http://dkpittsburghsports.com/2015/11/12/matts-stats-too-much-time-on-their-hands/
From one of the writers at dksports.. rest of the story is at the link
From one of the writers at dksports.. rest of the story is at the link
Using Pro-Football-Reference.com’s superb Win Probability Calculator, the Steelers had about an 8 percent chance of getting a victory after Cincinnati’s Reggie Nelson intercepted Ben Roethlisberger and returned the ball to the Pittsburgh 26-yard line.
After Mike Nugent‘s field goal and a touchback on the resulting kickoff, the Steelers’ win probability actually rose to 15 percent, for the simple reason that they had possession. But what would that number look like if Tomlin had burned all three timeouts before the two-minute warning?
If we assume that the Bengals would have run the ball three times, that would take about five to seven seconds per play. Let’s say that Nugent’s field goal goes through at 2:20 and the Steelers accept the touchback. Their win probability at that point is 16 percent, so it appears Tomlin would have put his team in a slightly better position to prevail by being more aggressive with timeouts.
Could Ben Roethlisberger have been put in a better position to win against the Bengals? — GETTY
But, one thing the Win Probability Calculator doesn’t account for is the value of a timeout. Clearly, having one timeout with the ball puts a team in a better spot than having none, which was part of Tomlin’s postgame argument.
Last year, football analyst Brian Burke (no, not that one) studied what a timeout might be worth in the context of win probability. Burke concluded that 3 percent is a good estimate, although there are other contingencies that could make that value slightly more or less.
For simplicity’s sake, though, let’s accept that calling a timeout costs a team 3 percent win probability. If we apply that number, a team that has zero timeouts with 2:20 remaining and trailing by six points is roughly 2 percent worse off than a team that has one timeout and 1:47 to go.
However, the two-minute warning provides a clock stoppage, which benefits the team trailing more, especially if that team is on offense. Maybe it’s a wash, at least until around the 1:15 mark, when win probability begins to decline sharply for a team down six. Until then, a timeout is worth more than the 30 seconds or so the Steelers let run off the clock.
All of which leads us back to Tomlin’s concern about the Bengals getting the ball back with enough time to counterpunch. If the Steelers call all three timeouts before the two-minute warning and drive down the field in 1:20 — to pick a random number — that gives Cincinnati one minute to kick the winning field goal.
The Steelers’ win probability in that situation is 73 percent, so it’s not like they would be in a bad position. However, if the Bengals’ two-minute offense is as good as Tomlin feared it would be, perhaps the result would be closer to a coin flip.
Still, 50-50 is better than not scoring at all and losing the game. Obviously. Even in this era of increased offense, at no time during a game is the trailing team more likely to win than the one in front. I feel this part of Tomlin’s explanation to be misguided.
One thing is clear to me after completing this exercise: These controversial late-game decisions are not nearly as clean-cut as some of us may believe.
The Steelers hired Carnegie Mellon professor Karim Kassam earlier this year to handle analytics and research, and I would be highly surprised if the team hasn’t looked at time-management optimization in depth.
This is merely one example. I’m not saying every call the Steelers make is the right one, but it’s probable there’s more than one brain at work when it comes to controlling the clock.