quite simple really, take a block or window of last 18 draws, see how many times each number hit, then classify them as hot, avg or cold based on 3 hits or more (hot), 1 or 2 (avg) and 0 hits (cold).
The chart shows how many numbers were Hot/Avg/Cold prior to each draw and then how many hits came from each category for that draw.
ie: just prior to draw 1445 there were 22 Hot nrs ... 4 of those hit
Perkis is going against the status quo here as 10 draws is the cut off for hot/cold normally, neither are average nrs singled out. His way of classifying leads to the 'Incredible Super Wheel' ... I haven't got as far yet, want to see first how BC49 compares to his example (Florida 6/53 ?), and possibly along the way stumble onto something usefull. Thought I share my 'discovery' trip as I go so other ppl can point out things I miss ... glasnost approach
not at all sure,so this:
8 boards 2, 17, 39, 40
8 boards 4, 17, 39, 40
then 23, 24, 28 48
16 boards playing 6 #'s between 15 and 36
40 boards where i started with
then on each subsequent board moved one number up
or down or diagonally eg, the next board i played five #'s
the same, and moved the 23 to 12....except for 48 and 40,
stayed away from the 40's............huuraa
It may be possible to spot some short term trends that way as you point out.
This H/A/C sorting is designed to help with line making in that one should not use more than 4 Hot, or 3 Cold (rare but happens) ... I gather that some kind of mix of H/A/C in every line is desirable. It is possible to get all 6 from the Avg column, tho not recommended to play them exclusively.
I think that even when using wheels other than what Perkis suggests one should crosscheck with the the H/A/C chart and change lines that are skewed with too many Hot nrs, etc. (this may be hard to do without changing the wheel cover).