'Session' bankroll ROR

[Qh4Mate]

You can't do the standard deviation your way. First of all, when you spread, rather than flat bet, the s.d. per hand is much larger than 1.14. So you have to simulate to get the correct (larger) value. Second, you're using UNITS for all the inputs. You can't switch to using $5 for the s.d. calculation. That has to be in UNITS as well. So forget the $5.

Use the methodology on page 20, not 16. Don't bother with single session RORs. You can't calculate them properly anyway, until you get the s.d. correct.

OK. I looked at the tables on page 20 and read pages 18-21. So now I am convinced this book is way too advanced for me right now. I guess what I need is a beginner's primer on Blackjack Standard Deviation.

[DSchles]

OK. I looked at the tables on page 20 and read pages 18-21. So now I am convinced this book is way too advanced for me right now. I guess what I need is a beginner's primer on Blackjack Standard Deviation.

No you don't, because if you spread your bets, there is no other way to calculate s.d. than the way described on p. 20. So, if you get a beginner's book, it won't help you to answer the question you've asked. Page 16 will be the same as the beginner's book, but I've already told you that you can't answer your question using that methodology.

Einstein once said: Everything should be as simple as possible, but not simpler.

Don

[Qh4Mate]

No you don't, because if you spread your bets, there is no other way to calculate s.d. than the way described on p. 20.

Ok, I am planning to use a bet spread of 1 - 8 units. Looking at the 6 deck game on Table 2.1 (page 20) the "Bet Squared (units)" column indicates a spread of 1 - 6. The bottom formula says "4.70 = one-round s.d. in units". How is this converted to Std. Dev. for 100 hands?
Looking at Table 2.3 (6-deck col) I find the S.D. per round of 4.70. S.D. per hour is 24.32. 26.77 hands played per hour. So I would need about four hours to play 100 hands. Do I take the S.D. per hour and multiply it by four? 24.32 x 4 = 97.28. Is this my S.D. for 100 hands?

[Dog Hand]

Ok, I am planning to use a bet spread of 1 - 8 units. Looking at the 6 deck game on Table 2.1 (page 20) the "Bet Squared (units)" column indicates a spread of 1 - 6. The bottom formula says "4.70 = one-round s.d. in units". How is this converted to Std. Dev. for 100 hands?
Looking at Table 2.3 (6-deck col) I find the S.D. per round of 4.70. S.D. per hour is 24.32. 26.77 hands played per hour. So I would need about four hours to play 100 hands. Do I take the S.D. per hour and multiply it by four? 24.32 x 4 = 97.28. Is this my S.D. for 100 hands?

Qh4Mate,

SD increases as the square root of the number of rounds played. So, if the SD per round is 4.70 and you play 100 rounds, your SD will be 4.70*(100)^0.5 = 47.

Using the numbers from Don's example, if you play 26.77 rounds per hour with a per-round SD of 4.70, your hourly SD will be 4.70*(26.77)^0.5 = 24.3177..., which rounds to the value Don gave.

Note, though, that for your 1-8 spread, your SD per round will NOT be 4.70: you have to calculate it using either the method Don presents or software.

Hope this helps!

Dog Hand

[Qh4Mate]

4.70*(100)^0.5 = 47.

Dog Hand,

I appreciate your attempt to walk me through this, I really do. But I cannot recreate your math because I do not know what the "^0.5" means. I looked it up and it means square root. However, if I multiply 4.7 x 26.77 I get 125.819. The square root of 125 is about 11.2, not 24.3177. So obviously I am missing something in my current understanding of statistics (especially as applied to blackjack).

[PinkChip]

...
But I cannot recreate your math because I do not know what the "^0.5" means. I looked it up and it means square root. However, if I multiply 4.7 x 26.77 I get 125.819. The square root of 125 is about 11.2, not 24.3177. So obviously I am missing something in my current understanding of statistics (especially as applied to blackjack).

The problem here does not lie in statistics but in arithmetics. You must first take the square root of 26.77, which is about 5.17397, and then multiply it with 4.7, ending up with 24.3177. Rather than writing ^0.5, I strongly prefer sqrt(), like in programming languages. So you compute 4.7 * sqrt(26.77). Functions like sqrt with brackets round their input take precedence over multiplication/division, which take precedence over addition/subtraction.

[Qh4Mate]

So you compute 4.7 * sqrt(26.77).

