Index for 8,8 v 9 anyone?
- Thread starter UK-21
- Start date
No takers? Obviously not in anyone's repotoire then?
I've extracted the win/loss probabilities for 8,8v9 from my 6 deck table and get this:
8,8 vs 9
Standing Hitting Doubling Splitting
-0.53889 -0.50571 -1.01142 -0.38920
This is based on OTT of a newly shuffled shoe. So for every £/$ wagered, splitting is the best move as it would mean you'd (only?) lose £/$ 1.95 for every £/$ 5.00 bet. Splitting is clearly a "defensive" move and aimed at reducing the long term loss from being dealt a bad hand - clearly the best thing to do for a BS player flat betting.
Now for those who count, what would be the shift in these figures if the TC was at +4, and the liklihood of the dealer drawing a 10 increased from 16/52 to 19/52? Also, going back to not doubling certain hands at high counts to reduce variance (sorry) which would be better - standing or doubling?
Assuming that the above figures don't shift too much (and as such let's use them as they are) then on a 1-8 betting ramp with £/$ 5.00 units and at TC+4;
Split
£/$ 5.00 x 8 x 2 x 0.38920 = £/$ 31.14
Stand
£/$ 5.00 x 8 x 1 x 0.53889 = £/$ 21.56
Reduction in loss by standing works out at around 2 units, or roughly the hourly return for APs who screw everything down. If you were a black chip player, doing this would cost £/$200 a time over the longer term. Why would anyone who counts cards in an effort to glean an advantage want to do this?
So does it therefore follow that splitting 8s against a dealer 9 at TC+4+ effectively cancels out the last hour's graft at the felt?
You could also apply the same consideration where the dealer has a 10 showing, although in the UK you never split 8s then due to the NHC rule.
Comments anyone ?
I've extracted the win/loss probabilities for 8,8v9 from my 6 deck table and get this:
8,8 vs 9
Standing Hitting Doubling Splitting
-0.53889 -0.50571 -1.01142 -0.38920
This is based on OTT of a newly shuffled shoe. So for every £/$ wagered, splitting is the best move as it would mean you'd (only?) lose £/$ 1.95 for every £/$ 5.00 bet. Splitting is clearly a "defensive" move and aimed at reducing the long term loss from being dealt a bad hand - clearly the best thing to do for a BS player flat betting.
Now for those who count, what would be the shift in these figures if the TC was at +4, and the liklihood of the dealer drawing a 10 increased from 16/52 to 19/52? Also, going back to not doubling certain hands at high counts to reduce variance (sorry) which would be better - standing or doubling?
Assuming that the above figures don't shift too much (and as such let's use them as they are) then on a 1-8 betting ramp with £/$ 5.00 units and at TC+4;
Split
£/$ 5.00 x 8 x 2 x 0.38920 = £/$ 31.14
Stand
£/$ 5.00 x 8 x 1 x 0.53889 = £/$ 21.56
Reduction in loss by standing works out at around 2 units, or roughly the hourly return for APs who screw everything down. If you were a black chip player, doing this would cost £/$200 a time over the longer term. Why would anyone who counts cards in an effort to glean an advantage want to do this?
So does it therefore follow that splitting 8s against a dealer 9 at TC+4+ effectively cancels out the last hour's graft at the felt?
You could also apply the same consideration where the dealer has a 10 showing, although in the UK you never split 8s then due to the NHC rule.
Comments anyone ?
Reduction in loss?
If (as you've done) you assume the same EV percenatges, unadjusted for the +4 TC, then the loss from standing is the same % of your original bet, whether that bet was £5 or £40.
Specifically, you should say -
Split
£/$ 5.00 x 8 x 1 x 0.38920 = £/$ 15.57
,not
Split
£/$ 5.00 x 8 x 2 x 0.38920 = £/$ 31.14
If the latter were correct, then you should always stand.
If (as you've done) you assume the same EV percenatges, unadjusted for the +4 TC, then the loss from standing is the same % of your original bet, whether that bet was £5 or £40.
