Development - Wheel Generator 1.9

Any latest news will be posted here...
User avatar
lottoarchitect
Site Admin
Posts: 1635
Joined: Tue Jan 15, 2008 5:03 pm
Location: Greece
Contact:

Development - Wheel Generator 1.9

Post by lottoarchitect » Thu Nov 19, 2009 11:26 am

Wheel Generator 1.9 development has started. As usual, feel free to make your suggestions on how to improve further Wheel Generator to your benefit.

WG 1.9 development so far targets to improve the performance of the optimization. Indeed we have an increase of a minimum 20% in overall computation speed and a much better focus when developing FD coverings, which translates as better FD coverings produced much quicker! This also applies to imported mother coverings, since these are FD coverings as well. However, there is a limitation to this new focused approach when we deal with FDs having millions of blocks to pick from and therefore FD construction reverts back to WG 1.8 approach in that case.
As an example of this performance enhancement on FDs, you probably remember the covering 16,6,5,6 = 138 FD 100% during the development of WG 1.8 at viewtopic.php?f=2&t=119. The new FD engine has managed this in 134 blocks (the covering is available at the constructions by Wheel Generator for the licensed users).
More things to come include :
  • a tickets checker with database history of what a player has played, the actual winning numbers, hits produced and cost of playing/earnings.
  • auto-save covering on improvement.
  • Keep a log on improvements as requested.
  • Filter : sum of last digit as requested.
  • The ability to individually enable/disable the active status of filters which contain sub-filters. For example, the number groups have a global active status however someone might wishes to temporarily disable a particular group among all defined number groups.
  • The ability to rearrange the order of the number groups filters along with the inclusion of a comment so to allow a hint on what each group is about.
  • A new covering improvement mode that tries to eliminate as much as possible the lowest winning divisions of the covering! This translates as we'll fall more often to higher winning divisions = more earnings besides the improvements we can expect because of the formation of our blocks due to the filters in normal mode! I suspect this new feature will be a topnotcher and will boost WG to a new dimension. This mode can work alone (i.e. for those not wishing to use filters and prefer unrestricted constructions) or in conjunction with the filters so to boost as much as possible the winning outcome. You can understand visually this feature when you observe the hits analysis of WG. This feature tries to eliminate the lines that offer the lowest payout (which are displayed at the bottom of the hits analysis) and transpose them to higher winning divisions.
cheers
lottoarchitect

User avatar
lottoarchitect
Site Admin
Posts: 1635
Joined: Tue Jan 15, 2008 5:03 pm
Location: Greece
Contact:

Re: Development - Wheel Generator 1.9

Post by lottoarchitect » Sat Nov 21, 2009 1:24 pm

You may also recall the test case construction back in WG 1.7 development 18,6,4,6 = 38 FD 100% which can be found at viewtopic.php?f=2&t=84. The new FD approach has managed further improvements:

My 18 selected numbers are: 3,6,7,8,12,15,19,23,24,25,26,28,30,34,35,39,43,46
Filters: Sums accept 80-160, Odd/even accept 2-4 and a number group accept 2-3 where the set is: 6,7,8,24,25,26,28,30 which now all set to be guides for the covering. The produced outcome in the WG 1.9 process via the new FD approach is:

c(18,6,4,6) = 37 FD 100% (missing 1044/18564 unconditional 4if6 = 94.37%)

03 06 12 30 35 39
03 06 15 24 28 34
03 06 19 25 26 46
03 06 23 28 34 43
03 07 12 25 28 46
03 07 19 30 34 39
03 07 23 24 35 46
03 08 12 30 43 46
03 08 15 26 39 46
03 08 19 23 30 39
03 08 19 28 35 43
03 08 24 25 34 43
03 15 25 26 34 35
03 15 26 28 30 43
03 19 24 26 28 39
06 07 08 35 39 46
06 07 15 23 30 46
06 07 19 26 34 43
06 08 15 25 43 46
06 12 19 24 35 43
06 12 23 25 26 39
06 12 25 30 34 39
06 15 19 24 28 39
06 19 23 25 30 35
07 08 12 30 34 43
07 12 15 24 39 43
07 12 15 26 28 35
07 12 19 23 28 46
07 12 23 26 30 34
07 19 23 24 34 46
08 12 19 23 26 43
08 12 19 25 34 35
08 12 23 24 28 39
08 15 23 24 26 35
08 15 23 25 28 34
08 15 28 30 34 35
12 15 19 24 25 30

cheers
lottoarchitect

relowe
Advanced
Advanced
Posts: 63
Joined: Tue Feb 05, 2008 12:17 am
Location: Melbourne

Re: Development - Wheel Generator 1.9

Post by relowe » Sat Nov 21, 2009 10:51 pm

Hi LA,

The list of improvements and additions you propose for WG 1.9 are very impressive - this puts some very bright polish onto this already great tool.

The database history of what the player has played and presumably also including a database history of the lottery drawn data is very desirable. I know I for one have wondered about my previous wheels and whether I would have been better keeping on playing them or not - in fact I do keep a wheel long term - but you give them away eventually for another wheel after a while because you think you can do better. Being able to check back will allow one to review this theory of - is it better to keep wheels or use fresh ones?

And the ability to remove low yielding blocks which presumably means that better yielding blocks remain sounds sensational.

