I guess I would start with adding HWD filter to go along with LWD filter.
Then find the coldest LWD pair, and the coldest HWD pair for each game.
By pairs I mean LWD pairs:
(0,1) (0,2) (0,3)
(1,2) (1,3)
(2,3)
Pair (2,3) should be weighted higher because it represents fewer combos
Pairs (1,2) and (1,3) should be weighted higher than pairs that contain 0 because it represents fewer combos.
Then you would do the same for HWD pairs
(9,8) (9,7) (9,6)
(8,7) (8,6)
(7,6)
Then do something similar for Key Digit Pairs
Using these 3 weighting scores you can rank each state game as the best game to play for that day.
I do something similar for Pick 4 but its more complex.
So I would say it's taking your alert system to the next level where you analyze all the games and score them. This will get you more into the area of AI regression when you start dealing with weighting.