BR model

Goal: 10K bankroll. Initial roll: 500$. Assume winning player @10-200NL. Furthermore, assume a increasing winrate at higher levels (in absolute $/hr terms, not bb/100).

Strategy 1:
Moving up when 25 buy-ins for next level; never step back. So moving to 50NL with 1250$.

Strategy 2:
Moving up when you have 4 buyins for the next level above 400. Lose 2, move back.
For example, start at 500 with 20NL, move to 50NL when you hit 700, move back to 20NL below 600; move up to 100NL with 900, move back to 20NL under 700. Etc.

Both strategies are low-risk, in the sense that you'll have a minimum of 25 buyins at all times.

Pros/cons?

I really need to make a BR post.
I'll do this this weekend.

Neither is optimal.
bb/100 is more important regarding BR than $/hr.
If you are making 10BB/100 at 50NL and 5BB/100 at 100NL,
The $/hr is the same, but your variance will he higher at 100NL

Standard deviation determines everything.
The first one is closest to what you should be doing.
The second actually increases risk. (and is nowhere near "low risk")

Quote:

---

Originally Posted by Demiparadigm

I really need to make a BR post.

---

Yes. Yes you should.

Does anyone else have any BR specific questions that they think I should address?

I am currently playing with 71 buyins at NL100. Should I move up?

Yes, I know the answer to this...I have taken BR management to the next level, I do believe.


P.S. And yes, I really do play NL100 with close to 71 buyins...I dabble in NL200 occasionally, and have made a profit there. I just haven't made the jump yet. For what it's worth, I am almost always 2-3 tabling NL100 at all times when I play.

And another BR post (rilla has one) would be cool to read.

Quote:

---

Originally Posted by johnny_fish

Goal: 10K bankroll. Initial roll: 500$. Assume winning player @10-200NL. Furthermore, assume a increasing winrate at higher levels (in absolute $/hr terms, not bb/100).

Strategy 1:
Moving up when 25 buy-ins for next level; never step back. So moving to 50NL with 1250$.

Strategy 2:
Moving up when you have 4 buyins for the next level above 400. Lose 2, move back.
For example, start at 500 with 20NL, move to 50NL when you hit 700, move back to 20NL below 600; move up to 100NL with 900, move back to 20NL under 700. Etc.

Both strategies are low-risk, in the sense that you'll have a minimum of 25 buyins at all times.

Pros/cons?

---

Demiparadigm sounds like he knows way more about this than I do, so listen to him.

But I say, go for it! If you're a good player at least. I like the second strategy. It sounds like you can build a roll pretty quickly that way, but what do I know?

I'm studying statistics in college. One of the units includes standard deviation. I could not quite grasp the concept so I asked for clarification. The teacher could only say that standard deviation is sort of like mean deviation except a little more complex. I got the feeling that she has no idea either, as she mentioned "I wish the text would explain it in more detail."

Can someone please explain this?

Demi it would be good if you could explain how standard deviation affects your role.
Try and explain some of the different ways of using your role eg for sngs and ring or ring and mtt etc
these questions get asked a lot in the begginers circle.
Also id be interested in your views on downswings in relation to your br, if you think you should take a week out or how if possible to recognise a downswing as it starts

standard deviation (according to a site i use)

10. What is stdv (or standard deviation)?
Short Answer:
Basically it is a measure of variance. The bigger the stdv, the bigger your swings are.

Long Answer:
If you assume that your performance is normal (looks like a bell curve) then, 68% of your samples will fall within one standard deviation of the mean. About 95% will be within two standard deviations of the mean. For example, if your average hourly rate is +$10 and your standard deviation is ±$50, then 68% of your hours will be between -$40 and +$60 . Please Note: the stdv value in the charts is only an estimate of your "true" standard deviation. The more sessions you have, the more accurate this estimate should be. I use a weighted standard deviation, which means I take into account the length of the sessions when I calculate standard deviation. Also, as with all the statistics generated on this website, your past performance is NOT a guarantee of future performance. Future performance can be affected by other factors, including, but not limited to, changes in your playing style and that of your opponents.

But if its varience then id like someone like demi to explain and show what high and low rates of varience are.

Quote:

---

Originally Posted by Demiparadigm

Does anyone else have any BR specific questions that they think I should address?

