How LottoDesigner works
Nick Koutras
I have post this response to RGL and to this Yahoo eGroup
You can find LottoDesigner eGroup at:
http://groups.yahoo.com/group/LottoDesigner
and LottoDesigner web site at:
http://LottoDesigner.5u.com/
First "the concept behind LottoDesigner is to cover as more sets
as possible".
You found this statement false.
I'll prove you wrong below, using all the designs you indicated
except the C12,4,3,6=7 which I was not able to find or create with 7
sets.
But you can do the test yourself after you understand this post.
Back to the concept of LottoDesigner.
I use the live example of your first design below C(15,5,3,6)=8
The THING (LottoDesigner) generated with just one pass the following
design.
1 2 4 8 14
1 5 12 14 15
1 6 7 10 13
2 3 6 11 14
2 8 10 12 15
3 5 8 9 13
4 5 10 11 12
4 7 9 11 15
and indicated a UWR of 76.295
Using Ininuga or Covermaster we can arrive to a design as below
1 4 7 11 14
1 5 6 11 14
2 3 8 9 12
2 8 12 13 15
3 4 10 13 15
3 5 6 7 9
4 5 6 7 10
4 9 10 13 15
If you can import this design to LottoDesigner you will get
a UWR of 152.570
At this point we have two designs which, one covers the parameter 3if6,
the one by Ininuga or Covermaster and one, from LottoDesigner, that
does not cover the parameter 3if6 in 8 sets.
Now comes the moment of truth about which design is better for
the Lotto Player to use.
Do the following test.
1. Strip Ininuga's design from the top text
2. Load the design to Covermaster
3. Use Tools/Report/Quick and get the report
4. Sum up all the Uncovered sets
5. The result using the Ininuga's design given above, that covers, shall
be 13903
total uncovered sets.
This number, means to the Lotto Player that, if he uses this design
13903 sets
of possible winning combinations will not be included with this
design
Here is the report: I've done the arithmetic for you.
T if M Tested Covered % Uncovered %
-----------------------------------------------------------------
45 2 On 2 : 105 60 57.14286 45 42.85714
9 2 On 3 : 455 446 98.02198 9 1.97802
- 2 On 4 : 1,365 1,365 100.00000 0 0.00000
- 2 On 5 : 3,003 3,003 100.00000 0 0.00000
- 2 On 6 : 5,005 5,005 100.00000 0 0.00000
382 3 On 3 : 455 73 16.04396 382 83.95604
705 3 On 4 : 1,365 660 48.35165 705 51.64835
492 3 On 5 : 3,003 2,511 83.61638 492 16.38362
- 3 On 6 : 5,005 5,005 100.00000 0 0.00000
1,326 4 On 4 : 1,365 39 2.85714 1,326
97.14286
2,622 4 On 5 : 3,003 381 12.68731 2,622
87.31269
3,396 4 On 6 : 5,005 1,609 32.14785 3,396
67.85215
5 On 5 :
4,926 5 On 6 : 5,005 79 1.57842 4,926
98.42158
13903= Uncovered sets
Now do the same test for the design that the THING (LottoDesigner)
generated.
Here is the report:
T if M Tested Covered % Uncovered %
-----------------------------------------------------------------
32 2 On 2 : 105 73 69.52381 32 30.47619
8 2 On 3 : 455 447 98.24176 8 1.75824
- 2 On 4 : 1,365 1,365 100.00000 0 0.00000
- 2 On 5 : 3,003 3,003 100.00000 0 0.00000
- 2 On 6 : 5,005 5,005 100.00000 0 0.00000
375 3 On 3 : 455 80 17.58242 375 82.41758
587 3 On 4 : 1,365 778 56.99634 587 43.00366
300 3 On 5 : 3,003 2,703 90.00999 300 9.99001
29 3 On 6 : 5,005 4,976 99.42058 29 0.57942
1,325 4 On 4 : 1,365 40 2.93040 1,325
97.06960
2,595 4 On 5 : 3,003 408 13.58641 2,595
86.41359
3,188 4 On 6 : 5,005 1,817 36.30370 3,188
63.69630
5 On 5 :
4,925 5 On 6 : 5,005 80 1.59840 4,925
98.40160
13364 = Uncovered sets
The design by LottoDesigner leaves Uncovered 13364 sets vs 13903 of
Ininuga
Which design is better to use day in day out?
...the user shall consider this concept very carefully.
LottoDesigner's Design covers 539 more possible winning combinations
than Ininuga's
or about 4% better.
-- LottoDesigner's Designs include the T=2 into the UWR for better
diversification
Below is a complete table of comparison between LottoDesigner's designs
and the one's you indicated and cover.
Design UWR Uncovered Ininuga's UWR Uncovered Difference % Better
LD(10,5,4,6)=7 101.539 -495 126.676 -527 32 6.5%
LD(11,5,4,7)=5 39.371 -1358 63.481 -1440 82 6.0%
LD(12,4,3,6)=7 42.155
LD(13,6,4,7)=5 49.394 -3524 93.539 -3932 408 11.6%
LD(15,5,3,6)=8 76.295 -13364 152.57 -13903 539 4.0%
On all cases, except the one you will check, LottoDesigner
produces better designs between 4% and 11.5%.
