What is the chance, on a per-game basis, that a starting running back will suffer an injury that affects his future earnings? ACL tears don't seem to have a long-term effect anymore. It would have to be a torn Achilles (and even that's not as bad as it once was) or a compound fracture or dislocated ankle or a life-changing concussion or something along those lines.
There are 512 RB starts each season. How often does a running back suffer an injury so bad that it affects his future earnings? It happened to Robert Edwards 20 years ago, but that wasn't even during an NFL game. It happened to Napolean McCallum 25 years ago. Off the top of my head, I can't think of other injuries to RBs that have been career-ending (since ACL reconstructions have become routine). If we expand to non-RBs, Alex Smith counts, as does Joe Theismann. Darryl Stingley, Dennis Byrd... (But expanding to non-RBs makes the denominator grow as well.)
I'm sure I'm missing some, but let's say there have been 10 examples in the last 20 years. That's about a 0.1% chance per game.
So skipping a game for $330,000 would be worthwhile if Gordon had a 0.1% chance of having his future earnings reduced by ... $330 million.
Even if we bump it up to a 0.5% chance per game, the risk to his future earnings would have to be over $60 million to justify sitting out.
I don't think that's realistic. In purely monetary terms, I think it's pretty clearly -EV to miss games for $330,000 each.