---

I'd like to see someone talking about getting the majority or most of your bankroll on the table at a time. aokrongly talks about this, I'm generally around 50% of my BR (essentially a sethunting strategy)

If you've got 600, there's a big difference between opening up 5 100NL tables and 10 50NL tables. Big difference

Quote:

---

If you've got 600, there's a big difference between opening up 5 100NL tables and 10 50NL tables. Big difference

---

Yes, because of Standard Deviation. It will be much lower on the lower buy in tables.

Quote:

---

But if its varience then id like someone like demi to explain and show what high and low rates of varience are.

---

I am not sure I understand this question.
The explanation of Standard deviation above is good.

Here's another example:
What is the chance of you being down $100 after 3 hours with an hourly rate of $10 and standard deviation of $50?

Your expected result is +$30 after 3 hours... So you are $130 below your expected result.
Standard deviation increases by the sqrt of the number of hours.
So $50(sqrt3) = $86
Now divide $130/$86 = 1.5 which is the number of standard deviations you are away from your expected result.
Assuming the normal curve of distribution, this will happen 7% of the time.

The above example should be a good one to explain why we need to maintain an adequate bankroll.

Quote:

---

Originally Posted by LeFou

Quote:

---

Originally Posted by Demiparadigm

Does anyone else have any BR specific questions that they think I should address?

---

I'd like to see someone talking about getting the majority or most of your bankroll on the table at a time. aokrongly talks about this, I'm generally around 50% of my BR (essentially a sethunting strategy)

If you've got 600, there's a big difference between opening up 5 100NL tables and 10 50NL tables. Big difference

---

I second this as a topic to discuss. You may have 25 buyins but it's different if you have 10 tables open at once. How does that affect the number of buyins you should have?

Stats so far.

I average $16.8 per hour per table..

I play 5 tables so that means that I should expect that I make $84 per hour, on average. My standard deviation per table is around 150... so that means that 95% of the time I am between +$166.8/hour/table and -$133.2/table/hours.

Its scary to think that I can be losing like $666 per hour playing poker (and it HAS happened and IT IS scary :P) but stats usually tell the truth (actually they always do...). Being properly bankrolled I have no problem with this anymore.

Quote:

---

Originally Posted by m3laNcholy

Stats so far.

I average $16.8 per hour per table..

I play 5 tables so that means that I should expect that I make $84 per hour, on average. My standard deviation per table is around 150... so that means that 95% of the time I am between +$166.8/hour/table and -$133.2/table/hours.

---

68% of the time.

Quote:

---

Originally Posted by m3laNcholy

Its scary to think that I can be losing like $666 per hour playing poker (and it HAS happened and IT IS scary :P) but stats usually tell the truth (actually they always do...). Being properly bankrolled I have no problem with this anymore.

---

Your standard deviation increases with the number of tables but not like this.

Your variance goes up as a multiple of tables
StdDev is sqrtVAR

So your variance for 1 table is 150^2 = $22,500
Your variance for 6 tables is then 22500*6 = $135,000

Since StdDev is sqrtVAR, Your stdDev for 6 tables is
sqrt(135000) = $367

Quote:

---

Originally Posted by Miffed22001

... or how if possible to recognise a downswing as it starts

---

Unless you believe in karma and think it "was meant to be" you can never predict a downsing "as it starts".

A downswing is a series of unlucky events (assuming you played right) like losing 4 times in row a 60/40 in you favour after going all in on the flop ---> 4 buy-ins lost. This has a 1 in 256 chance that will happen (and trust me it will). But if you have lost 3 "60/40" in a row, that doesnt mean you should fold the fourth one(again assuming you are right that you are a 60/40 favorite to win the hand) because you "are on a downswing", this has nothing to do with it. Results are indepentant, if you are 60/40 to win a hand, regardless of past results, you are well... 60/40 to win that hand...

Quote:

---

Originally Posted by Demiparadigm

Since StdDev is sqrtVAR, Your stdDev for 6 tables is
sqrt(135000) = $367

---

Even scarier...

I think a discussion on what varience is and how to recognise it at the most basic level would be good. How extreme of a swing can a player expect in a certain length of time or number of hands. These are things I used to wonder about alot.

If an experienced player drops 2 buy-ins in 100 hands he can blow it off because he knows that he might be up 3 buyins over the next 100. This has happened him loads of times before. A noob on the other hand hasen't got this reassurance and this could tilt him.

Formally the buyin requirement is the solution of the following model:

Define X(i) as the bankroll after hand i, i=0,1,2... Define B=X(0) as the starting bankroll.
Now define R(B)=P(min X(i)=0) as the risk of ruin. It's the probability you will ever lose your whole bankroll.

