AxiomOfChoice
AxiomOfChoice
Joined: Sep 12, 2012
  • Threads: 32
  • Posts: 5761
October 21st, 2012 at 5:11:31 PM permalink
Quote: AceCrAAckers

It is equally likely that a natural will happen on the player side as it is the banker side.



That has nothing to do with it, though. When naturals are likely enough, both sides will be profitable. Bet them both simultaneously.

Quote: AceCrAAckers

I have not read anywhere about Dragon Bonus being countable.



Then you should read what Stephen How wrote at discountgambling.net. Both the Dragon and the Dragon-7 are countable.

Quote: AceCrAAckers

Will there be a situation where knowing what few cards are left gives a player an edge? I doubt it but would like to see if there is AP possibility once it gets installs.



Clearly, there are some deck compositions where the player will have an edge. The question is, how often will they come up in practice? For an extreme example, in a deck with all 10's and 9's (split 50-50), half the hands will be naturals, and you will have a 9-9 tie about 1/4 of the time. Of course, this will not come up in practice, but other deck compositions with a smaller edge might. The question is, how often?

I'd guess that, 8's, 9's, and 10's are good and all other cards are bad (because 10-9 and 10-8 are 4x more likely than other naturals.

Now, I have not bothered to do any simulations or anything, so I don't know if it's actually countable in practice, but that's the first thing that I'd look at.
AceCrAAckers
AceCrAAckers
Joined: Jul 12, 2011
  • Threads: 30
  • Posts: 377
October 21st, 2012 at 8:20:36 PM permalink
I have read discountgambling.net about the Dragon bonus. How talks about the Player side Dragon Bonus which has a HE of 2.65% He does not even touch on the Banker side Dragon Bonus with a HE of 9.37% How also mentions that it is not worth counting the Dragon Bonus on the Player side. The moment that How shows that Dragon Bonus as being countable on the Banker side, let me know. The house edge is too great for counting to overcome for the Natural side bet that I have.

As for the Dragon 7 being countable, it has been reported by Dr.Eliot Jacobson also.

Thorpe talked about the baccarat side bet for natural that paid 9 to 1. The chance of a natural is 1 in 5.27. So this previous payout was clearly wrong. Any baccarat bonus that has payouts for naturals above 5.27 has a player's edge.
Edward Snowden is not the criminal, the government is for violating the constitution!
AxiomOfChoice
AxiomOfChoice
Joined: Sep 12, 2012
  • Threads: 32
  • Posts: 5761
October 21st, 2012 at 8:49:13 PM permalink
I don't think he says that the Player Dragon is not worth counting. I think he says that the Dragon-7 is not worth counting. Not because the edge isn't big (it is) but because the variance is just too high. You need a monstrous bankroll, relative to your bet size. Even though you have a decent edge, you can wait a long time between those 40-1 payouts.

Player Dragon, on the other hand, is a goldmine. You don't have to wait for those 30-1 payouts; the smaller payoffs are fairly common and keep you afloat for a long time. Remember that, when applying the Kelly Criterion, you need to divide the edge by the variance, and I think that the variance in Player Dragon is around 1/3 that of Dragon-7 (I don't remember exactly; I did the math on that a long time ago)

The reason that the Banker Dragon is not worth counting is just that the banker stands on 2 cards way too often. This has nothing to do with your bet, since yours is symmetrical. For yours, if one side is countable, both sides are (you would never bet one but not the other. It lowers the variance, too, which is nice)

You state that the house edge on your bet it too great for counting to overcome, but you make this statement with no proof. Is there any particular reason that you believe this? Have you run some simulations, or done some math? A high house edge alone doesn't imply that the bet is not countable; you also have to look at how sensitive it is to removal of certain cards.

Note that I'm NOT saying that you're wrong (I haven't done the math myself, so I don't know). All that I'm saying is that if you haven't done the math, you don't know either. To me, it "feels" countable, but of course my gut feeling could be way off.

When it goes live, are you going to tell us which casino it's at?
7craps
7craps
Joined: Jan 23, 2010
  • Threads: 18
  • Posts: 1977
October 21st, 2012 at 10:00:40 PM permalink
Quote: AceCrAAckers

Thorpe talked about the baccarat side bet for natural that paid 9 to 1.

It seems you do not understand that there were four different side bets way back in the 1960s.
Dr Thorp:
"A winning strategy is developed for the nine to one side bet on a Banker natural nine.
...
Similar strategies are developed for the side bets on
Banker natural eight,
Player natural nine, and
Player natural eight."

There were 4 different side bets that could be wagered on individually or any combination of the 4 or all 4.
Quote: AceCrAAckers

The chance of a natural is 1 in 5.27.

Correct.
Any natural
Quote: AceCrAAckers

So this previous payout (9 to 1 or 10 for 1) was clearly wrong.

The house advantage for the 4 side bets is stated in section 4.
5.10 per cent for a Banker natural 9
5.10 per cent for a Player natural 9
5.47 per cent for a Banker natural 8
5.47 per cent for a Player natural 8


So, that makes your conclusion erroneous.

Your side bet is countable, again,
is it worth the effort?

