Delete
Unless the TC indicated an insurance/even-money call, I would play it out.
On the other hand, one might play based on "risk averse" (or adverse?) insurance TC, meaning that he would insure a BJ before the TC was met. If I had a SUPER MAX bet out, I might insure something like 1/4 of my bet [but then again, if I had a super-max out previous shoe, I might be out the door already........].
"Everyone wants to be rich, but nobody wants to work for it." -Ryan Howard [The Office]
Originally Posted by moses
The only reason I would ever have a max bet out at the start of the SD deck is by accident*.
From your other posts, you usually play heads-up single deck.
Mathematically, it doesn't make sense to take even-money (specially with 2 aces and only 1 "ten" out).
But in this case, the max-bet was a huge mistake. I got lucky and now have a chance to keep it and gain another max-bet?
I'm taking even-money! It's just plain old "risk aversion".
* - If the "max-bet off the top" was for cover purposes, why not make another cover move, and take "even money" on a highly-negative count.
Last edited by Math Demon; 08-26-2013 at 04:36 PM.
.
To NFL newbies: Please perform your own analysis. Confirm any stats presented. Draw your own conclusions.
Handicapping is EXTREMELY hard! All statistical evidence (and game insights) may indicate strongly a specific outcome, winner, or continuing trend; but a turn-over, a missed field goal, an erroneous call, a key injury, etc. can easily change the outcome, the margins, and/or the totals. Division rivalry games and games with playoff implications are highly unpredictable.
.
Hello ZenMaster_Flash.
I am a relative newbie to this game, but I'm trying to prepare myself as much as I can before going into battle.
There's still a lot that don't understand. There are some aspects that I don't know that I don't know.
There is a 32-page math-laden paper by MathProf that I still have to digest. It is all about Risk-Averse Insurance.
My naive question is this: "Specially on a large bet spread, why would one take a guaranteed partial win, when even-money can guarantee one a max-bet?"
Thank you very much. I have learned a lot from your posts.
.
To NFL newbies: Please perform your own analysis. Confirm any stats presented. Draw your own conclusions.
Handicapping is EXTREMELY hard! All statistical evidence (and game insights) may indicate strongly a specific outcome, winner, or continuing trend; but a turn-over, a missed field goal, an erroneous call, a key injury, etc. can easily change the outcome, the margins, and/or the totals. Division rivalry games and games with playoff implications are highly unpredictable.
.
Once the bet is out and the cards are dealt, your mistake is made. Why compound it with another mistake? 1 max bet isn't going to make or break you. I'm going to assume a heads up single deck game, since that's what you play. I agree that 15 times you'd win insurance and 34 times you'd lose, which means taking full insurance would cost you about 4% of a max bet in EV. So not taking insurance leaves you 4% of a max bet better off than taking insurance, and not taking insurance essentially is risking 1 max bet to try to win 1.5 max bets (meaning relatively low variance). Why wouldn't you risk 1 max bet for a 4% gain? You risk 1 max bet for half that edge in normal play.
Very well said. You made a mistake in your extremely recent past and got lucky. It boils down to a simple risk-reward equation. EV is higher by refusing even money, and thus an EV maximizing AP is mathematically required to risk one max bet for a very good chance at winning 1.5 max bets. Ignore emotion, make the play.
Actually I misspoke slightly, as you are in effect risking 1 max bet to win .5 max bets, not to win 1.5 (relative to the position you are in if you take even money), but you still have a 4% higher EV this way than by taking even money. You should still be quite willing to bet your usual max bet with a 4% edge and a standard deviation of less than 1.
Even money takes the variance out of the dealers hole card. Without insuring you win either 0 or 1.5 bets. Even money you win 1 bet. A smaller insurance bet like for example 1/4 your bet you either win 0.5 bets or 1.25 bets. There is no way to lose both bets. The risk averse move is to limit variance to some degree. You can either eliminate it with full insurance or limit it with partial insurance or maximize EV by maximizing variance, the latter wins either 0 or 1.5 bets. When your biggest bets are out there, limiting variance is helpful to anyone with a real RoR.
Proper use of insurance is a great variance reducer. You just want to avoid losing both your insurance bet and your main bet as much as possible. I recently insured two 20's for about 1/3 max bet so I didn't lose 2 max bets to a BJ. The count was high but not at the index yet. It was at 3/4 the index. I knew the likely outcomes where winning 1.67 max bets to a BJ or losing 1.33 max bets with a very small chance of losing 2.333 max bets. The dealer flipped a BJ and I saved 2/3rds a max bet. I don't do this too often but for money management in a session it has its places for use.
