1,000 hands or 100 million hands

[Rainmaker]

If you play a thousand hands, people would say the sample size is far too small to mean anything. But no one is going to play a trillion hands. So at what point can I look at my results and say "I'm playing with an advantage, this is more than just luck". Is the answer never? If so, what's the point?

[DBJT]

Is the answer never? If so, what's the point?

Certainly not! The answer you're grasping at can be found in statistics, such as N0 (pronounced "enn-zero").

A quick Google search turned up a few threads discussing it:

http://www.blackjacktheforum.com/sho...e-score-N0-etc

http://www.blackjackinfo.com/bb/showthread.php?t=18006

http://www.blackjackinfo.com/bb/showthread.php?t=15951

-DBJT

[hardin county boy]

If you play a thousand hands, people would say the sample size is far too small to mean anything. But no one is going to play a trillion hands. So at what point can I look at my results and say "I'm playing with an advantage, this is more than just luck". Is the answer never? If so, what's the point?

this is a great question on two levels:

first, variance = 100%, luck and the effect of this is LARGE. advantage play = STATISTICALLY TRUE BUT THE EFFECT OF THIS IS......small, compared to variance . bet-sizing helps bridge the gap but only "some" and this finds plenty of advantage players waiting a LOOOOONG time to realize a secure gain and, with large spreads and bankrolls, the hole can be considerable ---we all know, the longer the advantage player perseveres, the more likely he is to enjoy being a winner BUT, he remains practically subject to negative variance till his lifetime win is SO VERY LARGE that negative variance is UNLIKELY to return him to even, or worse. of course, with such a win, many will choose to enlarge their starting unit bet size so that their possibility of a DEEP hole is, again, a realistic prospect.

second, i think the dedicated professional counter, with many hands per month that turns into years, will almost surely reach winning status that grows and becomes secure. i doubt the percentage of lesser-playing counters who play "A LOT" and reach ever higher unit sizes, such that they will NOT increase it any more---i doubt the percentage who get there with less than full time play is very encouraging. (those of us who get 1000 hands a month are playing "a lot", compared to some of our friends but our prospects of achieving the fruit of "the long run" by anything other than luck + counting is low--variance is far more significant). so, those who play A LOT (much more than 1000 hands a month) but not full-time, can expect an experience in between and if we screw up the count, bet sizing, bankroll sizing etc, make that more toward the negative expectation, especially if a good experience finds us re-setting the bankroll clock to zero with larger starting units.

[DBJT]

...but our prospects of achieving the fruit of "the long run" by anything other than luck + counting is low...

The TL;DR version of your post? "Playing Solo As A Simple Straight Old-Fashioned Card Counter Is A Tough Grind".

One possible remedy is TEAM PLAY

-DBJT

P.S. Of course, there are others, too...

[hardin county boy]

Certainly not! The answer you're grasping at can be found in statistics, such as N0 (pronounced "enn-zero").

-DBJT

thanks DBJT---i looked at the threads and they were helpful though, for those who do not fully follow it, could i re-state the original question as: assuming whatever counting approach YOU choose, executed perfectly, with rules that are "good enough" and available enough, how many hands would need to be played for 68% of players to be up 5000 units?

[Three]

we all know, the longer the advantage player perseveres, the more likely he is to enjoy being a winner BUT, he remains practically subject to negative variance till his lifetime win is SO VERY LARGE that negative variance is UNLIKELY to return him to even, or worse. of course, with such a win, many will choose to enlarge their starting unit bet size so that their possibility of a DEEP hole is, again, a realistic prospect.

The nature of the statistics can be looked at in 2 ways that seem to create a paradox as both are in fact correct:
1) If you have experienced large losses in the past and are in the hole quite a bit you are not any more likely to win than in the beginning.
2) You have an EV for your play and after enough rounds you will approach that EV for the time you played.

Both of these statements are correct but they don't seem to go well together. The truth is you are never "due" for a winning run and your future results have nothing to due with past success or failure. But a long enough sampling of your play there will be large losing runs. The fact that you started with one of these doesn't mean you won't approach your EV after enough play. This doesn't mean you are due but only that you have experienced one of many downswings that occur in the long run and will eventually be erased by winning runs if you have a +EV. It seems both statements are mutually exclusive but both statements are true.

[hardin county boy]

well said, Tthree.

[Rainmaker]

The nature of the statistics can be looked at in 2 ways that seem to create a paradox as both are in fact correct: 1) If you have experienced large losses in the past and are in the hole quite a bit you are not any more likely to win than in the beginning. 2) You have an EV for your play and after enough rounds you will approach that EV for the time you played.Both of these statements are correct but they don't seem to go well together. The truth is you are never "due" for a winning run and your future results have nothing to due with past success or failure. But a long enough sampling of your play there will be large losing runs. The fact that you started with one of these doesn't mean you won't approach your EV after enough play. This doesn't mean you are due but only that you have experienced one of many downswings that occur in the long run and will eventually be erased by winning runs if you have a +EV. It seems both statements are mutually exclusive but both statements are true.

I hate variance. It would be easier if they just mailed me a check.

[Pacman]

Casinos are mailing EV checks now? Finally! >.<
Can this be done anonymously? Or atleast pay me in cash when I leave?

~Pac

[ZenKinG]

EV checks LOL

[bigplayer]

There's a difference between a sample size big enough to know if you're playing a winning game and a sample size large enough to prove a scientific point down to two or three decimal points. You're confusing the two. Depending on the game, 50,000 to 100,000 rounds should be enough to give you a good chance of being near break-even or ahead with a fairly wide variation from huge winner to small winner. If you play 100,000 hands and are a significant net loser you're doing something wrong.

[Norm]

What he said.

[mofungoo]

To guarantee results within 10% of EV will take over 1,000,000 hands, the exact number of hands depends on the rules and quality of the game you play.