Now choose 0<alfa<1, depending on risk preference. Estimate the distribution of the X(i) and solve for B.

edit: I just read Demi's new BR post, the text above adds nothing to the discussion now :cool: Btw, great post Demi.

But getting back to the first post. I still haven't read any disadvantages of the second strategy. The advantages are obvious: The end goal will be reached sooner than the convential strategy 1. Variance will be higher too, but since the risk of ruin essentially is zero that doesn't matter.

The risk of going broke is "essentially zero" but the risk of moving up, and having to move back down is real.
In your example, you are playing 50NL at $700 and 25NL when you drop down to $600.
What you are doing is playing 50NL with an effective bankroll of $100.
substituting into the risk of ruin formula, you will see that your probability of moving back to 25NL is very large.
Chances are you will spend a lot of time jumping back and forth between 25 and 50, never really moving up to the next level.
If you ever do make it to 100NL, your chances of going back to 25 are still very large.

Quote:

---

Originally Posted by Demiparadigm

The risk of going broke is "essentially zero" but the risk of moving up, and having to move back down is real.
In your example, you are playing 50NL at $700 and 25NL when you drop down to $600.
What you are doing is playing 50NL with an effective bankroll of $100.
substituting into the risk of ruin formula, you will see that your probability of moving back to 25NL is very large.
Chances are you will spend a lot of time jumping back and forth between 25 and 50, never really moving up to the next level.
If you ever do make it to 100NL, your chances of going back to 25 are still very large.

---

Lol. I take back what I said about going for it. It's a horrible horrible idea.

Okay, I tried to read your posts and it was way over my head. So here are just a few basic questions I have about Standard Deviation.

1. For NL what Standard Deviation would be considered high, and what would be considered low?

2. In Poker Tracker, is the Standard Deviation/Hour based on all tables or should I multiply it by the number of tables I play.

3. If two poker players are driving at 50 MPH in opposite directions, which one will reach Vegas first?

Thanks.

1: 20/100 in my opinion is low. mine is 50/100. That is probably high.

2: use your SD/100 hands.

3: The one heading toward Vegas.

Quote:

---

Originally Posted by Demiparadigm

Chances are you will spend a lot of time jumping back and forth between 25 and 50, never really moving up to the next level.

---

There will surely be a lot of jumping back and forth. What's the problem with that?

Maybe next week I'll do some simulations to see what the expected number of hands, jumps, hourly wage etc. in both strategies is.

in response to lefou's earlier post. isnt getting 50% of your br on the tables like breaking the bankroll management rules big style?
if you dump a few buy ins here (lets say varience) it could constitute 20% of your total br, i dont like that! :(

Great threads!

While I understand the principles, I don't quite get the math. I'm not the kind of person that asks someone to "do the work for me", but I get the feeling that you could, in a matter of seconds, answer my question. So here goes:
I have tracked 5200 hands (exactly) for NL$50
My win rate is 16.22BB/100 or 9.92BB/Hr.
My Standard Deviation is 53.69BB/100 or 43.53BB/Hr

I've been playing since 9/1/05 and am improving every day. These stats include my learning time also, so I may be better than they reflect, or I may just be running hot over the last 2 months. My bankroll is $1050 from a $100 initial buyin ($950 profit)

Now, the question:
I've been itching to move to NL$100. Can I, and more importantly, Why?

Many thanks,
cryptyk

I wouldn't recommend it, for a couple reasons.
You really don't have enough hands for a good idea about your winrate (this is the least important)

-Most important is that you don't really have the bankroll.
BB in Pokertracker is 2xbb
your bankroll for 100NL would be 525BBs
Using the risk of ruin formula:
r = exp(-2uB/sigma^2)
= exp[-2(16)525/(54)^2]
= 0.003 (0.3%)
Which is actually very good if you can maintain the same winrate.
This is unlikely however.
Also, when will you go back to playing $50NL? The above formula assumes you will play 100NL until you go broke. If you plan on moving back to $50NL when you hit $750 (15buy ins) You are playing 100NL with an "effective" bankroll of $300. Substituting for a reduced winrate and effective bankroll:
exp[-2(10)150/(50)^2]
=0.3 or 30% chance of losing your $300, and having to grind it out at $50NL.
This is very similar to the question asked by the OP.