Specifically, you should say -
Split
£/$ 5.00 x 8 x 1 x 0.38920 = £/$ 15.57
,not
Split
£/$ 5.00 x 8 x 2 x 0.38920 = £/$ 31.14
If the latter were correct, then you should always stand.
What you have picked up on is the fundimental difference between flat betting BS and ramping bets up according to the count. Although the long term return figures used don't change with the level of the bet, the amount that you stand to lose will increase by betting more - splitting rather than standing - as will the variance.
I don't understand why you've used "£/$ 5.00 x 8 x 1 x 0.38920 = £/$ 15.57" as this would assume that you don't need to double your bet if you take the option to split, and the loss (neg return of 0.38920 is based on the original 8 unit bet??
I don't understand why you've used "£/$ 5.00 x 8 x 1 x 0.38920 = £/$ 15.57" as this would assume that you don't need to double your bet if you take the option to split, and the loss (neg return of 0.38920 is based on the original 8 unit bet??
All EV percentages are based on the original bet. Otherwise comparing them would make no sense in determining basic strategy.
You seem to be suggesting that splitting with your £5 bet will cost you a total of (£5+£5) * 0.38920 = £3.89, while standing costs only £5 * 0.53889 = £2.69. It would be a perverse kind of basic strategy that directed you to throw away £1.20 like that, when you could save the additional £5 for the next hand where it will only cost you about 2.5 pence.
You seem to be suggesting that splitting with your £5 bet will cost you a total of (£5+£5) * 0.38920 = £3.89, while standing costs only £5 * 0.53889 = £2.69. It would be a perverse kind of basic strategy that directed you to throw away £1.20 like that, when you could save the additional £5 for the next hand where it will only cost you about 2.5 pence.
newb99,
I was about to post a reply to this topic but my bloody computer froze so I will post up the reply. I agree with London Colin in his analysis. 8,8 vs 9 is a losing hand no matter how you put it - our aim is to lose less money by making the correct decision. Splitting has been shown to reduce less money compared to the other decisions.
Because of ENHC, you would hit 8,8 vs 10 and Ace rather than split and you would also hit your A,A vs 10 or Ace rather than split.
Splitting 8,8 vs 9 creates more variance and more fun esp. if the game allows DAS. You could draw a 2 or three and then double down. I checked Katarina Walker's book on 8,8 vs 9 in both Pontoon games and Spanish 21 games and in both cases, the correct strategy is to split. I know that her book is Pontoon related and not BJ related but I think because the index for 8,8 vs 10 is TC = 8 , I would assume that the index for 8,8 vs 9 would be a TC greater than 10. Its very rare for TC of > 10 to occur in a shoe game (similarly with 16 vs 7, you would stand if the TC > 9 - which is rare).
How about surrendering if the game offered you the opportunity to surrender? 16 vs 9 is one of the Il 18 index whereby if the TC = 4 or above, you surrender. However what about the case of 8,8 - if the TC was 4 and you are dealt 8,8, would you surrender?
I was about to post a reply to this topic but my bloody computer froze so I will post up the reply. I agree with London Colin in his analysis. 8,8 vs 9 is a losing hand no matter how you put it - our aim is to lose less money by making the correct decision. Splitting has been shown to reduce less money compared to the other decisions.
Because of ENHC, you would hit 8,8 vs 10 and Ace rather than split and you would also hit your A,A vs 10 or Ace rather than split.
Splitting 8,8 vs 9 creates more variance and more fun esp. if the game allows DAS. You could draw a 2 or three and then double down. I checked Katarina Walker's book on 8,8 vs 9 in both Pontoon games and Spanish 21 games and in both cases, the correct strategy is to split. I know that her book is Pontoon related and not BJ related but I think because the index for 8,8 vs 10 is TC = 8 , I would assume that the index for 8,8 vs 9 would be a TC greater than 10. Its very rare for TC of > 10 to occur in a shoe game (similarly with 16 vs 7, you would stand if the TC > 9 - which is rare).