Surrender is another variance reducer that has zero variance. It can be a +EV play at the same time. Some of the hardest parts for a real RoR Counter is the swings at max bet. These and other risk averse plays are great tools to smooth the swings at max bet out a bit. Some resist or refuse + EV doubles if the gain is not enough for the added risk. Some of these moves just don't have your count as reasonably correlated to the EORs for the play. This makes it a little better than guessing as to whether it's a good time to put more money out.
Splitting 22v8 is such a play. It is slightly +EV for level one counts at a high enough true count but the correlation is so bad that you have trouble pinning down an index through sims. Each sim will vary by several TC on the EV maximizing index. The drivers of the advantage in this play are the concentrations of 8's and 9's. Side counting these cards get you a strong index play but few main counts even count them.If they do their tag weighting will be way off. All other cards act in opposition except the 7 which is almost neutral The low cards are somewhat significant (less than 1/2 the magnitude of the 8 and 9 at best but act the same direction as the T's and A's. Level 2 counts have a little better correlation but still are very risky splits without the 8 and 9 intel. The best fit is the ratio of 8's and 9's to 5's, 3's, 4's and 2's with a balanced approach.
Anyway choosing how to modify your plays to limit variance is increasingly desirable as your RoR increases. The BR you save may be your own.
Last edited by Three; 08-27-2013 at 02:12 PM.
If you take even money, you will win 1 max bet, with zero variance. I consider this the baseline scenario for this decision. Not taking even money (or any amount of insurance) means risking the 1 max bet you would have won, in hopes of winning 0.5 max bets more. Risking 1 to win 0.5 will give you a standard deviation (and variance) less than 1. The variance in this case is approximately 0.5. In this case, whatever amount you risk (whatever amount you don't insure) will be risked at a 4+% edge. If you aren't willing to risk 1 max bet (typically 1% or less of your bankroll) with a 4% edge and variance less than 1, this isn't the game for you. With any bet up to bankroll * k * .04 / .5, you should not buy any insurance. With any bet larger than that amount, you should insure any excess beyond that number. Since that works out to k*8% of your bankroll, you will likely never make such a big bet when only counting. There is no excuse for taking any amount of insurance here.
Nyne your math is perfect but there is more to this thing we do than just the math. Being a math robot will get you banned quick. If every insurance bet you make is plus EV then you make their job easy. The other option is more expensive, never take insurance. By picking what negative EV insurance helps you most in order to have longevity and all the plus EV insurance is each persons chore. I prefer to have a larger top bet by using risk aversion. Thereby making more money on all my other max bets combined from the negative EV bet. Other's choose to insure for less or insure a couple min bet hands when they have an audience. Insurance is a valuable index, almost as valuable as longevity. Just pick when you have an audience or when it helps your risk profile to make bigger max bets at the same risk.
I hear you. But getting nervous when betting max-bet most probably is a sign of over-betting. Some authors state that the bankroll should be 1,000 units. Some say BR should be 100 max-bets. But there is also this aspect that the max-bet was a mistake.
The thing is the games I play are mostly 6D shoes, requiring a spread of 10:1, 12:1 or up. Now a $60 gift due to as mistake might be easy to risk for a $5 player. For a $25 player, that's a $300 gift from a mistake. For $50, a $600 gift. For $100, a $1200 gift.
From my post #3, I stated that taking even-money here is mathematically-wrong. You are giving up EV of 0.038.
__
I think it's just my inexperience that's making me lean towards risk-aversion. I'll know and understand better in a few months.
.
To NFL newbies: Please perform your own analysis. Confirm any stats presented. Draw your own conclusions.
Handicapping is EXTREMELY hard! All statistical evidence (and game insights) may indicate strongly a specific outcome, winner, or continuing trend; but a turn-over, a missed field goal, an erroneous call, a key injury, etc. can easily change the outcome, the margins, and/or the totals. Division rivalry games and games with playoff implications are highly unpredictable.
.
Thanks Tthree, Nyne, ZenMaster_Flash, RollingStoned, and moses.
I'm still early in the training process, and these discussions are helping solidify concepts, and correct errors in my thinking.
Hopefully, in 6-9 months, I can balance the math, longevity, risk-aversion to find the proper actions.
Hopefully, in the future, when faced with hard decisions, I wouldn't be like Colonel Slade (Scent of a Woman):
"... I always knew what the right path was. Without exception, I knew. But I never took it. You know why? It was too damn hard."
.
To NFL newbies: Please perform your own analysis. Confirm any stats presented. Draw your own conclusions.
Handicapping is EXTREMELY hard! All statistical evidence (and game insights) may indicate strongly a specific outcome, winner, or continuing trend; but a turn-over, a missed field goal, an erroneous call, a key injury, etc. can easily change the outcome, the margins, and/or the totals. Division rivalry games and games with playoff implications are highly unpredictable.
.
Posting Permissions
-*-
Reply With Quote
Bookmarks