Tourney Timing of Tourneys (1 Viewer)

[rjbf65]

What is the most reliable formula you all have found to predict the end of a poker tournament? Assuming no chopping happens. I've read that when the BB is more than 5% of the total chips in play is typically when it ends.

Also, whats a good predictor on when a tourney will get down to final table? Say you expect 60 players for a tourney and would like the final table to form at a certain time. Would you say an average stack of 15 or 20 bigs is about when that occurs on average?

 

[CraigT78]

In two years of hosting a two table tournament we have always been finished within +/- 1 level of 5% of the big blinds. Can't speak to when to expect a final table - someone more experienced with the math of it all might be able to. @BGinGA

 

[v1pe]

I don't have an answer, but I'm interested in responses...

Your 15-20 BB estimate sounds about right to me.

 

[BGinGA]

Most tournaments will rarely go past the level in which there are a total of 20 big blinds in play.... and most events won't even reach that level, especially if there is a really big stack sitting at the table (which can shorten play by several levels). A bunch of super-nits that don't understand short-handed play can occasionally drag it out longer.

When a final table occurs (or how long it will last) is dependent upon many factors -- field size, starting stack size, re-buys/add-ons, blind structure specifics, and payout structure. No single or easy formula due to those variables, but it can usually be reasonably estimated for a given set of conditions, although in reality, the range of the answer is usually quite large and not really useful.

 

[BGinGA]

Would you say an average stack of 15 or 20 bigs is about when that occurs on average?

No. That is when the tourney typically starts to become a shovefest, not when it becomes 9- or 10-handed.

 

[BGinGA]

I hate typing on my phone.

 

[Poker Zombie]

5% the total stacks in play is right. Like BG points out it can go longer or shorter depending on stack equality and the nerve/ability to use it.

I have a spreadsheet and graph that shows when players get KO'ed. I built a new filter just for this exercise that shows which level players are eliminated, forming the final table...

The X axis (horizontal) shows the level, Y axis shows the number of players eliminated. Red line is actual data, the green line is a trendline. This covers 26 events, and I already filtered out some of my more unusual blind structures for "uniform" results.

The reason they are spread out so much is because sometimes it takes 1 or 2 eliminations to make the final table. Sometimes we have no rebuy events. Sometimes you have to eliminate 14 players to make the final table. It's just all over the place, and I'm not sure reverse engineering anything will develop a reasonable formula.

If you give me an "average" number of players I might be able to separate out some reasonable data, but it's probably still all over the board. Do you see that spike at level 14? Breaking out that data point, I see that came from 7 different events, with 13 player,15, 17, 18(x2),and 19 (x2) player events. Level 13 occurs when players the starting stack = 8.3x the current BB. That means the spike occurred with between 108 BB and 158 BB on the table.

TL/DR: You can't just tell your cash players to show up at x time when the tournament is only taking up one table. Make them show early and either play the tournament, or make themselves useful dealing/serving drinks.

 

[rjbf65]

That is some sweet data. I ask the question as the local Knights Tourneys can have 40 to 80 people show up. I'd like it to end close to the same time each time. Sounds like the easiest thing to control is to shorten the blind times if we get a big crowd and lengthen if it's a small crowd.

 

[BGinGA]

the local Knights Tourneys can have 40 to 80 people show up. I'd like it to end close to the same time each time. Sounds like the easiest thing to control is to shorten the blind times if we get a big crowd and lengthen if it's a small crowd.

Three variables control the overall length of time that any given blind structure will run -- blind level times, starting stack size, and number of players.

If the number of players changes, you can compensate by changing the blind level times, changing the size of the starting stacks, or a combination of both. You can also change the effective starting stack size by adding or deleting levels at the beginning of the structure.

 

[DJ Mack]

All good stuff! I’ll also suggest considering your players and their attitudes. If it’s a regular tournament try to find out if your players want consistent end times or consistent structures.

If levels are being tinkered up until the last minute every week some may not like that. On the other hand if there is a hard stop on the facility then that’s what you have to do.

 

[Poker Zombie]

Shortening the levels is the easiest, rather than changing 80 stacks last minute. You can also run a typically reasonable tournament through all the levels needed for a good length 40 player tourney, and then have bigger jumps between levels for the final levels that you would only encounter in 60-80 player tournaments. This way you don't have to mess with the levels each tournament.

 

[rjbf65]

The stop time is 1 AM with a 7 PM start time. 8k starting stacks with an optional add on of 2k or 4K. No rebuys.

If a players buys everything available they are invested $70 with $42 to the prize pool and $28 to the Knights. My thoughts are to give the players the best bang for their buck as $70 is a decent amount to shell out for some. Don't want them feeling short changed with a super fast structure. But also has to end at the appropriate time.

 

[Poker Zombie]

1am is a hard stop? I'm sure BG has a structure for 80 players that would never exceed 6 hours, but would last 5.25 hours if it were only 40 players...

 

[rjbf65]

1am is a hard stop? I'm sure BG has a structure for 80 players that would never exceed 6 hours, but would last 5.25 hours if it were only 40 players...

