Mike H: Unbalanced true count?

[Mike H]

Can anyone explain what is meant by an unbalanced count in true count mode? Brett Harris talks about it with his BRH systems but I assume it can be applied to any unbalanced count. Is it similar to dividing by remaining decks with a balanced system?

[Parker]

> Can anyone explain what is meant by an unbalanced
> count in true count mode? Brett Harris talks about it
> with his BRH systems but I assume it can be applied to
> any unbalanced count. Is it similar to dividing by
> remaining decks with a balanced system?

Essentially, yes. For example, let's take a look at KO. There are several ways to do it, but the simplest is to determine the IRC with the formula:

IRC=(# of decks) x -4

This standardizes the pivot point to 0 for any number of decks. So, for a six deck shoe the IRC would be -24.

Then we simply divide by the number of unseen decks, just as with a balanced count. The resulting "true count" is roughly equivalent to a Hi-lo TC -4. In other words, a KO TC of -4 would be about the same as Hi-lo TC 0, and a KO TC of 0 would have about the same advantage as Hi-lo +4.

Hi-lo indices could be used by subtracting 4 from them, although it would be better to sim a custom set.

True-counted KO is sometimes referred to as "TKO."

[Mike H]

So with Red7 the IRC is -2(#Decks) and a pivot point of 0 would be close to a Hi-lo TC of +2? I assume you would use the true count for bet sizing as well as the play of the hands? Basically if the running count isn't close to the pivot point then it's not giving you accurate information, especially early on in the shoe.

[Parker]

> So with Red7 the IRC is -2(#Decks) and a pivot point
> of 0 would be close to a Hi-lo TC of +2? I assume you
> would use the true count for bet sizing as well as the
> play of the hands? Basically if the running count
> isn't close to the pivot point then it's not giving
> you accurate information, especially early on in the
> shoe.

Correct on everything. The weakness of all unbalanced counts used in running count mode is that they tend to underestimate the edge early in the shoe and overestimate it late in the shoe.

[Fred Renzey]

> The weakness of all unbalanced
> counts used in running count mode is that they tend to
> underestimate the edge early in the shoe and
> overestimate it late in the shoe.

While this is true, there's less inaccuracy in the range where you're just beginning to accrue an advantage among systems whose pivot is +2 true -- such as Red 7, KISS or UBZII. That makes it easy and accurate to wong in, or begin ramping up your bets in running count mode. Those very same systems do produce a greater inaccuracy at high counts such as +4 or +5 true (where KO is quite accurate), but by then, you're usually already betting your max and making nearly all your index plays.

With the "+2 true" systems, it's quite easy to "true fudge" your RC at high counts to obtain virtual true count accuracy. You just need to know your fudging scale in advance.
For example, since the KISS index number for 6 deck Insurance is 25RC, you know you'll insure at "25" near the middle of the shoe, but at "26" if it's early and at "24" if it's late. It's been determined in advance that this will be pretty darned accurate, so there's no multiplying or dividing to be done.
You could easily "true fudge" in the same way with Red 7 or UBZII.

[Parker]

Hi Fred, always a pleasure to have you posting here.

> While this is true, there's less inaccuracy in the
> range where you're just beginning to accrue an
> advantage among systems whose pivot is +2 true -- such
> as Red 7, KISS or UBZII. That makes it easy and
> accurate to wong in, or begin ramping up your bets in
> running count mode. Those very same systems do produce
> a greater inaccuracy at high counts such as +4 or +5
> true (where KO is quite accurate), but by then, you're
> usually already betting your max and making nearly all
> your index plays.

All very true. Any count is going to be most accurate at its pivot point, where indices and bet ramping are unaffected by penetration.

Of course, some players might find it comforting to know that their count is at its most accurate at times when they are shoving max bets out on the table. :-)

[Mike H]

That's funny- I just bought your book yesterday out of the blue and here you are responding to my post. Anyone serious about this game should pick up a copy of Blackjack Bluebook II. I especially enjoyed the sections on hand interaction, "true fudging", camouflage, and the true count conversion chart. Thanks for an excellent book!

