Oscar's grind stats

[Joker21]

Hi everyone,

New to the forum after having a look around. Seems to be a wealth of knowledge on here and hoping someone may be able to answer a question I have.

Has anyone ever done a sim or are there available stats on how many times an OG sequence will hit 5 units down over an extended period i.e. 2000 sequences?

 

[shadroch]

Can't help you with sims, but I believe I've used OG in live play as much as anyone here. Getting down 5 units is not uncommon in the least. In a two hour session, it may happen several times.

 

[Joker21]

Hey thanks for the input. And 2hrs being roughly 30-40 sequences? So we could assume 10% of the time?

And I also read on here how you take it to a 40x stop loss. Any idea of rough percentage of time you will hit this limit within a specified timeframe? Sorry for all the questions being that you have actually put this to play live would be interested in your experience.

 

[shadroch]

Let me say, right off the top, that if you are playing a game that is can be counted, counting is a much better way to go.
I play using OG at one casino on an electronic game that you can't count on, but gives points similar to a slot club. I play there only for comps, and try to break even on the game itself. For me, in those circumstances, it's the right play. I don't advise playing BJ using OG in 99% of the circumstances.
The machine I play starts at $1, so the semi-rare situations where I find myself down 10-15 units, I'm only down $10-$15. On even a $5 table, being down 15 units means you are down much more.
My own circumstances are playing the machines I've described, I am up a few dollars after several thousand hands( perhaps as many as 10,000).
More importantly, to me, I've gotten well over $1,000 in room comps, a couple hundred in free play and about the same in food and show comps.

 

[Joker21]

I am up a few dollars after several thousand hands( perhaps as many as 10,000)

Thanks for sharing. I thought I read somewhere that you were trying something out using OG 40x stop 20x win on $5 table. Apologies if this was not you. Using above quote as a guide approx how many times did you reach your 40x stop limit?

 

[Baseball John]

The following stats are actual dealt hands. The game was 6 decks DAS LS
S17. House Advantage on this site lists -.36%.

814 shoes. 42,859 hands dealt excluding pushes. 5,738 sequences. Every hand was exactly according to Basic Strategy. No insurance or indices were used. All play was head to head against dealer. Penetration was not important since counting was not involved.

I dont know how many times the sequence reached -5 units but the following figures show the low point for each shoe. When a shoe ended I did not stop the sequence , I continued where the last shoe finished. The actual results and disadvantage I played showed (-.44%) after almost 43,000 hands. this is a little above average but the sample is only 43,000 hands. After about 250,000 hands the disadvantage went to (-.375%) which is much closer to
(-.036%) The Avg. Bet was 9.2 units for these 43,000 hands.

UNITS SHOES
01-10 .... 202
11-20 .... 182
21-30 .... 98
31-50 .... 98
51-100 .... 82
101-200 .... 70
201-500 .... 42
501-1000 .... 17
1001-2000 .... 11
2001-5000 .... 12
5001-10000 .... 0
over 10000 .... 0

REMEMBER THESE ARE THE LOW POINT OF EACH SHOE.

 

[Joker21]

814 shoes. 42,859 hands dealt excluding pushes. 5,738 sequences.


UNITS SHOES
01-10 .... 202
11-20 .... 182
21-30 .... 98
31-50 .... 98
51-100 .... 82
101-200 .... 70
201-500 .... 42
501-1000 .... 17
1001-2000 .... 11
2001-5000 .... 12
5001-10000 .... 0
over 10000 .... 0

REMEMBER THESE ARE THE LOW POINT OF EACH SHOE.

So by those figures if you were running a stop loss of 50x units you would bump that about 70% of the time give or take!

 

[k_c]

Hi everyone,

New to the forum after having a look around. Seems to be a wealth of knowledge on here and hoping someone may be able to answer a question I have.

Has anyone ever done a sim or are there available stats on how many times an OG sequence will hit 5 units down over an extended period i.e. 2000 sequences?

From time to time Oscar's Grind Progression seems to come up so in Oscar's honor I wrote a program that simulates the Grind and uploaded it to my website in this post. It's still there. Link to download is at bottom of post.

It's not much of a program. It needs to be rerun for every set of inputs.

Inputs are:

Number of times out of 32768 grinder expects to win a single bet. An input of 16384 means grinder has a 50% chance of winning like in a fair coin flip. An input of 16153 means grinder has about a 49.295% chance of winning, which is close to a craps pass line bet. 32768 would be a sure win. 0 would be a sure loss.

Bank - number of units grinder is willing to risk

Goal - if grinder reaches this number of units he succeeds. he can celebrate or do whatever but when he continues he will restart the grind with the same bank. if he gets to a point where his bank has dwindled to where he cannot pony up the next bet he fails. at this point he may want to do some serious drinking because he needs to figure a way to replenish his bank in order to restart grinding.

