A question of bankroll

[Milkybarkid]

I have read loads about what cash games you should play in relation to how big your bankroll is.

I have been thinking about bankroll and tournaments.

What percentage should be risked on a single tournament? How much would you risk?

Say hypothetically my bankroll is 100k (Its nowhere near that big)

Is a £5000 tournament too big for my bankroll? (5%)

[Sheriff Fatman]

As a 'one-off shot' then I'd say its well within your bankroll.

If you were intending to exclusively play £5000 buy-in tournaments from now on then it would be a different matter but, assuming that the loss of £5k from a £100k bankroll doesn't significantly affect your ability to play your 'bread and butter' limit/game, then I don't see that its a problem.

Sheriff

[Colchester Kev]

One thing that I have learnt while following the circuit, is the "Expenses" involved, over and above the cost of the buy-ins.

Im sure when many people see that player X wins 50k here and 20k there, they immediately sell the family silver and embark on the dream of getting themselves on the poker gravy train, what they dont see is the fact that these journey men pro's virtually live out of a suitcase, travelling from town to town, country to country, paying hotel bills, travel costs, food, drink etc. all this while still having to account for their normal living expenses such as household bills and mortgage etc...and if you arent drawing in these events on a regular basis the cost can be frightening... Making a living playing live tournament poker is difficult, and unsponsered players that can do this have my utmost repect.

So when thinking about money needed for 1 event, you can add all your extra expenses to the buy-in, and calculate where that leaves you in terms of overall bankroll, then multiply that by the number of events you want to play over the course of a year.... Add that little lot up and then imagine you dont get a few top prizes....I think you would be frightened to death.

This is why sponsorship is becoming more and more necessary for many players, even if their sponsorship just covers the buy-in, it still makes a huge difference to their costs when calculated over the year.

[Milkybarkid]

I take your point about including expenses in the calculation for tournaments.....

Take Barcelona the last time i was over there.

9 days... including flight, hotel, food e.t.c it was about £1500

[RED-DOG]

Great answer Kev, this is why I think it's easier to make a living playing online, as a bread and butter earner, and picking and choosing your live comps

[Royal Flush]

Ben, i would say that 5% is the max you would ever wish to risk in a tournament. 1 thing that is rarely considered in this, which i think it is important is the amount of runners in the tournament.

[matt674]

Great answer Kev, this is why I think it's easier to make a living playing online, as a bread and butter earner, and picking and choosing your live comps

And even better if the live comps you choose to play have online satellite qualifiers where you can win the whole expenses paid package.....

[Buzzer]

Couldnt agree more with Kev, as I have $80.000 in live tournament winnings the last 12 months yet am only breakeven with buyins and costs.   

With Sheriff I can not really agree. The question is, "Am I overbetting my bankroll?". Yes is he would be, hence he is making a mistake. But just because he does not intend to repeat that mistake, doesnt change the fact that he is making a mistake by overbetting his bankroll.

I am also into Sportsbetting and there is the so called "Kelly system" which helps you calculate your optimum bet size for each bet. Apart from your bankroll size, the other main factor is the "EV", the Expected Value. Do you estimate that your 10/11 shot has a 54% or 58% chance of winning for example?
I think this also applies to poker. Do you feel you have an edge over the field in the MTT you plan to enter? If no, why play it in the first place? If yes, do you feel it is significant even taking costs into account, and hoe big is it?

Of course there is also a massive difference in quality from one 5K event to the other. For example the CPC 2004 was so unbelievable weak in relation to the money (well yes I did finish unplaced as I got outfished ) yet other events are pretty strong. Another example is that I am happily playing $5.000 STTs at PP against "Joe Sixpak" etc. but would not be interested to fill the 10th on a table that already includes Marcel Luske, Humberto Brenes and 7 other pros for the same buyin. I am trying to say, pick your spots!

Finally, I advise you to maybe find a backer (or backers) that buy(s) a stake in you for a big tournament. As I think you have a solid reputation in the poker community you should be able to find someone.

[ontilt]

Here's a slightly different way of thinking about the original question.  There is a lot of discussion on the bankroll required for big bet games, generally the consensus is that you need at least 30 buy-ins and possibly a lot more to feel comfortable in high variance games like a loose omaha game. This would mean risking less than 3% of your bankroll in any one game.

Obviously tournaments differ in the amount of variance, which is almost impossible to work out, but I think it would be reasonable to suggest that it was on average quite a bit higher than cash games. Therefore your max exposure in any tournament should be lower than in cash games which makes 5% too high.

I guess it all depends on your objectives in managing your bankroll, whether you are trying to maintain a steady income or make a big score, or avoid the possibility of going broke.

[Sheriff Fatman]

With Sheriff I can not really agree. The question is, "Am I overbetting my bankroll?". Yes is he would be, hence he is making a mistake. But just because he does not intend to repeat that mistake, doesnt change the fact that he is making a mistake by overbetting his bankroll.

You're missing the point of what I said.  Conventional wisdom for a limit holdem cash player is that a 'safe' bankroll needs to be in the order of 300BB.  Therefore if my 'bread and butter' limit is anything upto £150/300 (or the dollar equivalent) then I have a 'safe' bankroll with £100k.  Speculating £5k of this on a one-off tournament does not affect my ability to play this limit.  Doing this every week would likely wipe it out pretty quickly.

