 3 (or optionally 2) losses are eliminated. The random Byes offer no advantage, as they don't apply to a player's W/L record. All rounds are a new random draw vs similar records but with no repeated opponents.

Example:
Rd 1 -- random draw
Rd 2 -- 1-0 vs 1-0 players and 0-1 vs 0-1 players
Rd 3 -- 2-0 vs 2-0, 1-1 vs 1-1, and 0-2 vs 0-2
Rd 4 -- 3-0 vs 3-0, 2-1 vs 2-1, and 1-2 vs 1-2
Rd 5 -- 4-0 vs 4-0, 3-1 vs 3-1, and 2-2 vs 2-2
etc.

Eventually you get down to one undefeated (the winner), several players with one loss (who play off down to a 2nd/3rd tie), and players with 2 losses (who play off for 4th/5th tie). Tiebreaks are previous heads-up winner (if applicable) or otherwise chop the $$.

Imo, it's much better than RR, but requires a lot more management to track and implement. It was a popular format on the backgammon circuit.

I was this close to getting into this. My brother was an accomplished chess player and would dabble in tournaments where swiss system is popular.

 

[Moxie Mike]

Thanks for all the feedback and ideas. Unfortunately, the idea wasn't very popular with the core members of the Moxie Poker constituency, so I probably won't be pursuing this.

 

[BGinGA]

the idea wasn't very popular with the core members of the Moxie Poker constituency

Interesting. My group enthusiastically embraced the concept of a heads-up tournament.

The experienced players liked the idea (because they felt they had an edge), and the weaker players thought it would be a great way to improve and get valuable heads-up experience (which one rarely gets to play when you're not very good).

 

[Moxie Mike]

The experienced players liked the idea (because they felt they had an edge), and the weaker players thought it would be a great way to improve and get valuable heads-up experience (which one rarely gets to play when you're not very good).

I thought it would be a neat thing to try. It may have just been the timing - I don't host tournaments during the summer since most of my crew will be in Vegas at various points in the summer for the WSOP. I may revisit the idea in the fall.

 

[JustinInMN]

Thanks for all the feedback and ideas. Unfortunately, the idea wasn't very popular with the core members of the Moxie Poker constituency, so I probably won't be pursuing this.

The nice thing about heads up is there is no waiting. You are always involved every hand instead of waiting around and playing whatever-percentage of hands is dictated by your starting strategy. That's how I would "sell" it to players if I were to run this.

Tournament formats are fascinating to me so I really enjoyed this thread, hopefully it's a good one for future readers.

 

[Poker Zombie]

I've tried multiple times to run a HU tournament, but have nixed the format every time.

• Seeding can appear biased - especially if a bye is involved
• Random seeding is starting out completely random, which is opposed by many (if not most) skilled players - especially if a bye is involved.
• Double elimination (my preferred system, as I have players that drive for hours to play) requires very short rounds, or a very long night. My average attendance of 18 players would require 11-12 rounds to complete. I just don't see a way to make a HU blind structure that elimination is likely in 30 minutes without feeling like it's a flip-a-ment.
• Except for the first full round, somebody is always sitting out. In fact with 18 players, a double-elimination would require someone to sit idle for 3 consecutive rounds.

 

[Dodger]

Just a thought…but if each round is already a best of 3 to advance, why not a single elimination tourney? I know, I know…it’s not ideal, but it could help on the timing of the event as a whole.

And then, like any other tourney night, eliminated players can start a regular cash game on another table.

 

[Poker Zombie]

Just a thought…but if each round is already a best of 3 to advance, why not a single elimination tourney? I know, I know…it’s not ideal, but it could help on the timing of the event as a whole.

And then, like any other tourney night, eliminated players can start a regular cash game on another table.

As I see it, there are a couple challenges with this

• Table space. Trying to give space for each player to play, & hopefully have a dealer (or 2 ) for each HU match requires 3-4 chairs per match.
• If eliminated (or waiting) players are playing cash games, the HU players are more likely to have to shuffle/deal for themselves. That would mean even fewer hands in the already short round.
• Cash games can often present even bigger prizes than a tournament. A skilled player may tank the tourney just to keep playing cash, especially if the skilled player had to either shuffle or deal every single hand in what is already going to be a tournament that is pushing luck to a premium.

 

[JustinInMN]

I've tried multiple times to run a HU tournament, but have nixed the format every time.

• Seeding can appear biased - especially if a bye is involved
• Random seeding is starting out completely random, which is opposed by many (if not most) skilled players - especially if a bye is involved.
• Double elimination (my preferred system, as I have players that drive for hours to play) requires very short rounds, or a very long night. My average attendance of 18 players would require 11-12 rounds to complete. I just don't see a way to make a HU blind structure that elimination is likely in 30 minutes without feeling like it's a flip-a-ment.
• Except for the first full round, somebody is always sitting out. In fact with 18 players, a double-elimination would require someone to sit idle for 3 consecutive rounds.

The more I think about it, the more I think @BGinGA in post #13 has the right approach to get over the field size objections. It's actually not that hard to learn how to manage the swiss system and as @bigdonkey pointed out there is software that will help. It's easy to guarantee a minimum number of matches. Determining how to get to the final stage is tricky, but seems workable.

 

[bsdunbar1]

This thread is great, I have it bookmarked for future use, so even if Moxie doesn't go forward with it all the ideas and experiences are valuable info.

 

[Poker Zombie]

The more I think about it, the more I think @BGinGA in post #13 has the right approach to get over the field size objections. It's actually not that hard to learn how to manage the swiss system and as @bigdonkey pointed out there is software that will help. It's easy to guarantee a minimum number of matches. Determining how to get to the final stage is tricky, but seems workable.

I can see how that may be fair for the players, but I'm inclined to think that managing it would be a full-time task for the host. If your players are able to self-manage and find their next match, it might work, but alcohol could be a wrench in that system.

There's probably not a lot of drinking at chess tournaments.

 

[bigdonkey]

I can see how that may be fair for the players, but I'm inclined to think that managing it would be a full-time task for the host. If your players are able to self-manage and find their next match, it might work, but alcohol could be a wrench in that system.

There's probably not a lot of drinking at chess tournaments.

The software makes it pretty easy. Basically enter names and hit start round 1. Software creates matches. Enter winners and losers and hit start round 2. Software creates matches based on records, etc.

The biggest drawback to Swiss is that every match in a round has to be completed before the next round can be started, so if one match is over in 10 minutes those players are waiting around for the hour-long match to finish. Whereas in round robin, the next matches can start pretty much as soon as players come available.