Was it correct to double ?

Minimum bet is 100 and maximum bet is 500 and bets in increments of 1 can be made. No doubling for less allowed and no surrender.

It is hand 9 out of 10 and I am BR2 with 598 chips betting first 108 chips and my opponent BR1 701 chips matching my bet with 108.

I am dealt a pair of 5`s (but for the sake of this example lets say is only a 6 and a 4 right now) and my opponent a hard 14, dealer shows a 6 up card.

I am doubling with the hope that I will not lose this hand and then I will have about 75% win chance by just matching my opponent bet on the last hand.

OK, but what it bugs me, is that if I lose this double, then my opponent will have a near-lock on me on the last hand by betting between 164 to 218.

And for this particular hand, I have another question: should I had split those 5`s instead to double down them ? (I think so, but I am not 100% sure.)

 

My one-minute-if-it-were-me-at-the-table assessment is to go ahead and double for the reason that you gave - a chance at a strong advantage on the final hand. Your opponent actually handed you this opportunity by not covering your possible double with his bet. If you, instead, go for the swing you have a much lower chance of gaining that advantage.

True that if you lose your double, you could be in a tight spot, but it's perhaps not as bad you might think. He would likely either stand on his 14 or double it (which I think he should). If he stands, he cannot win while you lose and if he doubles, I think it is unlikely that he will win while you lose.

Sometimes (usually?) the strategy which yields the best chance of success also increases the chance of disaster. But it doesn't matter whether you lose by a hair or by disaster, you still lose.

As for whether to double or split the 5's, for the goal of winning two bets, I would take the double every time.

 

Thank you Gronbog, and sorry but I still have one question and so I will better ask you now about it too:

- The goal of winning two bets, seem equal with the goal of winning one hand and losing the other one ?

And one more questions sorry: do you think I should have bet 208 instead of 108 ? (I am again, not sure)

 

I know your questions are really for Gronbog.

Your opponent should have bet at least 114 to start with. That way if you win a double he only has to win to still be ahead.

As for keeping the low, he can only do that if you have like results or better(w/w, l/l, p/p or he wins and you do not win a DD) since you bet enough to take the high if he loses or pushes and you win.

You have about a 60% chance of winning a DD plus about 8% push. However, if you lose then your opponent will also lose unless he hits or DD (unlikely for either to occur) with 14 vrs a 6 unless you catch a good card. If so he can double for less (if able to do so) to try to keep the lead. As long as he does not drop more than a 1/2 max bet behind you it is not going to matter much in the % chance he has to win.

As to the split. You will be hitting two 5's. Even against a 6, your odds are only about 49% to win starting with 5. I don't believe you would want to take this play vrs the nearly 70% chance of winning or pushing with the double.

 

No need to apologise for asking questions. We all learn from asking and attempting to answer them.

In this situation,. the two goals are not equal. If you win two bets, you have a good shot at taking the lead simply by winning. Your opponent would have to double his 14 in order to take that away from you. Even so, you still have a better chance at winning your double on 10 than he does doubling his 14.

If you split you do still have a reduced chance chance to win two bets, but your probability of a push does increase (many splits increase the odds of a push), which helps you should your opponent lose. So the question is whether your overall probability of taking the lead is better. My gut still says that the increased chance of swinging him with a push is less than what you give up by not doubling. My simulator can provide the answer.

In this situation, no matter what you bet, your opponent can approximately match it and can cover your bet and your double (because you can only bet 299 or less and still double) without giving up the low. (This is incorrect!!! See my later post below!!) So I don't see any improvement in betting 208 over 108. You could bet 403 or more and beat him with a blackjack, but that's not a good trade against the ability to double.


Given that, I would prefer the lower bet, since it reduces the level of catastrophe, should you get swung, leaving you with a small chance .

