I'm just starting to get into playbook editing, but I noticed some things to think about here, if you haven't tested them.
I noticed there are no values identified specifically for fourth and long, 4th and medium, or fourth and short. the computer may just use the same parameters as if it's going for a 3rd down conversion, or maybe tag #36 is 4th and long etc?
Also, there are no values yet defined for when the computer is in "4th quarter comeback mode" which I don't believe it steals from the 2 minute drill tags.
And finally, the computer will also read being down in the 4th quarter like this: Is it 4th quarter? If yes, then is the offense down by 3 or less? If yes, then is the offense inside the opposing 40 yard line (field goal range)? If the answer is yes to all of these questions, it will usually break out of 4th quarter comeback and go to running plays. This is most probably clock control run, but maybe it has a different value?
I didn't test anything extensively, but I made a test playbook to problem solve with. In the in-game editor, i chose 10 different formations that were easy for me to tell the difference between at a glance, i chose 1 pass play and 1 run play from each and created a playbook of 20 plays. Then i went into the playbook editor and gave each play a specific value unique to itself. This way i knew that if the computer chose a run play from the 3wr split back set, it was using value #1 etc. When there is more than one play in any given value, it is chosen randomly, but the computer does not choose an actual parameter randomly. As far as I can tell, it chooses the same parameter in every identical situation and if there is some deviance, then there was another trigger for a different parameter that took precedence.