Determine the bankroll at which you will move down if you hit it,
Then determine the amount of risk you are willing to accept of moving down.
The required bankroll to move up in limits is your"base" bankroll, + the bankroll determined from the formula.
B = -(sigma^2/2u)ln(r)

You are doing very well at $50NL now. Build up a bankroll that can withstand a bad run when you move up. It won't take long at your rate.

Excellent. You have a knack for explaining this stuff.

I wouldn't have a problem playing 100NL until I was broke, but I do understand that I dont have the hands required to make effective long-run decisions.

The biggest thing for me is that it seems to take so freakin long to make money at 50. I've made about $500/Month for 2 months (short stretch, I know) and that amount doesn't really make a difference in my big financial picture, especially. $1000 monthly is starting to become "worth the time" and 2k monthly would defineately keep the wife off my back =)

I guess I'll try to curb my ambition until I can support my estimated win rate by showing more hands and I have 20 buyins.

Thanks!

Quote:

---

Determine the bankroll at which you will move down if you hit it,
Then determine the amount of risk you are willing to accept of moving down.

---

Let's assume I don't care about moving up or down (I just want to maximize EV). What is the optimal strategy now?

Quote:

---

Originally Posted by johnny_fish

[q]Determine the bankroll at which you will move down if you hit it,
Then determine the amount of risk you are willing to accept of moving down. [/q]


Let's assume I don't care about moving up or down I just want to maximize EV). What is the optimal strategy now?

---

http://www.flopturnriver.com/phpBB2/...=190566#190566

Good luck :)

Quote:

---

Originally Posted by r8ed

Quote:

---

Originally Posted by LeFou

Quote:

---

Originally Posted by Demiparadigm

Does anyone else have any BR specific questions that they think I should address?

---

I'd like to see someone talking about getting the majority or most of your bankroll on the table at a time. aokrongly talks about this, I'm generally around 50% of my BR (essentially a sethunting strategy)

If you've got 600, there's a big difference between opening up 5 100NL tables and 10 50NL tables. Big difference

---

I second this as a topic to discuss. You may have 25 buyins but it's different if you have 10 tables open at once. How does that affect the number of buyins you should have?

---

If your StdDev/100 hands stays constant, your StdDev/hr will increase as the square root of the number of tables you have open.
so if you are playing on 9 tables it will increase by a factor of 3.
Your winrate per hour will increase by a factor of the number of tables you have open so, by a factor of 9.
Thus for the same risk of ruin, your banroll must increase by
sigma^2/2u
=3^2/2(9)
= 1/2

i.e. you would need 1.5 times the bankroll required for 1 table if you wanted to play 9.

If your bankroll for 1 table was 20 buy ins, you would need 30 for 9 tables.
Thus you could theoretically be playing with almost 30% of your bankroll in play at once.

Interesting.. Kelly bankroll for sigma=30$, u=4$ equals 225$.. So I'm not that stupid playing 50NL with a 500 roll, it's just 1/2 Kelly :wink:

I could even play 100NL according to this.. (assuming u>7 which is approx. bb/100 > 2.4)

Quote:

---

Originally Posted by johnny_fish

Interesting.. Kelly bankroll for sigma=30$, u=4$ equals 225$.. So I'm not that stupid playing 50NL with a 500 roll, it's just 1/2 Kelly :wink:

I could even play 100NL according to this.. (assuming u>7 which is approx. bb/100 > 2.4)

---

Yes... I am waiting for the backlash from my last 2 posts.
(playing based on kelly criterion, and playing with a large portion of your bankroll)

I want to point out I don't reccommend either of these strategies.
The kelly bankroll for its mathematical risk, as well as why I don't believe you should have a large portion of your bankroll in play: Your winrate is not constant in poker. Many things can affect your ability to play the game. If you are off your game, and swing your winrate toward the negative, it can greatly affect your results. If you have a large portion of your bankroll in play, It could result in a devastating loss.
But, I think it is important to point out how the math works. And also the real loopholes in bankroll management as opposed to contrived ways to cheat like Dwarfy's challenge.

One VERY VERY important thing I want to point out is that playing outside of the Kelly bankroll actually DECREASES your Expected Value.
Once you have a larger portion of your bankroll in play than SD^2/2u You are effectively LOSING MONEY.

I'm still a bit foggy on this. Right now my bankroll sits at $160 for online NL. Giving me 16 buy-ins to $10Max NL tables. I like to play only two tables at once, are you saying that $160 is insufficient? I acctually thought I was erroring on the safe side, but perhaps I'm misleading myself.

Just wanted to say Demi is a God!

