I definitely wouldn't totally re-do all the all-madden sliders because of one human game. Defensively, the game against the Eagles was the best the CPU looked in weeks against me (ignore the box score, I had three broken long plays that skewed yardage - mckinnon 84 yard rush, 64 yard pass to eifert, and I think a long broken tackle play to calvin). CPU offense still needs some help though.
if the set the guy made are 15 accuracy, we had it at 30, and bueno had a bunch of bad throws, I'd play a few more games out before making a total readjustment. Bueno, how often %-wise were you throwing it deep vs short in that game? I'll typically complete like 80-85% of my underneath passes, but only like 25% of my long passes with a ton of bad throws, so if i throw it more often longer than short, it kills my completion % and usually causes more INTs (just like real-life)... I lean more toward short-medium, and only throw it long 5ish times a game, so my accuracy is usually in the 60ish range... last game I think it was 58% which was perfect (considering I threw a little more longer than usual). If we up accuracy, I'll start completing 75%+ against the CPU.
Were the INTs short or deep?
I do see that human vs human with sliders can cause issues with other things such as pass rush, running, tackling, etc.... but QB accuracy shouldn't be affected at all by human vs human, it should be because of sliders.
Bueno, keep in mind this was your first human game... When Rob and I played back when we had all-pro sliders, both our QBs were under 50% compeltion (wilson 7 for 23, stafford 13 of 27, combined 4 INTs), but there was some wind (at seattle)... when Irish and I played with little or no wind, both games, both our QBs were over 50% completion, (7 TDs 3 INTs). Maybe I've seen less issues because half my games are indoor.