Yes, 1 AM but a couple extra minutes isn't a disaster by any means.

6000 staring stack
2000 "bonus chips" at buyin. (Everyone pretty well buys this)
2000 or 4000 add on (after level 4 200-400) Atleast half of remaining players buy this.

Other considerations, this tourney has always used one deck. I'm campaigning hard for 2 decks. They have had a fast structure with 25 minute blinds in the past. (Most level jumps are doubling of blinds). Top heavy payout which induces a lot of chopping. I'm trying to get them to change it to 20 minute blinds with more of a BG type structure and 2 decks. They have a committee of 3. One of which ever plays poker so they have a hard time making these changes. They aren't convinced that there is a better way since they don't know any better.

This crowd is typically extremely nitty. For instance I made the final table last time played and was the chip leader with 13 big blinds. 130k with 5k-10k blinds. It was chopped 8 ways just a few hands into it. Blinds were going to go up to 10k-20k in a few minutes when that chop happened. You would of had to get atleast 2nd place to get more than the chopped amount. So it's very top heavy as mentioned before, which is another thing I'd like to change but I'm more concerned about the structure and 2 decks.

An official BG structure would be awesome. I've learned so much from seeing his (and others) responses since I've been on here.

 

[Poker Zombie]

While a 33-50% increase style structure is often preferred by many "expert" players, not all players are "expert".

There is nothing wrong with a 100% blind increase, as long as the levels are long enough. Although it doesn't come out very often, I have a structure where the blinds increase from 50-100% each level. Why? because this is what the vast majority of <$100 tournaments in Las Vegas play like. If players are buying in for $70 (or less), you don't need to stress over giving them a $350 structure. They'll likely find it odd (25/75? what kind of structure is this?).

The advantage to big level increases is that the tournament is far more likely to end at the predetermined time. If it squeezes over to the next level, the blinds are again doubled and there is virtually no chance of it running over.

We often harp on improved blind structures. From what I've seen at meet-ups, most of the players here are expert-level players, so it makes sence that they want the best structure possible to minimize the luck factor. But remember, not everyone is looking for the tournament that just pays out the best players week after week. A lot of the lucky SOBs want a shot too.

 

[rjbf65]

Valid points. And maybe that is what the majority would prefer. The push for 2 decks has been my main goal. I certainly would think that most players would prefer that.

 

[GIANTDustySquid]

It shouldn't be too hard to get two decks involved. I just said to my entire league last game "we are playing with two decks this week!"
18 player league, pretty much everyone loved the idea instantly. And we have been playing the same league for 4 years with just one deck.
Tell them it's less boring. Most people will agree with that

 

[BGinGA]

Shortening the levels is the easiest, rather than changing 80 stacks last minute.

Absolutely true, although it's just as easy to eliminate early blind level(s).... changing the effective size of those 80 stacks with a single action. As reduced-length blind level times start approaching 10 minutes or less to meet fixed time limits, it's probably better to explore other options to save time.

There is nothing wrong with a 100% blind increase, as long as the levels are long enough.

Again, totally agree. In fact, there are some 100% increase structures that actually play ~slower~ than some average 40% increase structures, if viewed from an "average blind increase per hour" perspective.

Although it doesn't come out very often, I have a structure where the blinds increase from 50-100% each level.

I prefer keeping the range of increases a bit narrower than that, but I do have a popular format that uses increases from 50% to 67% (averaging 59%). One major advantage it has over a 40% structure is that you can use longer blind level times and still finish in a relatively small time window.... yet still not feel like you're getting ravaged by massively huge blind increases.

The stop time is 1 AM with a 7 PM start time. 8k starting stacks with an optional add on of 2k or 4K. No rebuys.

structure for 80 players that would never exceed 6 hours, but would last 5.25 hours if it were only 40 players...

I'll take a look at this later today.

 

[BGinGA]

Why? because this is what the vast majority of <$100 tournaments in Las Vegas play like.

Only comment where I take exception -- I think that's an incredibly awful reason to do anything..... because the 'way it's done in Vegas' is ALWAYS better for the casino, not the players.

 

[Poker Zombie]

Only comment where I take exception -- I think that's an incredibly awful reason to do anything..... because the 'way it's done in Vegas' is ALWAYS better for the casino, not the players.

In the case of poker, I think the casino's reason for big level jumps is to wrap it up in 4-5 hours and run another tournament. This is not unlike the OPs desire to wrap a game at a specified time with little to no advance knowledge of how many players will attend.

Sometimes what's in the casino's best interest is not opposed to the luck-box dependent player's best interest.

In the case of the OP, I would favor 25-33% increases early to give players their monies worth. Then 50% increases mid to late and get the field down. The tournament should end at the 40 player level with no more than 50% increases, a little before 1am. Then 70-80% increases to end the tournament at 1am for 60 players, ending with 100% increases afterward to wrap up an 80 player event in 1-2 levels after 1am.

 

[Mojo1312]

What is the monetary breakdown for the chips? $50, $10 and $10?

