Wheels
- lottoarchitect
- Site Admin
- Posts: 1635
- Joined: Tue Jan 15, 2008 5:03 pm
- Location: Greece
- Contact:
Re: Wheels
The naming format is explained at the help file; this 99%-1-576 means 99%+ coverage achieved, with 1 uncovered combination left at 576 blocks.
For a 4if6 construction, in order to evaluate the coverage (or uncovered combinations) we essentially test combinations of 6 numbers x1, x2, x3, x4, x5, x6 (all the drawn posibilities), if they have at least 4 correct numbers in any of the blocks of the covering - in that case of L= 30, this 6-number block must be covered by at least 30 12-number blocks of the covering (it is covered by a block if any 4 numbers of it exist in that covering block). So, 1 uncovered combination means there is one 6-number block that isn't covered by at least 30 blocks (might be 29, 28 or even 0). Any 4-number combination does not need to be covered at least L times to offer the guarantee, instead any 6-number combination must be covered L times.
The problem in replying late at night is it can be easy to write something wrong. In the previous post I said every 4-number combination will be covered at least 31 times. This is only true in t=m but in t<m this is not true. The correct definition and what happens is at the above paragraph.
lottoarchitect
For a 4if6 construction, in order to evaluate the coverage (or uncovered combinations) we essentially test combinations of 6 numbers x1, x2, x3, x4, x5, x6 (all the drawn posibilities), if they have at least 4 correct numbers in any of the blocks of the covering - in that case of L= 30, this 6-number block must be covered by at least 30 12-number blocks of the covering (it is covered by a block if any 4 numbers of it exist in that covering block). So, 1 uncovered combination means there is one 6-number block that isn't covered by at least 30 blocks (might be 29, 28 or even 0). Any 4-number combination does not need to be covered at least L times to offer the guarantee, instead any 6-number combination must be covered L times.
The problem in replying late at night is it can be easy to write something wrong. In the previous post I said every 4-number combination will be covered at least 31 times. This is only true in t=m but in t<m this is not true. The correct definition and what happens is at the above paragraph.
lottoarchitect
Re: Wheels
hi lottoarchitect
thanks for the explanation, very help ful, so i do need to get this to 100%, which 578 blocks
now any suggestion how to get the 4 in 1 block occur as well in pick 6 at least 1 time?(with fewest block) cause obviously i have 30 of them already
a regular 4if4? unconditional serotic?
kind regards
thanks for the explanation, very help ful, so i do need to get this to 100%, which 578 blocks
now any suggestion how to get the 4 in 1 block occur as well in pick 6 at least 1 time?(with fewest block) cause obviously i have 30 of them already
a regular 4if4? unconditional serotic?
kind regards
- lottoarchitect
- Site Admin
- Posts: 1635
- Joined: Tue Jan 15, 2008 5:03 pm
- Location: Greece
- Contact:
Re: Wheels
if you mean all the 4-number combinations to be covered at least once in a pick 6 game, then a 4if4 will do it.
Also did you need 578 blocks for a 36,12,4,6,L=30 covering? This is what you got by WG? If so, try increasing the left slider (expansion) bias to 800 and the right slider (tension) to 500 and let it improve this further. I have made this 36,12,4,6,L=30 in about 550 blocks in just a few minutes and seems it can go quite lower than that. Use this setting instead of the default to reduce further the blocks needed. Generally, the higher 1-block indicator is, the higher the initial bias should be. At later optimization stages (as per the suggestion at the engine tutorial) we reduce these.
lottoarchitect
Also did you need 578 blocks for a 36,12,4,6,L=30 covering? This is what you got by WG? If so, try increasing the left slider (expansion) bias to 800 and the right slider (tension) to 500 and let it improve this further. I have made this 36,12,4,6,L=30 in about 550 blocks in just a few minutes and seems it can go quite lower than that. Use this setting instead of the default to reduce further the blocks needed. Generally, the higher 1-block indicator is, the higher the initial bias should be. At later optimization stages (as per the suggestion at the engine tutorial) we reduce these.
lottoarchitect
Re: Wheels
wow cheers for that, do you think with right settings, wg can get it to 504-512? oh can i send you a file, i'm doing an experiment on a pick6/36, somehow i need another eyes and brain to confirm my finding and hopefully an input on how can i improve this
- lottoarchitect
- Site Admin
- Posts: 1635
- Joined: Tue Jan 15, 2008 5:03 pm
- Location: Greece
- Contact:
Re: Wheels
I can't tell if it can make it to 512 blocks but it is already around 530+ blocks. I try to reduce this even further. Feel free to send me your file along your findings.
cheers
lottoarchitect
cheers
lottoarchitect
Re: Wheels
hi lotto architect
did you get my email yet?
did you get my email yet?
- lottoarchitect
- Site Admin
- Posts: 1635
- Joined: Tue Jan 15, 2008 5:03 pm
- Location: Greece
- Contact:
Re: Wheels
I got it thanks.
Re: Wheels
hi lotto architect
definitely need to put eliminate duplicates blocks if occur feature on in the next update, when is the next update? just curious
definitely need to put eliminate duplicates blocks if occur feature on in the next update, when is the next update? just curious
- lottoarchitect
- Site Admin
- Posts: 1635
- Joined: Tue Jan 15, 2008 5:03 pm
- Location: Greece
- Contact:
Re: Wheels
As soon as the new GAT version is out I'll work on WG for this feature along some other enhancements.
lottoarchitect
lottoarchitect
Re: Wheels
hi lottoarchitect
how do i know that new version is out?
and have you take the time to look at the files i sent you?
kind regards
how do i know that new version is out?
and have you take the time to look at the files i sent you?
kind regards
Who is online
Users browsing this forum: No registered users and 10 guests