Trials - number of times grind is restarted

Outputs are:

Number of bets won if instead of grinding a flat bet of 1 unit was made
Number of bets lost if instead of grinding a flat bet of 1 unit was made
+/- if instead of grinding a flat bet of 1 unit was made

Number of times grind succeeded in reaching goal
Number of times grind failed with insufficient funds to continue
+/- Using the grind

 

[Joker21]

Thanks guys. Great setup sort of worked it out lol. I'm not sure if I'm doing it right however, the sim I would like to run is with a bankroll of 5k betting $10 units, how many times would you hit 40x (so $400 down) stop over say 2000 sequences.

 

[Baseball John]

Joker21
yes . Each shoe needed to go beyond 50 units about 27% of time. As I said before these numbers are for each shoe of about 50 hands not for each betting sequence.
When you implement stop losses this is a losing progression until you get very very high losses and recoup them. Well into the thousands.

 

[Machinist]

Please

Friends dont let friends, Grind Oscar.... I know a good charity...
Or i could possibly take your action...:laugh:
Just do a whole bunch of sims..... and figure your losses... Then take those losses, reach into your wallet.... go into the bathroom and flush that money down the toilet... Then come back on here and tell us how you feel....:cool2:
Now of course if this is a life and death situation... say you are gonna die in a week........... Hell ya baby let her ride!!!!!!!!!!!!!!!!! Probably beat the short run,,,,not so much to the long run!!!!


Machinist

 

[Joker21]

:laugh: I hear you .............. just something I'm playing around with that's all. I think we all know how I'd feel :cry:

 

[Machinist]

whew

I had hoped you would take my post light heartendly ... lol:laugh::laugh::laugh:

Machinist

 

[Baseball John]

Oscar's Grind

Joker21 :
here are the sequence stats for the first 2132 sequences:
Units ..................... Sequences

zero .............. 980
1-10 ............ 840
11-20 ........... 141
21-30 ........... 47
31-50 .......... 46
51-100 .............. 27
101-200 .............. 29
201-500 ............. 16
501-1,000 ........... 3
1001-2,000 ........... 1
2001-5,000 ............ 2
5001-10,000 .......... 0
over 10K ............ 0

This was 14,840 hands. Avg bet of 8.4 units. House advantage for these hands was (-.54%).

 

[Joker21]

I had hoped you would take my post light heartendly ... lol:laugh::laugh::laugh:

Machinist

:laugh: Of course. Any resistance to advice from those who have been there before is foolish IMO

 

[Joker21]

Joker21 :
here are the sequence stats for the first 2132 sequences:
Units ..................... Sequences

zero .............. 980
1-10 ............ 840
11-20 ........... 141
21-30 ........... 47
31-50 .......... 46
51-100 .............. 27
101-200 .............. 29
201-500 ............. 16
501-1,000 ........... 3
1001-2,000 ........... 1
2001-5,000 ............ 2
5001-10,000 .......... 0
over 10K ............ 0

This was 14,840 hands. Avg bet of 8.4 units. House advantage for these hands was (-.54%).

Thanks John,

So over the 2132 sequences you would have to hit over 50 units down 78 times or roughly 3.7% of the time. If you ran this sim over and over how varied would the results be? Thanks again for the input.

 

[Baseball John]

Joker21 .
Yes that is true.
This is not a sim. I don't know of any software capable of betting in that sequence. I have about 250,000 hands and the results pretty much remain consistent. Just remember if you play the sequences with a stop loss you will lose.

 

[k_c]

Thanks guys. Great setup sort of worked it out lol. I'm not sure if I'm doing it right however, the sim I would like to run is with a bankroll of 5k betting $10 units, how many times would you hit 40x (so $400 down) stop over say 2000 sequences.

I think you're saying you are willing to risk 40 units, so your bank=40. Also you are saying you want to restart the grind 2000 times, so number of trials=2000

Now 2 other inputs are needed:

1) What will be your quitting point for success? Will you be happy with 41 units (1 unit profit)? 42 units (2 unit profit)? 100 units (60 unit profit)? to end a sequence. You have defined each unit = $10 above.

2) What is your win rate 50%, <50%, >50%? (16384 is 50%), (< 16384 is < 50%), (> 16384 is > 50%). 50% win rate = 0% ev; 49% win rate = -2% ev; 51% win rate = +2% ev

Once you've determined the 2 missing inputs you will be able to sim your 2000 sequences, although 2000 isn't a large number. You would be better off to input 20000000 instead of 2000 and then divide results by 10000 to get average result for each 2000 sequences.

I think you'll find in general that EV, which is determined by win rate, is what is important. The program simulates a fixed ev, which means win rate is the same for every w/l opportunity and also assumes all wins are paid at 1 to 1 odds.