Rationale for 9/9 Split or Stand

[ArcticInferno]

Why do you stand on 9/9 against 7, but split against 8 or 9, and then stand for 10 or A.
The pattern doesn't make sense.
According to "Blackjack Bluebook II" page 62, a hard 18 will beat a dealer's 9 only eight out of twenty.
According to the book, two 9's against a dealer's 9 is better than one 18 against 9.
I'm assuming that the explanation is similar for an 18 against a dealer's 8.
What about an 18 against a 10 or an Ace?
I would think that the dealer's 10 or Ace is stronger than a 9, so if you have to split 9/9 against a 9,
then you should certainly split 9/9 against 10 or Ace.
What about an 18 against 7? That makes even less sense.

 

[Blue Efficacy]

You stand with 9's against seven because your 18 is a favorite against a dealer 7, and you don't want to throw that away at neutral counts.

Splitting against 9 is a defensive split, you do it to lose less. It is still negative EV, but less so than standing.

It is not a good idea against A or especially 10, because they could very well have a strong hand already made with 10 showing, or could easily make one with an Ace, and you do not want to lose twice.

The ability to double after a split also makes splitting more attractive against 9 or 8, you could get a 2 and double for 21, lose the other one and you're still a net winner. An ace gives you 20, also..

 

[shadroch]

The bottom line is that 18 is simply not a very good hand. You will lose more hands with 18 than you will win.
If the house offered a game where you got 18 every hand,and the dealer had to play with normal rules,would you play it?

 

[zengrifter]

Why do you stand on 9/9 against 7, but split against 8 or 9, and then stand for 10 or A.
The pattern doesn't make sense.
According to "Blackjack Bluebook II" page 62, a hard 18 will beat a dealer's 9 only eight out of twenty.
According to the book, two 9's against a dealer's 9 is better than one 18 against 9.
I'm assuming that the explanation is similar for an 18 against a dealer's 8.
What about an 18 against a 10 or an Ace?
I would think that the dealer's 10 or Ace is stronger than a 9, so if you have to split 9/9 against a 9,
then you should certainly split 9/9 against 10 or Ace.
What about an 18 against 7? That makes even less sense.

Bottomline: the only logic is based on statistical probabilities derived from high-speed computer simulations, otherwise known as BASIC STRATEGY. zg

 

[rookie789]

Not to deny your theory of when basic stratagy developed but I think it was prior to hi speed computer simulations although computer simulations (either hi or low speed) confirmed prior evaluations.

 

[shadroch]

Not to deny your theory of when basic stratagy developed but I think it was prior to hi speed computer simulations although computer simulations (either hi or low speed) confirmed prior evaluations.

What was hi-speed at the time would be laughable these days, but they were state of the art at the time.

 

[rookie789]

Hi Speed

I agree but ZG's post appears to state basic strategy is a recent development based on simulations produced by hi speed computers, I believe BJ basic strategy was known and published prior to HP, Sony, Dell etc.

 

[shadroch]

BS was developed on the best computers around at the time. As slow as they are by todays standards, they were considered hi speed at the time as they could achieve it a few days what it would take a human a lifetime to achieve. All todays higher speed computers have done is pretty much confirm the pioneer work of the 50s and 60s.

 

[fredperson]

Basic Strategy

I can't find my copy of "Beat the Dealer" but I thought Thorp said that
basic stategy was based on mathematical analysis.
Computer time in the 50s was way to precious to waste on simulating games.

 

[shadroch]

I can't find my copy of "Beat the Dealer" but I thought Thorp said that
basic stategy was based on mathematical analysis.
Computer time in the 50s was way to precious to waste on simulating games.

Not if you are the US Army, it wasn't. Thats how it was developed and then Julian Braun fine tuned it using an early FORTRAN(sp?) system. BS as we know it comes for Brauns work, as best I can tell. Everything proir to his work contained small errors.

 

[fredperson]

Not if you are the US Army, it wasn't. Thats how it was developed and then Julian Braun fine tuned it using an early FORTRAN(sp?) system. BS as we know it comes for Brauns work, as best I can tell. Everything proir to his work contained small errors.

Actually, I worked for the Air Force computer center in the basement of the pentagon. We had a 704, 705, and were getting a 709 when I left.
Computer time was estimated to be worth $1000.00 per hour.
Program debugging was only allowed on the 3rd shift, and as programmers we were allowed one debug shot per night. Frankly, I can't imagine anyone developing or running game simulators in those days.

Here's the real story of how the original basic strategy was developed. Braun didn't got involved until the mid-60s while working as an IBM employee.

"In 1956, Roger Baldwin wrote "The Optimum Strategy in Blackjack". This paper was published in the Journal of the American Statistical Association. It was the first recognized effort to apply mathematics to Black Jack. Using probability, statistics and calculators, Baldwin wrote of different ways to substantially reduce the house advantage. This paper consisted of 10 fairly mathematical pages. Although entitled "Optimum Strategy", the theory still needed a computer to refine the system, thus making it not such an "Optimum Strategy" during it's time."

