Question about Down Under Blackjack
https://wizardofodds.com/games/down-under-blackjack/
The situation is BLUE, meaning the down card is 2, 3, 4, or 5. If you have a total of 16, the strategy line instructs to STAND against 5 or 6, otherwise hit. I'm confused.
If the dealer's upcard is a 5, he either has a total of 7, 8, 9, or 10. If his upcard is a 6, his total is either 8, 9, 10, or 11.
Seems to me those are the kinds of totals you would want to HIT the 16, not stand.
And further, say the dealer's upcard is a TEN. His total would be 12, 13, 14, or 15. Yet now I'm supposed to HIT the 16. Seems to me I'd be better off standing.
What am I missing?
Thanks in advance.
Quote: bobbartopHere's a link to Wizard's strategy. Allow me to direct you to a part of the strategy that I do not understand.
https://wizardofodds.com/games/down-under-blackjack/
The situation is BLUE, meaning the down card is 2, 3, 4, or 5. If you have a total of 16, the strategy line instructs to STAND against 5 or 6, otherwise hit. I'm confused.
If the dealer's upcard is a 5, he either has a total of 7, 8, 9, or 10. If his upcard is a 6, his total is either 8, 9, 10, or 11.
Seems to me those are the kinds of totals you would want to HIT the 16, not stand.
And further, say the dealer's upcard is a TEN. His total would be 12, 13, 14, or 15. Yet now I'm supposed to HIT the 16. Seems to me I'd be better off standing.
What am I missing?
Thanks in advance.
You're missing Rule 4.
"The dealer peeks for blackjack when an ace or any 10-point card is the dealer's face-up card."
The dealer only uses the reader (and reveals the range of the down card) when they have an A or 10-value card.
So, if the dealer's up card is NOT one of those, you would revert to the strategy correct for the other/house rules of the game. DAS, DA2, 6:5 or 3:2, whatever.
Um.... The dealer uses the reader for EVERY down card.Quote: beachbumbabsThe dealer only uses the reader (and reveals the range of the down card) when they have an A or 10-value card.
The old version used specially designed cards with symbols on the back to identify the card’s range. Therefore, with those cards, the peek was used only for potential BJs.
But the new reader can identify the range for any card, making custom decks unnecessary.
Bottom line: Every dealer down card IS scanned for the range.Quote: Rule 5Using the peeking device, the dealer shall then reveal to the players whether the hole card is small, medium, or high.
Quote: DJTeddyBearUm.... The dealer uses the reader for EVERY down card.
The old version used specially designed cards with symbols on the back to identify the card’s range. Therefore, with those cards, the peek was used only for potential BJs.
But the new reader can identify the range for any card, making custom decks unnecessary.
Bottom line: Every dealer down card IS scanned for the range.
Thank you, both Babs and DJTeddyBear for your replies. I did not know there had been a change from an original version. But yes, now it is as DJTeddybear describes.
So I am finding it very difficult for myself to memorize Basic Strategy. To me it is difficult, but I am sure others will find it easier. It's like there are three different sets of Basic Strategy.
A specific hand is bothering me, as I mentioned above. In BLUE Mode (small card), Wizard's strategy says to hit hard 16 against a dealer's 10 upcard. That means you are hitting it against a 12, 13, 14, or 15. Does this seem correct to you?
Interested to see if this can catch on like a Free Bet, or BJ Switch did in the past.
Quote: WizardI didn't actually analyze that game but copied the strategy from the BMM report. I will try to confirm it based on infinite decks. I'll be out of town over the weekend so this will have to wait until early next week.
Thank you. I did play a couple hours yesterday while I got my car serviced in Fresno. The second hand I was dealt was the particular hand I was questioning, in fact it came up rather often. I decided to go against the strategy and stand on the 16.
Quote: SM777Has this game actually gone live anywhere yet? It's been 5 months since the conference, are any of those winning casinos up and running?
Interested to see if this can catch on like a Free Bet, or BJ Switch did in the past.
