A regular analysis of strategy, decisions and calls that impacted the week of NFL play.

I'll admit it. The NFL's change to its point after touchdown rule last season, which once seemed too much of a compromise to affect outcomes, has cascaded in ways that were never anticipated. It helped decide two games on Sunday during what was the most entertaining nine hours of the season, generating new strategies in an industry that rarely incentivizes on-field innovation.

We've now had 27 weeks of spotting the extra point at the 15-yard line and making it a "live" play, meaning defenses could recover a loose ball -- via fumbled snap or a block -- and return it for a two-point conversion. (The ball previously was placed at the 2-yard line.)

The shift has accomplished its primary, if modest, task. The extra point is no longer automatic; its success rate has dropped from 99 percent to 94.4 percent. But two other by-products are more interesting to me.

The rate for two-point attempts has nearly doubled, spurring games like Sunday's matchup between the Dallas Cowboys and Pittsburgh Steelers. The Steelers, confident in their high-powered offense, went for two after all four of their touchdowns and missed each time in a five-point loss.

I was in favor of moving the two-point conversion to the 1-yard line, which would have encouraged a massive shift toward what is undeniably a more interesting and less automatic play. Still, teams have modestly embraced it from the 2-yard line as a scoring tool beyond fourth-quarter math. We've seen seven attempts in the first quarter of games since the start of the 2015 season, including five by the Steelers, and another 13 in the second quarter.

The other development has revealed itself more recently. Teams are now strategizing ways to block extra points, hoping to capitalize on the shift to a live play. Their 11 blocks this season have nearly matched their total for the entire 2015 season (12), aided by a method that seems relatively foolproof and might require a response from blocking schemes.

When the play is designed and executed properly, it is perfectly legal for a player to jump over the long-snapper, land in the backfield and block the kick from point-blank range. On Sunday, the NFL approved a wrinkle to that approach: Senior vice president of officiating Dean Blandino said it was legal for the Denver Broncos to push down on the back of New Orleans Saints long-snapper Justin Drescher to make the jump easier.

Justin Simmons blocked what would have been a go-ahead extra-point attempt from the Saints' Wil Lutz, and Will Parks returned it for a game-winning two-point conversion with 1 minute, 22 seconds remaining. It was the third such return since the rule was changed, but the first that affected the outcome of a game.

The end result? If nothing else, smart NFL coaches have some fertile ground in the otherwise picked-over terrain of conventional football strategy. In Week 7, as you recall, Seattle Seahawks linebacker Bobby Wagner twice hurdled Arizona Cardinals long-snapper Aaron Brewer to affect kicks, careful to avoid anything other than legal incidental contact in the process.

Someone asked me on Sunday why teams aren't jumping over long-snappers on every play. There is some risk of a penalty if the player mistimes the leap and lands on the long-snapper, but otherwise it appears that coaches have found a competitive loophole. Now you see why Cardinals coach Bruce Arians, a member of the league's competition committee, was so upset when the Seahawks did it to his team.

Pulaski (Arkansas) Academy coach Kevin Kelley, who is known nationally for never punting, among other strategy innovations, suggested Sunday night that teams might need to create a "fullback" position in their kicking alignment. That player could either stuff a leaping blocker or deter the call altogether.

When the NFL passed the rule, I thought it would produce more annoyance than entertainment. I still don't enjoy watching kickers miss extra points. It has happened 107 times since the start of last season. But intentional or otherwise, the rule has added a few additional layers to the game. That's never a bad thing, and on Sunday its forces collided to produce drama and entertainment at levels that might not otherwise have occurred.

More adventures with Triplette

Referee Jeff Triplette has produced some of the most visible moments for NFL officials this season. He added two more to the list in Week 10.

Triplette ejected Tennessee Titans offensive lineman Taylor Lewan for making contact with back judge Brad Freeman in the first quarter at Nissan Stadium in Nashville, Tennessee. Lewan was the sixth player Triplette has ejected from a game this season, accounting for more than half of the NFL's season total. It was also Triplette's third ejection in the first quarter, presumably a conscious attempt to set a disciplinary tone early in games.

It didn't really do the trick, however, as the Titans and Green Bay Packers briefly brawled in the third quarter after Titans cornerback Perrish Cox hit Packers quarterback Aaron Rodgers late after a touchdown. That play did, however, produce a tortured explanation from Triplette that required two corrections and nearly four minutes to sort out.

Initially, Triplette announced: "After the score, unnecessary roughness, No. 20, late hit on the quarterback in the end zone. Which instigated the following things: That 15-yard penalty will be assessed on the kickoff."

Triplette then returned to say: "Correction on the previous announcement. Tennessee, correction, Green Bay, has elected to take the half the distance to the goal on the try. Moving the ball to the 1-yard line and going for two."

Then with an apparent chuckle, Triplette returned one more time: "Correction, correction. That is not something they can do. So we'll move it back to the 15 and penalize the 15 [yards] on the kickoff. They've elected to go for two from the 2."

The NFL has a limited policy that allows both the replay official and supervisors in New York to advise the referee on correct application of the game rules via wireless communication. So it's possible that Triplette got some help, and that the system worked, to avoid allowing the Packers to move the line of scrimmage for their two-point conversion.

But the sequence hardly inspired confidence that the game was being officiated well, much less at the crisp pace the NFL demands. Commissioner Roger Goodell said last week that the league was looking for ways to shorten game times in response to a decline in television ratings. Sunday's contest in Nashville took 3:36 to complete, almost 30 minutes longer than the league average, and Triplette's announcement adventure was a small part of that.

According to my amateur stopwatch, 3:52 passed in between Rodgers' touchdown and the snap of the ball for the two-point conversion. It wasn't as long of a delay as we saw in Week 5, when referee Gene Steratore needed 8:36 to sort through a review of an Oakland Raiders touchdown pass, but it was the kind of sequence that would prompt a non-diehard football fan to find something else to watch.

If Ezekiel Elliott had taken a knee

In theory, the Cowboys could have eliminated much of the fourth-quarter drama in their game against the Steelers. It would have required a simple but counterintuitive decision, one that probably won't ever hit mainstream practice no matter how much sense it makes.

The approach required rookie running back Ezekiel Elliott to take a knee at the 1-yard line, rather than score on a 14-yard run, with 1:55 remaining. Had Elliott done that, according to ESPN senior analytics specialist Brian Burke, the Cowboys could have drained most of the clock and then converted a 19-yard field goal to take a 26-24 victory.

Instead, the Cowboys left enough time for the Steelers to drive and take a 30-29 lead before Elliott scored the game winner with nine seconds remaining.

I'm sure it would go against every fiber in Elliott's body, not to mention every fantasy player's dreams, to eschew a touchdown with less than two minutes to play. And there is always the chance that a penalty, turnover or freak field goal miss would upend the best of intentions.

But Burke's win probability formula, which takes those possibilities into account, would still have given the Cowboys a better chance to win (97 percent) had Elliott taken a knee at the 1:55 mark rather than score immediately (87 percent).

I doubt that competitive coaches and players will ever teach this approach, much less put it into action. But if nothing else, the numbers provide a look at the impact of leaving too much time on the clock at the end of a close game.