Graduating to UBZ2 (from KO)

I recently decided to explore graduating from KO to Unbalanced Zen II.

I purchased and read the UBZII booklet ($20, 20 pages). I found it woefully lacking on its own, but having read KO Blackjack I understood it quite well.

For those interested in purchasing it, it is 13 pages about the system (the other 7 pages are stories and junk :/ ), it is a printed booklet, with sparsely filled pages and a strong bias towards DD games. The book covers SD, DD, 4D and 6D games, no 8D info (fine by me!).

There are several things about UBZII that are unclear to me and maybe someone else could help me out. I have a perfect understanding of KO, with a good understanding of how KO correlates to a TC via a fair understanding of hi/lo.

Concern #1:
UBZ2's pivot point is set at 0
so for 6D IRC -24, pivot 0. take insurance at +1 (RC = 25), some indicies at 0, most at +4 (RC = 28), bets increase modestly at 0. George C. says at 0 you have roughly a 1% edge (but what would the TC be?)

vs. KO
6D IRC -20, pivot +4 (same RC of +24 as UBZ2), though insurance is taken at +3 (RC = 23), with most indicies at +4 (RC = 24), with max bets at +4. (KO +4 = TC +4).

with KO you'd be betting a 2 unit bet at an RC increase of 16, with UBZ a 2 unit bet comes out at RC increase of 24. Based on the tags, both counts are unbalanced by 4 units per deck.

does this really Jive? Since UBZII is a level 2 count, does that explain the differences in both betting and playing deviations between it and KO?

Concern #2:
The bet ramp tables are really weak! How could one get a more precise or more felxible bet ramp for UBZII? Would you need something like CVCX?
what if you wanted a 1-4 DD spread, or a 1-12 6D spread? UBZII doesn't give details.

Concern #3:
are the sims and results in BJA (latest ed) run using UBZII exactly as written, with the same indicies?

Concern #4:
Does BALANCED Zen look like UBZII in the same way that Hi/Lo looks like KO? e.g. indicies, insurance, bet ramp, TC?

any thoughts would be of great help as I make this transition.

What are the "default" bet spreads for UBZII for DD and 6D?

A 1% advantage would seem to correlate to a hi-lo TC of approx +3, using the gross rule of thumb that each TC is .5%, and you need to overcome the house edge of .5%.

It seems strange to me that the point of first bet increase (RC +10) is so close to the point of insurance (+1), and yet also correlates to a 1% advantage. Then again, UBZII may be tweaked in such a way so that it is badass at detecting insurance opportunities.

i don't have the book right here, the spreads are really generic, and the DD spread is something like 1-6, the 6D 1-8 !? weird.

what George C says is that at a RC of 0 (for 6D this is +24) you have a 1% edge, minus the house 0.5% total edge = 0.5%. but it reads really fuzzy, it doesn't seemed backed up by anything

That would roughly correspond to a Hi-lo TC of +1 then. Which seems like a good part to increase betting at, but a slightly crazy point to take insurance at (again, not knowing the nuances of UBZII).

You've already demonstrated that you understand the two different mechanisms which link full rank unbalanced counts (KO) and half-rank unbalanced counts (Red 7, KISS) to the true count.

Actually, UBZ II works the same as a half-rank unbalanced count, in that its pivot is at +2 true. Just divide all the card tags in half to see this. Hence, if your IRC is "-24" with six decks and your pivot and key count are both "0", that's +2 true in cards-per-deck terms. With level two counts, since all the running count integers are doubled, the same has to be done with the betting ramp and index plays.

I have not seen George C's booklet, but it surely would be better to use a key count 22 points higher than the IRC. This, just like Red 7 and KISS, will always be a true count of somewhere between +1.3 and +1.8

It's true that when your R/C rises 24 points, you have an advantage of 1%, minus the initial 0.45% house edge, plus the index play gain at that point, roughly equaling +0.7% net.

I think your UBZII would run silky smooth if you used for six decks, an IRC of "18" and "40" for your Key Count. No negs to deal with. Insurance would be at "50", as well as doubling 8 vs. 5; doubling 9 vs. 7; standing 12 vs. 2; and doubling A/8 vs. 4. Standing 15 vs. 10 would be at "54", and standing 16 vs. 9 would be at "60". In short, you'd be taking all the KISS III numbers right off of page 155 and doubling them. Same with the betting ramp.

I know after looking this over, you'll be able to figure for yourself that R/Cs of "42" = +2.0 true (always)
..."48" = +3.0 true (on average)
..."50" = +3.3 true (on average)
..."54" = +4.0 true (on average)
..."60" = +5.0 true (on average)
etc.

Fred,

Thanks so much!

