thought error – still do a model?

I am working through models related to overeating Halloween candy.
One model:
C- Halloween Candy on table
T – “Just one more”
F – conciliatory
A – there’s always one more, overeat, looking for one more, avoiding judgment by just taking another
R – lie to self/not honoring self

Another T, similar: “I need to eat this to get it out of here.”

OK, both T’s are problematic. “Just one more” is a lie, and “eat to get it out of here” is a Thought-error. I question if identifying the Thought-error aborts doing a model on it — do you do models on thought errors or just identify that it is bunk? Both thoughts just feel like a vortex/never ending cycle — is this because they are entering into thought loop territory? Or is it the nature of overeating where you think things that go on auto-play without an end point (meaning, just one more = keep looking for more)?