KO Preferred vs. KO Full set

Adam N. Subtractum

Well-Known Member
The following is a re-post of an answer of mine to a question posted on AP.com, where it got no responses. Figured someone here may find it of interest.

>

Unfortunately, the KO book doesn't really provide a good comparison of the gains that can be made by using an extended set vs. the Preferred set.

In the book the respective expectations (unfortunately they do not include a risk-weighted comparo) for Preferred and Full are .62 and .63 (6D,DOA,S17,75%pen,1-8). This comparison shows an increase of a mere 1.61% to our expectation. Obviously, someone looking at these figures would deduce that the gains are probably not worth the effort. It should be noted though, that the "illeged" KO Full set is comprised of only 44 plays...not quite a huge mental feat to put to memory.

Now let's make a more fair comparison, using a more complete "Full" set, generated by some up-to-date, quality software. (The figures I am using for this particular case come from Dr. Org's "Understanding Knockout Blackjack" article, and were generated with the condition of risk-aversion.) This time, using around 80 indices and the same exact conditions as in the KO book, except for a 1-10 spread, we achieve expectations for Preferred and Full of .73 and .78, respectively. This shows us that we increase our expectation by 6.85% by using the Full set, over 4 times the increase from the Full set in the KO book!

This works out to about $1.75 extra on your $25/hour win rate...is it worth it? That's a relative question, and one only you can answer for yourself.

If your still not moved by these figures, an increase of over 10% could be attained (using the same amount of indices) with only a MINIMAL increase in difficulty...see my upcoming post at cardcounter.com, to be entitled "21st Century True-Counting".

ANS

ps: Note that the improvement would've been even more noticable if the 1-8 spread used in the first example was employed.

>

Note also that with a more optimal system and indices, such as those developed by T-Hopper, the effects will be even greater.

ANS
 

CanKen

Well-Known Member
1) You mention KO Full comprising 44 plays. My 1998 edition of the book, p.164,5 only shows about 22-25 depending on number of decks. What am I missing?

2) Where can I find Dr. Org's "Understanding Knockout Blackjack" article?

3) Ive been using KO (quite successfully) for about 2 years in the local 8-deck game, and have customized to suit myself. One thing I've done is develop my own method for indices near TC=0 for balanced systems. I assume that these counts would be roughly equal to a neutral count in KO. But being a long way from the pivot point, they depend greatly on the number of decks already played. So I've learned to estimate the number of decks played and from that and the running count determine whether the count is at, above, or below neutral. Then I play hands like 16vs10, 12vs4 etc. accordingly. Does this make sense?
 
Top