Risk of Ruin help

[blackjack avenger]

Those Formulas and a Few Others

There are also formulas for only playing until x amount is won and for only playing until y time is reached and even if considering both variables.

 

[blackjack avenger]

The Short Term Is The Long Term

For ****'s sake, I've explained this three times, and I'm getting sick of explaining it. Whatever the hell formula you're using is wrong, because eternal risk of ruin is always infinite. ALWAYS. Nothing to argue about, because the chance of any sized negative fluctuation is 100% given an infinite number of hands.

The number you're quoting for "lifetime risk of ruin" is either based on a wrong formula, or it's assuming a number of hands that you aren't aware of. Has anyone even posted the formula they're using?

It's no different than trying to state the lifetime risk of ruin for General Motors. What are the odds that General Motors will ever go bankrupt? THERE'S NO WAY TO KNOW! You can guess, over the next year, next thirty years, next 100 years. But there's no magical number for eternity.

Simple SD formula

this is based on a real game
6 deck h17 wonging, flat bets, 10g bank, fixed Kelly betting (13.53 ror)

variance times square root of number of hands
1.50 times square root of number of hands 1000 (10 hrs. short term)
sd=47


times a flat bet of $50
10 hour SD is bet times SD (47 times 50) = $2350
hourly win $25
win for 10 hrs = $250


$250 + or - 1 sd (68% of time) -$2100 to $2600
2 sd (95% of time) -$4450 to $4950
3 sd (99.7% of time) -$6800 to $7300

Hopefully this shows that the short term is the long term. If someone has a 10g bank and is playing in this game then they can lose a very large part of their bank in just 10 hrs. This is for fixed Kelly betting if you bet twice Kelly these numbers go up and as I stated previously you double your win rate while more then trippling your ROR which is not a good idea.

This also illustrates my previous statement that if you are playing for a weekend and betting fixed Kelly then you need to bring about half your bank. The short term variance is so so so high that the lifetime ROR numbers are quite appropriate.

The general ROR formula is based on sound math that has been known for decades.

 

[InPlay]

:grin:

It's just a convention so a back-counter, or anyone who does not play every hand, can compare his hourly win rate to other bj games since he stands around so much counting but not playing. Also useful for people who play every hand and wonder if they can make more money per hour if they back-count.

But if you just want to make it hands physically played, you can do that.

Apparently you don't bother comparing hourly win rates for diff games.

90% of the time I am playing the same game so I don't have to compare. It might sound like I don't know but I know excately the hourly win rate where I am playing. :grin:

 

[Kasi]

:grin:


90% of the time I am playing the same game so I don't have to compare. It might sound like I don't know but I know excately the hourly win rate where I am playing. :grin:

LOL.

So is the game you play 10% of the time better or worse lol?

Are you really strictly a play-all player?

No big deal lol.

 

[Kasi]

Simple SD formula

this is based on a real game
6 deck h17 wonging, flat bets, 10g bank, fixed Kelly betting (13.53 ror)

variance times square root of number of hands
1.50 times square root of number of hands 1000 (10 hrs. short term)
sd=47


times a flat bet of $50
10 hour SD is bet times SD (47 times 50) = $2350
hourly win $25
win for 10 hrs = $250


$250 + or - 1 sd (68% of time) -$2100 to $2600
2 sd (95% of time) -$4450 to $4950
3 sd (99.7% of time) -$6800 to $7300

Hopefully this shows that the short term is the long term. If someone has a 10g bank and is playing in this game then they can lose a very large part of their bank in just 10 hrs. This is for fixed Kelly betting if you bet twice Kelly these numbers go up and as I stated previously you double your win rate while more then trippling your ROR which is not a good idea.

This also illustrates my previous statement that if you are playing for a weekend and betting fixed Kelly then you need to bring about half your bank. The short term variance is so so so high that the lifetime ROR numbers are quite appropriate.

The general ROR formula is based on sound math that has been known for decades.

I'm not disagreeing with your conclusions but some of your methodology seems possibly a little funky. Maybe lol. Maybe not too lol.

Should you want to pursue in any way let me know. As much for my benefit as for yours lol.

Are you figuring these results out yourself or from a sim maybe?

Maybe a specific spread, like at what point you enter and leave flatbetting $50 might help me. Is this a 6D game with DAS or LS etc kind of thing.

 

[InPlay]

LOL.

So is the game you play 10% of the time better or worse lol?

Are you really strictly a play-all player?

No big deal lol.

Since most of the time what I see is a standard 6 deck game the other 10% is usally a better gsme and thats only when I see a 2 deck game.

 

[InPlay]

LOL.

So is the game you play 10% of the time better or worse lol?

Are you really strictly a play-all player?

No big deal lol.

Just wondering you seem to post a lot of theory and you seem very smart and knowlegable how do you actually do when it comes to playing for real money? I can honestly state that after playing for 7 years 2 months I am on the plus side of $35,800 and thats after haveing 2 loseing years in a row. Some people have a lot of book smarts but when it comes to makeing money NADA !

 

[Kasi]

...how do you actually do when it comes to playing for real money?

About as expected lol.

Absolute dollars won or lost doesn't mean too much to me lol.

But I hear you - boring I know lol.
Pprobably started when I developed an interest in trying to measure results, etc. I guess from the golden era of the internet when I didn't trust them too much while playing my million+ hands and measuring profit by the mouse-click lol.

So I hope your losses over the last couple are small and well within expectation and may you win another whatever over the next 7 years.

 

[InPlay]

About as expected lol.

Absolute dollars won or lost doesn't mean too much to me lol.

But I hear you - boring I know lol.
Pprobably started when I developed an interest in trying to measure results, etc. I guess from the golden era of the internet when I didn't trust them too much while playing my million+ hands and measuring profit by the mouse-click lol.

So I hope your losses over the last couple are small and well within expectation and may you win another whatever over the next 7 years.

Just the kind of answer I thought I would get lol. :laugh: :laugh: :laugh: :laugh: :laugh: :laugh: :laugh: :laugh: :laugh: :laugh:

 

[sagefr0g]

Just the kind of answer I thought I would get lol. :laugh: :laugh: :laugh: :laugh: :laugh: :laugh: :laugh: :laugh: :laugh:

it's an enigma lol.:devil:

 

[Kasi]

Just the kind of answer I thought I would get lol. :laugh:

See how nice it is when you get your EV. And with no standard deviation lol.

Only posting to say I screwed up and meant I wagered over a million not that I played a million hands sorry.

But at 10 cents a mouse-click (and I know what you're thinking but I didn't actually count mouse-clicks lol), it worked out OK lol.

 

[InPlay]

See how nice it is when you get your EV. And with no standard deviation lol.

Only posting to say I screwed up and meant I wagered over a million not that I played a million hands sorry.

But at 10 cents a mouse-click (and I know what you're thinking but I didn't actually count mouse-clicks lol), it worked out OK lol.

You are truly an enigma !

 

[sagefr0g]

You are truly an enigma !

:1st: InPlay.

 

[blackjack avenger]

Some Details

I'm not disagreeing with your conclusions but some of your methodology seems possibly a little funky. Maybe lol. Maybe not too lol.

Should you want to pursue in any way let me know. As much for my benefit as for yours lol.

Are you figuring these results out yourself or from a sim maybe?

Maybe a specific spread, like at what point you enter and leave flatbetting $50 might help me. Is this a 6D game with DAS or LS etc kind of thing.

Flat betting, h17, das
I dont remember if entering on TC1 or TC2 and only playing that TC or above.

The variance number is from a sim, the hourly win is from a sim.

If you play a better game the numbers won't be so severe.
If you play a worse game the numbers will be more severe.

I did not cherry pick a particular game to prove a point. I picked a game that was quick and easy to look at.

The methodology is sound

The swings in BJ are very very very high. Think about it, all the stories of people winning and losing thousands in an hour. If you play a couple bad negative hours then your bankroll can take a heavy toll.

 

[Kasi]

Flat betting, h17, das
I dont remember if entering on TC1 or TC2 and only playing that TC or above.

The variance number is from a sim, the hourly win is from a sim....
The methodology is sound

No big deal here but I was surprised flat-betting a 200 unit roll ($10K/$50) in a 6D game could even be a full-Kelly bet.

And a win of .5 units/hd just seemed a little high maybe.

