6516 hands of BJ played Concerned with my results

[JohnCrover]

Hello, I have been playing with the basic strategy trainer on this website and I have been losing more than usual and quicker than usual and I thought it was strange so I decided to track my results on Excel. I have played 6,516 hands so far and the results have been dismal, I'm down 587.5. (The blue line is hands played and the orang one is won/lost) These are the parameters for the game: 6 deck, H soft 17, DAS, DA2, No RSA, One card per Ace, split to 4 hands, Late surrender, 1 deck cut off from play, Hi-lo with the Illustrious 18+ a few extra indexes but lets just call it the I18 for simplicity, for the most part no wonging, 1 deck cut from play, the bet spread : 5 dollars TC 0 or less, TC 1 : 15, TC 2:30 TC3: 45 TC 4:60, and 60 is the max bet.
Does anyone know the chances of these dismal results occurring? With this many hands played this is easily a 25 hour down swing.

 

[21forme]

My browser doesn't support Flash, so can't look at it. However, are you sure it doesn't reshuffle after every hand or use an infinite deck?

 

[JohnCrover]

If you asked me a year ago I would say "Yes, i'm sure it doesn't use infinite decks or an automatic shuffler" because I have beaten it in the past but now i'm starting to get skeptical because my results have been so bad. It's not supposed to use an automatic shuffler.

 

[ZenKinG]

6000 hands and you're concerned? Wait till you try to do this for any significant amount of time and go through 300 hour losing streaks. Even better try going through that in vegas playing half hour sessions and putting in over 200 hours in simply walking only to still be down. Then get back to me. If you haven't played at least 20k hands, results are irrelevant, even after 20k hands results still are irrelevant. Get to 40,000 - 60,000 hands then your results should be somewhat near EV. Unless of course you're me and have played over 60,000 rounds and you're down 1SD and at nearly half your EV playing a flawless game and getting one outed to 21 time and time again with two 20s on the felt against a stiff with max bets.

 

[JohnCrover]

6000 hands and you're concerned? Wait till you try to do this for any significant amount of time and go through 300 hour losing streaks. Even better try going through that in vegas playing half hour sessions and putting in over 200 hours in simply walking only to still be down. Then get back to me. If you haven't played at least 20k hands, results are irrelevant, even after 20k hands results still are irrelevant. Get to 40,000 - 60,000 hands then your results should be somewhat near EV. Unless of course you're me and have played over 60,000 rounds and you're down 1SD and at nearly half your EV playing a flawless game and getting one outed to 21 time and time again with two 20s on the felt against a stiff with max bets.

I have 20k hands played in my previous simulation and i'm up about 5.6k but I lost the data since then. I have began to run a new simulation (the one you see now) because I have been playing the trainer recreationally and have noticed abnormal results.
Thank you for your constructive input though, it's always a pleasure Zenking.

 

[JohnCrover]

Okay so i'm beginning to understand. I just came across an amazing post online explaining standard deviation here: http://www.gamesblackjack.org/strategy/standard-deviation.php
I can answer my problem by taking the square root of the numbers I have played and multiplying it by the standard deviation of the game, which is roughly 1.1418.
The square root of the number of hands I have played multiplied by 1.1418 will give me a X and X is the amount of units I can expect to win 68% of the time over the amount of hands I have played. That's known as one standard deviation. 95% of the time is 2 standard deviation which is just X +X so if X were 5 then 5+5=10 and I would win or lose 10 units over the amount of hands I have played 95% of the time and 3 standard deviations would just be X+X+X and I will win that many units 99.7% of the time and that is 3 standard deviations.
Question to the math guys out there: How do I take in to account the house edge, or player edge? If i'm assuming a 1$ player edge then how do I add that to the equation?

 

[Joe Mama]

The HE or player EV is at the center of the bell shaped curve. If, for example, EV is $1/hand and you have played 6516 hands then the center is $6516.

 

[DSchles]

Okay so i'm beginning to understand. I just came across an amazing post online explaining standard deviation here: http://www.gamesblackjack.org/strategy/standard-deviation.php
I can answer my problem by taking the square root of the numbers I have played and multiplying it by the standard deviation of the game, which is roughly 1.1418.
The square root of the number of hands I have played multiplied by 1.1418 will give me a X and X is the amount of units I can expect to win 68% of the time over the amount of hands I have played. That's known as one standard deviation. 95% of the time is 2 standard deviation which is just X +X so if X were 5 then 5+5=10 and I would win or lose 10 units over the amount of hands I have played 95% of the time and 3 standard deviations would just be X+X+X and I will win that many units 99.7% of the time and that is 3 standard deviations.
Question to the math guys out there: How do I take in to account the house edge, or player edge? If i'm assuming a 1$ player edge then how do I add that to the equation?

No, you can't do it this way. The 1.14 works only for flat betting. You aren't betting the same amount on each hand. You're spreading your bets, betting different amounts at each count, and the s.d. will be much higher than the way you're suggesting. Nor can you use 1.14 with the AVERAGE bet over all hands. That doesn't work either and isn't high enough.

I explain how to do this properly on pp. 19-20 of BJA3.

Don

 

[JohnCrover]

Ahh right, I understand my mistake, the standard deviation is obviously going to be higher on a per hand basis when the bets are larger than the previous one. I'm going to read more on it then get back to you guys if I have any questions.
Thank you