Somewhere in the description it says that Barona is spreading it, and I think somewhere in the state of Washington. But I have personally seen it at two places in central California, Eagle Mountain has had it for a few months and now Table Mountain started dealing it maybe three weeks ago. Eagle is a small casino but the game does get play. They offer it at a lower minimum which attracts more players. They have five decks in a continuous shuffler. At Table Mountain, it is a six-deck shoe. Table Mountain is a much larger casino servicing a large city like Fresno. I asked and it has been received well there. Yesterday it was $5 to $300. A lot of action while I was there in the afternoon. Naturally, Friday and Saturday nights are busy nights at both places.
This is a very interesting game, in my opinion. And it should be good for the casinos because I can't imagine the average recreational player ever figuring out how to play it properly. It's confusing. But for the more determined player I think it has much potential.
Quote: DJTeddyBearUm.... The dealer uses the reader for EVERY down card.
The old version used specially designed cards with symbols on the back to identify the card’s range. Therefore, with those cards, the peek was used only for potential BJs.
But the new reader can identify the range for any card, making custom decks unnecessary.
Bottom line: Every dealer down card IS scanned for the range.
You're right, thanks. It is every hand regardless of upcard. I read the wizards page only, and rule 4 says, dealer uses the device to peek for bj. Rule five says, dealer then uses the device to determine the range. I took that to mean the device was only in use when the up card was a 10 or A.
Wizard explains it better in the updated video from G2E with the inventor, here.
https://wizardofodds.com/video/down-under-blackjack-2017
Mike, can I suggest you link that video to the strategy page in WoO, or perhaps include in rule 4-5 something like " with any upcard, the dealer uses the reader to determine the range of the hole card".
Quote: bobbartopSomewhere in the description it says that Barona is spreading it, and I think somewhere in the state of Washington. But I have personally seen it at two places in central California, Eagle Mountain has had it for a few months and now Table Mountain started dealing it maybe three weeks ago. Eagle is a small casino but the game does get play. They offer it at a lower minimum which attracts more players. They have five decks in a continuous shuffler. At Table Mountain, it is a six-deck shoe. Table Mountain is a much larger casino servicing a large city like Fresno. I asked and it has been received well there. Yesterday it was $5 to $300. A lot of action while I was there in the afternoon. Naturally, Friday and Saturday nights are busy nights at both places.
This is a very interesting game, in my opinion. And it should be good for the casinos because I can't imagine the average recreational player ever figuring out how to play it properly. It's confusing. But for the more determined player I think it has much potential.
Good info.
Thanks !
Quote: SM777Has this game actually gone live anywhere yet? It's been 5 months since the conference, are any of those winning casinos up and running?
Interested to see if this can catch on like a Free Bet, or BJ Switch did in the past.
Table Mountain Casino outside of Fresno California has it running.
It looks like I made some transcription errors in creating the old strategy. Here is a corrected one, based on 4 to 8 decks.
Comments?
There are also at least two important errors on the Medium Hole Card chart. if there is any interest, I will post that as well.
Quote: gordonm888I calculated this. For small hole card , this includes double 8v10 and some doubles on soft hands. There are some minor differences on the split portion of the chart as well.
There are also at least two important errors on the Medium Hole Card chart. if there is any interest, I will post that as well.
YES, there is interest. Please post.
Thank you, and thanks to Wizard. Thanks everyone.
Quote: gordonm888There are also at least two important errors on the Medium Hole Card chart. if there is any interest, I will post that as well.
Thank you for your help. I'm interested.
Quote: WizardSorry this took so long. ChesterDog was kind enough to send me a hard total strategy with a small hole card, which I agree with.
It looks like I made some transcription errors in creating the old strategy. Here is a corrected one, based on 4 to 8 decks.
Comments?
I notice there are a few differences between Wizard's (Chester Dog's) hard totals and Gordon's. I appreciate that this is a lot of work for all of you and I didn't expect to see any more posts in this thread so quickly. Thanks again to all.
I'll be offline for a couple days, I am looking forward to your findings and taking the knowledge up to Table Mountain.
Where is Romes in this thread? He's got a good brain, he should like this stuff.
Did Gordon say "double hard 8 v. Ten"?
Quote: bobbartop
Did Gordon say "double hard 8 v. Ten"?
I meant to say Double "10 v 8 + small."
The two bad mistakes in the Medium Hole Card charts both involve the player having a hard 17.
You should definitely HIT H17 v A+medium. Fer Chrissakes, the dealer has, with equal probability, an 18, 19 ,20 or soft 17 and you advise we stand on H17? . I get for "H17 vs Ace+medium": Stand = -0.8202 and Hit = - 0.5927. That's a large difference for a situation in which the player never stands or doubles on anything below a hard 17. This single correction is worth almost 0.06% in House Edge.
Also, you should not Hit a "H17 vs 2+medium. I get Stand= -0.491 and Hit = -0.5949, so its not even close. This is another case where the player will get a H17 over 13% of the time because he never stands or doubles with less than H17.
Also, the posted strategies for 10 vs 2-A+Med and 5,5 pair vs 2-A+Med were very different. Such an inconsistency was a flag that there were still more problems so I calculated the following strategy for medium cards. On the chart the dark toned squares are the differences form the posted strategy:
I am not really sure about 10,10 vs 7 (or 6), because I was unsure of the rule interpretation: if you hit a split 10 with another 10 whether that 20 will win against a dealer 22. If it does win against a dealer 22 then you should split 10,10 vs 7 +medium. However, the value in splitting 10s is that you will frequently resplit to a total of four hands and the dealers 7+medium will bust with a frequency greater than 50%. I think someone needs to look at that particular hand very carefully -with a composition dependent code or with a sim model, because it is such a wild hand.
I think, all together, that over 80 squares in the posted strategy charts were in error!
Interestingly, on this chart, is that you should double on a soft 20 vs 5,6 + high. Not a lot of games where you double on a soft 20, but a 5+high means the dealer has a hard 15 80% of the time and a soft16 the other 20% of the time.
Also, I calculate that you should split A,A pair vs 8 + High. Of course, i could be wrong, but i spent a lot of time checking and double checking that one.
House Edge
So what is the ding-dong House Edge for this game? An alarming fraction of the posted information on this game was wrong (much of it originating in the apparently-botched math report) and so there is no particular reason to trust the posted values of House Edge. I know of three efforts by forum members (one of them me) to calculate the House Edge for Down Under BJ using my Vers. 2.1 strategy -the one which I've just posted. Obviously, the prevailing view is that this does not look like not an AP opportunity - in fact, my suspicion is that the HE is not as low as the 0.5% that the game owners claim - even with optimal strategy.
Let me make it clear that I am not criticizing Mike Shackleford - he did not do these calculations and this is not his dumpster fire.
Quote: gordonm888Let me make it clear that I am not criticizing Mike Shackleford - he did not do these calculations and this is not his dumpster fire.
Thanks, but I'll take some heat for this. I made some transcription errors from the BMM report, but I think that may be in error too. Let me at least post my own infinite-deck strategies. Stay tuned.
THANK YOU.
Here are my values:
| Player | EV Stand |
| 16 | -0.327764 |
| 17 | -0.146722 |
| 18 | 0.160033 |
| 19 | 0.394391 |
| 20 | 0.608178 |
| 21 | 0.813650 |
Color met flattered bobbartop =)... to be honest I was investigating this privately with another member for our own personal greedy reasons =). Was kinda hoping this thread just died at the time lol.Quote: bobbartop...I'll be offline for a couple days, I am looking forward to your findings and taking the knowledge up to Table Mountain.
Where is Romes in this thread? He's got a good brain, he should like this stuff.
Did Gordon say "double hard 8 v. Ten"?
Quote: Wizard...
Player EV Stand 16 -0.327764 17 -0.146722 18 0.160033 19 0.394391 20 0.608178 21 0.813650
I've only had a quick look and agree your 16 and 17 figure. It looks as if I have an error with my higher values (also I'm ignoring some 20s winning vs 22s at this stage) and get the following...
0.899671098
0.608178243
0.394390647
0.160033043
-0.146722236
-0.327764348
Quote: WizardGordon, we have differences all over the place. May I ask for your EV numbers for a player 16 to 21 standing against a dealer 2 with a small hole card?
Here are my values:
Player EV Stand 16 -0.327764 17 -0.146722 18 0.160033 19 0.394391 20 0.608178 21 0.813650
I appreciate the opportunity to compare.
My dealer probabilities for 17-22 are not infinite deck. They are based on 6 decks such that a dealer hand with multiple cards of the same rank has a lower probability than with an infinite deck. This may lead to small differences between our numbers. My numbers are:
| Player | EV-Stand |
| 16 | -0.3260876 |
| 17 | -0.145049 |
| 18 | 0.161052 |
| 19 | 0.394613 |
| 20 (not TT) | 0.6078688 |
| 20 -TT | 0.694298 |
| 21 | 0.8995276 |
The big difference appears to be in player 21. I wonder if you may have overlooked the rule that dealer 22 loses (not pushes) against player 21 and 20 (TT).
Also, you appear to have posted the same strategy for Small and Medium Hole Cards?
Quote: RomesColor met flattered bobbartop =)... to be honest I was investigating this privately with another member for our own personal greedy reasons =). Was kinda hoping this thread just died at the time lol.
That's funny, I did actually think to myself, "Gee, I hope I'm not messing up anyone's opportunity to beat the crowd on this". Sorry about that.
Correcting that, I now agree that when dealer has a Small Hole Card that player should HIT "3-3 vs 2" and "3-3 vs 3." I had previously recommended that player should split those hands.
However, correcting that error doesn't seem to flip any of the other decisions in my strategy charts.
First, yes, I didn't count that a 21 wins against a 22 in my Small Card strategy nor that table of EV's. In my defense, it looks like I did do it correctly for the other two strategies.
Second, I had a titling error, which caused the Small Card strategy to be given the file name for both the Small and Medium card strategies. Stupid.
For what little it is worth, I think I stand by my previous High Card strategy.
Let me make my due corrections. First those expected values for standing against a 2 and Small Card in the hole.
| Player | EV stand |
| 16 | -0.327764 |
| 17 | -0.146722 |
| 18 | 0.160033 |
| 19 | 0.394391 |
| 20* | 0.608178 |
| 10,10 | 0.694199 |
| 21 | 0.899671 |
* Any 20 other than two tens.
Here are my new basic strategies. You may need to close and reopen your browsers to clear the old ones from the cache because I overwrote the previous strategies.
So, meaning no disrespect, but I still think there are prima facie issues with your results.
For example, with a Medium Hole Card, a dealer 7 is a 13,14,15 or 16. "7+ Medium Hole Card" is the situation with the weakest dealer cards on the Medium Hole Card chart, yet your strategy does not reflect that. In fact, your strategy for "9 + Medium" is what I calculate for a "7 + Medium."
Going further, on the Medium Hole Card "Hard" chart, if the columns that are currently labeled 4-10 were instead labeled 2-8, I think we would be in agreement.
By the way, I agree with everything on your High Hole Card charts, except I recommend Split on A-A vs 7 and 8. I'll take a closer look at that.
Quote: gordonm888By the way, I agree with everything on your High Hole Card charts, except I recommend Split on A-A vs 7 and 8. I'll take a closer look at that.
Just basing it on what I remember from exact known HC strategy, you dont split AA vs 17, 18, 19 except AA vs 17 when RSA is allowed. I don't know if the Push 22 rule changes things though either.
As a mildly interested gambler, but very entertained spectator, I’ll just say that I’m enjoying watching from the sidelines.Quote: gordonm888God, I hate to be publicly disagreeing about strategy with the Wizard of Odds! It's like playing a televised tennis match against Roger Federer.
So, meaning no disrespect...
Disrespect? Nah. It’s just a meeting of the minds where the problem is causing a lot of head scratching and second gussing from both parties.
