Thanks for clarifying. I haven't played SP21 in the US or Canada but I think that our rules are slightly more favourable for the player.
The NHC rule allowing players to hit their hand before the dealer takes a second card, possibly beating a dealer natural with a 21 and possibly also with bonuses is certainly beneficial. I've never encountered it in my area of Canada or anywhere in the U.S. that I've played.
Both NHC and HC Spanish are offered by legislation in Alberta. To my knowledge, only 1 SP21 table remains in the Province, and it is NHC.Originally Posted by Gronbog
https://aglc.ca/sites/aglc.ca/files/...Handbook_0.pdf
Spanish starts at section 10, about page 30 or 31.
In Australia, our casinos also allow back betting which if done correctly reduces an already low house edge even further, however, it requires a large bankroll and very few players use this strategy but it's good for comps.
Casino Enemy No.1
I would recommend trying to find "Secret Monkey Count" on this site. It isn't secret anymore, but was when it first came out (prior to the release of Katarina's book, as I recall). It was created by a former poster here who had "Monkey" in his name- hence the name of the count. It is better than Katarina's count (KatCount), which is unbalanced, and doesn't give the A the power it deserves in Spanish 21. The tags for Secret Monkey are simply KatCount, but using the A as -2 instead of -1, and starting with a running count of zero, as it is balanced. The index plays may even be around here somewhere..... Automatic Monkey (the creator) had mentioned that he doesn't mind if it is given out (he uses a far more advanced counting system, and hasn't used SMC in years), and while there is so much further to go, that is definitely a big step beyond Katarina's count.
I know five people (T3 is one of them), all of whom used to post online often (I don't believe any currently do) that switched to counting Spanish 21 as their primary game- but all five of them use the S17 game, and not the H17 redoubling game. All of them use counts more advanced than Secret Monkey (which itself is more advanced than KatCount, as mentioned). When I go to casinos (not my full time job- I do have a day one, in the STEM area), I count the S17 game as well. I use a balanced playing count, and add a balanced level-3 side-count of aces (the ace tagged as -3, to reflect that aces are far more valuable for betting purposes than ten-valued cards, and in S17 games, are around three times as valuable), but it isn't one I can share- I worked on getting it, along with the appropriate index numbers, with other players, and can't give the information out. T3 was the one who used to post the most about this game- I do miss reading his input.
For more information about counting Spanish 21 in general, you can also do searches for threads that have "Spanish" in them, confirm they are related to SP21, and read through them. One can also try looking at threads that I've posted in- it seems like I have under 90 posts since I created my account in 2012, and almost all of them, and the threads they are in, likely relate to Spanish 21, which is the only game I've been counting for many many years.
Hey All Clear I read your posts on SP21 and wanted to pick your brain real quick.
I have been analyzing the H17 redouble variant with Cvdata. We have some things in common in that I had also been playing BJ halves but SP21 has peaked my interest since I'm in the pacific northwest where the game is available.
I wanted to apply a level 1 playing count with a level 2 BASC, ace tagged as { -2 }. The ace EOR is closer to 1.5x the 10 in H17 redouble. Youve done the same albeit with a -3 for the ace since you play S17. I used the MGP analyzer to generate the EORs and checked them through simulation with cvdata. I had to compromise a little with the playing count tags for the 2 { 0 } and 9 { -1 } to balance the count. 2 thru ace .. { 0, 1, 1, 1, 1, 0, 0, -1, -1, 0 }
If you dont mind could you share the playing tags you use, or if they are the same? Cvdata index generation does not support redoubling so it takes a bit of finess to find those appropriate indices. Also the 678 bonus indices are pretty finicky too.
As for a BASC this is new to me but wondering what card rank do you choose to balance the ace and if it matters? I had chosen the 7 in practice since it was already neutral. I assume you then count the 7 and ace as { 2, -2 }.
Do you use this ace density info to change any of your index plays? I.e. 10 double vs 9. I suppose even the 7 density could be used to alter 678 hit decisions.
Do you use indices at the 4, 5 and 6 card levels? Kat gave rough approximations of what these should be and cvdata cannot generate this directly. Though I could through trial and error I wonder how much there is to gain.
Again if you dont mind sharing, how much of the house edge can you reduce just through index play alone? Three had mentioned eliminating the HE all together though I imagine his system was more complex.
I have Kats book and scoured BJTF for all the Three Zenmaster and Firewalker threads and learned a lot.
Thanks!
Last edited by Super Natural; 12-09-2020 at 10:13 PM.
Hi Super Natural,
I know someone (met him on this site- I think he still posts on occasion) who simmed a S17 SP21 level two playing count with a level five side count, and we got a house edge of .0971% via flat betting on the S17 game, when the Super Bonus pays 200-1 (so, when betting $5 or $25), or even something slightly lower, if playing at a full table with the envy bonus. That is something that dealer errors can easily overcome, and counting with even a small spread can beat quite easily.