Quote:

---

Originally Posted by dzl

I'm still a bit foggy on this. Right now my bankroll sits at $160 for online NL. Giving me 16 buy-ins to $10Max NL tables. I like to play only two tables at once, are you saying that $160 is insufficient? I acctually thought I was erroring on the safe side, but perhaps I'm misleading myself.

---

Can someone else answer this?
I am starting to feel like I am doing a poor job of getting my point across.

If you are playing within your bankroll for 1 table, you are probably still within your bankroll for 2 tables.
Just remember your Standard Deviation per hour indreases by the square root of the number of tables open.

Quote:

---

Originally Posted by Demiparadigm

Quote:

---

Originally Posted by dzl

I'm still a bit foggy on this. Right now my bankroll sits at $160 for online NL. Giving me 16 buy-ins to $10Max NL tables. I like to play only two tables at once, are you saying that $160 is insufficient? I acctually thought I was erroring on the safe side, but perhaps I'm misleading myself.

---

Can someone else answer this?
I am starting to feel like I am doing a poor job of getting my point across.

If you are playing within your bankroll for 1 table, you are probably still within your bankroll for 2 tables.
Just remember your Standard Deviation per hour indreases by the square root of the number of tables open.

---

My apologies if you already answered it, I've apparently mis-used the search function. I have read the majority of your threads though (I thought) and did not find a clear answer to my question.

Regardless, appreciate the time to respond.

bump bump
demi
i had another question, not sure how relevant it is ?!
i can track the total amount of cash i have won, gross winnings and gross loses, then i have a net total winnings (think ive use the rigt terms i hate math)
anyway, do the relationships of these stats to each other hold any value?
eg
gross winnings 2.5k
gross losses 1.65k
net winnings 0.85k

do they have any relation or can they tell you anything about your play or is it al just fun stats that mean little?

I just stumbled accross this thread thanks to Miffed' bump.

One point to consider - and I havn't time to read all of Demi's BR posts yet - I just skimmed this thread.

1) I don't think anyone has mentioned that while multitabling decreases your variance (with respect to just playing at higher limits) this assumes that MULTITABLING HAS NO EFFECT ON YOUR WINRATE. I think if this is true - then you are either some kind of multitasking god, or you could be playing better. I find that if I 2-table SNGs - I don't feel like I am playing as well (no sample size, sorry). And SNGs are the least read dependent for a given limit.

2) Playing lower limits and building BR slowly also assumes that:
a) You have an infinite amount of time to play poker
b) Doubling your BR (or increasing it enough to move up to the next limit) is a significant amount of real money to you.

E.g., my BR is about $1000. So I am rolled to play NL50 or $25 SNGs, or low-limit MTTs. But I only play like 4 hours a week or something pathetic like that. So, I will basically never double my bankroll without going on a huge rush.

In this case, I would argue that I might be better off playing slightly above my head, hoping to catch a good run and doubling my BR - BUT dropping down if I lose, so for example playing 100NL with an effective BR of say, $250.

Note that this is a HORRIBLE idea if you actually want to ever cash, or pay bills or whatever. I don't forsee getting to play much more than a few hours a week for a long, long time - so I am just trying to become a better player while not going bust (and maybe if I get lucky, I make some money).

Quote:

---

Originally Posted by Miffed22001

bump bump
demi
i had another question, not sure how relevant it is ?!
i can track the total amount of cash i have won, gross winnings and gross loses, then i have a net total winnings (think ive use the rigt terms i hate math)
anyway, do the relationships of these stats to each other hold any value?
eg
gross winnings 2.5k
gross losses 1.65k
net winnings 0.85k

do they have any relation or can they tell you anything about your play or is it al just fun stats that mean little?

---

Not really relevant except for tax purposes. Standard Deviation is really your driving factor, as opposed to gross totals.

Quote:

---

Originally Posted by zenbitz

I just stumbled accross this thread thanks to Miffed' bump.

One point to consider - and I havn't time to read all of Demi's BR posts yet - I just skimmed this thread.

1) I don't think anyone has mentioned that while multitabling decreases your variance (with respect to just playing at higher limits) this assumes that MULTITABLING HAS NO EFFECT ON YOUR WINRATE. I think if this is true - then you are either some kind of multitasking god, or you could be playing better. I find that if I 2-table SNGs - I don't feel like I am playing as well (no sample size, sorry). And SNGs are the least read dependent for a given limit.

2) Playing lower limits and building BR slowly also assumes that:
a) You have an infinite amount of time to play poker
b) Doubling your BR (or increasing it enough to move up to the next limit) is a significant amount of real money to you.