You have one expert offering his service to you to show the numbers when the bet goes live.
Super!
IMO,
I suggest for you to either accept his very nice offer or see how countable it really is on your own.
Time will tell.

I do not think any one here is against your new bet and wants to see it fail,
just to bring up the point that it could fail because the bet could be highly countable.
Are you willing to gamble on it??
winsome johnny (not Win some johnny)
AxiomOfChoice
AxiomOfChoice
Joined: Sep 12, 2012
  • Threads: 32
  • Posts: 5761
October 21st, 2012 at 10:04:19 PM permalink
Quote: 7craps


I do not think any one here is against your new bet and wants to see it fail,
just to bring up the point that it could fail because the bet could be highly countable.
Are you willing to gamble on it??



Just because it's countable doesn't mean that it will fail. The Dragon bets are still alive and well. So is the game of blackjack. And all the countable blackjack side bets.

Personally, I'm hoping that it's countable and it succeeds. But then, I'm a greedy bastard :D
AxiomOfChoice
AxiomOfChoice
Joined: Sep 12, 2012
  • Threads: 32
  • Posts: 5761
October 21st, 2012 at 10:18:30 PM permalink
Quote: AceCrAAckers

The anlaysis was based on 4,998,398,275,503,360 possible baccarat hands with an eight deck shoe.



I have to ask... what???? That is a very big number (5,000 trillion. Is there a word for a thousand trillion?) You only need to look at the first 4 cards.

Even if you don't take any shortcuts, ie, you differentiate between all the different cards of the same rank, and you don't ignore order of cards within a hand, there are STILL only 416 x 415 x 414 x 413 = less than 30 billion possibilities. Are you looking at possible draws too?
AxiomOfChoice
AxiomOfChoice
Joined: Sep 12, 2012
  • Threads: 32
  • Posts: 5761
October 21st, 2012 at 11:09:22 PM permalink
Ok, I was curious, so I just did a back of the envelope calculation. I didn't run any simulations, and I made some intentional "rounding errors" to simplify the math, but I don't believe that they introduce any large errors.

My results show that this bet is almost certainly countable. I was totally wrong about the 10s; there are so many of them that removing one actually helps you get a natural.

It looks like 8s and 9s are very good for the player, 4s are somewhat good for the player, and everything else is bad. You would probably capture most of the value with the following index numbers:

A, 2, 3, 5, 6, 7, 10: +1
4: 0
8, 9: -5

The 4 is actually closer to -0.75 and the 8s and 9s are closer to -4.625. I suspect that the easier numbers above are close enough.

Each +1 to TC should be worth about 0.9% or so the the player, so the bet should be close to break-even at around TC = +9. Again, this was totally back-of-the-envelope, so it could be somewhat off, but I think it's in the ballpark. To get exact numbers I'd have to correct my intentional rounding errors and run some simulations, which I probably will never bother to do unless (a) the game comes to a casino that I frequent, and (b) no one else does it first. These numbers are probably, at the very least, a good starting point to run a simulation on before you tweak it.

[edit: Fixed the signs on the index numbers, since, of course, we count removals, not cards left in the deck]
AceCrAAckers
AceCrAAckers
Joined: Jul 12, 2011
  • Threads: 30
  • Posts: 377
October 22nd, 2012 at 9:29:37 AM permalink
Quote: 7craps

It seems you do not understand that there were four different side bets way back in the 1960s.
Dr Thorp:
"A winning strategy is developed for the nine to one side bet on a Banker natural nine.
...


Your side bet is countable, again,
is it worth the effort?



I highly doubt that the count can overcome the HE and if such edge ever present itself, it would be so infrequent that it is not worth waiting for.

Quote: 7craps

You have one expert offering his service to you to show the numbers when the bet goes live.
Super!
IMO,
I suggest for you to either accept his very nice offer or see how countable it really is on your own.
Time will tell.



No one is turning down his generous offer. I welcome this analysis. It will not be done till this goes live in a brick and motar casino so we need to wait til then. It should be real soon though. Remember, the natural pays out at only 2 to 1. Not the 9 to 1 that was being offered when Thorp made his report.
Edward Snowden is not the criminal, the government is for violating the constitution!
MathExtremist
MathExtremist
Joined: Aug 31, 2010
  • Threads: 88
  • Posts: 6526
October 22nd, 2012 at 9:48:42 AM permalink
Read more about how bacc side bets based on naturals are beatable:

http://www.jstor.org/discover/10.2307/2282821?uid=3739856&uid=2129&uid=2&uid=70&uid=4&uid=3739256&sid=21101296931781
(includes the abstract from Thorp's paper which 7craps quoted)

http://greenbaizevamp.hubpages.com/hub/Edward-Thorps-baccarat-adventure
"In my own case, when it seemed to me after a long illness that death was close at hand, I found no little solace in playing constantly at dice." -- Girolamo Cardano, 1563
Wizard
Administrator
Wizard
Joined: Oct 14, 2009
  • Threads: 1343
  • Posts: 22215
October 22nd, 2012 at 9:51:19 AM permalink
I'll work this out as soon as it has a confirmed casino placement. Can someone please let me know if and when this happens.
It's not whether you win or lose; it's whether or not you had a good bet.

  • Jump to: