Software for Progression Systems?
- Thread starter justmaybe
- Start date
With a 500 unit roll and a 200 unit goal, you have met your goal before even starting. That's what the program is reporting.:grin: I had a feeling that would be a source of confusion, but I was too lazy to program that in. Just pretend a message appears saying you already have met your goal any time bank >= goal.
k_c
lol.
So, if I put in a 499 unit roll with a goal of 500 units I will get how often I will win 1 unit?
If so, I won 299 out of 300 so that's not bad!
Cool. And thanks. And then some.
Amazing how much longer the losing sessions are. Maybe if you don't win early, you won't win at all.
With 3.5 throws for a pass-line decision, and 100 throws an hour, at least it seems I could be standing at that table for a pretty long time before I changed my unit size
Next thing you know, you'll be the foremost authority on even-money betting systems!
If you can take the heat of the reputation of knowing the most about worthless betting systems
Not that there's anything wrong with that, as Seinfeld used to say:grin:
Risk of ruin is .002=.2% on a fair coin toss flat bet for a bank of 499 and goal of 500. I'm not sure what it is for Oscar's Grind, but it is probably more since bets can be more than 1 unit. I guess the way to find out by simulation would be skip notifying the user after each trial and let the computer do its thing for a long time, while recording each result. It woudn't be that hard to change the program to do this if you want.
k_c
My 5000 attempts had 11 failures. Why did you stop at 300? Finger tired of holding down the enter key? You can set a glass on it and come back 5 minutes later. :grin:
New version
New version allows input of larger numbers and computes without stopping. Bank, goal and number of trials can be <=32767. Larger numbers can take a long time, but you can cancel and display results up to cancel.
I ran a bank of 499 and goal of 500 units and 10000 trials about 4 or 5 times. Results were 18-26 failures per 10000 trials.
k_c
New version allows input of larger numbers and computes without stopping. Bank, goal and number of trials can be <=32767. Larger numbers can take a long time, but you can cancel and display results up to cancel.
I ran a bank of 499 and goal of 500 units and 10000 trials about 4 or 5 times. Results were 18-26 failures per 10000 trials.
k_c
Oscar's paradox?
I input Bank=10, Goal=20, EV = 0 (same as fair coin toss,) Trials=10000 several times into my program and it outputs about 46% successes and 54% failures. This means 46% of the trials 10 units is won and 54% 10 units is lost. This seems like a paradox since this is for a fair coin toss where EV=0.
So, I wrote a program in C++ to check this and here an output for a million random trials:
As you can see, total wins and total losses are about 50-50 as expected. When I started this post I thought that this might indicate Oscar's Grind is worse than flat betting, but now I think I see what is happening. Each of the successes realizes exactly a 10 unit profit. However, each of the failures realizes a maximum loss of 10 units but sometimes less because it is a failure if you don't have enough for the next bet. So if you have 2 units left and the Grind requires a bet of 3 units, that is a failure but you have only lost 8 units.
There is no paradox, after all.
k_c
I input Bank=10, Goal=20, EV = 0 (same as fair coin toss,) Trials=10000 several times into my program and it outputs about 46% successes and 54% failures. This means 46% of the trials 10 units is won and 54% 10 units is lost. This seems like a paradox since this is for a fair coin toss where EV=0.
So, I wrote a program in C++ to check this and here an output for a million random trials:
Code:
Bank: 10
Goal: 20
EV (-1 to 1): 0
Trials: 1000000
Grind successes: 462947
Grind failures: 537053
Wins: 14927403
Losses: 14921601As you can see, total wins and total losses are about 50-50 as expected. When I started this post I thought that this might indicate Oscar's Grind is worse than flat betting, but now I think I see what is happening. Each of the successes realizes exactly a 10 unit profit. However, each of the failures realizes a maximum loss of 10 units but sometimes less because it is a failure if you don't have enough for the next bet. So if you have 2 units left and the Grind requires a bet of 3 units, that is a failure but you have only lost 8 units.
There is no paradox, after all.
k_c
Thanks k_c.
Since the definition of a sequence is to only win 1 unit, I was wondering how that worked with winning 10 units. Like you win 1 unit 9 times, lose 8 units the next time, and then win 9 more times would that still count as failure?
I know I'm not making myself very clear lol - stay tuned tomorrow maybe lol.
If u or miplet ever get really really bored it would be interesting to know how doubling a 10 unit roll, or any of this stuff, compares to flat-betting lol.
For Kasi
I added gross results. In order to be unbiased, I recorded 10 consecutive 1,000,000 trial runs. It looks like flat betting outperformed the grind on this 10M but I don't think that proves anything. Maybe on the next 10M the grind might outperform flat betting. I'm sure that for a very large number of trials both the grind and flat betting would tend to break even since EV=0. That's just the math of it.
k_c
I added gross results. In order to be unbiased, I recorded 10 consecutive 1,000,000 trial runs. It looks like flat betting outperformed the grind on this 10M but I don't think that proves anything. Maybe on the next 10M the grind might outperform flat betting. I'm sure that for a very large number of trials both the grind and flat betting would tend to break even since EV=0. That's just the math of it.
k_c
Code:
(Run 1)
Bank: 10
Goal: 20
EV (-1 to 1): 0
Trials: 1000000
Successes: 461976
Failures: 538024
Wins: 14917292
Losses: 14914265
Gross (flat bet): 3027
Gross (using Oscar's Grind): -3349
(Run 2)
Bank: 10
Goal: 20
EV (-1 to 1): 0
Trials: 1000000
Successes: 461814
Failures: 538186
Wins: 14911579
Losses: 14919134
Gross (flat bet): -7555
Gross (using Oscar's Grind): -6519
(Run 3)
Bank: 10
Goal: 20
EV (-1 to 1): 0
Trials: 1000000
Successes: 461883
Failures: 538117
Wins: 14915525
Losses: 14912562
Gross (flat bet): 2963
Gross (using Oscar's Grind): -4810
(Run 4)
Bank: 10
Goal: 20
EV (-1 to 1): 0
Trials: 1000000
Successes: 462434
Failures: 537566
Wins: 14928693
Losses: 14927278
Gross (flat bet): 1415
Gross (using Oscar's Grind): 5577
(Run 5)
Bank: 10
Goal: 20
EV (-1 to 1): 0
Trials: 1000000
Successes: 462306
Failures: 537694
Wins: 14925329
Losses: 14919091
Gross (flat bet): 6238
Gross (using Oscar's Grind): 2576
(Run 6)
Bank: 10
Goal: 20
EV (-1 to 1): 0
Trials: 1000000
Successes: 461534
Failures: 538466
Wins: 14906342
Losses: 14906394
Gross (flat bet): -52
Gross (using Oscar's Grind): -9875
(Run 7)
Bank: 10
Goal: 20
EV (-1 to 1): 0
Trials: 1000000
Successes: 462742
Failures: 537258
Wins: 14925590
Losses: 14918911
Gross (flat bet): 6679
Gross (using Oscar's Grind): 10505
(Run 8)
Bank: 10
Goal: 20
EV (-1 to 1): 0
Trials: 1000000
Successes: 461834
Failures: 538166
Wins: 14928653
Losses: 14932867
Gross (flat bet): -4214
Gross (using Oscar's Grind): -6102
(Run 9)
Bank: 10
Goal: 20
EV (-1 to 1): 0
Trials: 1000000
Successes: 462910
Failures: 537090
Wins: 14912163
Losses: 14907565
Gross (flat bet): 4598
Gross (using Oscar's Grind): 13921
(Run 10)
Bank: 10
Goal: 20
EV (-1 to 1): 0
Trials: 1000000
Successes: 461944
Failures: 538056
Wins: 14907818
Losses: 14910043
Gross (flat bet): -2225
Gross (using Oscar's Grind): -3325