Yesterday featured three different overtime games in the NFL. The Browns were the victims in one of them, after a very questionable overturn of a first down gain by Carlos Hyde late in the game that would have allowed Cleveland to run out the clock. The other two featured rookie AFC South coaches in Frank Reich and Mike Vrabel put to the test in some crucial decisions.
Let’s start with Reich since his decision to go for the first down rather than punt it away and almost assuredly guarantee a tie game ended up resulting in a Texans’ victory at the end of overtime. Indianapolis had a 4th and 4 at their own 43 with 27 seconds remaining in overtime. A failed attempt, of course, puts the ball somewhere around the 43, then (could be a sack, or a tackle short as well as an incompletion). That would leave Houston just outside of a long field goal attempt, time for one play to the middle of the field (which they successfully accomplished) or some sideline passes to get in field goal range.
A conversion, meanwhile, would leave Indianapolis in an almost similar situation. 4th and 4 situations have been converted at around a 43% rate, and the median gain is 9 yards. That would leave Indianapolis on average just inside Houston territory. They had one timeout left, and would need about 15 yards for a decent field goal opportunity.
Because the chances of conversion were less than 50%, you might be tempted to say that it was a poor decision, but the timeout differential and the possibility that the Colts break a play on the 4th down that gets them into field goal position right away I think balances it out. It was an aggressive decision, and maybe it was a negative net decision by 1-2% (though we see coaches make far bigger errors regularly and they don’t get called the worst decision of the year).
I base that on the score not being automatic by the Texans (as many of the commentary I have read seem to assume) but more like a 50/50 shot with no timeouts. Six of 13 teams in last decade with a 1st down in similar part of field and similar time left have successfully scored. Teams do miss field goals, get tackled and waste time, etc. I also base it on a timeout being worth about 15-20 seconds, which would increase the Colts’ chances of scoring relative to the Texans, by about 10% in a similar part of the field. If you think there is a material difference between Adam Vinatieri and Kaimi Fairbairn as kickers then you might even close the gap even more.
There’s also the question of the impact of a tie. We don’t know how the season is going to play out, and teams later in the year have an advantage of more specifically knowing how a tie will impact them. Teams that have gone 9-7 in the last decade have made the postseason about 43% of the time, while 10-6 teams have made the postseason 82% of the time. There aren’t many actual 9-6-1 cases but a hypothetical 9-6-1 team would have made it about 60% of the time over the last decade depending on what division and conference they were dropped in. The Colts were already 1-2. 1-2-1 would still need them going 8-4 just to have a realistic chance, while 2-2 would give them a far better chance. I suspect when we look back the Colts and Reich were right to assume they need to be very aggressive in trying to get a win.
Reich said after the game that he would play for the win rather than the tie 10 times out of 10. Now, I’ll point out that going for it made a tie less likely, but it was still at least a 50/50 chance despite his attempt (because either team could have failed to get in field goal range or missed the kick).
Now, there is an earlier decision that you might want to question if you want to enforce the “we don’t play for ties” mentality consistently. You see, on 4th and 2 from the Houston 25 earlier in overtime, Reich chose to take the 3 point lead. That decision isn’t a crazy one in a vacuum. Teams do convert a 4th and 2 about 58% of the time, and the Colts would have been a much heavier favorite to win in OT had they gone and converted. Kicks aren’t automatic (though Vinatieri converted, he is about 81% from that distance in recent years). Add in the greatly increased likelihood that the kick would result in a response by the Texans, and decent chance that the teams would be tied again with limited time left, and I think it’s fair to say that going for it on 4th and 2 on the first possession of overtime would be the aggressive move to make to better prevent a tie.
Mike Vrabel, as it turns out, had a similar 4th and 2 call in overtime against Philadelphia. He went, the Titans converted, and Tennessee beat the defending Super Bowl champions with a late touchdown a few plays later.
The situations, though, were different, and those differences should have made Vrabel’s decision an absolute no-brainer. That we need to praise him for making the correct call shows the state of current conventional wisdom. In fact, if a coach in Vrabel’s position had opted for a 50-yard field goal attempt to simply try to tie the game in OT, with only 1:17 remaining, to then give the ball back to the Eagles, it would actually qualify for the worst decision this year (like some articles about Reich had claimed).
Vrabel was already playing with house money because the Titans had converted two fourth-down attempts already to get to that point, down 3 in overtime. But it should have been a fairly easy decision because:
- 50-yard field goals are far from automatic (about 73% in recent years);
- 4th and 2 attempts are successful 58% of the time;
- the time remaining dictated that the Eagles would have been heavy favorites to win if the game was tied, relative to the Titans’ chances of winning.
Given the time remaining after a conversion, the Titans would have had about a 37% chance of winning outright, a 17% of still tying with a shorter field goal at the end of OT, and 4% chance of not scoring. Add in the 42% chance of losing outright, and count tying as a half-win, and the Titans expected win percentage by deciding to go for it was 45.5%.
Meanwhile, based on similar situations, the Titans would have had about a 3% chance of winning in OT by kicking the field goal (basically the chances of getting a turnover or quick punt and scoring off it with little time left), while the chance of losing was 50% (along with a 47% chance of a tie). Using the tie as half-win calculation, that would make the expected win percentage by kicking a long field at 26.5%.
Like I said, it would have been a massive error of 19% in win probability, and Vrabel made the clearly correct decision. And yes, the size of that error is influenced by a long field goal not being a given. But even if we assume that the kick is made, the win percentage chances are 10% greater by going, because of the clock situation and the very small likelihood that the Titans could then still win the game.