How about surrendering if the game offered you the opportunity to surrender? 16 vs 9 is one of the Il 18 index whereby if the TC = 4 or above, you surrender. However what about the case of 8,8 - if the TC was 4 and you are dealt 8,8, would you surrender?
Yesterday I had $200 on each of two boxes (TC +6) and I was dealt 8, 8 on my right hand and 19 on the left; the dealer's card was an Ace (no surrender against A) and I took insurance to cover $300. After thinking for a while I decided to split the 8s (considering that if the dealer had bj I would't have lost the split bet) and I got 18 in both. At the end the dealer had blackjack and I lost $100 (could have been worst). I was reading your post and I was thinking if I made the right move... 
88 vs. 9 is a legitimate surrender play, you'd surrender it before you'd stand but you'd split before you surrender. 88 vs. 10 is something you surrender in any positive count, if surrender is available. Exactly what the index is I don't know, and it's going to depend quite a bit on what count you're using, and whether or not you can double after splitting.
Quite right. And I think I have misinterpreted the table I have. But it still looks to be another aspect of sacrificing some long term EV in order to reduce variance. Choice, I know.
The difference between the cost of standing and splitting on this hand is c0.15/£$, and at the top of a 1-8 spread at £/$5.00 per unit this would equate to an average (longer term) increase in the loss of c£/$ 6.00. But in order to reduce this we split and put out another £/$40.00, which doesn't change the expectation from a loss to a win (or even a push) but just reduces the loss. Where else in life would you risk £/$ 40 in order to reduce an expected loss from £/$21.56 to £/$15.57. If you have the option of surrendering the hand most would take it and give up £/$20.00. If the game is NS, most won't stand at a cost of £/$1.56 more, but risk another £/$40.00 in an effort to reduce the expected loss - risking another 8 units to reduce the expected average loss by just over 1.
If you didn't have the numbers behind the BS chart available to refer to, would you consider this a good deal?
The difference between the cost of standing and splitting on this hand is c0.15/£$, and at the top of a 1-8 spread at £/$5.00 per unit this would equate to an average (longer term) increase in the loss of c£/$ 6.00. But in order to reduce this we split and put out another £/$40.00, which doesn't change the expectation from a loss to a win (or even a push) but just reduces the loss. Where else in life would you risk £/$ 40 in order to reduce an expected loss from £/$21.56 to £/$15.57. If you have the option of surrendering the hand most would take it and give up £/$20.00. If the game is NS, most won't stand at a cost of £/$1.56 more, but risk another £/$40.00 in an effort to reduce the expected loss - risking another 8 units to reduce the expected average loss by just over 1.
If you didn't have the numbers behind the BS chart available to refer to, would you consider this a good deal?
"you would also hit your A,A vs 10 "
You sure about that ? according to BS on this website and BJA it says spilt AA vs 10
http://www.blackjackinfo.com/bjbse.php?numdecks=6+decks&soft17=s17&dbl=all&das=yes&surr=ns&peek=no
Thanks
Ming
You sure about that ? according to BS on this website and BJA it says spilt AA vs 10
http://www.blackjackinfo.com/bjbse.php?numdecks=6+decks&soft17=s17&dbl=all&das=yes&surr=ns&peek=no
Thanks
Ming
Ming, I apologize, you are correct - I guess splitting Aces is not that costly even if the dealer get a natural cos you will lose only twice the amount that you wager. This is assuming you cant draw or double to split aces because if you had the chance to draw or double to split aces, the ENHC could suck more money out of you.
bj21abc said:
I never use this index - so ran a sim on it just now (6D DOA).
Always split 88 v 9. Chart shows you the respective EVs.
Surrender (risk-averse at that) at TC +13
(good luck ever seeing this count
D.
Always split 88 v 9. Chart shows you the respective EVs.
Surrender (risk-averse at that) at TC +13
(good luck ever seeing this count
D.
Again, I shall refer back to Mr S's hand frequency chart to see how often this hand shows every 100,000 hand on average, and work out the loss in EV for standing rather than splitting - effectively calculating the "cost" of reducing variance by not splitting.