Simply count red 2's and 7's as plus one and you have effectively turned hi-lo into a more efficient count system that is virtually the same as the RPC count,yet your count still remains a balanced one level system. Learn a few index changes, and you have now upgraded hi-lo into a powerhouse.
> Simply count red 2's and 7's as plus one and you have
> effectively turned hi-lo into a more efficient count
> system that is virtually the same as the RPC count,yet
> your count still remains a balanced one level system.
> Learn a few index changes, and you have now upgraded
> hi-lo into a powerhouse.
If you count red sevens instead of black twos you would have a worsening of the betting correlation because the EOR's of sevens are only 0.28 compared to 0.38 for the twos.
Francis Salmon
Yes that gets you close to RPC. The problem I have with such methods is ease of use. I'm not convinced it's any easier than RPC because cancellation becomes less obvious. When I see a Ten and a Seven, I automatically think -1. To me it's much more difficult to see it as -1 if the Seven is one color and -2 if it is another. It's worse with a Two and a Seven. I automatically see this as +2. But by using a Red 2-7 system, a Two and a Seven can be +2, +1 or 0.
> Simply count red 2's and 7's as plus one and you have
> effectively turned hi-lo into a more efficient count
> system that is virtually the same as the RPC count,yet
> your count still remains a balanced one level system.
> Learn a few index changes, and you have now upgraded
> hi-lo into a powerhouse.
> Simply count red 2's and 7's as plus one and you have
> effectively turned hi-lo into a more efficient count
> system that is virtually the same as the RPC count,yet
> your count still remains a balanced one level system.
> Learn a few index changes, and you have now upgraded
> hi-lo into a powerhouse.
NO! Counting red 2's and red 7's is NOT the same as counting 2's as 0.5 and 7's as 0.5. Let me show you why:
Example: 6D,S17,DAS
A = -1
R2 = 0.5
B2 = 0.5
3 = 1
4 = 1
5 = 1
6 = 1
R7 = 0.5
B7 = 0.5
8 = 0
9 = 0
T = -1
The above tags are known as the RPC system (level 2) which has the following parameters:
BC = 0.9804
IC = 0.7847
Now, what if we count the red 2's and red 7's as +1?
A = -1
R2 = 1
B2 = 0
3 = 1
4 = 1
5 = 1
6 = 1
R7 = 1
B7 = 0
8 = 0
9 = 0
T = -1
Here we have a level 1 system with some poor parameters:
BC = 0.9556
IC = 0.7647
Let's compare the above against Hi-Lo:
BC = 0.9648
IC = 0.7647
As you can see Hi-Lo does perform better.
The same analysis goes with Red 7 counting all 7's as 0.5 and Red 7 counting the red 7's as +1:
A = -1
2 = 1
3 = 1
4 = 1
5 = 1
6 = 1
R7 = 0.5
B7 = 0.5
8 = 0
9 = 0
T = -1
Level 2 system which has the following parameters (in TC mode):
BC = 0.9800
IC = 0.7852
Compared against:
A = -1
2 = 1
3 = 1
4 = 1
5 = 1
6 = 1
R7 = 1
B7 = 0
8 = 0
9 = 0
T = -1
Which is a level 1 system (in TC mode):
BC = 0.9683
IC = 0.7757
In RC mode we should expect lower values.
Hope this helps.
Sincerely,
Cac
Have you run sims between Red7 and Sevens as .5 to see if the differences are as expected by the BC difference? I did many years ago but frankly it was so long ago I've forgotten the results.
> Have you run sims between Red7 and Sevens as .5 to see
> if the differences are as expected by the BC
> difference? I did many years ago but frankly it was so
> long ago I've forgotten the results.
Yes, but by that time I didn't use a SCORE to make an apple-to-apple comparison and counting the sevens as 0.5 was far superior. BTW, the program was a mess after that modification
Anyway, I wouldn't doubt about the BC differences when the systems are compared in TC mode. The differences are probably narrowed when we use RC mode.
Sincerely,
Cac
I did my sims pre-SCORE. Have to look at doing it again. And making another mess
> Yes, but by that time I didn't use a SCORE to make an
> apple-to-apple comparison and counting the sevens as
> 0.5 was far superior. BTW, the program was a mess
> after that modification
> Anyway, I wouldn't doubt about the BC differences when
> the systems are compared in TC mode. The differences
> are probably narrowed when we use RC mode.
> Sincerely,
> Cac
I guess this ends the myth that counting red sevens as +1 is virtually the same as counting all the sevens as +0.5.
When calculating the betting correlations,I get slightly higher values than yours for all the systems you mentioned and they match the figures on bjmath.
Did you use different EORs than the ones mentioned by Griffin in ToBJ on page 44?
Francis Salmon
> Did you use different EORs than the ones mentioned by
> Griffin in ToBJ on page 44?
He undoubtedly used the superior ones in BJA3. :-)
Don
> I guess this ends the myth that counting red sevens as
> +1 is virtually the same as counting all the sevens as
> +0.5.
> When calculating the betting correlations,I get
> slightly higher values than yours for all the systems
> you mentioned and they match the figures on bjmath.
> Did you use different EORs than the ones mentioned by
> Griffin in ToBJ on page 44?
Yes, I use the ones in BJA3 which are far better. For S17,DAS you can use the following:
(A to T, m) -0.5794 0.3809 0.4339 0.5680 0.7274 0.4118 0.2823 -0.0033 -0.1731 -0.5121 0.1735
Take into account that they are Single-Deck EORs and the efficiencies calculated were for 6D.
Sincerely,
Cac
I guess that sooner or later I will have to buy that famous Schlesinger book!
Francis Salmon
> I guess that sooner or later I will have to buy that
> famous Schlesinger book!
What?! You don't have it yet?
Cac
Thanks for the depressing information.) I assumed (in the long run)the 2 and 7 would cancel out,thus giving them a value of .5,did you use hi-lo index numbers when running sims? As you know using the upgraded count will change some index numbers.
Thanks,
Brick
Posting Permissions
Bookmarks