Thank you for getting down to my level. I think I finally got it. So 4.7 x sqrt of 26.77 = 24.3177. This is my S.D. per 26.77 hands. If I want S.D./100 hands (which is what the online calculator wants) I go 4.7 x sqrt of 100 = 47. If I now (hopefully) understand this my ROR calc using my original post data and the online calculator will be as follows:

400 'Trip' Bankroll units
2.0 Win Rate in units per 100 hands
47 Standard Deviation per 100 hands
1,500 Hands played (10 sessions at 150 hands each for 'Trip')
1.9% Risk of Ruin (this seems just fine)


40 'Session' Bankroll units
2.0 Win Rate in units per 100 hands
47 Standard Deviation per 100 hands
150 Hands played (1 session)
47% Risk of Ruin (this seems too high)

50 'Session' units yields 37% ROR per session with corresponding reduction in available 'sessions' from 10 to 8.
60 'Session' units yields 28% ROR with corresponding reduction in available 'sessions' from 10 to 6.

Just intuitively the 28% seems acceptable since by definition a 'session' is only a small fraction of your total 'trip' bankroll and you would never expect the 'session' ROR to come close to the 'trip' ROR. Heck, maybe even the 37% makes sense using this logic. I do not know what an ideal 'session' ROR should be or even if calculating one is ever done. However, since having only 6 'sessions' of bankroll (of 60 units ea) seems like to few for a 'trip', it looks like I need to increase my 'trip' bankroll (even though for the overall 'trip' my ROR is fine at 1.9%). Seems like a bit of a balancing act between 'Trip' & 'Session' bankrolls. Does this make any sense?

BTW, I know this is all for a 1 - 6 spread and I previously stated I was planning 1 - 8 which I am sure changes everything but at least I am hoping to get the basic idea down. I will leave the calculation confusion of changing to a 1 - 8 spread (I'm sure it's in BJA Table 2.1, page 20 somewhere) to another day.

Thank you for your patience.

[DSchles]

All the math is correct. Why do you care if you lose a session BR if you bring ten such ones for the trip? The ROR for the 150-hand session has little importance. You're going to lose the 40 units with regularity. You just aren't going to lose 10 of them on the short trip.

Don

[ZeeBabar]

Discussing Session Bankroll is fine but I, for one, really end a session if I go down by $2k at a 6 deck table, $3k at a $25 min., add game. Losing more than that effects my day.

In theory, I take $10k+ as a trip BR but in practice, I would quit playing on that trip if down over $6k (feeling I am not going to recover it on that trip). Generally, after a session or trip loss, I have the need to take a few days off.

I am aware this is wrong and it’s why I have such great admiration for the pros. It ain’t easy.

[refinery]

Wrong in this instance is subjective. Loss stops lower your hourly EV, but they also lower ROR, so there is a justification for using them if you choose. But, when I do use them, I do not use them per session, I use them per shoe at >=-1 S.D. (units, not $$$) per shoe for the specific ramp I am using and number of players.

Bad advice.

[refinery]

Not giving any advice, it is clearly up to the player, not me.

Then type your opinions into notepad and hit exit instead of posting them here. Everything in these threads is tantamount to advice because of the nature of conversation. It's the same garbage other posters use to say "here's what I do, but in no way am I saying this is advice." Putting it into the ether is advice on here.

Just because "it's up to the player" just means it's their decision whether or not to take your advice. Your stop loss advice is nonsense, voodoo, and dangerous.

[Dog Hand]

Wrong in this instance is subjective. Loss stops lower your hourly EV, but they also lower ROR, so there is a justification for using them if you choose. But, when I do use them, I do not use them per session, I use them per shoe at >=-1 S.D. (units, not $$$) per shoe for the specific ramp I am using and number of players.

Wave,

Can you explain how a stop loss lowers your RoR? The only way I can see that working is if you are being cheated... is that your justification?

Dog Hand

[ZeeBabar]

Then type your opinions into notepad and hit exit instead of posting them here. Everything in these threads is tantamount to advice because of the nature of conversation. It's the same garbage other posters use to say "here's what I do, but in no way am I saying this is advice." Putting it into the ether is advice on here.

Just because "it's up to the player" just means it's their decision whether or not to take your advice. Your stop loss advice is nonsense, voodoo, and dangerous.

Why do you assume you know how everyone thinks and acts? You may think it’s advice but who gives a shit what you think. I post what I do and how you or anyone else takes it is not something I worry about. People knock me for what I do and from their posts, I learn and I decide if I want to make changes, to things differently.