manny
manny
  • Threads: 1
  • Posts: 1
Joined: Feb 21, 2010
February 21st, 2010 at 12:39:57 PM permalink
Having recently taken a huge beating at the Borgata (I know, I know, sample size of one) I was wondering about the wisdom of resplitting 8s against a 10, 9 or even an 8. I understand and accept the math that in the long run splitting cuts your loss in half compared to hitting a 16, but in any single trial is there a point where you've got too many units out there at a negative expectation to make the strategy worthwhile?

In my example I was playing an 8-deck shoe and ended up splitting 8s to five pairs (a sixth was 8-3, which I dutifully doubled and of course drew a 4). The whole time I'm putting new units on the table I'm staring at that 10 and thinking that I'm throwing away money. And sure enough, the dealer turns over another 10 and all that cash is gone.

Happily, the entire rest of the shoe was something of a statistical miracle for the whole table so my beating merely turned a massive win into a small loss, as opposed to completely ruining my night.

But the question remains. Has someone done work on re-splitting 8s against high cards?
Wizard
Administrator
Wizard
  • Threads: 1508
  • Posts: 26886
Joined: Oct 14, 2009
February 21st, 2010 at 6:59:02 PM permalink
If splitting is the right play for splitting once, then you should keep re-splitting as long as you still can. I know that it is natural to question splitting eights when you re-split them to four hands, and then lose all of them. At those times, think of the times you split eights to get 18s, and then the dealer got a 17 or busted. As I've said hundreds of times, a combinatorial analysis or computer simulation will prove you should always split eights under American rules every time.
"For with much wisdom comes much sorrow." -- Ecclesiastes 1:18 (NIV)
pacomartin
pacomartin
  • Threads: 649
  • Posts: 7895
Joined: Jan 14, 2010
February 22nd, 2010 at 7:44:25 PM permalink
Assuming 6 decks, dealer hits a soft 17:

A player who has double 8's against a dealer 10 if you stand or hit you have expected values of -0.537 and -0.535. It's not a surprise since you should surrender a hard 16 against a 10.

The expected value are -0.475 or -0.486 depending on if you can double after you split or not. So the improvement is measurable but is still pretty bad.
FleaStiff
FleaStiff
  • Threads: 265
  • Posts: 14484
Joined: Oct 19, 2009
February 23rd, 2010 at 2:14:34 AM permalink
Quote: manny

Having recently taken a huge beating at the Borgata (I know, I know, sample size of one) I was wondering about the wisdom of ...


Its that "huge beating" that has the emotional impact. Despite the sample size of one, it sure sticks in your mind and takes on a great weight of its own.

So it doesn't matter too much if its "the wisdom of splitting 8s" or "the wisdom of declining insurance" or the wisdom of anything else. We tend to give undue weight to instances that have such emotional impact.

I don't even remember the details of one time but I wound up with four hands yet that dealer wound up with 21. I remember the whole table groaning. I think even the dealer regretted that he won. Anytime the entire table groans in sympathy its going to be memorable but that doesn't make it anything other that proper strategy just the same.
  • Jump to: