Splitting ten's
- Thread starter squeeks
- Start date
Martin Gayle
Well-Known Member
Splitting 10's
Splitting 10's drastically increases your variance.
Split 10's always if you can get away with a huge spread. It is not that bad of a move, it is always +EV vs 5,6 at neutral or positive count.
Splitting 10's drastically increases your variance.
Split 10's always if you can get away with a huge spread. It is not that bad of a move, it is always +EV vs 5,6 at neutral or positive count.
zengrifter
Banned
That is a somewhat misleading statement. zgMartin Gayle said:
Not really; the indexes are +5 and +4. You still might be +EV, but you're definitely much less so with this move, and it will hurt your aggregate EV (and variance).Martin Gayle said:
Many bad plays can be overcome with additional spread/time (see Andersen), but I'm not sure this one makes sense. Maybe a couple of times with a small bet if you wanted to make a big show of it. Still doesn't seem worth it to me.
I took out my trusty CVDATA for this debate and the results are in. Since hilo is the board favorite I used compelte hilo indices w/ 1dtc.
4.5/6d, s17, ls, das, play all
top sim: hilo with no splitting tens or DD s20.
btm sim: hilo complete full indices
BJC
4.5/6d, s17, ls, das, play all
top sim: hilo with no splitting tens or DD s20.
btm sim: hilo complete full indices
BJC
Attachments
Last edited:
I agree, but if you're clearly doing it at bad times (such as off the top of the shoe) one could argue that it may mark you as an idiot/desirable player in their eyes. But it's expensive camo, and probably not all that effective considering the risks. Further, considering the sim posted above, the ability to safely split tens later on (when appropriate) seems to be of very small utility anyway.bjcount said:
No point, IMO.
moo321
Well-Known Member
But at the margin, you could really hurt yourself not splitting tens. If you have a max bet out and the count is +9, you're sacrificing an hour's expectation by not splitting.
I use a higher index for splitting, but it needs to be done. Use free camo, like playing shorter sessions, not expensive camo like strategy changes.
I use a higher index for splitting, but it needs to be done. Use free camo, like playing shorter sessions, not expensive camo like strategy changes.
nightspirit
Well-Known Member
Can you pleas also run a sim where the player is always splitting tens vs. 5 & 6? Just for comparison. But please more rounds, try a billion. thanksbjcount said:
Blue Efficacy
Well-Known Member
Lets not forget the value of clearing off a table by splitting tens! :grin:
nightspirit
Well-Known Member
Thank you. The standard error gets smaller the more rounds you run.bjcount said:a billion DOLLARS, oh you said rounds? I thought you owned a bank, car dealership, or insurance company too?
WISH GRANTED
BJC
But I'm wondering why the SD stays almost the same though we split tens all the time. Maybe the amount of indices used compensates the penalty.
nightspirit
Well-Known Member
Well, since I was somehow confused by the sims of BJCount I ran them again. Maybe we misunderstood.
Though I used another system, spread etc. than BJCount the principle stays the same. Like squeeks wanted to know, in second sim the player was always splitting tens (so it's basic strategy) vs. 5 & 6 and in the first he was only using the play at or above the correct index number.
Sim parameters: 1 player, 6D S17 DAS, 75%, Spread 1 to 20, play all, Hi-Opt II w/ Ace SC, 2 Billion rounds.
In case I did no mistake I would say always splitting tens is a no no! :whip:
Though I used another system, spread etc. than BJCount the principle stays the same. Like squeeks wanted to know, in second sim the player was always splitting tens (so it's basic strategy) vs. 5 & 6 and in the first he was only using the play at or above the correct index number.
Sim parameters: 1 player, 6D S17 DAS, 75%, Spread 1 to 20, play all, Hi-Opt II w/ Ace SC, 2 Billion rounds.
In case I did no mistake I would say always splitting tens is a no no! :whip:
Attachments
Last edited:
bj bob
Well-Known Member
The bottom line?
Now let me get this straight. The CVDATA calculations show that there is only a minuscule $.65/hr. difference between correct hi-lo full index play and simply splitting 10's v. 5,6 willy-nilly without counting? Sounds pretty low to me. Also, what min. max. bets were you using here?
Now let me get this straight. The CVDATA calculations show that there is only a minuscule $.65/hr. difference between correct hi-lo full index play and simply splitting 10's v. 5,6 willy-nilly without counting? Sounds pretty low to me. Also, what min. max. bets were you using here?
nightspirit
Well-Known Member
I was also confused by it, but I think bjcount simply forgot to edit the strategy. Can happen.bj bob said:
Maybe there is some misunderstanding among us all.nightspirit said:
Since Hilo seems to be the most popular, not the one I use in play, I ran the sims for the majority.
The setup was: 4.5/6d, s17, das, LS, S up to 4 hands, One card on split aces no resplitting.
The betting strat was a 1-12 spread play all. TC<=0=1u, tc1=3u, tc2=6u, tc3=8u, tu4=10u, tc5=12u . $10. unit
Going to the first pair of sims I posted in this thread (top & bot are two different sims explained below):
The first SIM is using "Complete HILO" with all the split ten indices removed as well as all the "Double Down Soft 20 indices" and changed to "Stand" as I stated in my prior post that the heat all these plays would generate (in HL and high stakes) are not worth using and I placed in my "Never Attempt" file.
The second Sim is using the "Complete HILO" as provided with CVDATA using all indices, None removed, so all indices for splitting tens were used not just against dealers 5 or 6. None of the Sims used BS or "willy nilly" counting.
BJ BOB, I think there was a mix up some where along the way, but I hope the above straightens it out.
The single sim is the billion hand sim run using "Complete HiLo" all indices with the same rules, bets, etc. above.
BJC
Last edited:
bj bob
Well-Known Member
OK, I get the picture now. You were comparing all the split/ no split 10's assuming a count with a $10 min. bet and a $120 max which then tells me that the vast majority of these decisions made at the $120 level, and therefore puts into perspective the poor risk/ reward ratio.bjcount said:
nightspirit
Well-Known Member
Thanks for the clarification bjcount!
With the parameters you used your sims are correct, my request was to run a sim where the player is always splitting tens, to compare the result to a player who only splits at the right time (like sqeeks asked). Maybe we get lost there, never mind.
My betting spread for Hi-Opt II was 1 unit <= TC +1, 3 units @ 2, 7@3, 10@4, 13@5, 17@6, 20@7 and above. True count is floored. Unit size was €25.
I think squeeks has now all the information he needs to decide what to do with the pair of 10's. But I agree with you, I would simply forgo the split, especially at high stakes.
With the parameters you used your sims are correct, my request was to run a sim where the player is always splitting tens, to compare the result to a player who only splits at the right time (like sqeeks asked). Maybe we get lost there, never mind.
My betting spread for Hi-Opt II was 1 unit <= TC +1, 3 units @ 2, 7@3, 10@4, 13@5, 17@6, 20@7 and above. True count is floored. Unit size was €25.
I think squeeks has now all the information he needs to decide what to do with the pair of 10's. But I agree with you, I would simply forgo the split, especially at high stakes.