Little help with bet sizing/spred I should be using.

**DSchles** · 06-17-2015, 08:23 AM

"A 13.5 RoR is a high RoR but offers optimal BR growth (full kelly)."

This is a somewhat misleading statement and needs to be clarified. Optimal BR growth is obtained by playing full Kelly, all the time, which means that, as the BR fluctuates, so do the bet levels. That style of play does NOT lead to 13.5% ROR. The latter occurs when you start with a specified BR level, FIX your bet sizes, and stick with them, regardless of BR fluctuation, until you either win all the money in the world or go broke trying. Surely, you know this, so be careful as to how you express these notions.

Don

**DickFer** · 06-17-2015, 08:41 AM

Hmmm I found this cvcx software and I think it answers all my questions. You plug in your game type and it runs a sim and tells you everything you need to know.

**DickFer** · 06-17-2015, 09:26 AM

https://www.youtube.com/watch?v=IyATmCJf4fc

good video on Kelly criterion, he simplifies it well. (beginner explanation)

**DickFer** · 06-17-2015, 10:21 PM

Originally Posted by Shadowless

With a $1,000 bankroll, your full Kelly unit is $3.75.
Since you want to keep RoR as low as possible, you might consider doing half or quarter Kelly.
80% Kelly would be $3, which is an easier number to work with.

Equation would be: (TC-1) * $3
So at a TC of 6, you would bet $15.

Basically $3 per true count over 1.

can you tell me how you came up with Kelly unit = $3.75 please

(I wanna be able to recalculate my bet sizes as my BR fluctuates)

**Shadowless** · 06-18-2015, 12:31 AM

Originally Posted by DickFer

can you tell me how you came up with Kelly unit = $3.75 please

(I wanna be able to recalculate my bet sizes as my BR fluctuates)

Bankroll * 0.005 * 0.75
0.75 is to account for variance (learned that the other month from Don).
0.005 represents the approximate 0.5% advantage per true count.
And remember to account for the house edge by subtracting 1 from the TC before multiplying by your betting unit.

If you plan on quitting and never playing again if you lose your $1,000, then you should stick strictly to your current betting unit.
If you plan on reinvesting into your bankroll, or if you could access more funds once you became more serious and confident in your game, you should resize accordingly.

**DickFer** · 06-18-2015, 12:52 AM

Right now what I would have to do is probably start with $1000 BR and be able to add maybe $100-$150 a month to it if I have to. If I'm showing positive results at $5 dollar table for a while building my bankroll then I would probably start increasing my Kelly unit? correct? So I build my BR up to $2000, I would then do 2000*0.005*0.75= $7.5/Kelly unit. SO if TC is 6, I do 6-1*$7.50=$37.50. So I would bet $37.50 on a TC of 6 if my Kelly unit was $7.50.

How confident are you in that formula you presented by the way? How accurate is it? Dead on? Thanks for the help man.

**Shadowless** · 06-18-2015, 01:16 AM

Originally Posted by DickFer

Right now what I would have to do is probably start with $1000 BR and be able to add maybe $100-$150 a month to it if I have to. If I'm showing positive results at $5 dollar table for a while building my bankroll then I would probably start increasing my Kelly unit? correct? So I build my BR up to $2000, I would then do 2000*0.005*0.75= $7.5/Kelly unit. SO if TC is 6, I do 6-1*$7.50=$37.50. So I would bet $37.50 on a TC of 6 if my Kelly unit was $7.50.

How confident are you in that formula you presented by the way? How accurate is it? Dead on? Thanks for the help man.

The formula is accurate and effective because the amount you are betting is in relation to your percent advantage for the given hand and accounts for variance.
It is correct and has been written in a few reputable books by reputable authors. Burning the Tables in Las Vegas by Ian Andersen to name one.

