Also, maybe check your deviation numbers in the Playing Strategy table.
Also, maybe check your deviation numbers in the Playing Strategy table.
I’ll take a look soon, but Wong says in his book that the average win is about $20 an hour with a S.D.number. of about $400. Don knows this better.
At 100 million hands; 6 deck, s17, $10/$25/$50/$75/$100/$125; tc 0 or below/+1/+2/+3/+4/+5, no rsa, das, no surrender, 75% penetration, tc floored:
Sweet 16:
$16.07 per 100 hands; SD = $377.57; Win of 0.656% of total $ bet; 0.747% of initial hand $ bet
Basic strategy, insurance at tc +3 floored:
$12.02 per 100 hands; SD = $362.26; Win of 0.501% of total $ bet; 0.569% of initial hand $ bet
Phil
I’ll take a look soon, but Wong says in his book that the average win is about $20 an hour with a S.D.number. of about $400. Don knows this better.
If you can get $18 with basic strategy and insurance, would much prefer that to $12!
Doubt that it's $18. More like $14 or $15. Why don't you pick one of the games and bet spreads listed in Chapter 10 of BJA3, or in the new book, to see if you can replicate what was done there?
Don
Doubt that it's $18. More like $14 or $15. Why don't you pick one of the games and bet spreads listed in Chapter 10 of BJA3, or in the new book, to see if you can replicate what was done there?
Don
I have been running table 10.4.3 (4.5/6 S17 DAS) from Blackjack Attack 3 page 236 with Play All Practical 1-12 with Illustrious 18 from page 213. I am getting W/L % of 0.807% and W/100 of $18.87, SD of $432.327 average initial bet of $23.38
I am likely somewhat off as the table in the book has results of W/L % of 0.83% and W/100 of $19.49, SD of $431.11, and average bet of $23.49.
Phil
Did you follow ALL of the guidelines for the sims, which precede the tables? For example, do you have four players at the table, with the sim player playing last? That makes a difference.
Don
I was running it heads up, thank you for alerting me to this. I think I have all the other guidelines in. I will run it as the 4th of 4 players.
But I am seeing some interesting (higher $) results as player 4 of 5 players with the game in 10.43 Play all Practical 1-12.
Am getting W/L % = 0.861%, W/L 100 = $20.36, SD = $439.03, avg bet = $23.65.
Currently the extra players in my simulator just receive two cards and don't do anything else. It was this way to save speed. With this I can see I am getting a fairly significantly higher proportion of hands at higher counts than with heads up.
I will run it as the 4th of 4 players as it currently is and eventually with the extra players using basic strategy.
Thanks,
Phil
Did you follow ALL of the guidelines for the sims, which precede the tables? For example, do you have four players at the table, with the sim player playing last? That makes a difference.
Don
Would you expect the returns to be lower heads up for table 10.43 Play All Practical 1-12 than with 3 players to the right for this case? I will probably still be running mostly just one on one for a while so was wondering what might be expected.
Thank you,
Phil
But I am seeing some interesting (higher $) results as player 4 of 5 players with the game in 10.43 Play all Practical 1-12.
Phil
I do not understand this paragraph at all! Also, this multiple-players effect was investigated by this simulator here three years ago, but it came up again?
Did you follow ALL of the guidelines for the sims, which precede the tables? For example, do you have four players at the table, with the sim player playing last? That makes a difference.
Don
Would you expect the returns to be lower heads up for table 10.43 Play All Practical 1-12 than with 3 players to the right for this case? I will probably still be running mostly just one on one for a while so was wondering what might be expected.
Thank you,
Phil
Yes, the extra players to the right allow you to see more cards before making the decision on our hand, effectively increasing penetration before you make the play.
Don
Did you follow ALL of the guidelines for the sims, which precede the tables? For example, do you have four players at the table, with the sim player playing last? That makes a difference.
Don
Would you expect the returns to be lower heads up for table 10.43 Play All Practical 1-12 than with 3 players to the right for this case? I will probably still be running mostly just one on one for a while so was wondering what might be expected.
Thank you,
Phil
Yes, the extra players to the right allow you to see more cards before making the decision on our hand, effectively increasing penetration before you make the play.
Don
Ok, yes, thanks.
Phil
But I am seeing some interesting (higher $) results as player 4 of 5 players with the game in 10.43 Play all Practical 1-12.
Phil
I do not understand this paragraph at all! Also, this multiple-players effect was investigated by this simulator here three years ago, but it came up again?
I was getting some higher returns in this case, possibly higher than expected, was just posting that. A while back I think we investigated counts around the table.
Thanks,
Phil
Before we compare different simulators, we need to clarify the definition of variance. The variance of X is equate to the mean of the square of X minus the square of the mean of X. In blackjack, it’s more complicated than this though.
Let’s use this hand, T,6 v A, as an example to demonstrate. Here, X=EV, the gain from a hand. At TC=0, player bets $10 on the main and then loses it, so the variance component from this hand is (10x1)^2=100 minus the square of the mean of EV without insurance.
At TC=+4, player bets $100 on the main and $50 on insurance and then loses all, so the variance component from this hand is (100x(0.5+1))^2=22500 minus the square of the mean of EV with insurance.
The mean of EV is usually tiny, about $10x0.01, so the square of it is 0.01, negligible to the total variance. In your simulator, do you set it as a fixed value of zero? If not, how do you choose the mean of EV with insurance? And without?
Before we compare different simulators, we need to clarify the definition of variance. The variance of X is equate to the mean of the square of X minus the square of the mean of X. In blackjack, it’s more complicated than this though.
Let’s use this hand, T,6 v A, as an example to demonstrate. Here, X=EV, the gain from a hand. At TC=0, player bets $10 on the main and then loses it, so the variance component from this hand is (10x1)^2=100 minus the square of the mean of EV without insurance.
At TC=+4, player bets $100 on the main and $50 on insurance and then loses all, so the variance component from this hand is (100x(0.5+1))^2=22500 minus the square of the mean of EV with insurance.
The mean of EV is usually tiny, about $10x0.01, so the square of it is 0.01, negligible to the total variance. In your simulator, do you set it as a fixed value of zero? If not, how do you choose the mean of EV with insurance? And without?
I don’t know how insurance always affects variance. Also, I don’t choose EV, it depends on the game, betting, playing. After all the bets and results are recorded, EV, variance and standard deviation are calculated. Also, sometimes insurance is won resulting in 0 won or lost overall on a hand.
In the majority of these situations, player will lose both in the same hand.
However, at TC=+4, player may surrender T,6 v A after losing insurance and thus loses one half of the $100 main bet, so the variance component from this hand becomes (100x(0.5+0.5))^2=10000 minus the square of the mean of EV with surrendering. What is the mean of EV now?
.
