Optimal length

[PygmyHero]

I've been thinking about this topic a little bit following my end of the month recap / analysis. Basically, it seems like a trend may be starting to emerge showing how my win rate is related to the length of my session.

From the limited data I have, it seems like my optimal session length is either one hour, or more than twice that. That is, I tend to do poorly with short session and, to a lesser extent, sessions of 1.5 hours - 2 hours. Oddly, if I play for 2+ hours, then I seem to do fine.

I rarely play less than an hour, but it would also be hard to make any statements about too short a session because individual hands would have too much of an impact on the session (i.e. I play really poorly in 30 second session since I tend to post a BB and don't win that hand too often).

Here are my thoughts: if I am playing a short session (let's say 30 minutes or so), I feel rushed and tend to make sub-optimal plays. I believe it is hard for me to escape the mindset that I have very little time, and will not see too many hands, so I better make the most of each one. This can lead to overly aggressive play.

If I play more than an hour, my results seem to drop off. Maybe I am falling into the trap of not varying my play or betting patterns.

If I play for longer, it seems that I regain some edge - maybe I do start to vary my play again. Maybe I refine my reads on other players. Maybe my table image is strengthened and I am in better position to take advantage of it.

I am wondering if others keeps track of this sort of stuff, and what they think their optimum length is, especially if they have many playing sessions to back up their statements. I would also be very interested in hearing why certain session lengths may be good / bad for different players, especially if someone found a way to relate them to a certain playing style (e.g. a LAG may do well in short session because they try to run the table over and leaves before people catch on and start to fight back).

[numeri]

Do you go in planning on a certain length of a session?

This whole thing is faulty logic. The table should determine the length of your session, and therefore the table determines the quality of your session.

[PygmyHero]

[ QUOTE ]
Do you go in planning on a certain length of a session?


[/ QUOTE ]
Part of what I'm hoping to figure out is if I should go in planning on playing for a certain length. But to answer your question, no, I do not, with the exception that there is usually some upper maximum (e.g. I have to go to dinner in 2 hours, but I can play for any amount less than that).

[ QUOTE ]
The table should determine the length of your session, and therefore the table determines the quality of your session.

[/ QUOTE ]
I agree with this, however, it seems to ignore the fact that I am part of the table, as is my table image, my reads on other players, their reads on me, etc., and all of these things affect the quality of the table from my perspective. Maybe I was not clear in my OP - if there is a time when the factors that help me are the strongest (my reads, image, attentiveness, etc.), and my opponent's are the weakest (their read on me), then that is the optimal time for me to be sitting (or, in your words, the table 'quality' is at it's zenith). If that point existed, then my question has an answer, and does not alter the fundamental truth of your statement.

[bozlax]

[ QUOTE ]
If I play more than an hour, my results seem to drop off. Maybe I am falling into the trap of not varying my play or betting patterns.

[/ QUOTE ]

I'm going to pluck out this gem, and beat you up over it. This is not your problem. 19 out of 20 micro-limits players do not pay enough attention to what you are doing for what you are doing to impact the way they play against you. Actually, I suppose, 15 out of 20 don't pay attention, and another 4 pay attention but mis-use the information.

As to the original question, I find that it's very dependent on the number of tables I'm playing, as, for me, it's all related to my bleariness-factor (when I actually have to start concentrating on focusing on the screen). With 1 table going I can go for hours while I do other stuff. If I'm 4-tabling to clear a bonus, I usually try to cut myself off after 90 minutes or so, and get up and move around for a little while.

[PygmyHero]

Thanks for your input.

[ QUOTE ]
I find that it's very dependent on the number of tables I'm playing

[/ QUOTE ]
This is a good point, and I maybe should have made some effort to incorporate this into my OP. I almost always play exactly 2 tables, so it is not really a factor in my play, but is obviously highly relevant. To all future posters - please try to factor in how the number of tables you play affects your win rate, session length, etc.

You stated some playing guidelines for yourself, citing weariness as the factore that stops you. What I'm wondering is if your results reflect this. That is, if you 4 table for 2 or more hours, is there a noticeable dropoff in your win rate? Do you have a large enough sample to make a statment confidently?

[ QUOTE ]
19 out of 20 micro-limits players do not pay enough attention to what you are doing for what you are doing to impact the way they play against you.

[/ QUOTE ]
I think this is a little overused as a reason to not vary your play, but I'm fine with just agreeing to disagree on it.

[bozlax]

[ QUOTE ]
Do you have a large enough sample to make a statment confidently?

[/ QUOTE ]

I don't have exact statistical analyses, but I am confident enough after a year of self-observation that my play falls off after 90 minutes of 4-tabling that I'll stop no matter how well I'm running.

[ QUOTE ]
I think this is a little overused as a reason to not vary your play, but I'm fine with just agreeing to disagree on it.

[/ QUOTE ]

Clarification: I didn't say you shouldn't vary your play. I said that your falling into patterns that are being noticed and exploited by your opponents is not the reason that your winrate drops off after a certain period of time.

I think varying your play is critical, especially if you're playing the TAG style espoused on these boards, or close to it. If you're playing micros as a TAG, and playing ABC poker, I don't see how you can prevent your brain from melting and running out your ears no matter how many tables you're playing. There's just not enough challenge.