 

- I am very grateful if you can post what your simulator says in this case. (because I think this scenario can be used as a rule of thumb)

I find this interesting too, just beacuse my oponent did a quite bad bet by exactly matching my bet. If I bet 208, my opponent should bet at least 210. But since he matched my 108 bet, I have reasons to think that he would also exactly match my bet with 208 too. - Well, if I am right on this, then a blackjack for me would force him to double if he does not get a blackjack also. So, I would see a small gain for me with a 208 bet instead of 108 ?

 

Surprise!

Well, so much for my gut -- at least against a skilled opponent.

I ran two sims, one vs a basic strategy opponent and another vs an optimal opponent.

The basic strategy opponent will stand on his 14, meaning that the situation is as outlined by Larry. In this situation you take the lead by doubling 68.2% of the time as opposed to 53.9% by splitting (and playing optimally).

Vs an optimal opponent, you take the lead 46.1% of the time by splitting and only 39.4% by doubling. Your opponent's optimal response against your double or split is to double regardless of the totals you end up with. It seems that the increased opportunity for a push is valuable indeed in the face of the reduced effectiveness of doubling.

Hitting to 17 or better gets you a swing 31.9% of the time vs either opponent.

 

Good point! You would give him an opportunity to make the same mistake but with a better result for you.

 

Thank you so much ! :cheers: - I really never thought this scenario could go so deep. Well, so as a rule of thumb if I know/think I am playing vs a very good player I should split. But if I know/think I am playing vs a basic strategy player, then I should double. Good to know, and yes, I hink it can be a rule of thumb.

And I think that if after splitting I will get a hand 8 to 11 on either one I should double ? (considering doubling after split is allowed)

 

The above is incorrect. Sorry for any confusion!!

It's even better. A bet of 207 or more is an example of the rule of 2 (see Ken's book for the rules of 2, 4 and 5). With a bet of 207 or more (twice the lead plus a chip), your opponent cannot cover your double without giving up the low. I'm liking this bet more and more.

 

Hmmm, just thinking a bit more about it: (this situation is soooo complex) - if only 46.1% chances for me to take the lead now on this hand, then I am not better if I only hit and hope for a swing on this hand taking into account that on the next hand (last hand) I will be acting last and I will be able to take the high ?

 

Gronbog, thanks for running your sims.

I find it an interesting result between doubling and splitting against an optimal opponent. Even though you may have the lead 8% more often with the split, it makes me wonder how much of a lead? Are you going to be better off overall going to the last hand? Will you gain enough % in this scenario to offset when you will really be worse off considering the next hand?

I assume the optimal play splitting the fives would have you double on A5, 9, 10 or 11 and not busting. These situations could really put you in the hole.

Larry

 

The optimal strategy for playing the first split hand is to double any hard total up to and including 12 (includes another pair of 5s) and stand otherwise. The difference between double/stand on 13 is within 0.1%. Also double soft 16 (A,5 -- the only soft total you can be dealt from a 5). There is a separate optimal strategy for playing the second hand for each possible result of the first one. Too many to list here.

 

I am still on this: it seems to me that against an optimal player I get 46.1% swing chances by splitting (only if I play correct my split thereafter! - which is also NOT always easy) and if I double I get only 39.4% swing chances.

Thinking again about it, and knowing that if I only hit once my pair of 5`s I still stand about a 44% win chance acting last on the last hand in case that a swing will not luckly happen right now, I am not better to only hit instead of anything else ? (only assuming that I play against an optimal oponent - otherway double)

 

We have been considering the best possible way to play your hand should you decide to double/spit in order to go for the high. It is true that in order to decide whether this is the best play overall with two hands remaining you must consider each possible outcome of each play (hit/split/double) combined with your opponent's possible responses and outcomes and their probabilities along with your probability of prevailing on the final hand from each of these positions.

I did such an analysis for someone here once. It required many simulations and a spreadsheet to put it all together.

 

Time is precious, I know. I just did not know that this situation could be so complex. I did not meant to put anyone on a great deal of effort in order to help me. (am sorry if that has already happened) - I have got a lot of useful insights until this point which should help me a lot, and I can only thank you a lot for this !

 

Not to worry. No extreme effort has been expended and, as I said before, we all learn from these teasers. Please keep them coming!