Norm Wattenberger: REKO - Another simplified KO strategy

[Norm Wattenberger]

Since most play these days is with shoes, I thought I?d create a variation of KO specifically for six decks. For lack of a better name I?ll refer to it as REKO (Ridiculously Easy KO.) I believe this strategy requires substantially less time to learn than KO Preferred. First there are 16 indexes (incl. Surrender) instead of 19. Not much difference here. But, more importantly, KO Preferred has three different index values (+4, -4 and +3.) REKO has only one index value: +2. This is more of a difference. With KO you need to memorize the index values assigned to the 19 indexes. With REKO, you need to know what the 16 index plays are. But, you do not need to memorize the index values. They are all +2, including Insurance. A Flash Card drill would have 16 cards with +2 on every cardJ

The indexes are:

Insurance
16vT
16v9
15vT
12v2
12v3
11vA
11vT
10vA
9v2
9v7
8v5
8v6
Surrender:
15v9
15vT
15vA

As I said the indexes are always +2. However there is one change. The initial running count is ?20 for no surrender and ?19 for surrender.

Results

I compared REKO with KO for four situations. In all cases, there are four players, three to your right, S17 and DAS and a 1:16 optimal spread. Following are the SCORES:

	KO	REKO	% of KO
NSr 4.5/6	23.5	23.0	98%
NSr 5.0/6	38.2	37.7	99%
LS 4.5/6	35.1	35.4	101%
LS 5.0/6	55.0	54.0	98%

REKO results are 98% to 101% of the results of KO Preferred for Surrender and No Surrender at the simmed spread and penetrations. I also ran the same four sims for HiLo with optimal betting by True Count with half-deck resolution and 24 indexes (The Catch 20 and Fab 4.) REKO with 16 indexes all set to +2 had SCORES of 93% to 96% of HiLo with 24 indexes.


CVCX Online

[Norm Wattenberger]

Most HiLo folk these days seem to use the Sweet 16 indexes. That is, the Illustrious 18 without the ten splits. Also most people estimate remaining cards by full decks. So I simmed this along with the Fab 4 Surrender indexes.

	HiLo	REKO	% of HiLo
NSr 4.5/6	23.1	23.0	99.6%
NSr 5.0/6	37.3	37.7	101.1%
LS 4.5/6	35.1	35.4	100.9%
LS 5.0/6	54.5	54.0	99.1%

REKO with 16 indexes all set to +2 using running count performed almost exactly the same as HiLo, True counted, with 20 indexes from ProBJ. I created a chart for all penetrations, but didn't bother posting it since the lines were all on top of each other As always, the sims are designed to match strategies as actually played. Comparing a simple strategy to a crippled version of another strategy that no one would use does not make for an honest comparison.

[Don Schlesinger]

> 11vT
> 10vA

Did you mean the top one to be TvT?

Don

[Norm Wattenberger]

[Cacarulo]

Hi Norm,

Very interesting study!
1) How many rounds were simulated?
2) Could you add an additional sim that uses a CBS strategy + INS? The plays you can use for the CBS could be 16vT=S, 12v3=S and maybe a couple of soft doublings.

Thanks in advance.
Cac

[Norm Wattenberger]

> 1) How many rounds were simulated?

One billion in each sim.

> 2) Could you add an additional sim that uses a CBS
> strategy + INS? The plays you can use for the CBS
> could be 16vT=S, 12v3=S and maybe a couple of soft
> doublings.

Good idea.

[Cacarulo]

> One billion in each sim.

Hmmm, I guess there's too much SE yet. I think 5 billion would be more reliable.

> Good idea.

Great!

Sincerely,
Cac

[MJ]

[Norm Wattenberger]

[MJ]

Interesting study Norm!!

Howcome I don't see REKO on CVCX online? Please add it if it is not there already. I wonder how it compares to Arnold's Red 7.

You ought to contact Ken Fuchs and Olaf Vancura...I'm sure they would be glad to know that you had taken there system to an easier (and just as powerful level).

What if you had used Hi-Lo with 1/2 deck resolution in your comparison? Sorry but most people do not estimate the discard rack with full deck resolution. I try and estimate it to the nearest 1/4 deck! A player on a well known team once said "The TC is more sensitive the closer you get to the end of the shoe". For example, if there are 4.25 decks in the discard rack and the RC = +10, then the TC = +5.7.
But what if you try and round the discard to the nearest 1/2 deck? If you cautiously round the discard to 4 decks then the TC = +5.

If you are aggressive and round the discard to 4.5 decks, then the TC = +6.7. Now look at the difference 1/4 deck makes in estimating the discard between 4.25 decks and 4.5 decks. A difference of an entire unit (6.7-5.7)!!! That is why I am surprised by the chart you posted a few months ago comparing full, half, quarter, and exact card resolution at various levels of penetration. All the lines were right on top of each other. It defies logic!! As my example clearly demonstrates, the further you get into the shoe the more important accuracy in discard estimation becomes!!

MJ

[MJ]

[Norm Wattenberger]

For increased remaining card estimates to be of full value, you need to generate the indexes using the same accuracy.

[Don Schlesinger]

You can always concoct examples that show that, under certain contrived circumstances, one approach is much better than the other. In fact, it isn't.

In your examples, you would have had your max bet out no matter which resolution you used, and it would have made little or no difference.

Quarter-deck resolution in the shoe game is, more or less, overkill and quite useless. Half-deck might slightly outperform full-deck. Norm and I did some studies on this recently. I think half-deck won 1-2% more, tops.

Don