Let's define a session (finally!)

[bj bob]

Since joining this forum I have noticed the term "session" has been used and abused to the point where it has no definite meaning to anyone. It can range from a one hour "sit-down" at a particular table to an entire weekend in Vegas or AC.
It is my opinion that the term "session" can be defined as to have some mathematical relevance. Here are my general thoughts on this idea, without indulging in the specific mathematical computations, of which I am sure that the cognoscenti of this esteemed forum can indulge more thoroughly.
My initial idea, as I have stated, is to define the term that has some sort of mathematical pertinence. I would suggest that it would be the lenght of time that it would take a player to put his entire bankroll "in play" one time through at a standard bet spread at 1.0 x Kelly. Therefore, in a SD game my hunch would be about 2-2.5 hrs, DD maybe 3-4 hrs and shoe about 5-6 hrs. All this, of course must be prefaced by a general consensus of what standard bankroll would be, an acceptable "typical" bet spread and # of hands/hour.
In the end, I would imagine that the resultant "session" could be an actual increment of N0, and thus, when added up, could actually be graphed as such on a spread sheet to visually demonstrate the "long term".
I'm certain the with the brain power residing within this forum that we can arrive at a sound definition and am looking forward to your further imput on this subject.
PS- I have to knock off starting threads right after reading Cicero in the original.:grin: :grin:

 

[EasyRhino]

Okay, first of all, stop reading Latin. That's the most -EV thing in the universe.

Second, my ghetto-definition of a session was always how long I actually played at one casino in one day. Which is usually between 45 minutes and whenever I get backed off (3 hours).

To make sure I have things clear you're talking about total amount wagered matching your ENTIRE bankroll? For instance, if I had a $30k bankroll, and my average wager was $150, and I played 100 hands/hour, then I'd be looking at 2 hours?

Interesting, although it would be highly variable depending on speed of play (full tables, etc).

 

[bj bob]

Okay, first of all, stop reading Latin. That's the most -EV thing in the universe.

Second, my ghetto-definition of a session was always how long I actually played at one casino in one day. Which is usually between 45 minutes and whenever I get backed off (3 hours).

To make sure I have things clear you're talking about total amount wagered matching your ENTIRE bankroll? For instance, if I had a $30k bankroll, and my average wager was $150, and I played 100 hands/hour, then I'd be looking at 2 hours?

Interesting, although it would be highly variable depending on speed of play (full tables, etc).

Yeah, you're on the right track. I am assuming that there will be time differences between pitch games and shoes, but I do think that a standard can be established for each. Also, as you well know, a "sit-down", especially in SD, DD can be very short(due to full tables) and shouldn't even qualify as a "session".

 

[Kasi]

It is my opinion that the term "session" can be defined as to have some mathematical relevance.

And I thought it was Italian lol.

Another approach might be to break up a trip roll into a series of session rolls. Like maybe you brought 150 trip units and maybe you then decide to have 5 30 unit session rolls.

So you leave a session if you lose all 30 units or after an hour or so anyway to not wear out your welcome. If you win 30 units in less time than that, you also leave and play another table.

Then you record the session results in as much detail as you want, players at the table, time played, etc which shouldn't be too hard because you haven't played that long and any estimates are likely to be more accurate than doing same after 3 or 4 hours.

Obviously this might require some planning ahead of time.

 

[21forme]

As my play is usually from an afternoon until the following morning (with sleep thrown in) I consider my session the time traveling east on the AC Expwy until I get back on it in the other direction.

 

[bj bob]

To make sure I have things clear you're talking about total amount wagered matching your ENTIRE bankroll? For instance, if I had a $30k bankroll, and my average wager was $150, and I played 100 hands/hour, then I'd be looking at 2 hours?

Interesting, although it would be highly variable depending on speed of play (full tables, etc).

I'm not quite sure that if your BR was 30K @ full Kelly that your average bet would be $150, it would also assume a standardized spread . As far as speed of play goes, I believe that 100/hr. is considered "standard".
My underlying point, though was to make whatever definition of a session a mathematical measurement of N0

 

[Kasi]

My underlying point, though was to make whatever definition of a session a mathematical measurement of N0

Then just keep track of number of hands played, EV and SD whether sessions are short or long.

For the same game of course.

 

[Kasi]

My underlying point, though was to make whatever definition of a session a mathematical measurement of N0

Just to expound in case of whatever, since I don't quite understand the magic of NO anyway, the most important thing to me would be to make sure that all "sessions" were for the same game, bet the same way with the same penetration.

In other words, N0 is defined by things like spread and penetration. It is fixed the moment you choose to play a certain game a certain way. And it can fluctuate wildly.

In other words, as an example, a 4.5/6 game might have 68000 hands for an N0 spreading 1-8 but 40000 hands if you spread 1-16.

The exact same game might only have an N0 of 24000 hands if you back-count and spread 1-4. And that would change depending on entry and exit points.

Add an extra half-deck penetration for the exact same game and the same 1-8 spread above might now only be 40000 instead of 68000. And the 1-16 might only be 25000 instead of 40000. And the 24000 might only be 16000 instead of 24000.

Take away a half-deck, making it a 4/6 game, the original 68000 might be more like 118,000.

So there you are playing the exact same game with the same rules, with only a max difference of 52 cards in penetration and the N0 could be anywhere from 16000 to 118000, all with the same ROR and bankroll, just depending how you choose to spread and when.

And, btw, if you change the rules or number of decks, all of that will change too.

I hope you get the idea here.

 

[blackjack avenger]

On Sessions Defined

I think you answered your own question. A session can be defined as the artificial period of time you designate for a given game. As an example, if you are playing a 6 deck game with similar rules for 2 days then it is up to you when you define a session. However, as soon as you start to play an 8 deck game it's probably best to consider that a new session. Of course it would be good to know how to play the different games properly. When facing different cuts in the same session you would probably face a standard cut and face that most of the time so your results will mostly reflect that. If you are in one casino and bounce back and forth between several games then it is up to you how to define that session, but again the most important thing is to have some idea of how to play each game. This should not be difficult.

 

[bj bob]

Yes and "N0"

I do agree that the N0 can fluctuate dramatically due to variables such as you have described, however I still maintain that a "session" can be quantified mathematically given certain parameters. In the SD world, there are fixed protocol followed almost to the letter by pits everywhere, i.e Rox which in reality average out to Ro6, since I will not touch a game offering only Ro5. The bet spread also is pretty much locked in at 1:4. These factors, therefore make it very easy to calculate an accurate quantity of "session".
In the DD game, the same set of scenarios holds pretty much true as well. Since I have a personal threshhold for min. pen. and the utilize the same spread cosistently, I believe the same holds true there as well. The only possible pitfall in this consitency would be in shoe games where pen, speed and wonging could dramatically affect accuracy here.
In an interesting coincidence Norm posted a graph the other day in a thread entitled "Is it true?" where he displays the results of trip wins and in essence visually dramatizes basically the same fundamental concept I'm trying to get at here, but in terms of win ratios rather than units of N0.

 

[blackjack avenger]

Response to Yes and "NO"

If you are looking for a winning amount to determine a session then that could be NO which I believe is the number of hands to double bank if you are at near perfect expectation, or a fraction of NO if that number seems to high. There is research around involving the time to double bank etc. A problem you may face is the shorter the time duration for a session the more meaningless it is because you are just blowing in the winds of SD.

 

[bj bob]

If you are looking for a winning amount to determine a session then that could be NO which I believe is the number of hands to double bank if you are at near perfect expectation, or a fraction of NO if that number seems to high. There is research around involving the time to double bank etc. A problem you may face is the shorter the time duration for a session the more meaningless it is because you are just blowing in the winds of SD.

