How did the game play out? What are the abilities of the players involved? Those are my only questions.
Did the CPU just out-play me on the ground? My LBs were trying to fly around, the run blocking for the CPU wasn't godly, and the HB at least had some ability? Then I'd leave it for now and try to call a better game. If I can make adjustments, even if it means having to "gamble" a little or worry about the run more often, then I would think my sliders there are doing good. I might just need to get better players or a better scheme.
On the other hand, if it's just a bunch of long runs because the HB broke 6 tackles each time, then I might would address that. (Again, depends on player ability too - if it's Blount and my guys are tired - I probably don't do anything). If some scrub HB is putting up 5-6 yds/carry against my elite run defense, a flag goes up. If it happens constantly where my top level run D is giving up that kind of yards and it's not because of what I'm doing or having to respond to (I have to play pass a lot so I can't load up the box), I likely change it.
For INTs, if the CPU is constantly throwing 2 INT a game, even in an otherwise good day, I would likely address that, unless I have the most ballhawking secondary. If they are good at coverage but have a bunch of 50 CTH ratings, I probably lower it. Coverage does not have to equal INT.
But stats are just half the equation, especially over just a few games. I look at how the player behavior is responding on the field. LBs not attacking? I deal with it. Too many blown assignments in the secondary? I'll address that. Any old offensive line is manhandling my defense? I'll address it. My OL not sucking like they should in pass protection? I'll make them suck like they should.
For me, it's about player behavior and having ratings express differences in ability whenever possible at the end of the day.
Because of this approach, MySkill didn't really do a lot for me. One of the big problems is that it was just raw stats focused. If the CPU didn't run, the game assumed I was awesome against the run, not that the other team ran just 12 times for 48 yards (which is an average 4.0 per carry, not elite). MySkill was a good idea - it just needed much more considerations than just raw numbers, imo.