Quote: gordonm888God, I hate to be publicly disagreeing about strategy with the Wizard of Odds! It's like playing a televised tennis match against Roger Federer.
Trust me, I don't deserve such respect. This is far from the first time I've made a mistake. There are plenty of people who are on par or a little better than me at this. I just had good timing.
Quote:So, meaning no disrespect, but I still think there are prima facie issues with your results.
Trust me, none taken.
Quote:For example, with a Medium Hole Card, a dealer 7 is a 13,14,15 or 16. "7+ Medium Hole Card" is the situation with the weakest dealer cards on the Medium Hole Card chart, yet your strategy does not reflect that. In fact, your strategy for "9 + Medium" is what I calculate for a "7 + Medium."
Yeah, I had a bad formula. I am doing this in three separate spreadsheets and sometimes I think I'm in one but actually in another. That was the case here.
So, I updated my Medium Card strategy again. It should say 3:45 PM at the bottom.
Quote:By the way, I agree with everything on your High Hole Card charts, except I recommend Split on A-A vs 7 and 8. I'll take a closer look at that.
Here are my numbers on that:
AA vs. 7 & High hole card
Split: 0.246153846
Hit: 0.334146847
AA vs. 8 & High hole card
Split: -0.061538462
Hit: 0.134532007
Hit 10 vs 8+small (not Double)
Split 2,2 vs 3+small (not Hit)
So here is my status on the strategy (I call this Version 2.2). Again, dark shaded boxes represent areas where I differ from the posted WOO strategy.
Areas of disagreement with the Wizard's latest strategy:
Small Hole Card, Hard hands: None, 100% Agreement!
Small Hole Card, Soft hands: I recommend STAND on S18 vs 6,7 and A, Wizard says HIT
Small Hole Cards, Pairs: My chart says to SPLIT on 6,6 vs 3, Wizard says HIT. I've just checked that and Wizard is correct.
Medium Hole Card, Hard hands: None, 100% Agreement! (Was a new version of Wizard's chart edited in?)
Medium Hole Card, Soft hands: S18 vs 2,3 I say STAND and Wizard says HIT
S18 vs 5 I say DOUBLE, Wizard says STAND.
Medium Hole Cards, Pairs: I say split 2-2, 3-3 and 4-4 vs 10, Wiz says to HIT those. I will check that further. Also Wizard says to SPLIT T-T vs 7, I said STAND pending an interpretation of the rules.
High Hole Card, Hard and Soft Hands: None, 100% Agreement!
High Hole Card, PAIRS: I still calculate: SPLIT A-A vs 7,8 and Wizard says HIT. Damning evidence from Double Exposure indicates Wizard is probably correct.
So, lots more agreement, now.
Quote: gordonm888Areas of disagreement with the Wizard's latest strategy:
Small Hole Card, Soft hands: I recommend STAND on S18 vs 6,7 and A, Wizard says HIT
Soft 18 vs 6 & Small
Stand -0.19532474
Hit: -0.166058834
Soft 18 vs 6 & Small
Stand -0.221362893
Hit: -0.191135341
Soft 18 vs A & Small
Stand -0.103928497
Hit: -0.092409389
So, these are kind of close but too far apart for the number of decks to be the issue, I think. What do your numbers look like?
Quote:Medium Hole Card, Soft hands: S18 vs 2,3 I say STAND and Wizard says HIT
Soft 18 vs 2 & Medium
Stand: -0.19532474
Hit: -0.166058834
Soft 18 vs 3 & Medium
Stand: -0.221362893
Hit: -0.191135341
Quote:S18 vs 5 I say DOUBLE, Wizard says STAND.
Soft 18 vs 5 & Medium
Stand: 0.041361125
Hit: 0.01831607
Double: 0.036632141
Quote:
Medium Hole Cards, Pairs: I say split 2-2, 3-3 and 4-4 vs 10, Wiz says to HIT those. I will check that further.
I'll leave those alone for now.