Also, it's not always practical to bet EXACTLY the "correct" amount.
In the case of betting $37.50 at TC 6 once you are at $7.50 betting unit, you might find yourself betting $35 or $36 instead, for simplicity, since getting a blackjack on $37.50 isn't a number that dealers are used to paying out and could slow the game slightly. ($56.25, in which you'll be shorted a 25 cents).

If you're more of a risk taker and plan on adding to your bankroll, you could round up to $40, but most people will advise against that since players tend to slightly overestimate their edge, or the true count conversion and/or deck estimation may not have been entirely accurate.

**DickFer** · 06-18-2015, 01:21 AM

Also, lets say TC is 0. I do 0-1*$3.75. =-$3.75. So is this saying I shouldn't be making any bets at TC0 because my expected result is negative? Also if TC=1 it would recommend I bet $3.75. So is this saying betting $5 is too much? So technically according to this I should be waiting to make any bets until TC is near 2 or 2 or higher? Which means I need to be back counting, and sitting out as much as possible when the shoe isn't in my favor, correct?

**DickFer** · 06-18-2015, 01:43 AM

Sounds good!! Yea I wouldn't make ridiculous bet sizes like that lol, I'm just trying to get a feel for what my bet sizes should be. You've helped me immensely. I think I have my bet sizing down thanks to your help and formulas. Now I just need to keep practicing on counting and learning the index numbers for hi-low.

**Shadowless** · 06-18-2015, 01:44 AM

Originally Posted by DickFer

Also, lets say TC is 0. I do 0-1*$3.75. =-$3.75. So is this saying I shouldn't be making any bets at TC0 because my expected result is negative? Also if TC=1 it would recommend I bet $3.75. So is this saying betting $5 is too much? So technically according to this I should be waiting to make any bets until TC is near 2 or 2 or higher? Which means I need to be back counting, and sitting out as much as possible when the shoe isn't in my favor, correct?

If possible, do not bet at all when the advantage shifts to the house's favor (at TC 0 and below).
The only disadvantage to this is that the casino will eventually catch on to what you're up to.
Most casinos won't mind too much if you're just a recreational player who doesn't do heavy betting, but an advantage is still an advantage and some places will back you off even if you're not a really a "threat."

If you can dedicate $1,334 for your bankroll, you can use a $5 betting unit instead of $3.75 at $1,000.
It doesn't mean that you have to immediately have that amount available since it's not wise to take your entire stake with you for every session.

Correct, backcount and only play when you have a positive expectation.
Again, keep in mind the possible repercussions of backcounting and know your casino and its tolerance.

Originally Posted by DickFer

Sounds good!! Yea I wouldn't make ridiculous bet sizes like that lol, I'm just trying to get a feel for what my bet sizes should be. You've helped me immensely. I think I have my bet sizing down thanks to your help and formulas. Now I just need to keep practicing on counting and learning the index numbers for hi-low.

Keep in mind that S17 and H17 indices are slightly different, and will also vary if the game allows double after split or not.
Learn the correct indices for the game you plan on playing.

**DickFer** · 06-18-2015, 01:52 AM

Thanks again, this thread has really clarified things a lot. Now I know how big of advantage it is to start with a bigger bankroll. You wouldn't have to back count and you could even make min bets when you're at a disadvantage because your larger bet sizes when you have an advantage would offset the losses. Obviously it would still be wise to not make any bad bets when TC0 or worse but if you're trying to stay off the radar it would help to play through a little when you have -EV(maybe?).

This was just random thought and really had nothing to do with thread lol. Anyway back to practicing with casino verite software

**Shadowless** · 06-18-2015, 02:16 AM

It is recommended to not play another round once the shoe becomes a TC of -1.

You don't always have to justify your reason for leaving the table or sitting out. If you feel the need to, be creative. For example, if the dealer happens to pull a 21, you lose a double down, you push on a 20, or something to that degree. Or you could try to sound superstitious and say you are "changing the order" of the cards (many people, especially gamblers, say that).