I think you are missing my point here. I wasn't looking for a winning amount to define a session, rather a # of hands (time frame) that is a meaningfully mathematical increment of the finite term of N0. If N0 is indeed a finite # of hands to overcome SD's, then it should logically follow that this number can be further broken down into increments called sessions. My intuition is telling me that a session would come out to approx. 2-3% of N0 and therefore, if 50-66 session results were graphed together you would therefore have a graph of the complete picture.

 

[ihate17]

I will stick with my original defination

A session for myself is: whatever I determine it to be because it is all just one big session.

ihate17

 

[Automatic Monkey]

I think you are missing my point here. I wasn't looking for a winning amount to define a session, rather a # of hands (time frame) that is a meaningfully mathematical increment of the finite term of N0. If N0 is indeed a finite # of hands to overcome SD's, then it should logically follow that this number can be further broken down into increments called sessions. My intuition is telling me that a session would come out to approx. 2-3% of N0 and therefore, if 50-66 session results were graphed together you would therefore have a graph of the complete picture.

That sounds like a reasonable usage for this kind of a number- record keeping. It's going to change from game to game so if you are doing any graphing of results it would be nice to space your data points at an increment that makes sense relative to the NO of the game you are playing. I'm not a believer in keeping win/loss records myself because it can compound the illusions we face with winning and losing, but it's useful for many and an AC shoe player and a SD/DD player are going to need to look at their records in a different way.

 

[Kasi]

I I wasn't looking for a winning amount to define a session, rather a # of hands (time frame) that is a meaningfully mathematical increment of the finite term of N0. If N0 is indeed a finite # of hands to overcome SD's, then it should logically follow that this number can be further broken down into increments called sessions. My intuition is telling me that a session would come out to approx. 2-3% of N0 and therefore, if 50-66 session results were graphed together you would therefore have a graph of the complete picture.

I'm still a little in the dark of what you are trying to do or why, N0 is just a theoretical number for when it's reasonable to conclude your results are more than luck. Some say it's when EV-1 stan dev=0, others choose to say it's 2 stan dev.

So, if you want to relate each session to EV for that session and the SD for that session, both of which soley depend on the number of hands played, (assuming pen never changes etc - big assumption), that's cool. It won't matter how much you won or lost for that session then. But after the N0 hands have occurred you might to be able to compare how much you actually won to how much you were expected to win.

If you want to relate it to wagering bankroll so many times, you need to know your average initial bet first.

 

[bj bob]

I'm still a little in the dark of what you are trying to do or whyIf you want to relate it to wagering bankroll so many times, you need to know your average initial bet first.

I'm simply attempting to ballpark a general definition of the term based on some valid mathematical theory, such as N0. so, therefore if someone were to plan a trip to LV for 4 days and plan to spend 25 hours of net table time, he would have a good idea of how much to take (or have available) with his bet spread and table conditions at hand. According to my hypothesis, if he were to play 6D, ideal rules and pen, he would have to be prepared to "recirculate" his overall bankroll approx. 6 times, so he would have to calculate how much of his BR he would need to bring in order to do this. Since avg. bets are easily obtainable as to min/max units, the rest of the calculation should be a breeze. It's that simple.

 

[Kasi]

if someone were to plan a trip to LV for 4 days and plan to spend 25 hours of net table time, he would have a good idea of how much to take (or have available) with his bet spread and table conditions at hand.

OK - maybe we're getting somewhere now lol.

It sounds like all you want to do is calculate a Trip ROR - the amount of money to bring assuming so many hours of play at some risk of losing it all you are comfortable with.

Short answer - probably more than you think. I'd guess 50% of total roll anyway for 25 hours might get you to around a 10% risk. But that's just a guess so don't put too much stock in it. When in doubt, bring more.

But I think all you need is the EV of the trip in units and the standard deviation of the win to calculate the units needed for various levels of risk.

How many units of your total bankroll are you coming up with that you should bring for your trip? And what risk do you think it is?