[AutomaticMonkey]

> Correct on everything. The weakness of all unbalanced
> counts used in running count mode is that they tend to
> underestimate the edge early in the shoe and
> overestimate it late in the shoe.

While balanced systems underestimate (slightly) your advantage late in the shoe due to the floating advantage. BRH-I I've used for quite a while and the unbalance is such that it does a reasonably good job of tracking the floating advantage.

But what I didn't like about using it in unbalanced mode is it takes away the power of the mighty 16 vs. 10 playing index, which is nowhere near the pivot point.

[Fred Renzey]

> What I didn't like about using it in unbalanced
> mode is it takes away the power of the mighty 16 vs.
> 10 playing index, which is nowhere near the pivot
> point.

With unbalanced counts, just become familiar with what a "normal" (neutral) running count would be at your current depth level -- then stand whenever the actual RC is above that. Also, hit 12 vs. 4 if the RC is below that.

[AutomaticMonkey]

> With unbalanced counts, just become familiar with what
> a "normal" (neutral) running count would be
> at your current depth level -- then stand whenever the
> actual RC is above that. Also, hit 12 vs. 4 if the RC
> is below that.

Well, sure. But if I'm going to pay any heed to the depth, I might as well just true count adjust the whole thing. It's the deck estimation I find more onerous than the division.

The thing I really like about running count mode systems is that if you walk in any time after the first hand you start counting and playing exactly as if it was the beginning, same IRC and all. That's nice in big, busy stores where you are looking for speed and volume, get your bets down any time and any place you can.

[fatcat519]

"For example, since the KISS index number for 6 deck Insurance is 25RC, you know you'll insure at "25" near the middle of the shoe, but at "26" if it's early and at "24" if it's late. It's been determined in advance that this will be pretty darned accurate, so there's no multiplying or dividing to be done.
You could easily "true fudge" in the same way with Red 7 or UBZII."

I've just been rereading Blackjack Bluebook II. When using using a +4 pivot point system such as KO, fudging is a little different from the above. With unbalanced counts, equivalent TC's rise with shoe depth on the high side of the pivot and fall with deck depth on the low side, whether you fudge up or down depends on which side of the pivot the RC index happens to be. For example, with KO the insurance index is below the pivot, so if you were going to fudge it, you would fudge down early and fudge up late in the shoe, just the opposite of the above. (But then, the KO insurance index is so close to the pivot that fudging probably won't get you much.)
As Fred R. suggests, it's easy to recognize a neutral count for each deck level and use it for 16vs10, 12vs 4, 13vs2. You can also learn KO Key Counts for the first few decks and use those for 9vs2, 11vsA, A8vs5,6 etc.
If any of this doesn't make sense please let me know.

[Mike H]

> The thing I really like about running count mode
> systems is that if you walk in any time after the
> first hand you start counting and playing exactly as
> if it was the beginning, same IRC and all. That's nice
> in big, busy stores where you are looking for speed
> and volume, get your bets down any time and any place
> you can.

I didn't realize you could do that with running counts. With true counts you can do the same thing but you imagine the cards that have already been played as being behind the cut card. So if half a deck sits in the discard tray when you start counting imagine that it's not there. Just add that half deck to your # Decks Remaining estimation. Picking up the count late like this decreases penetration so I wouldn't recommend it unless you see a bunch of small cards on the table.

[AutomaticMonkey]

> I didn't realize you could do that with running
> counts. With true counts you can do the same thing but
> you imagine the cards that have already been played as
> being behind the cut card. So if half a deck sits in
> the discard tray when you start counting imagine that
> it's not there. Just add that half deck to your #
> Decks Remaining estimation. Picking up the count late
> like this decreases penetration so I wouldn't
> recommend it unless you see a bunch of small cards on
> the table.

Right, and with using a RC-mode count you don't even have to imagine, the calibrating effect of an unbalanced count does it automatically for you. It's pretty obvious: let's say you are starting with an IRC of -32 4 decks into a 8D shoe cut down to 1.5 decks. The chances of you getting a count in the +EV zone in those next 2.5 decks are very low, just as if you were using a balanced count and had to imagine an extra decks behind the cut card.