Everyone has noted the resilience of this TCU team.  And I don't want to discount it, but here is what resilience looks like statistically. 

Since week 5, TCU has just a 40% offensive success rate and 6.0 YPP, and has trailed in every game except the Iowa State game and Fiesta Bowl.  This is 4% and and .5 yards below their season average on offense. Although, the Horned Frogs did run up a 50% offensive success rate and 7.0 yards per play and never trailed in the 51-45 win over Michigan in the Fiesta Bowl.

Looking at the defensive trending in the second half of the season.  In the last six game, they are also trending higher than season average in yards allowed, but have improved on the success rate allowed.  At close to 6 yards per play, yet a 39% success rate, this very much looks like a "bend and don't break defense"

In this stretch of the season, the fact that TCU has allowed 6 of 9 opponents a higher than 5.4 yards per rush. This might be where Georgia attacks should the game script go as Georgia would like it from an offensive standpoint.  As we know, TCU has come back in several of these games.  Here are some post week 5 splits when trailing.  153 plays and allowing more yards per play and a higher success rate... that is a lot of bending.

Yet, the offense doesn't seem to improve..  Yard average holds but offensive success is lower.  Clearly the number of plays advantage is a factor favoring TCU. But, I really can't explain this with math..  In fact, when trailing the TCU defense is allowing 65% success rate in the red zone and only a 30% offensive success rate in the red zone.  Bonkers... 

So, don't stop scoring Georgia if you get the lead.