 

[shadroch]

I believe thats the WIKI version. As you can see,it also states that the article contained errors and it wasn't until computers were used, that BS as we know it was developed. Baldwins strategy,although an improvement on the way most people played,was riddled with errors. HE didn't even call it Basic strategy.
I guess it's like arguing over who discovered America. Columbus gets much of the credit,even though the Norse and the Irish were the first Europeans to reach it, and besides the fact he never landed on American soil. Nor did it need discovering.

 

[PrinceDragon]

Why do you stand on 9/9 against 7, but split against 8 or 9, and then stand for 10 or A.
The pattern doesn't make sense.

09/09/09

It does make sense if you start to do it tomorrow.

P.D.

 

[shadroch]

Strange.
I posted a post where it said I had found the article I was lookiking for, and while my memory of it being four US Army Analyst was right,it turned out the original work was done on desktop calculators, not computers.
That said, BS as we know it wasn't done until Braun did it with the high speed computers of the early 60s.

 

[21gunsalute]

Why do you stand on 9/9 against 7, but split against 8 or 9, and then stand for 10 or A.
The pattern doesn't make sense.
According to "Blackjack Bluebook II" page 62, a hard 18 will beat a dealer's 9 only eight out of twenty.
According to the book, two 9's against a dealer's 9 is better than one 18 against 9.
I'm assuming that the explanation is similar for an 18 against a dealer's 8.
What about an 18 against a 10 or an Ace?
I would think that the dealer's 10 or Ace is stronger than a 9, so if you have to split 9/9 against a 9,
then you should certainly split 9/9 against 10 or Ace.
What about an 18 against 7? That makes even less sense.

You stand on 9,9 against a 10 or an Ace because you are the underdog whether you split or not. If you do split you are likely to be making 2 losing hands out of one. Even if you hit each 9 with a 10 you're only slightly improving your hand and will not beat a dealer 20. It's simply not worth the risk.

You split 9,9 against a dealer 9 because you have 18 and the dealer is looking at having 19. You're the underdog if you stand. If the dealer has a 10, J, Q, K or A you lose w/o the dealer even drawing. If the dealer has a 2 down for 11 she'll draw again and nearly half the deck will beat you. The only "safe" card you'll beat is if the dealer has an 8 down for 17. A nine down will push you, and any other down card will give the dealer at least 3 cards she can draw to beat you and one to push you. It's a defensive measure. You're probably not going to win both hands but you have an excellent opportunity to win one and lose one for an overall push or push both hands. And you may even hit 1 or both hands with a 2 for double down possibilities. In this case the risk is worth it.

You split 9,9 against an 8 because now you are the favorite, not the underdog. You have an excellent chance of winning both hands, whereas if you stand you're more likely to be pushed if the dealer has a 10 down. Better to win 2 hands than push with one. You aren't likely to bust on both hands, so in the worst case scenario you should still be able to win one and lose one for an overall push. Of course nothing is certain, it's possible to bust both hands or even for the dealer make 21, but the odds are in your favor.

Standing on 9,9 against a 7 is a bit more tricky and harder to understand. Part of the reason you stand is because your 18 will beat the dealers expected 17 hand. While it may be tempting to split this hand and hit both of your 9's with 10's for a pair of 19 hands, you don't want to take 10's away from the dealer in this situation. You may end up busting one hand and take what would appear to be an almost certain winning hand into an overall push if you split. Worse yet, if the dealer doesn't have a 10 down, and you take 10's the dealer would have drawn you may well be taking cards away from the dealer that would have busted her. A seven card up gives the dealer a lot of outs. And if you split your 9's and don't hit them with 10's you may well bust. So it's better to stand on 9,9 against a 7 even though it may not seem too logical on the surface.

 

[zengrifter]

I agree but ZG's post appears to state basic strategy is a recent development based on simulations produced by hi speed computers, I believe BJ basic strategy was known and published prior to HP, Sony, Dell etc.

Computers in the late 50s and early 60s that confirmed and perfected BS include Honeywell (Wilson) and the IBM-704 (Thorp, Braun). zg

IBM-704

 

[zengrifter]

Strange.
I posted a post where it said I had found the article I was lookiking for, and while my memory of it being four US Army Analyst was right,it turned out the original work was done on desktop calculators, not computers.
That said, BS as we know it wasn't done until Braun did it with the high speed computers of the early 60s.

Hand crank calculators. Thorp is believed to be the first use of computer mainframe, but even his improvement of the four Army scientists' work had numerous errors. Perfection was achieved in '66 at the IBM Data Center in Chicago. zg

 

[ScottH]

Bottomline: the only logic is based on statistical probabilities derived from high-speed computer simulations, otherwise known as BASIC STRATEGY. zg

Right. You can try come up with logic to figure out why that's the correct move to help you remember the correct play, but at the end of the day the only reason is because of combinatorial analysis. I do think if you try come up with a rational reason why you do it, it will help you to remember it, but those reasons are only memory aids.