E.g., my BR is about $1000. So I am rolled to play NL50 or $25 SNGs, or low-limit MTTs. But I only play like 4 hours a week or something pathetic like that. So, I will basically never double my bankroll without going on a huge rush.

In this case, I would argue that I might be better off playing slightly above my head, hoping to catch a good run and doubling my BR - BUT dropping down if I lose, so for example playing 100NL with an effective BR of say, $250.

Note that this is a HORRIBLE idea if you actually want to ever cash, or pay bills or whatever. I don't forsee getting to play much more than a few hours a week for a long, long time - so I am just trying to become a better player while not going bust (and maybe if I get lucky, I make some money).

---

1) This is true, you need to consider your winrate multitabling lower limits as opposed to single tabling the higher limits. Also, variance is not necessarily a bad thing as long as your bankroll can withstand it, but lowering variance lowers bankroll requirements for a given winrate, so you could theoretically win more multitabling for a given bankroll.

2) This is true and it is what economists refer to as "utility" A great example is buying a lottery ticket. while clearly -EV, there is marginal utility in the $1 it costs to buy one. If it were taken out of your paycheck each week, you would probably not notice. However, if you happened to win one of those weeks the effect would be life-changing great. This is the main reason a number of people gamble in all of the -EV games at the casino. A small loss is the cost of entertainment, whereas a big win is a rush and possibly life altering experience.
So, while it may be -EV to play outside your bankroll, many non serious players would rather do this than grind out a profit that has "marginal utility"

Quote:

---

Originally Posted by Demiparadigm

Quote:

---

Originally Posted by zenbitz

I just stumbled accross this thread thanks to Miffed' bump.

One point to consider - and I havn't time to read all of Demi's BR posts yet - I just skimmed this thread.

1) I don't think anyone has mentioned that while multitabling decreases your variance (with respect to just playing at higher limits) this assumes that MULTITABLING HAS NO EFFECT ON YOUR WINRATE. I think if this is true - then you are either some kind of multitasking god, or you could be playing better. I find that if I 2-table SNGs - I don't feel like I am playing as well (no sample size, sorry). And SNGs are the least read dependent for a given limit.

2) Playing lower limits and building BR slowly also assumes that:
a) You have an infinite amount of time to play poker
b) Doubling your BR (or increasing it enough to move up to the next limit) is a significant amount of real money to you.

E.g., my BR is about $1000. So I am rolled to play NL50 or $25 SNGs, or low-limit MTTs. But I only play like 4 hours a week or something pathetic like that. So, I will basically never double my bankroll without going on a huge rush.

In this case, I would argue that I might be better off playing slightly above my head, hoping to catch a good run and doubling my BR - BUT dropping down if I lose, so for example playing 100NL with an effective BR of say, $250.

Note that this is a HORRIBLE idea if you actually want to ever cash, or pay bills or whatever. I don't forsee getting to play much more than a few hours a week for a long, long time - so I am just trying to become a better player while not going bust (and maybe if I get lucky, I make some money).

---

1) This is true, you need to consider your winrate multitabling lower limits as opposed to single tabling the higher limits. Also, variance is not necessarily a bad thing as long as your bankroll can withstand it, but lowering variance lowers bankroll requirements for a given winrate, so you could theoretically win more multitabling for a given bankroll.

2) This is true and it is what economists refer to as "utility" A great example is buying a lottery ticket. while clearly -EV, there is marginal utility in the $1 it costs to buy one. If it were taken out of your paycheck each week, you would probably not notice. However, if you happened to win one of those weeks the effect would be life-changing great. This is the main reason a number of people gamble in all of the -EV games at the casino. A small loss is the cost of entertainment, whereas a big win is a rush and possibly life altering experience.
So, while it may be -EV to play outside your bankroll, many non serious players would rather do this than grind out a profit that has "marginal utility"

---

great
this means i need a hug bankroll for nl considering my standard deviation is up near $150 but then i have had huge up and down swings last 2 weeks or so both at limit and no limit.
At least i can see the math behind it now

Quote:

---

Originally Posted by Miffed22001

i need a hug

---

Yeah, we've all been there.

Remember limit poker has more variance for a given winrate.
It also requires a greater skill set to beat the game than NL. (for those of you who I offend by this, pretend I said said "different" instead of "greater." Then learn to play limit.)

If you are afraid of variance, you are best off nut camping the lower stakes No Limit tables.