why spread betting confuses me

**ohbehave** · 11-06-2014, 12:19 PM

Originally Posted by moo321

I've done it. Got a blackjack, too.

Yes but you are the MOO!

**black221** · 11-06-2014, 01:08 PM

Originally Posted by KJ

I gotta say, that I have never understood this either, Don. Seems to me that if you were to bet say $5 in all negative EV situations and say $500 in +EV counts, no matter how small, you ought to mathematically do pretty good (casino tolerance being a separate issue).

Math is often easy to think about when you consider the extremes. Let's say you have a 51% chance of doubling your money, you have 500 dollars, and you bet 500 dollars. You're pretty much at a coinflip of going home bankrupt. If you take this bet 10 times in a row, betting everything you have each time, it's almost like you're betting on getting a conflip right 10 times in a row. Super rare and it's easy to see how you'll eventually go bankrupt when that 49% happens. Of course, you don't want to bet nothing either. <insert complicated math> so you bet kelly.

**Exoter175** · 11-06-2014, 01:34 PM

Originally Posted by Bodarc

You will never have a large enough bankroll because kelly is based on the size of your bankroll. If you over bet kelly it is possible to turn a + EV into a - EV. That is why Don said you are certain to go broke.

Bingo.

**Nikky_Flash** · 11-06-2014, 01:43 PM

i agree with all this , in my post i seperated real life and math theory, cited variance as the reason not to go crazy with small advantage ,
anyways , the op has decent bets out in small advantage situations , his bankroll sounds like it can take it , if that's the case , good cards to you Sir.

**SiMi** · 11-06-2014, 03:49 PM

Hi theredcore!

I put the following into CVCX: HiLo Sweet 16/Fab 4; 6D; 5/6 pen; S17 DAS; Spread from $5 to $1,000; Custom bets: TC<=1 = $5 TC>1 and TC<=5 = $150 TC>5 = $1,000

CVCX showed a win rate of 2.46% (WOW!) and a RoR of 51% with a $10,000 B/R and a 5% RoR with a $45,000 B/R. It took a $114,000 B/R to get the RoR to 0%. The SD was 3.65 so the Variance was over 13.33 (I'm used to seeing SD around 2 and thus Variance around 4 so this a hellacious amount of Variance!) The win per hour was $136.32

So, then, just for fun, I put in Custom bets as follows: TC<=1 = $5 TC>1 = $1,000. (The idea is that anytime you have an edge, you bet $1,000.)
CVCX showed a win rate of 1.7% and a RoR of 77% with a $10,000 B/R and a 31% RoR with a $45,000 B/R. It took a $114,000 B/R to get the RoR down to 5%. At $290,000 B/R, the RoR went to 0%. The SD was 2.63 so Variance was almost 7 (not as bad as I expected). The win per hour was over $314.

Interesting post! I live on the West Coast. If you're willing to share, what part of the Beaver State do you play mostly (North, Coast, South, East)?

SiMi

**Nyne** · 11-06-2014, 05:10 PM

And for comparison with SiMi's numbers, here are the results with optimal betting on the same game (I used $1 chip denom, so some bets will need adjusting for practicality):

$45,000 BR, 5% RoR
1-16 $109.38/hr
1-20 $121.82/hr
1-30 $137.56/hr
1-200 $188.43/hr
Notably, $25-700 gets you $140.70/hr at 5.7% RoR, and $5-1000 with optimal betting gets you $157.16/hr with a 3.7% RoR.

$114,000, 0.1% RoR
1-16 $123.13
1-20 $134.42
1-30 $149.48
1-200 $188.43

Your 1-200 spread has a SCORE comparable to more normal spreads, but you'll need a hefty bankroll and I'd expect quite a bit more heat with this approach than with a typical approach. If you've got the funds to back it up and you feel like this is the best way to get the money down, more power to you. It's not the strategy I would use.