 

[rjbf65]

25T, 50T, 100t, 500t, 1000t, 5000t

These are massed produced, very cheap plastic, blanks. Have over 1000 of atleast 4 different colors. I know, not ideal at all. Someone could easily buy the same thing on Amazon and sneak them in. I mentioned it would be good to get some sort of custom sticker on these to help prevent that from happening. The starting chips are all in a ziplock bag with a cheat sheet for what each color is worth.

I recommended for the 50T to not be included anymore as it was too many colors to deal with at once.

 

[rjbf65]

This tourney was played tonight with pretty much all the changes I have suggested being adopted. Went from 25 minute levels and 1 deck to 20 minute levels and 2 decks.
Got rid of the 50t chip which reduced confusion as we have all blanks to play with.

8000t starting stack.
Structure
25-50
50-100
100-200
150-300
Break - color up 25t sell add on. 2000 or 4000 t
200-400
300-600
500-1000
700-1500
Break - color up 100t
1000-2000
1500-3000
Quick pause- color up 500t
2000-4000
3000-6000 3 way chop happened here. I was one of them
5000-10000
7000-15000

Old structure was....
25-50
50-100
100-200
200-400
300-600
500-1000
1000-2000
2000-4000
3000-6000
5000-10000
10000-20000

Changes were well received for the most part. We had 37 players. I'm hoping the changes results in more players showing up next time. It ended around 11:30 or so which is a good time for everyone. Bigger crowds will push that back to after midnight which is good too. Fun night of poker for a good cause!

 

[TexRex]

BG and Zombie have both provided some great info. I'll add this.

BG said the key factors in tournament length are blind level times, starting stack size, and number of players. He's right, though I'd add starting blinds and average blinds increase to the factors. Alter any one of those 5 things and you alter the length of time a tournament will last.

I don't use any software to determine tournament structure, unless you count an Excel spreadsheet though I use it to calculate things when I change factors by hand. I use two formulas. When the total of all blinds and antes = 5% of the total chips, the tournament should end. A second formula (referred to earlier in this thread) is when the BB = 5% of the total chips, the tournament will end. Here are the two formulas spelled out:

A+SB+BB=#*$*.05 (Formula A)
BB=#*$*.05 (Formula B)

Key
A = Ante
SB = Small Blind
BB = Big Blind
# = Number of players
$ = Value of chips players receive
* = x (these formulas were pulled out of Excel so you could plug them in as is.

While those two formulas yield different results, reality should be between the two, 80-90% of the time, the time falls within that range. I use the two formulas to bracket the length of time a tournament should last. Sometimes it goes faster than Formula B. From there I identify a round where it should end, and by going +/-1, I know that it should end within 3 blind levels -- the target level and +1 or -1.

We had a tournament recently that despite more than normal chips and more than normal players, it ended much sooner than normal. I attribute that to an unusual number of hands where at least two players had really strong hands and good players were knocked out in situations where there wouldn't normally have been two players in the hand. I've also seen the opposite happen, but those are definitely the exception and not the rule.

I've developed several "model" blind structures with steady increases from 25% to 100%. Any of those models can be well-designed to reliably end a tournament at a particular time. Any can be designed to be very attractive to highly skilled players. I then take the model and develop a structure that fits the chip set. Our current structure has blinds increasing by 50-67% every round except Round 2 where it doubles. For some reason, my players don't like an increase between the first two that doesn't double. In the interest of giving them what they say they want, I double the blinds in Round 2 and adjusted the started blinds slightly upward to accommodate that.

If a tournament goes beyond the anticipated time, I say it "red lined," meaning it went over. Red line tournaments may happen 10% of the time. Eventually they will happen. Others are right about "nits" playing, if by that term they mean players who are very tight and not willing to take risks late in the tournament when it affects whether they get in the money or not.

Now let me use a couple of examples to demonstrate how factors work together. Suppose you want a 20 player tournament to last 4 hours (bracket 3:20 to 4:20), 20 minute blinds, with BB = SBx2, and where skilled players have the best chance (our model tournament):

A -- Using a Fibonacci sequence to determine the next blind level (Round 1 + Round 2 = Round 3; Round 2 + Round 3 = Round 4; etc.), your increase is going to average 1.62 the previous blinds. 250 BB to start yields that result. Decrease the increases to 1.5x the previous blind levels and you will add 2 blind levels to the end time, or (4:00 to 5:00).

B -- Use 1.4x increases, you could accomplish that time with with only 50 BB, but the luck factor increases considerably. To have the same skill factor, your tournament needs to be 2 rounds longer, or (4:00 to 5:00).

C -- Doubling the blinds every time requires 2000 BB to start.

If you have a hard stop on time, and you can reliably start at the planned starting time, there are many possible ways to do it, but I've found some structures are easier to design than others.

I am happy to share the spreadsheet I use with anyone. It provides an incredible amount of information about a tournament and will work for any size tournament.

 

[v1pe]

@TexRex would love to check out that spreadsheet!

 

[grantc54]

I am also interested in the spreadsheet.

Thanks
Grant

 

[TexRex]

PM me and give me an email address.