And it shouldn’t be the same. He beat 1 player per game vs. beating 2 players per game.
2 Likes
In general, the scoring for 3-player groups is fair and the best way to handle things. A 3-player group will be slightly higher variance though. If I was in a situation in Pinburgh where I needed at least 9 points in a round , I would probably rather be in a 3-player group. If I needed at least 5 points for a round, I would rather be in a 4-player group.
2 Likes
We’re trying something new in league this season. Three and four player groups. Software creates as many three player groups as possible. Four player groups score 4,3,2,1 and three player groups score 3,2,1. Here’s the kicker: Four player groups play four games, three player groups play five games. Sweep your three player group and get a bonus point. No bonus points for four player groups. Most points you can get in a night is 16, no matter what size group you’re in.
No one is complaining yet. The guy who came up with it is a really smart guy who’s been in league forever, so there’s a lot of trust there. Seems like you could use this format in a match play tournament, but you just know someone wouldn’t be happy. ‘I only got to play 4 games!’ ‘I had to play 5 games!’
2 Likes
/I’m/ totally with you. 
We used to do similar in our league. Four player groups played four games, three player played six. Forget what the scoring was but it added up evenly like it should. Mathematically it was ‘fair’, but I was never a big fan. Even though there was only 1-2 four player groups a night (rest were three) during one season I somehow ended up in the four player group 7/8 nights (that’s math vs reality for you…). The problem to me is, it’s much easier to blow up four games than six (assuming that, in general, whoever blows up the game wins). Sure you’ve got less people to beat, but if you blow up a three player game, chances are if there was a fourth person’s score in there you’d beat them too. And if the number of games stays the same for a three person group instead, it works the opposite!
I understand that with scoring like this it’s about as fair as you can be, but I don’t like people saying it’s ‘the same’…
For leagues where you are playing the same set of games for all groups (or at least a list of overlapping groups) TOPL uses the median score for the night as the fourth player in 3 player groups.
For people who the mean variance argument does not work, I would suggest people change their perspective, don’t think of the forth player as a random Joe off the street. Just always assume KME would have been the fourth player in your group, then ask if you would still rather be in a group of 4.
1 Like
Except it’s more likely to be a random Joe than KME.
In group knockout formats what is more common for a three player group: bottom two get a strike or only one strike issued for the group?
I know I never want to be in a three player group where the bottom two get a strike–all other things being equal a 66% chance of a strike versus a 50% chance in a 4 player group.
The one exception where I think bottom two get a strike added to the fun of the tournament was playing in a Critical Hit style group knockout ( thanks @kdeangelo!) where you had ways to get yourself out of the group or better yet put someone else into a three player group.
1 Like
It depends on how long you want your event to last. I personally use 3rd only gets a strike in 3 person group.
3 Likes
I do 4 strike tournaments and make 3 player groups get two strikes.
Most knockouts I play in are 1 strike for 3 player groups though.
1 Like
The “who’s in your group” issue also balances out. In some cases, you’ll be the “middle” skill person of 3; in some high, in some low. Likewise in groups of 4, you can be in any of the 4 positions. While in any specific instance it’s unbalanced, across the entire event, it is. For every person who gets the harder end of 3-player group, there’s another who gets the easier end … but the same is true of 4-player groups. If you’re #3 in a 4-group, you’re worse off than being #2 in a 3-group, so having that extra person there isn’t helping you.
As for having somebody tough there, as I recall, I did not take a 12 in that 3-player group round. But then KME didn’t take 12 in the 4-player round I was in with him, either. Having someone who is ranked above you, even well above you, guarantees nothing. You play your best, take what points present themselves, and move on.
1 Like
I have been doing one strike for three player groups on Match Play, because the way the software works, the three player groups are always the ones with the most strikes and I feel like it’s kind of pouring salt on the wounds of those players to make them also eject two out of a three player group. If it was always the players doing best who got the three player groups I would be much more inclined towards 2 strikes/3 players. (Then I would get to hear the bitching about how you shouldn’t be punished for doing well.)
6 Likes
Unfortunately, there doesn’t seem to be any way as far as i can tell for balancing 4 and 3 player groups in knockout tournaments. Either the three player group is going to be at an advantage or disadvantage. Id say if you are pairing players according to record, it would be better to get the three player groups the advantage, because those people are at the bottom of the standards all ready.
- E(0,0,-1,-1) = -.5
- E(0,-1,-1) = -.666
- E(0,0,-1) = -.333
1 Like
Clearly, the answer is that finisher 2 in a 3-player group gets half a strike. 
5 Likes
My database column is an integer column so that’s not going to happen 
1 Like
Come on! Use MATH! 3/4 in 4p get 2 strikes, 2 in 3p gets 1, 3 in 3p gets 2! Ezpeezy!
2 Likes
In Match Play I added “progressive strikes” and it’s good fun.
1st place: 0 strikes
2nd place: 1 strike
3rd place: 2 strikes
4th place: 3 strikes
No need to make special accommodations for three-player groups!
2 Likes
Pretty big advantage for three player groups!
But it makes people feel better!