Sounds as though it's coming right along. I always enjoy reading about what you're doing on this stuff. Have a great one amigo.
Jack

Hypersoniq
Hi!
A 1 combo per draw every draw system should give you no less than 1 straight hit per year to be really good or in 10 years no less than 10 straight hits even if it does not hit every year.
So 164 straight hits in 18 is really really good I think, that would be like 9 straight hits per year, Right ?
It does not need to be any better at all, but if you can then go ahead.
I think that you now have the system that you wanted.

my bad... 28 years, not 18

and 162 hits but the first 90 don't really count because rolling data was way too biased

here's the breakdown
1985-2005 20 years    expected 20 hits, got 27 hits!
1995-2005 10 years    expected 10 hits, got 9 hits.
2000-2005 5 years      expected 5 hits, got 7 hits!
year 2005 1 year        expected 1 hit, got 1 hit!
overall, 28 years, genuine hits = 72 expected hits = 28

best valid year was 1981 with 5

overall cost $9,878
overall return $36,000
net profit $26,122

so far, so good but I ned to root out the bad data at the beginning by reducing "pair" counts by 1 to remove info you didn't have before the draw, then I will let those results stand.

btw, the 18 year results
1987-2005 18 years    expected 18 hits, got 21:-)

Hypersoniq
That is still very good.
$26,122 even in 28 years, is till good.
Most people never make a profit in the long run, not after 28 years for sure.
Keep at it.
I will again try to make my own 1 pick per draw system soon, might take a long while to figure something out that works right.

I am stuck on the slower PC for the moment, but tonight I will get the P4 and recalculate the error out.
If all goes well, I will be doing that to PA evening pick4 data this week as well.