Everything became clear when you pointed out the correlation to KISS or 1/2 ranked counts. This post along with BJBB2 will be the ammuntion I need to refine the otherwise fuzzy numbers in the booklet - as well as the all important bet ramp! ScottH also pointed out to me that George C's definition of "pivot" might be mistaken for the key count, which appears to be the case, or rather that they are the same thing in 1/2 rank counts. either way, whatever you call the point at which you first have the edge is irrelevant to actually knowing when you have it, which as you pointed out would be at RC +22 (on average), or sooner earlier in the deck.

now time to get home and get some wongin/out points developed.

Thanks again for the jumpstart!

you're a legend.

The bottom line question is - is the added complexity over KO worth whatever additional advantage it creates?

the answer is yes, after going through the trouble of figuring out these details. the count itself is rather easy. a little harder than KO.
KO A through 10 -1,1,1,1,1,1,1,0,0,-1
UBZ A through 10 -1,1,2,2,2,2,1,0,0,-2

but the PE increases about 20-25%. I think the IC goes up as well. BE goes down 1% vs. KO.

PE vs BC

Hey Mim

Dont you mean Betting Correlation(BC)? If its a shoe game your playing, we know that BC is far more valuable than PE(Playing Efficiency)If all I am gaining is 20-25% on PE, and actually giving up 1% on my BC, Im not sure Its worth the trouble unless I am at least a green chipper with a good spread. But, I love your enthusiasm, and I wish you well. Keep us informed as you progress.

Depends on your Passion

The Count Sytem Performance Comparisons on pg 194 of BJBBII showed the EVs of UBZII to be +0.72% vs. +0.68% for Full Matrix KO. UBZII was tested using the exact format I suggested in my last post.

Is that enough to change? A personal call.

yeah actually it looks like UBZ loses 1% to KO in BC - the most important criteria for us. but overall it outperforms KO due to the other factors, PE and IC are two of them. but KO is something like 97% perfect with respect to BC, hilo, UBZII are like 96%. this % is really small from the theoretical max, but the EV gain for a level 2 count is significant, as mentioned by Renzey in his simulation/calculation.

This is why I use UBZ only on pitch games, where the spread is limited and you have more opportunities for index plays.

The set of risk-averse indices I use for the UBZ system tags outperforms the published one by about 20%. I'm getting ready to post them in this thread, hopefully I can format them in a readable way.

One thing I'm doing that George C. didn't was use an IRC. Theoretically it doesn't make a difference, but I think it makes the numbers more manageable.

UBZ indices- preliminary

In attached file.

If someone who has a simulator could plug them in versus the published index values for UBZII, that cross-check would be appreciated.

Thanks for all the replies.

I think i'm going to try to hack together a set of rounded indicies ala KO that are suitable for DD and 6D. I'm thinking of defining a deck dependent "Key count" a deck independent Pivot (TC = +2) and varrying the IRC.

i sadly lack the software, the hardware (have a Mac), and the knowhow to run sims or any sort of optimization at the moment.

I've made a TC-RC table for UBZII using Renzey's suggestion that the pivot is = TC +2 for UBZII. I'm not 100% this is the case as As,2s,7s are all equal to 1. I guess if you were to devide all the cards so that T,3,4,5,6 are = 1 and A,2,7 are 1/2 then you would have a pseudo 1/2 rank, or if you were to discount 2s or 7s and double the others value then only the As would be half ranked.
i.e. A-10: -1,0,2,2,2,2,2,0,0,-2 or -1,2,2,2,2,2,0,0,0,-2

i guess this would be equivalent to KISS, yes? no? help? :meeep:


So i'm not 100% positive UBZII is a perfect analog to KISS or Red7, i'm convinced it is very damn close (at the worst).
/end incoherence

I'm going to go through shortly some of the KISS, AM's SVUBZ, and the KO indicies to try and get a series of Neg indicies around the IRC, Keycount, and pivot indicies, along with insurance. That will serve 6D and DD well (hopefully with overlap for SD!)

at the very least i'd like to get some rounded index numbers for DD and 6D, even if they don't overlap with each other.

It might take me a few days/weeks, but if anyone is interested in helping me along the way (Like QFIT, e.g.) I would greatly appreciate it. :cat:

Here's a little work I did on some southern CA DD games.

On both games I used a very simple spread- it's a 1-8 spread where the bet equals the running count plus 5. So you will start spreading at RC= -3 and finish at RC= +3



Barona- DD, H17, DAS, LS, pen to 0.9 decks.

Win Rate= 2.48 units per 100 hands
IBA= 1.073%
SCORE= 44.43
N0= 22,509
Absolute RoR with 1000 unit bankroll= 2.77%


Pala- DD, H17, DAS, no LS, pen to 0.9 decks.

Win Rate= 2.05 units per 100 hands
IBA= 0.887%
SCORE= 29.09
N0= 34,382
Absolute RoR with 1000 unit bankroll= 5.85%

The game at Pala stinks. Frankly, at 0.9 pen the Barona game is pretty bad too, but you can see what an enormous difference just the LS rule makes.

I'd rather Wong a shoe with decent rules, but if you can find good pen in these DD games it would be worth playing.

Why in the hell don't you simply use ZEN for ALL GAMES? zg