I personally play with a level one playing count (all cards are either -1, 0, or +1), and a level three betting side count (aces are -3, three other cards are +1). Certain cards matter more in terms of deciding whether to hit/stand than others. For example, if you count the 2 and the 5 as the same +1, that is not only bad for betting purposes (per the EOR for each card), it is also bad when determining whether to hit or stand on, say, a 14 v 4 (while this is a very clear example, the same applies in most hit/stand situations, in which lots of 4 - 7 would lead to more hitting, and 2s don't have the same impact). If there are a lot of fives (and fours, and sixes, and sevens), you would want to hit. A surplus of twos won't have the same effect, and you'd be more likely to stand. One makes a decision of how much to bet, via how high the combination of the two counts are, just as in blackjack. But when making playing decisions, most of the index numbers don't take the ace into account, and instead use a ratio of tens (and sometimes nines as well) to "low" cards that affect your playing decisions, which are generally in the 3 - 7 range. There are S17 index numbers for which I use the (combined) count I use to bet, such as for doubling on a 9 or 10. In such cases, an ace should be treated as a "high" card for playing decisions. But since I am valuing the ace as 3x a picture, in general I need to remove that for index numbers regarding playing decisions.
Regarding a balanced ace side-count, that was always my only option. It became too hard to approximate quarters of 48 card decks, to adjust for ace surplus and deficiency. For players that do play blackjack on occasion, it would be even tougher to switch back and forth between 48 and 52 card decks. A balanced ace side count solves that issue.
For your case, the best thing would really be to find someone who can sim five or six billion rounds in the H17 game, and can give you a house edge under any conditions, along with index numbers for all reasonable situations. But from looking at the EORs in the H17 game (what I have doesn't include redoubling), a reasonable balanced high-level count to use for betting might be (A-T) -3,2,2,2,3,2,0,-1,-1,-2. And while I don't know what a good playing count would be in the H17 game, I get the feeling the only counting system you have index numbers for is KatCount. And if that is the case, I'd reccomend like (A-T) 0,0,1,1,1,1,0,0,-1,-1, so you can use similar index numbers. Counting the nine as a "high" card would screw up any index numbers for hitting/standing on a 12, but those don't come into play in Spanish (or almost never come into play). And counting a nine as a high card also screws up insurance correlation, but we don't take insurance in Spanish 21. (All the better playing systems in SP21 count the nine as a high card.) If that is the case, the betting side count would be the difference between the two, so that when combining them, you have a good combined count for betting purposes, which are close to the EORs. Of course, in the redouble game, the 2 might be a really important card, so maybe it should be valued higher. I dunno.
Yep, I also used Halves in Blackjack. And in the very rare times I play blackjack, I still do. I did enough examination, and the benefits of a side-count in regular blackjack just weren't there to be found. But in S17 Spanish 21, it really does help. I'm not sure enough about H17 Spanish to comment, though.
T3 uses four, five, and six card adjustments for every (reasonable) situation! But most of us, myself included, just use an adjustment. I adjust hitting/standing by +2 with four cards, and by +4 with a five card hand. I would think the adjustment Katarina makes would work for those cases.
I can't give out index numbers that I use- which were obtained via working with many others from this site, none of whom seem to post much anymore. But because they are for the S17 game, they wouldn't be as much help to you anyway, unfortunately.
Right! I found the traditional ace side count training described in Million Dollar Blackjack to be very good, but I felt like the difficulty factor was too high for 6/8 decks. A BASC feels much easier.
How would you compare the difficulty between Halves and a level 1 playing count + BASC?
I wanted to apply a BASC to H17 because I thought if I do attempt to play S17 it will be a simpler transition.
Thanks for the detailed reply, much appreciated.
Last edited by Super Natural; 12-19-2020 at 05:08 PM.
I have been studying the H17 redouble variant and found a few surprises through simulation.
Basic strategy for 22 and 33 are split 2-8. In CVData these defensive splits actually lose less when you hit instead of split.
22 v 2
33 v 2
22 v 8
EV from the MGP BJ CA confirms the WoO basic strategy..
22 v 2 Split: -0.0770 Hit: -0.0912
33 v 2 Split: -0.1350 Hit: -0.1432
22 v 8 Split: -0.0687 Hit: -0.0754
CVData sim results pictured below ..
22 v 2 Split: -0.1097 Hit: -0.0874
33 v 2 Split: -0.1665 Hit: -0.1433
22 v 8 Split: -0.0946 Hit: -0.0740
Any idea why the discrepancy?
Thanks!
Last edited by Super Natural; 12-19-2020 at 05:16 PM.
Keeping a second count was quite tough for me, and took practice. And since I don't go to casinos too often, there was some wondering whether it was worth it. I have prior posts on this forum in which I ask for advice on keeping a side count. I ultimately settled on using letters for the side count, but that means (for me) being able to count backwards in the alphabet, by threes (e.g. Z --> W --> T --> Q, etc.). Not intuitive, and not easy either. Halves (a single level three count) wasn't hard for me at all, in contrast. T3 uses a count where certain cards are positive for the playing count, but negative for the side count (so that they sum to zero for betting purposes, which is the goal). Automatic Monkey uses a count even more advanced than that. And since neither one posts anymore, I can call them crazy, without getting any push-back.
Posting Permissions
-*-
Reply With Quote
Bookmarks