With 1 minute, 40 seconds to play in Sunday's matchup with the Washington Redskins, the Seattle Seahawks trailed 10-8 with a first-and-10 on the Washington 30-yard line. On the ensuing play, Russell Wilson connected with a wide-open Doug Baldwin, who caught the ball at the 12-yard-line and ran untouched to the end zone to give Seattle a 14-10 lead with 1:34 to play.
The TD put Seattle in excellent position to win the game -- even with the failed two-point try the Seahawks' win probability was a sterling 88.2 percent. As it happened, the Redskins rallied when Kirk Cousins led them on a four-play, 70-yard touchdown drive in just 35 seconds, handing Seattle a 17-14 loss.
Going back to the Seahawks' go-ahead score, Baldwin likely gave no consideration to pulling up short of the end zone. Not only would it have been surprising if he had, you can bet Baldwin would have left himself open to criticism if he'd been thinking in terms of bleeding the rest of the clock and setting the Seahawks up for a game-winning field goal try that kept the ball out of Cousins' hands.
On a day when Blair Walsh -- the Seahawks' famously streaky kicker who authored one of the most notorious chip-shot misses in NFL history -- was already 0-for-3 on field goals, few could fault Baldwin or the Seahawks for thinking touchdown in that situation.
But win probability says Seattle's best chance to win the game would have been for Baldwin to give himself up on the 1-yard-line, which almost undoubtedly would have led to a Redskins timeout with 1:34 to play (their second).
Kneeling on the next two or three snaps, forcing the Redskins to use their final timeout, and further bleeding the clock to set Walsh up for what would have been approximately a 20-yard field goal with less than 20 seconds to play would have offered Seattle its best chance to win the game.
Really? Give up a sure touchdown to pin your hopes on Walsh and his scatter-leg? The percentages say yes.
It's true that Walsh had already missed from 44, 39 and 49 yards in this game. But this would have been a much shorter attempt, and Walsh is 35-for-36 on attempts from 20 of 29 yards for his career. Conditions were wet but winds were calm, and at that range conditions would have needed to be much worse for the kick's chances to fall far from the league average of 99 percent. Walsh's career record is a much, much better predictor of his accuracy from that range than the small sample of three much longer kicks, or one highly memorable 27-yard-miss in the playoffs with the Vikings.
Lest you believe such an unorthodox strategy is unthinkable for a team desperate to take the lead in the final minutes, it has been employed before. The tactic is well known in the analytics community -- something I call "the field goal choke-hold." The tactic was even on display in Super Bowl XLVI, when the Patriots intentionally invited the Giants to score the go-ahead touchdown in order to escape the choke-hold and have a chance to score in response.
Taking a knee rather than scoring the touchdown would have virtually assured Seattle victory rather than giving the Redskins an 11 or 12 percent chance to prevail. Instead, the door was left open for Washington to rally and cost the Seahawks a much-needed win that would have pulled them even with the NFC West-leading Rams.
For more from ESPN Analytics, visit the ESPN Analytics Index.