If this is not significant for the LottoPlayer then what is? To cover?
The bottom line, in my opinion, is as follows:
LottoDesigner is generating better designs for the Lotto Player since
these designs
do cover more sets.
I think that my statement:
"-- This program is implementing a different concept for construction
of wheels and the user shall consider this concept very carefully.
"Your design shall be able to cover as more sets as possible""
is correct!
Yes, the UWR is an excellent concept.
I'll not claiming that this Single Number for quality of a Design is
perfect
and that applies to all designs.
I'm the first to tell you that when B<=V it does not always gives the
best
answer. But I'm working to improve my equations to cover this problem
too.
Please note that UWR compares designs with equal B.
My statement for UWR was:
"-- The program does not try to cover the single parameter T if M,
but tries to find the biggest sum of all T's from 2 to M that are
covered.
This number is represented by the UWR (Universal Wheel Rating) "
This is the concept of LottoDesigner and its UWR number.
This program (LottoDesigner v1.02 B 02) is a PreRelease version
The release 1 will be a COM add-in with a supplementary User Interface
Sheet
and the following, release 2 will be an EXE program and will include
parameters of Unlimited B's.
The "rgd" equation is not a "blop".
If you learn to utilize the L parameter you may understand it's value.
Why you are so occupied with "cover"?
"To cover" is a waste of sets!
If you want to generate a design, your target of B's shall be always the
number
indicated by the "rgd".
Also the "Check Results" button under Sx indicates what the design
"wins" vs
a random test or the one you enter, and the summary of "Wins" is given
in Cells T5...Tn
But all the above is insignificant.
The concept of constraints that LottoDesigner is utilizing IS the
most important concept behind this program.
But this subject is still virgin in our minds
This is the area you should look at.
So prepare yourself for some significant Lotto Construction Architecture
which is finally here.
Happy Wheeling,
Nick Koutras
PS: My Front Page 2002 has screwed up my web site.
I'll hope that I'll fix that soon...
Orininal message below
=====================================
Hi Nick,
as promised I'm back with first results.They are pretty poor.
All I done was that I compared some Ininuga wheels with the THING
wheels.Some wheel that I could check with Covermaster .
The results
Ininuga : c15,5,3,6,8 was satisfied in 99,46%
c12,4,3,6,7 "" 97,97%
c10,5,4,6,7 "" 98%
c11,5,4,7,5 "" 97%
c13,6,4,7,5 "" 98%
I didn't try to figure out what will be the best ticket for the
perfect wheels just because I think that if the THING is not able to
hit these small resultats (on small wheels) whats the use of doing
conditions(constructions) and the UWR.
My oppinion is that the idea UWR is great but it has to be considered
that a less ticket wheel has a smaller UWR than the other bigger one.
The main criteria on caltulating the UWR factor should not be done on
comparing the the quantity of the tickets.It has to be something that
consider the V,K, and T ( again not the M ).You can't convince me
that a wheel that has 25 tickets more (and would cost me 25 something
more)(and even dosn't satisfie the main wheel criteria (V,K,T))than
the wheel that has bigger UWR is better.You know why, because with
that extra 25 tickets I could,should cover 2-4 more numbers;and if I
has a chance to cover even onenumber more I already covered more
sets.It's a simple logic that the wheel with one number more has to
and is covering more sets the the wheel with less numbers.
And thats the reason why
"-- This program is implementing a different concept for construction
of wheels and the user shall consider this concept very carefully.
"Your design shall be able to cover as more sets as possible""
is a nonsence.
The idea is great, but you have to change the algorithm that sort the
wheels.
The other thing that really makes me a big problem is, that RGD
button.
If you try to do a wheel 29,5,3,7,81 the RQD is 29(??????) in theory
yes,maybe...but with Iningua you need exactly 81 tickets and not
29.So the theory for RGD is a big blop as everybody knows that.
The third thing in the THING thats missing is a check the wheel.If
you don't have that and you don't know what is the perfect RQD for
wheels with constructions,or what is (could be) the perfect UWR(just
too satisfie you)how the hell could you know what kind of wheel did
the THING optimized.
Gal
FakeLake
FakeLake.
Paracelsus
FakeLake <nos...@me.com> wrote in message
news:9hsicq$sdp$1...@uranium.btinternet.com...
> English isn't your mother tongue is it?
>
That's maths graduates for ya.
Gerry
>
> English isn't your mother tongue is it?
>
> FakeLake.
Nice job!!!
You just insulted someone who knows more about
lottery combinations than you ever will ;-)
FakeLake
> You just insulted someone who knows more about
> lottery combinations than you ever will ;-)
Oooh! Please "Gerry", stop bothering me. This person may know everything
there is to kow in the entire universe, but unless he/she knows a language
in which to explain it to everyone else, the information is useless.
Apart from that, learn to recognise sarcasm when it smacks you in the
face....
PEACE <----------- NOTE THIS WORD!!!!!!
FakeLake.