The Ravens executed a regular season overtime strategy last night, by taking ball first, that I wrote about last year. At this point, I think it is the straightforwardly optimal path. I also think part of the confusion is (current) models probably miss this, because they don't account for the way the offense can manipulate the situation (despite acknowledging people who suggest models are missing something are wrong 99% of the time, because the models don't miss the thing). (It could also obviously be modeled.) And that's running your first drive as an intentionally clock-killing drive. It assumes success, but so does the strategy of taking ball second. (If both teams go three-and-out, you clearly wanted to have taken ball first.) I do understand the element of knowing if you need to go for it on a fourth down on the second drive. I think that is overstated relative to the strategic flexibility of controlling overtime pace when you go first. This is entirely because of the 10-minute hard constraint on regular season overtime periods. It would not apply to postseason overtime decisions. What happened last night is the Falcons had to use both of their overtime timeouts on defense to preserve time. They got the ball back with just under three minutes, and only the 2-minute stoppage, and it forced them to be one-dimensional. On the broadcast, Collinsworth suggested kicking to force the Atlanta QB situation onto the field quicker. What Baltimore did attacked the Atlanta QB situation more; if Atlanta got ball first, they could've tried a run-based offense. They really couldn't do that with 2:50 and no timeouts. Setting aside the specifics of Atlanta's QB situation, EPA models look at next-drive success, because that's crucial in football. What I'm suggesting is missed here is when a team executes like Baltimore did last night, the opponent's drive success takes a substantial hit because of the clock constraint. A similar thing happened to the Packers last year, in their game that ended up a tie against the Cowboys, where the clock became a real constraint and put them in a bind in the final minute. They essentially had to abandon the run, and nearly lost, but had one second to kick a game-tying field goal. Anyway, Baltimore executed this brilliantly last night, and more teams should take note. Kicking off gives your opponent the ability to execute this. The Falcons didn't necessarily run well last night, but they've been great on the ground in other recent games, and an alternate result last night where they had ball first could've been them executing a run-based offense that bled clock and put the Ravens in the situation they found themselves in, with not enough time to run a balanced offense and force to be one-dimensional and throw. We're seeing longer and longer drives in a modern NFL hellbent on taking away explosives, and this is increasingly how TD drives are put together. It's a crucial part of the overtime coinflip decision.