And could I suggest multiple Additional Coverages also for this release?

Looking forward to Ver 1.9

relowe

azevrec
Regular
Regular
Posts: 31
Joined: Fri Mar 27, 2009 11:43 pm

Re: Development - Wheel Generator 1.9

Post by azevrec » Sun Nov 22, 2009 12:12 pm

Congratulations for the new version.The new incorporations seem to be very interesting.

Certainly, do not forget your program GAT. we wait it very anxious.

Cheers

azevrec
Regular
Regular
Posts: 31
Joined: Fri Mar 27, 2009 11:43 pm

Re: Development - Wheel Generator 1.9

Post by azevrec » Wed Dec 09, 2009 12:50 am

Have you thought about this filters?:

1-Median- Its the Arithmetic mean of the numbers that each of the bets form to playing. The mentioned average is the sum of the numbers that compose the bet divided by the quantity of numbers.

2-Numeric Desviation- It is a concept narrowly tied to the average and is the result of dividing the sum of the differences of each one of the numbers that compose the smart one under examination, for the quantity of numbers that compose it.


Cheers.

User avatar
lottoarchitect
Site Admin
Posts: 1635
Joined: Tue Jan 15, 2008 5:03 pm
Location: Greece
Contact:

Re: Development - Wheel Generator 1.9

Post by lottoarchitect » Sun Dec 13, 2009 12:37 pm

Hi azevrec,

I know these filters however they are strongly connected to maths and so they are difficult to grasp by a user. I prefer more intuitive ones to be included. I could get the whole list of LA filters and include them as well however this is not the point, to have a vast collection of filters and diminish the ease of use and understanding of the system. These particular filters can be included but I don't think these can be a priority at the moment, unless if you have some analysis that shows they can offer a good edge in blocks outcome and an easy way to explain to a user how these could be used. Sums and odd/even for example are self-explanatory and what to set on these filters can be performed by an analysis everyone can do by himself or being available in various analysis programs (including LA of course). On the other hand, median and deviation do not fall in this category and unless someone can perform somehow such an analysis, he has no clue on what to set to these filters and this is my questioning actually.

cheers
lottoarchitect

User avatar
lottoarchitect
Site Admin
Posts: 1635
Joined: Tue Jan 15, 2008 5:03 pm
Location: Greece
Contact:

Re: Development - Wheel Generator 1.9

Post by lottoarchitect » Sun Jun 27, 2010 12:28 pm

Just a small update on what's going on in the development on WG.

The underlying engine has been upgraded to a more resilient mechanism which operates now dynamically to the current "stiffness" of the covering being developed. That means the engine automatically adjusts its bias to the needs of the covering so to bypass a possible stuck condition on its own. We are a step closer to a fully automated engine with an almost "set-and-forget" mechanism :) The variability bias and multiplier have been replaced by two other sliders. Their temporary names are "stuck tolerance" and "cool-off", which may change at the final release. The most important of those sliders is the stuck tolerance which practically decides on how to alter underneath the variability bias (which till now was a fixed setting by the user) so to outflank any on-demand situations during the development which can cause the engine to stuck for a long time at a given situation. The other parameter could be used on some very special situations but till now on all the testing experiments, the default value is more than sufficient. Some interaction by the user is still needed however, practically on deciding the initial values of these sliders but the good part is that the necessity to tweak them is not that frequent compared to the current engine. What matters here is that this new approach allows for a quicker build, and sometimes, bypass the performance of the current engine. Better coverings quicker :) The tests show a quicker build on t < m coverings where stuck situations would take quite some time for the current engine to bypass and also quite good improvements on filtered coverings, where the abnormality of the "candidates space" causes the current engine to enter in stuck conditions more regularly.
Also the building speed has been improved even further from the current 1.8.1 engine. Those previous months were devoted in improving the core of the system, and perform numerous optimizing tests, and the results are quite outstanding to be honest :)
Another development is a printing mechanism within WG so to print directly to gameslips. This is a revised version of the printing system used by LA but more easy to setup and more user friendly.

cheers
lottoarchitect

Skirrow
Advanced
Advanced
Posts: 53
Joined: Tue Jan 29, 2008 8:23 pm
Location: UK

Re: Development - Wheel Generator 1.9

Post by Skirrow » Sun Jun 27, 2010 1:33 pm

Excellent news , looking forward to 1.9 .............

luisgmira
Regular
Regular
Posts: 28
Joined: Sun Jul 13, 2008 3:29 pm

Re: Development - Wheel Generator 1.9

Post by luisgmira » Thu Nov 04, 2010 7:48 am

No news since long time ago.
Do you expect to release WG 1.9 soon?
We are eager to see it!!!

User avatar
lottoarchitect
Site Admin
Posts: 1635
Joined: Tue Jan 15, 2008 5:03 pm
Location: Greece
Contact:

Re: Development - Wheel Generator 1.9

Post by lottoarchitect » Thu Nov 04, 2010 6:27 pm

Indeed,
I'll try to make a release date soon but I can't announce something at the momment. I'm sure it will worth waiting for really :) The engine has become quite powerful in performance in all the tests made so far, which is the fine-tuning that made such a delay. I hope to have a release date to announce within December.

cheers
lottoarchitect

Post Reply

Who is online

Users browsing this forum: No registered users and 21 guests