Monkey Wrench

gronbog

gronbog

Top Member
Here is a situation that occurred yesterday at https://www.blackjacktournaments.com/tournaments/blackjack-21-tournament.1235/ and no, Monkeysystem was not there ;)

Final hand of the preliminary round, 2 advance, max 500, min 25, no surrender, DOA2 (except blackjack), DAS (except 21), double for less allowed, split to 4 hands, 1 card only on split aces.

I was in pretty good shape. The bankrolls and betting order were:

BR3: 2100
BR5: 1550
BR2: 2400 (me)
BR1: 3650
BR4: 1900

My thoughts were that BR4 and BR5 would be betting max and that BR3 would probably bet at least 300+ to cover BR4 but would probably also bet max to cover BR5's double. That would leave me free to also bet max so as to cover a tie with BR4's double by winning my hand and to possibly advance with a double for less. My next thought was, "man, I wish I had one more chip"!

That's when BR3 threw a monkey wrench into the whole thing. He surprised me by betting 100 while BR5 bet max, as anticipated.

My questions are:
  1. What do you think of my original plan?
  2. What do you think of BR3's bet?
  3. What would you bet now?
I'll update with what I actually did and what happened after a few responses.
 
I'm with S. Yama on the 175 bet. BR3's 100 is completely bizarre, I can find no mathematical reason for it anywhere. But we all know you can see some real clueless knuckleheads at the events sometimes.
 
1. Good plan.
2. Probably BR3 didn`t knew that that was the last hand ? LOL !
3. 175 (but took me about 30 seconds to decide between 175 and 375)
 
OK. I think we've received all of the responses that we're going to, and all of the reasonable ones have been suggested.

I called this thread "Monkey Wrench" because of BR3's bet. It's good to think ahead about what the other players might do so that we can formulate our own response with less time pressure, but when someone does something completely unexpected, we need to be able to reconsider things quickly. The first thing to think about is "why did they do that?", and that's why I asked what folks thought about BR3's bet. Whether they have made a clever play that you didn't think about, or a clueless one, we may need to change our strategy in order to protect ourselves or to take advantage.

In this case, I think that PlayHunter nailed it. It turns out that BR3 was pretty much clueless throughout. He was an elderly gentleman and didn't seem to be aware of the details of what was happening at the table, including the rules for betting, how we were betting in order, how we were playing in order, etc. He even made a very strange play early in the round where he declined to split his aces and elected to stand instead. He was Asian and didn't seem to understand English very well (if at all). It is quite possible, even likely, that he didn't understand that it was the final hand.

If BR3 had bet big, as I expected, I had two reasonable choices. I could bet max to cover a tie with BR4's double, and risk a potential tie via a 1/2 swing, or hold back a chip to avoid the 1/2 swing and rely on a double of my own to beat BR4. At the time, I was planning on betting max. Now, assuming that BR4 and BR5 would double, if necessary, and play for a swing otherwise and that BR3 would, at best, play basic strategy, simulation shows that I would have advanced or tied to advance with BR4 61.1% of the time with the max bet. It's unfortunate that my sim currently does not distinguish between winning outright or tying. It would be interesting to see the percentages of both and then to try an place a value on the tie result. By holding back a chip and betting 495 (The_Professional's original answer?) I would have advanced 52.7% of the time (no ties possible).

After BR3's surprise bet, I decided that, since I didn't want to give up the low to him, I might as well consider trying to lock him out. A bet of less than 100 would do that, even if he doubled. At the time, since I knew I was giving up on covering BR4's double, I decided that not covering covering BR5's double would be worth locking our BR3, so I bet 75 (I had no red chips). Another option would have been to bet 80, so as to cover BR5's double with a double. Other reasonable bets were suggested here. All of the suggested bets and their simulation results are as follows:

75: (Gronbog): locks out BR3: advances or ties 65.5%
80: (Gronbog): locks out BR3, cover BR5's double with a double: advances 66.9%
175 (S. Yama): covers BR5's double, risks losing while BR3 wins a double: chances of advancing: 66.4%
250 (The_Professional): covers BR3's push, covers BR5's double, can double to tie BR4's double: advances or ties 59.35%
275 (tirle_bj): covers BR3's push, covers BR5's double, can double to beat BR4's double: chances of advancing: 58.43%
375 (Sweet William): covers BR3's loss, can double to beat BR4's double: chances of advancing 58.83%

Once again, for the bets of 75 and 250, it would be interesting too see the ratio of wins to ties.
 
Last edited:
I forgot to post what happened.

BR3 was locked out when he received an unsplittable hand (so far so good)
BR5 was dealt a stiff, doubled it and busted (2 down, one to go)
My hand was now completely irrelevant. I was dealt 15 vs the dealer's 5 and stood.
BR1 had bet minimum and was out of range
BR4 bet max, was dealt 11, doubled and was dealt a face for 21 (ugggh, I need a dealer 21)
Dealer flipped a 3 for a total of 8 and then pulled an Ace for 19 (game over -- tilt)

It turned out that everyone lost their hands except for BR4.
 
