Tourney T- Base Preference for Tournaments (1 Viewer)

[JustinInMN]

To preface, I advocate for levels no shorter than 18 minutes. At 2 minutes per hand, everyone should get to have the button at least once per level.

With that said, i do prefer uniform level lengths, but in my published structure, I do have a reduction to 15 minutes for levels beyond the expected 20BB in play phase, that has never come into play.

 

[ValueTown]

I shorten from 17m to 15m in the late, late rounds. Keeps the game to 4.5-5 hr. Or the game would last 6+ hrs.

I use 2 decks.

 

[Mr Winberg]

I do have a reduction to 15 minutes for levels beyond the expected 20BB in play phase

I have done the same in the past, I called them the "get-the-f*ck-out" levels.

I shorten from 17m to 15m in the late, late rounds. Keeps the game to 4.5-5 hr. Or the game would last 6+ hrs.

I use 2 decks.

I don't think the difference would be as big as you think. Hypothetically, if a certain tournament ends precisely at the end of the level where there's 20 BBs left, reducing the last level from 17 to 15 would save you 2 minutes. Reducing the last two levels would save you 4 minutes. Last three saves 6 etc.

I don't know how many you mean by "late, late rounds", but I believe the result of that reduction is negligible.

However, a point could be made to reduce the late rounds just because you play more hands per level when you're short handed, so you don't need as much time. But in my experience a lot of time is spent tanking all-ins, so I still like to keep the levels equal.

 

[ValueTown]

I have done the same in the past, I called them the "get-the-f*ck-out" levels.


I don't think the difference would be as big as you think. Hypothetically, if a certain tournament ends precisely at the end of the level where there's 20 BBs left, reducing the last level from 17 to 15 would save you 2 minutes. Reducing the last two levels would save you 4 minutes. Last three saves 6 etc.

I don't know how many you mean by "late, late rounds", but I believe the result of that reduction is negligible.

However, a point could be made to reduce the late rounds just because you play more hands per level when you're short handed, so you don't need as much time. But in my experience a lot of time is spent tanking all-ins, so I still like to keep the levels equal.

2 mins isnt for last 2 rounds, it's for level 15 and beyond. 4/5 players, 2 decks...that's a lot of poker per level. Players feel a bit of pressure to play vs get blinded out. Not opposed to same levels throughout.

 

[BGinGA]

Shorter blind level times in the later stage of a tournament makes a lot of sense from a uniform "hands-played-per-level" perspective.
If earlier levels were 20 minutes (with 10-player tables), then typically 10-hands-per-level were getting played (using a 2 minute per hand estimate).

Towards the end of the tournament, a five-handed final table now with 15 minute blind levels would still be playing roughly 10 hands per level, since dealing and action for 5 players goes much faster than for a full table of ten. Same is true for heads-up -- basing blind increases on the number of hands played is far better than using a fixed time increment (although it can be estimated to provide the desired hands-per-level).

In reality, shorter levels late in the event doesn't make the event seem 'faster' -- it is a function of the number of hands played before blind pressure increases, not time.

in a perfect world, tournaments would have the same number of hands played during all levels, and increasing the blind amounts accordingly regardless of time. But minute timer technology is easier to use than hand counting technology.

 

[Poker Zombie]

Shorter blind level times in the later stage of a tournament makes a lot of sense from a uniform "hands-played-per-level" perspective.
If earlier levels were 20 minutes (with 10-player tables), then typically 10-hands-per-level were getting played (using a 2 minute per hand estimate).

Towards the end of the tournament, a five-handed final table now with 15 minute blind levels would still be playing roughly 10 hands per level, since dealing and action for 5 players goes much faster than for a full table of ten. Same is true for heads-up -- basing blind increases on the number of hands played is far better than using a fixed time increment (although it can be estimated to provide the desired hands-per-level).

In reality, shorter levels late in the event doesn't make the event seem 'faster' -- it is a function of the number of hands played before blind pressure increases, not time.

in a perfect world, tournaments would have the same number of hands played during all levels, and increasing the blind amounts accordingly regardless of time. But minute timer technology is easier to use than hand counting technology.

If you really wanted to balance the number of hands per level, you would do away with the clock altogether, and adjust by orbit or a preset number of hands. Not useful if you are running a MTT, but more accurate than adjusting level times.

However, if you wanted to use a clock to balance the hands per level, the data I have recently been collecting (hands per hour) would suggest that you would need multiple time changes. The data I've been collecting suggests that more hands are delt in the first hour than in the second hour.

Average hands per hour at my tournaments:
1st hour: 34.2
2nd hour: 26.3
3rd hour: 27
4th hour: 45
5th hour: (insufficient data)

Of course this also varies even more depending on the blind structure used. Deep stack tournaments seem to slow down play more in the earlier levels, but I still need more data on the topic to really draw that conclusion.

The 2nd hour and the 3rd hour appear to be slower as that is where blinds start to put pressure on players that roughly maintain their starting stack, making more players tank (even if only briefly). Later levels also see fewer limpers (there seems to be a lot of early level limping with deep stacks in my group).

Just some food for thought.

 

[BGinGA]

If you really wanted to balance the number of hands per level, you would do away with the clock altogether, and adjust by orbit or a preset number of hands. Not useful if you are running a MTT, but more accurate than adjusting level times.

However, if you wanted to use a clock to balance the hands per level, the data I have recently been collecting (hands per hour) would suggest that you would need multiple time changes. The data I've been collecting suggests that more hands are delt in the first hour than in the second hour.

Average hands per hour at my tournaments:
1st hour: 34.2
2nd hour: 26.3
3rd hour: 27
4th hour: 45
5th hour: (insufficient data)

Of course this also varies even more depending on the blind structure used. Deep stack tournaments seem to slow down play more in the earlier levels, but I still need more data on the topic to really draw that conclusion.

The 2nd hour and the 3rd hour appear to be slower as that is where blinds start to put pressure on players that roughly maintain their starting stack, making more players tank (even if only briefly). Later levels also see fewer limpers (there seems to be a lot of early level limping with deep stacks in my group).

Just some food for thought.

Can you apply/assign average stack size (in BB) to each of those hour ranges? I suspect that stack size relative to blinds is one of the driving forces of hand frequency, along with table size and bubble relevance.

 

[Poker Zombie]

Unfortunately, that would require a better camera or RFID.

 

[BGinGA]

Unfortunately, that would require a better camera or RFID.

No...... just knowing the number of players, total chips in play, plus the blind amounts at the beginning and end of each hour. Total chips divided by # of players divided by blinds = avg stack in BB.

 

[Poker Zombie]

Those averages may be thrown off as well, because I had not tracked the number of rebuys corelated to the data. Although I do have rebuy data available, when we break tables I do not know how many chips are going to each table. There are chip leaders and chip losers, and the data I have is from just 1 (of 2-3) tables, so the "average" may be wildly misleading.