Very High Improvements (Good or Bad)

Talk about anything Wheel Generator related which doesn't fit in the other forums
Post Reply
20mm
Casual
Casual
Posts: 13
Joined: Tue Mar 25, 2014 9:01 pm

Very High Improvements (Good or Bad)

Post by 20mm » Wed Apr 30, 2014 10:55 pm

Hi LA,
My main objective at this time is to fully comprehend the behaviour of the controls with WD and GAT before doing anything "outside the box". In conducting runs with WD, I attempt to maintain the manual's recommendation to keep "b" between 1% and 10% of "c" where Improvements is represented by a (b / c). This however I have found to be very difficult and improvements have been very slow and very low. In the past my improvements have not gone beyond 50. In my current run Improvements have gone pass 6 million and continuing and the Best Covering window is rippling as fast as the Current Covering. Even as I am writing it continues. Could you comment on this. Below is my story

Lotto 5 of 35
Covering Parameters: c (18,5,4,5)
L=1
L%=100
1-Block cover: 66 / 8568 (0.77031%) - Th.Min: 129.818
Four Groups in Matrix:
Group-1 : 0,0,0,0,0,0 - 3
Group-2 : 0,0,0,0 - 2
Group-3 : 0,0,0,0,0,0 - 3
Group-4 : 0,0,0,0 - 2
STRENGTH: Standard.
I started with 5 blocks progressively increasing by 5 when I observed no more improvements. I increased blocks by 5 up to 20 while incrementing the Expansion Bias up to +121 and the Tension to -741
I observed the improvements rapidly increasing reaching 5 million plus at this time.
Curious of this happening, I increased the blocks up to 50, incrementing by 10 each time. Doing this, the improvement was still rapidly increasing. An example of my progress window is below.

Improvements: 4832310 ( 3 / 9 )
Main Coverage : 6002 (6032 – 5268) / 8568 [29.470123%]
- Filtered : 1 (1) / 240 [99.1666667%]
Non-Guide Filters fit : 44 (44 – 82.00%)
Sum – 42 [42 – 84.00%]
O/E – 50 [50 – 100%]
Consecutive – 48 [48 – 96.00%]
Guide Filters active : 4
Group #1 to #4

YOUR COMMENTS WILL BE VALUABLE.

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

Re: Very High Improvements (Good or Bad)

Post by lottoarchitect » Wed Apr 30, 2014 11:23 pm

Hi 20mm, obviously getting that many improvements is impossible no matter what design you attempt. Something must have gone very wrong. The information you provide is puzzling given it is impossible to set this for example
Group-1 : 0,0,0,0,0,0 - 3

What are all those 0,0,0,0?? Because of that, I cannot test what you say here. I'd suggest to save the preset you use and send it to me for testing.

On the side note, regarding the rule "keep b between 1% and 10% of c", in reality this works in several coverings with t<m but many others will not work well with this guide. The best advice is really to keep a bias setting that does some fluctuation at the current coverage and moves away a bit (relative to the 1-block indicator) from the current best coverage found and bounces back at regular intervals, even equaling the current best coverage once in a while. Each construction is actually unique to the bias parameters needed during its construction and it is impossible to give a general advice on settings to use; experimenting with the bias sliders is the best route to find what works best. Apart from that, reaching the improvement count to 4832310 is impossible even if you run the engine for years! Send me the preset to test it.

20mm
Casual
Casual
Posts: 13
Joined: Tue Mar 25, 2014 9:01 pm

Re: Very High Improvements (Good or Bad)

Post by 20mm » Thu May 01, 2014 6:00 am

I thought something was a bit odd with that many improvements and the speed at which it progressed. Excuse my ignorance at this time, how do I send you the preset. I not seeing an attachment button to attach the file on this forum window.

20mm
Casual
Casual
Posts: 13
Joined: Tue Mar 25, 2014 9:01 pm

Re: Very High Improvements (Good or Bad)

Post by 20mm » Thu May 01, 2014 6:27 am

Couldn't attach file so I copied the Preset text to your message box. Thank You

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

Re: Very High Improvements (Good or Bad)

Post by lottoarchitect » Thu May 01, 2014 7:54 am

I got your preset. You could also send it by e-mail. I have disabled attachments support due to server limited space for external files. This includes photos too.

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

Re: Very High Improvements (Good or Bad)

Post by lottoarchitect » Thu May 01, 2014 6:41 pm

Hi 20mm, I see a variation between the preset I got and the information you post above. I see no guide filters active in the preset whilst your post indicates 4 number groups in guide mode? With the preset I got, I can't make it behave the way you describe it. Everything looks and works normal here.

20mm
Casual
Casual
Posts: 13
Joined: Tue Mar 25, 2014 9:01 pm

Re: Very High Improvements (Good or Bad)

Post by 20mm » Thu May 01, 2014 11:39 pm

I cant explain the discrepancies. However I eventually shut down the program and restarted. I tried at best to follow the same sequence of inputs, i.e adding of blocks, adjusting bias etc and WG did not react the same way. What did occur however, are improvements up to 200+ much increased from what I have seen before. I did a few practice runs and can say that it has to do with timing, i.e adding blocks and adjusting the bias at the right time. The multi million improvements occurrence may have been an isolated event.Thanks again for your commitment.

Betovn
Casual
Casual
Posts: 15
Joined: Sat Apr 18, 2015 3:30 pm

Re: Very High Improvements (Good or Bad)

Post by Betovn » Sat Apr 29, 2017 4:09 pm

man thats awesome i wish my immprovements were that high been trying to mess with everything that i can to accomplish but nothing

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

Re: Very High Improvements (Good or Bad)

Post by lottoarchitect » Tue May 02, 2017 9:03 am

Betovn, the improvements count indicated here are probably due to a glitch in the engine. It is normally impossible to have that high amount of improvements observed.

Post Reply

Who is online

Users browsing this forum: No registered users and 9 guests