One of the things, that I have been looking at,on and off, is what I call " Subtracting the Delta".
Now some of you have heard the term Delta before.
This is how it works:
Last draw 2080, 5,9,17,31,37,44. I do not include the bonus for
this one.
Line ONE
subtract 05 from 09=04.
subtract 05 from 17=12.
subtract 05 from 31=26.
subtract 05 from 37=32.
subtract 05 from 44=39.
Line TWO
subtract 09 from 17=08.
subtract 09 from 31=22.
subtract 09 from 37=28.
subtract 09 from 44=35.
Line THREE
subtract 17 from 31=24.
subtract 17 from 37=20.
subtract 17 from 44=28.
Line FOUR
subtract 31 from 37=06.
subtract 31 from 44=13.
Line FIVE
subtract 37 from 44=07.

Thats 14 numbers to look at for the next draw, numbers...
4,12,26,32,39,8,22,35,24,20,28,6,13,7.
We'll review how many of these came after the next draw.

Quite an interesting approach to the good old Delta theory. But as usual, my question is: Is it worthwhile?

Using each of the first 2080 drawing of Lotto 6/49, your method, Peter, generates on average 13.2 numbers due to duplicate delta values . It ranges from 15 numbers (the maximum) to only 5 numbers (with the quite improbable drawing #670 with numbers 22, 27, 32, 37, 42 and 47). This method of predicting numbers also has the disadvantage of not choosing number 49.

Verifying how many times these numbers came out in the next draw, we get on average 1.58 winning numbers. This is even a little under what could be expected, 1.62 (or 13.2/49*6).

So choosing random numbers might give as good results as this method.

interesting, thank you Gilles for testing this, as I said earlier it is not something I have pursued, just more or less looked at it every now and then.