[Goodnews]

The formula in the subject,

z(t) = c - tx^2 + bt + y is what I use to determine whether or not I should continue on to play or leave. This is not accurate, but it does help describe the trend of which I play.

c is my initial feeling, whether I am alert, restless etc. c is actually based on a personal benchmark (can be negative if I feel sleepy, tired etc).

t is time in minutes

x is the rate at which my weariness increases. This is related to the number of tables I play (where x increases as number of tables increases).

b is the rate of which I process and absorb my reads on players (this includes applying the reads). As time goes on the number of reads I obtain increases. It is important to note that b is inversely proportional to x, since x is proportional to the number of tables I play.

y is a free variable, it in fact is a combination of the skill level of the table and whether or not I judge that it will become easier or harder. This value can become negative if I sit at a table that is too tough.

As you can see, the optimal time of play is when t = b/2x (first derivative set to zero). From this equation we see that the time of play is related to the number of tables (since both b and x are related to the number of tables). From here we can say that if we cannot pick up any reads, then it is optimal we quit right at the beginning. Because my reads on the players is an edge, the greater the rate at which I apply this edge then the longer I should play.

Again, this formula is not perfect, but it does in fact incorporate the factors that affect our play and incorporate it to time.

This is a work in progress, any input is appreciated.

[bottomset]

for me logging hands is one of my biggest problems, and hopefully seeing a ton of 2+2ers with insane amounts of money to blow will make me play much more

i play based on how I feel, and if im still playing well, so it varies greatly in time

[BluEsiNsOuL]

I would say around 9 inches, but other people are saying it doesn't really matter.

Seriously though, most of my long sessions are winning ones, probably because I left bad tables pretty early. I also tend to leave a table after I received some bad beats. I don't do rebuy any more. I found it's really hard to recover the loss, probably because I lost the ability to win the pot without a showdown or something. I won't worry too much about my betting pattern. If the table is good and your image is solid, I will stay as long as I can.

grunching

7-9 inches

[PygmyHero]

That's actually really interesting that you have a forumla. Of course, we're online so I can't tell if you're being sarcastic, but I think not.

If you're serious, then I'm wondering how often you run the equation while you're playing.

Additionally, some extra info on how you're quantifying your variables would be enlightening.

To be honest, the formula strikes me as a little bit cumbersome, but I still find it of some interest.

Call me old fashioned...but how is this really any more accurate than looking at the clock, thinking of how many hands you want to play, or whatever else criteria to use?

If you don't feel in the mood to play and you know that you'll play poorly, don't play. If you like to play for a certain length of time before breaks and don't have that much time, don't play.

This is too much thought into something pretty simple. Just because you can solve polynomials, doesn't mean you should. Of course, you could be joking. /images/graemlins/smile.gif

I do think there is something to the OP's idea, but it's probably pretty individualistic. When I play, I tend have sessions of 2 or so hours. However, an hour is the shortest I'll go.

If you know that you're only going to play for an hour, you might want to push too hard to get the money. This is a tourament mode of play as you see that you only have a few oppontunities to get chips. Obviously, this is no good.

[Goodnews]

It is not necessarily more or less accurate than looking at a clock or thinknig of how many hands you want to play. Simply because it is something entirely different. When you consider time, or the number of hands you want to play, you leave out many variables.

From what I took from the OP, he wanted a way to optimize his play to find the correct amount of time to play before having his winrate taper off. By plugging in the numbers you get an idea of how long you should play.

Many will disagree with what I have proposed and even more will find it ridiculous. But gonig with your gut feeling is something that many of us had to fight (according to Psychology of Poker by Alan Schoonmaker). I concluded that if I've lied to myself, then whats stopping me from lying to myself in this aspect? Since I don't have Dr. Schoonamker's services to help me analyze this, I decided to quantify it. Because math cannot lie.

Of course, if I knew I was only going to play for only an hour, I would chcek if that would be maximizing my winrate, if not then I would have to reconsider because of the old adage 'money not won, is money lost' is something I firmly believe in.

[Goodnews]

I can run the equation as many times as I want to. It is very easy to compile something on MATLAB or C to run it over and over again. Of course sudden changes in values (like the variable y) will spring up right away and you will be able to detect tilt right away because you are addressing it immediately.

In an ideal world, I would be running it every hand.

The variables, I have stated are all relative to each other, with calm, normal etc as zero. Adding value to 'b' and 'x' are a little more difficult since I am doing it with trial by error. I like to think the more time I spend reading and playing the b and x values increase/decrease accordingly. If I have been been playing for 3 weeks fulltime, my b value may value might taper downwards, and after a break, spike up to a new height. Of course during the sessions I would take careful notes on any milestones that would occur (ie how long it took em to adjust to the maniac in seat 5).

[BluEsiNsOuL]

I would say it should be something like b*log(t) instead of b*t, it shouldn't be a linear function. The longer you play, the fewer read you obtain in a time unit.

[jaxUp]

[ QUOTE ]
grunching

7-9 inches

[/ QUOTE ]

This was my initial response as well, although it did make me feel sub-optimal (and not in a good way)