Similarly, a 'reasonable bankroll' for someone who solely plays single table SnG's is anything from 30-50 buy-ins.  I've seen some full time players recommend a bankroll as low as 20 buy-ins but I'd say this was pushing it.  Consequently, someone regularly playing £1000 SnG's could enter a £5k event without significantly affecting their ability to play this limit.

Your comments are perhaps valid for a player who exclusively plays Multi Table Tournaments, but a 'mathematically safe' bankroll (in terms of buy-ins) for such a player would likely never be attainable due to the huge variance involved.  Without calculating the figures, and assuming the MTT player was playing tournaments with an average field size of 100 players then, even with a bankroll of £100k, its likely he'd still be restricted to something like £20 buy-in tournaments to be within an acceptable confidence interval of not going bust.  In practice, how many players with £100k bankrolls are going to restrict themselves to tournaments of this size?

The point I was trying to make originally was that the bankroll requirements to be mathematically 'bust-proof' are astronomical ('Gambling Theory and Other Topics' by Mason Malmuth has some interesting analysis on this).  In the real world, most knowledgable players will have to accept a risk of ruin percentage that is acceptable to them - for some this might be 5%, for others this might be 20%.  300BB (for limit cash games) and 50 buy-ins (for single table SnG's) are reasonable examples of this and its perfectly possible to speculate £5k out of a £100k bankroll on an occasional basis without seriously denting these guidelines.

Sheriff

[The Baron]

Without calculating the figures, and assuming the MTT player was playing tournaments with an average field size of 100 players then, even with a bankroll of £100k, its likely he'd still be restricted to something like £20 buy-in tournaments to be within an acceptable confidence interval of not going bust.  In practice, how many players with £100k bankrolls are going to restrict themselves to tournaments of this size?

Are you talking about a 95% C.I. here?

[Sheriff Fatman]

Are you talking about a 95% C.I. here?

The MTT figure wasn't a calculation, just a guesstimate of the likely scale of the numbers.  In practice it'd be very difficult to calculate an actual figure because you'd need a meaningful win rate and standard deviation figure to do it.  For a player playing exclusively in 100-runner MTT's the sample size needed to estimate this with any reliability would be huge.

I've read much debate among single table SnG players about meaningful statistics.  The consensus seems to be you'd need a sample of around 500 SnG's before your ROI figure starts to have any meaning and even then a much bigger sample would be preferable.  Throwing an additional 90 players into each event would increase these requirements exponentially.

I doubt Doyle Brunson or TJ Cloutier has a big enough sample yet!

Sheriff

[The Baron]

Are you talking about a 95% C.I. here?

The MTT figure wasn't a calculation, just a guesstimate of the likely scale of the numbers.  In practice it'd be very difficult to calculate an actual figure because you'd need a meaningful win rate and standard deviation figure to do it.  For a player playing exclusively in 100-runner MTT's the sample size needed to estimate this with any reliability would be huge.

I've read much debate among single table SnG players about meaningful statistics.  The consensus seems to be you'd need a sample of around 500 SnG's before your ROI figure starts to have any meaning and even then a much bigger sample would be preferable.  Throwing an additional 90 players into each event would increase these requirements exponentially.

I doubt Doyle Brunson or TJ Cloutier has a big enough sample yet!

Sheriff

Ahh ok.

Do you know of any study looking at this hypothetically, maybe with dummy figures?

[Buzzer]

@Sheriiff "Speculating £5k of this on a one-off tournament does not affect my ability to play this limit"

You continue to be way off here! Buyin 5.000 lottery tickets for 1 pound each, may not "affect your ability to play your limit" but it still makes no sense and has -EV.

Investing 5% (adding fees+costs that might even be 6 or 7% unless the event is next door) of a 100K bankroll into a mega high variance event like an MTT, is overbetting your bankroll. Thats a fact.
You would have to feel that you have a massive edge over the field to even consider it in my oppinion, because then the +EV might be interesting enough to "justify" overbetting your bankroll.


"Similarly, a 'reasonable bankroll' for someone who solely plays single table SnG's is anything from 30-50 buy-ins.  I've seen some full time players recommend a bankroll as low as 20 buy-ins but I'd say this was pushing it. "

LOOOOOOOOOOOOOOOOOOOOL. I feel of my chair and rolled on the floor reading that. Try 8tabling STTs with a bankroll of 20 units. You could be busted within 2 hours. Common wisdom amongst STT pros is that you need a bankroll of 100 units to be safe (albeit you may very well "get away" with less units).

[doubleup]

One other issue that has only recently entered the equation is the size of the tournament.  A 90 person tourny with 9 paid is very much more predictable in terms of variance than the mega tourneys that have now appeared - they used to say that a top player would get into the top 10% around about 20% of the time - with smaller tourneys this means that the player could look forward to regular reasonable sized payouts with the occasional win - but just getting into the top 10% isn't really good enough in the mega tourneys, the payouts between 5% and 10% are usually fairly meagre relative to the buyin - so a good player might go broke before getting a big top 3 cash-in.