Quote:Also Wizard says to SPLIT T-T vs 7, I said STAND pending an interpretation of the rules.
I did some fine tuning of that, to account for a maximum of three splits, and still say to split that.
T-T vs 7 & medium
Split: 0.785422737
Stand: 0.654949056
Quote:So, lots more agreement, now.
Again, I really appreciate your help.
Quote: Wizard
AA vs. 7 & High hole card
Split: 0.246153846
Hit: 0.334146847
AA vs. 8 & High hole card
Split: -0.061538462
Hit: 0.134532007
For: AA vs. 7 & High hole card This is a funny one, because the original posted strategy on WOO from the math report did have this as a split, which I was agreeing with!
I get:
SPLIT: 0.246153846 (Infinite deck) and 0.2604355 with the two Aces and one 7 removed from a 6 deck shoe (because there is only an 18.64% chance that the high hole card is an Ace, as opposed to 20% in infinite deck theory.)
HIT: I had an error here. My HIT S12 algorithm was botched, but since you cannot HIT any 2 starting cards and get a S12 I never worried about it. However, it bit me in the butt when evaluating this decision on a starting A-A. I have now corrected the error.
So I now get Hit: 0.334146847 (Infinite Deck) and Hit: 0.333895 (3 starting cards removed.)
So, I confirm your strategy charts on these two decisions.
***********************************
Updated strategy. The time should say 8:30 PM. If it doesn't, clear out your image cache.
And yes, I agree-never split tens. For "TT vs 7 &Medium" I get Stand = 0.74089 and Split = 0.73678, even assuming that TT on a split pair will win against a dealer 22. And, to the extent that there are approximations in those values, I believe they work in the direction of STAND.
For "TT v 6 & High", I get Stand = 0.74591 and Split = 0.73884. Another close decision.
The funny thing about splitting tens is that your EV will be surprisingly low unless and until you have re-split to four hands and become eligible to have a 20 (T-T). Here are the EVs I calculate for splitting and re-splitting a "T-T vs 7+Medium" as a function of whether you wound up playing 2, 3 or 4 hands.
2 Hands = 0.45894
3 Hands = 0.68842
4 Hands = 1.2015
As to the house edge, I'm at 0.77% based on infinite decks. I tend to subtract 0.1% for a six-deck game, but if you have a true six-deck figure, I'm all ears.
Quote: gordonm888I have calculated house edge to be 0.766%, in good agreement with your value. My calculation is a hybrid - its a cross between 6 decks and infinite deck.
Perhaps it was a coincidence we were so close. I found a small error and have adjusted my number to 0.6185%. Here is more of a breakdown.
| Event | Probability | ER | EV |
|---|---|---|---|
| Dealer small hole card | 0.307692 | 0.062772 | 0.019315 |
| Dealer medium hole card | 0.307692 | 0.075949 | 0.023369 |
| Dealer high hole card | 0.337278 | -0.011184 | -0.003772 |
| Dealer BJ | 0.047337 | -0.952663 | -0.045096 |
| Total | 1.000000 | -0.006185 |
Quote: bobbartopFor what it's worth, I interpreted a rules card from a table last night as saying that 10-10 made AFTER a split is pushed by a dealer 22. The 21 wins, though.
That is my understanding too.
Quote: bobbartopI am unclear on 11 v. 5 in Blue Mode. Double, or Hit? Thanks.
I get: Double 11 v. 5 in Blue Mode (2-5 hole card), which is in agreement with the Wizard's strategy.
(For Double 11 v. 5 in Blue Mode, I get an infinite-deck EV of 0.194. And for Hit, I get 0.159.)
Quote: ChesterDogI get: Double 11 v. 5 in Blue Mode (2-5 hole card), which is in agreement with the Wizard's strategy.
(For Double 11 v. 5 in Blue Mode, I get an infinite-deck EV of 0.194. And for Hit, I get 0.159.)
Thank you very much, ChesterDog.
Just to be clear, did we ever figure out what the true house edge is after all the changes this thread went through?
thanks
