Last edited by BlackJackAgain; 04-09-2017 at 03:08 AM.
Min. bet exceeding Kelly ratio does not mean don't make the bet. It only means you are not betting optimally given the opportunity to make the proper Kelly-sized bet (which in this case you are not).
You need to figure your expected growth of the bet. If that is > 1, and you have no other better options, then you make the bet because your bankroll will increase.
I exploited a juicy sounding side bet. My normal EV is about $115/100 rounds. I average almost 200 rounds an hour so my normal hourly is around $200/hour for the main game. The EV maximizing approach would about double my EV/100 rounds by betting the $25 per spot per round on the side bet. But betting the same bet on both made for huge money flow problems. You tended to lose the side bet on both spots 3/4 of the time so without getting a bonus payout or a double or split the best you could hope for was an overall push on all 4 bets 75% of the time. So I reduced the side bet optimal bet some for a combined EV of about $170/100 rounds.The trouble was the game is always crowded and some patrons made an already glacial game even slower. They couldn't multitask to save their life so the game slowed to around 30 rounds/hr on average. So my actual hourly was about $50/hour if I targeted the side bet. So while my EV/100 rounds almost doubled my hourly was cut to 25% of my normal hourly. The swings were also crazy where the higher hourly without the side bet have tame results with my approach. It was hard to resist playing a 4.3% advantage side bet but the truth is I was much better off finding no side bet opportunity games that I could get a lot of rounds in per hour versus the glacially and torturously slow games that had the huge EV side bet.
Last edited by Three; 04-11-2017 at 08:17 AM.
No, this has to do with RISK. Not “optimal bet” but as a measure of RoR for current BR.
I would guess that this $5 bet, although nicely EV positive, has a variance high enough to cause an unacceptable RoR for a medium sized BR.
Back of the envelope – if the odds of hitting the big jackpot are 1:35,000, it is not at all unlikely you will run through 35,000 hands without winning it (1 in 3 or so ?). Assuming someone else does not win it first.
If you are giving up on average say $2 (bearing in mind that much of your EV comes from another high variance payoff) a hand , this means a fair chance of being down $70k++.
I wouldn’t play this.
I calculated a variance of ~30500 for this game, and you have a 54% edge, so the bankroll required to make $5 be a kelly-sized bet would be $5/(0.54/30500) = $282,000
If gain $2.7 in EV per bet and play 100 hands/hour then your EV per hour is $270. So the score is $270 * (10,000/282,000) = 9.5. A score of 9.5 is even worse than counting hi-lo with a 4:1 spread on a 6 deck with 1 deck cut off.
Better than 1 in 3.
On the first night I saw this game (before I calculated the odds), I sat down and place $5 on this side bet to start the evening just for the reason I want the pit boss think I am a ploppy. Then two others hit small payments on this hand and I have seen several seven's. So when the dealer urged me to bet again, I hesitated a bit then said no. Then immediately he dealt me three sevens of the same color on this hand. I would have won $19,000 mini Jackpot (1 in 2). Oddly enough, it doesn't hurt me at all because it is just not meant to be.
The OP stated odds of 1 in 35,000 for big jackpot for $160,000 payout, and return of $4.00. Actually, return would be $4.57.
Using the above assumptions for non-progressive payout (and a few more), bankroll for Kelly ratio of $5 bet are as follows:
Code:non-progressive payout Bankroll for $5 Kelly bet ----------------------- ----------------------- $100 $78,008 $250 $81,810 $500 $83,401 $1,000 $84,740 $2,000 $86,449 $4,000 $89,250 $5,000 $90,537
Bookmarks