But, regardless, assuming all that, regarding

"variance times square root of number of hands 1.50 times square root of number of hands 1000 (10 hrs. short term) sd=47"

doesn't seem quite right to me in that if 1.5 is the variance I'd think 1000 times that would be the total variance for 1000 hands. And taking the square root of that would be the SD.

Or take the sqrt of 1.5 to get the SD/hand and multiply that by square root of 1000.

 

[blackjack avenger]

OK, You Caught Me

No big deal here but I was surprised flat-betting a 200 unit roll ($10K/$50) in a 6D game could even be a full-Kelly bet.

And a win of .5 units/hd just seemed a little high maybe.

But, regardless, assuming all that, regarding

"variance times square root of number of hands 1.50 times square root of number of hands 1000 (10 hrs. short term) sd=47"

doesn't seem quite right to me in that if 1.5 is the variance I'd think 1000 times that would be the total variance for 1000 hands. And taking the square root of that would be the SD.

Or take the sqrt of 1.5 to get the SD/hand and multiply that by square root of 1000.

I admit, I have been told I am perhaps not the best writer. I have a friend who is a writer and editor, but what does he know. I think everyone else has a comprehension problem.:joker:

Ok, let me try again. It's the
square root of the number of hands
times
the variance
so
square root of 1000 hands (10 hrs) = 31.62
times variance of 1.5 = 47.43 (47)
I rounded down but I really should have rounded up. So the SD's would actually be a little higher.

Even if you think I am not quite right, or this game is not quite representative of real world bj ( I will say this game is quite common and this style of play is not perfect nor god awful ), then cut my SD numbers by 25% and they still show wide swings and needing a large part of the total bank in the short run.

 

[sagefr0g]

I admit, I have been told I am perhaps not the best writer. I have a friend who is a writer and editor, but what does he know. I think everyone else has a comprehension problem.:joker:

Ok, let me try again. It's the
square root of the number of hands
times
the variance
so
square root of 1000 hands (10 hrs) = 31.62
times variance of 1.5 = 47.43 (47)
I rounded down but I really should have rounded up. So the SD's would actually be a little higher.

Even if you think I am not quite right, or this game is not quite representative of real world bj ( I will say this game is quite common and this style of play is not perfect nor god awful ), then cut my SD numbers by 25% and they still show wide swings and needing a large part of the total bank in the short run.

seems like your not taking the square root of the variance , just the number of hands. so it'd be 1.22 times 31.62 = 38.6 or so..
i think that's what he was getting at sqrt var = std deviation ...

 

[Kasi]

seems like your not taking the square root of the variance , just the number of hands. so it'd be 1.22 times 31.62 = 38.6 or so..
i think that's what he was getting at sqrt var = std deviation ...

Yes O Wise Frog - exactly what I was getting at.

Either that or 1.5 * 1000= 1500 total variance. Sqrt of 1500=same.

He's overstating 1 SD by 8 units or $400. Still, though, a wide enough variance overall I guess which is what I meant that I thought his conclusions were OK.

All I know for sure is a game with a variance of 1.5 (units/hand I assume) played/seen for 1000 hands can't have both 39 or 47 units in one stan dev.

 

[Kasi]

Even if you think I am not quite right,

It's not whether I think it may be representative of a real world game, which I do wonder about as it happens, it's actually only about your methodology in arriving at your 47 units and I'm not talking about rounding up or down.

Maybe that methodolgy is off 25% this time, maybe a whole more under other circumstances.

God forbid you ever come across a game with a variance of 25 units/hand instead of 1.5 and play for 1000 hands. But if you want to think, then, a 750 unit loss is only 1 SD and you suffered a little expected "bad luck", fine with me. I'd likely think such results are either near-impossible or I'm doing something wrong in my "system" and rethink stuff lol.

Just trying to help you - nobody else seems to have had a problem with it so I guess why should I continue to try to lead a horse to water even though he won't drink once he gets there.

Hopefully a neutral arbiter here will break the tie between our methodolgies so you can disagree with them too.

 

[godeem23]

Can some authority figure please settle this? Is moo right? Do all non-resizing betting schemes have 100% RoR because the possibility of any sized negative flux is always there? Can Ken or Fred settle this? Anyone?