Tourney Any wisdom / formulas for when X% of field is knocked out? (1 Viewer)

[Larold]

One of the common topics discussed about tournament length is the math / adages behind when a tournament will finish. (i.e. "When the big blind is roughly 5-10% of the chips in play, the tournament will end."

I've been wondering about something slightly different though, and my research is coming up short. (I can go off of experience, of which I have a lot, but I'd like to hear from others and if you all have thought about this.)

Instead of structuring a tournament to last X hours (until the 2nd-place player is knocked out), I'd like to start structuring tournaments a few different ways for experimentation, such as:
- Structuring a tournament so that the final table will be forming roughly X hours in
- [Variation on the above]: Structuring a tournament so that roughly Y% of folks are still in play at about the X hour mark
- Structure so that we get down to the money at roughly X hour mark

A potential way to think about this:

Imagine a graph / plot of time along the x axis (say, a bar graph with each bar being blind level 1, 2, ... n), and the y axis being the number of players eliminated during that blind level...

Yes, I get it depends on many factors - rebuys, add-ons, $$ being played for, caliber of players, etc. But let's aim for gross generalizations.

First, I can't help but think if we could magically plot every tournament in history, we'd generally see a trend of bell-curves, with a heavier weight on the right than on the left. The left half of the graph would be sparse - few people bust out early and don't rebuy, unless unlucky or weird situation. In fact, I'd almost be tempted to look at these imaginary graphs only from the end of the rebuy period.

In my head, I'm wondering how far to the right of the entire graph the bell curve really ramps up and peaks.

The reason I'm thinking about this is player psychology: I truly believe that most players, if they end up in the money, don't mind playing a little longer. I'd thus like to structure a tournament length based on the amount of time only a majority of the field will be in play, not the final two players.

So, thinking over your own experience playing and hosting tourneys, can you think of some generalizations you've come to found hold true a lot of the time? Random examples: how far into a total timeline do you think 75% of players are knocked out? For a 30-person freeze-out, what percentage of the total tournament time does it take to get to the final 5?

That kind of thing.

 

[Santa123]

My approach would be based on the 5% / 10% rule. I think 5% is more accurate but it depends on player type. So with the 5% rule your tournament will end definetly when only 20 Big Blinds are left in total. The tournament will end because players are that short that all-ins will be that frequent that a winner will be very soon declared.

Based on that I would say you can estimate the maximum numbers of player by:
the total numbers of Big Blinds / 10 Big Blinds = maximum number of players

For example in Blind Level xy you have 43 Big Blinds in total, you shouldn't have more than 4 players left.

If you want a more realistic estimation you can use a Gaussian distribution to assume a realistic distribution of Big Blinds that each player has. Some players will be big stacks other mid and short stacked. An then try to estimate how likely it is that players will loose all there chips. But that seems quit complicated and it depends so much on player types and coincidence.

If I have to guess I would assume:
the total numbers of Big Blinds / 20 Big Blinds = ~number of players .....
So if you have for example in blind level xy 300 Big Blinds in total you have maybe 15 players left....

 

[Larold]

Thanks so much - I'm going to whip some sample schedules I've played and map this against the /10 BB amd /20 BB math, it may indeed align with what I've seen anecdotally.