Last edited:
gronbog said:
It's unfortunate that my sim currently does not distinguish between winning outright or tying. It would be interesting to see the percentages of both and then to try an place a value on the tie result.
Click to expand...
Of course, you all knew that as soon as I made a comment like that, I would be looking at fixing it :D

I looked at my code and it was even worse than "not distinguishing". It was actually favouring the player to the right. In the case of BR3 or BR5 vs me, it would place them ahead of me when tied. For me vs BR4, I would be deemed the winner of any ties. Clearly not a useful result. However, it was easily fixed. With that done, I re-ran the sims for my bets of 75 (double win ties with BR5's double), 250 (win could tie with BR4's blackjack, double win or loss could tie with BR4) and 500 (win or loss could tie with BR4). Here are the updated results:

BR3 bets big:
500: advances 49.51%, ties 11.60%
495: advances 52.7%

In this case, which bet is better depends on how you think you will do in the tie breaker. For example, if you think you have a 50/50 shot, then the value of the tie is 11.60% / 2 or 5.8%, making the total value of the 500 bet 49.51% + 5.8% = 55.31% which is better than the value of the 495 bet. In order for the 495 bet to be better, you would have to figure that your chances in the tie breaker were worse than 27.5%, i.e. (49.51% + (11.60% x 27.5%) = 52.7%

BR3 Monkey Wrench:
75: (Gronbog): locks out BR3: advances 64.05%, ties 2.72%
80: (Gronbog): locks out BR3, cover BR5's double with a double: advances 66.9%
175 (S. Yama): covers BR5's double, risks losing while BR3 wins a double: chances of advancing: 66.4%
250 (The_Professional): covers BR3's push, covers BR5's double, can double to tie BR4's double: advances 55.82%, ties 3.90%
275 (tirle_bj): covers BR3's push, covers BR5's double, can double to beat BR4's double: chances of advancing: 58.43%
375 (Sweet William): covers BR3's loss, can double to beat BR4's double: chances of advancing 58.83%

In this case, even if you were to succeed 100% of the time in the tie breaker, it would not be enough to beat the value of the 80 bet.
 
gronbog said:
Of course, you all knew that as soon as I made a comment like that, I would be looking at fixing it :D

I looked at my code and it was even worse than "not distinguishing". It was actually favouring the player to the right. In the case of BR3 or BR5 vs me, it would place them ahead of me when tied. For me vs BR4, I would be deemed the winner of any ties. Clearly not a useful result. However, it was easily fixed. With that done, I re-ran the sims for my bets of 75 (double win ties with BR5's double), 250 (win could tie with BR4's blackjack, double win or loss could tie with BR4) and 500 (win or loss could tie with BR4). Here are the updated results:

BR3 bets big:
500: advances 49.51%, ties 11.60%
495: advances 52.7%

In this case, which bet is better depends on how you think you will do in the tie breaker. For example, if you think you have a 50/50 shot, then the value of the tie is 11.60% / 2 or 5.8%, making the total value of the 500 bet 49.51% + 5.8% = 55.31% which is better than the value of the 495 bet. In order for the 495 bet to be better, you would have to figure that your chances in the tie breaker were worse than 27.5%, i.e. (49.51% + (11.60% x 27.5%) = 52.7%

BR3 Monkey Wrench:
75: (Gronbog): locks out BR3: advances 64.05%, ties 2.72%
80: (Gronbog): locks out BR3, cover BR5's double with a double: advances 66.9%
175 (S. Yama): covers BR5's double, risks losing while BR3 wins a double: chances of advancing: 66.4%
250 (The_Professional): covers BR3's push, covers BR5's double, can double to tie BR4's double: advances 55.82%, ties 3.90%
275 (tirle_bj): covers BR3's push, covers BR5's double, can double to beat BR4's double: chances of advancing: 58.43%
375 (Sweet William): covers BR3's loss, can double to beat BR4's double: chances of advancing 58.83%

In this case, even if you were to succeed 100% of the time in the tie breaker, it would not be enough to beat the value of the 80 bet.
Click to expand...
 
Gronbog, thank you for interesting info.
As we can see the bets are devided into the two groups:
First three ~ 66-67% (considering Tie)
Last three ~ 58-59% (considering Tie)
The key difference between those two groups of bets is the option to lock out (or not) BR3 by avoiding straight swing (BR2 lose, BR3 wins).
As we can learn from above example, it is better to eliminate one out of three opponents even by giving away protection by DD from both others DD.
One more note: between two bets 275(tirle_bj) and 375 (Sweet William) there is marginal difference which is the difference between getting BJ by BR2 to cover BR4's DD and half swing (BR2 loses, BR3 pushes).
After all this info in the real world I will consider the winner bet $175 by S.Yama's 10 second decision ;), since we all are less than perfect and it sims safer to me to have some minimal options.
 
You must log in or register to reply here.
Back
Top