Crazy thought of day: Cowboys should have missed extra point on purpose with 2:32 left so they'd be up 3, not 4

Are you drinking tonight?

see above.

What do you guys expect, he's a Cowboys fan.

Knew this would be stupid.

Score as many points as you can.

Since these games are all fixed to try to come down to the last play, the Cowboys had to make the extra point to go up 4 with plenty of time remaining. That way when the other team (who probably hasn't scored since the second quarter) all the sudden easily drives down the field and scores a TD, the Cowboys would still have a chance to tie the game with a FG. That is really what the network wants so the game can go into overtime and they can make a few million bucks on advertising. Every once in a while a team is too inept (even with the officials help) to do this as Detroit was in this game.

You should just be happy Dallas won and not worry about some probability that has nothing to do with proper decision making during a game.

If the other team has a better chance of winning by going for a touchdown instead of a field goal, why wouldn't they do that regardless of 3/4 points?

There might be someone somewhere who is stupider than rojo.

Great point, Rojo. The responses on here show why this type of tactical maneuvering is 10 years away. Maaaaybe 5. You literally show the 82%/74% stat, but still people just blast the idea. If he did miss on purpose, a coach would get fired for this, unless the owner is some kind of enlightened stats maverick. I'm guessing Jerry Jones isn't that guy. Maybe he is. More likely to find Mark Cuban supporting that move than Jerry Jones.

Once people start exploiting this tactic, I wonder if the league will actually give a team up by 3 some extra incentive to go for the point. Otherwise, a portion of the game will be clearly kind of pointless. I guess the extra point is already a pointless formality 99.5% of the time. So a team taking a knee, another common maneuver, will be totally fine. But, it would be interesting to ponder a rule like: for a team leading by 3 before the extra point, attempting the point will remove 'x' amount of time from the clock (only applies in the final 3 minutes, and probably needs to be pro-rated or a percentage of the clock).

Another thought: I think kicking the extra point or not actually depends on who the opposing quarterback is. If the opposing team has a great defense and Ryan Lindley, I think I'll take the extra point. Otherwise, I could get burned by some non-offensive special teams fluke, either in OT or regulation. If they're so-so on D but have Aaron Rodgers, I think I might consider tempting them into a field goal by missing the extra point, and then potentially taking another crack in overtime.

Of course, probably the first thing the league needs to do, if it really cared about being fair and adaptable in ways that greatly affected human lives, would be to address the head trauma and PED scandals that seem to intertwine enough to warrant serious soul-searching. You've advocated for HGH testing and the importance of NFL drug testing, which I applaud you for. So I only include this paragraph because even though I love your point, I think on some moral level it's important to not lose sight of the more marginalized but important issues that the league goes to great lengths to deny and cover-up.

comoneon wrote:

If the other team has a better chance of winning by going for a touchdown instead of a field goal, why wouldn't they do that regardless of 3/4 points?

Because calculating chance, in any endeavor, is perilous, and therefore humans make decisions in biased manners. An example of bias is 'risk aversion', which may apply to your particular question. Faced with 4th and 3 from the 25 yard line, with 36 seconds to go, down by 3, with the clock ticking and no timeouts after a pass across the middle, many teams are rushing the special teams unit out for the field goal. The advanced stats may say, statistically, 'go for it', but for now, most coaching brains say, "This could go either way, my brain isn't smart enough to calculate odds, and I have about 6 seconds to decide, and, oh yea, I want to avoid losing". Hence, risk aversion wins out. It's such a common tactic that we forget that there's even a choice. That's why the advanced stats are interesting. They're not a panacea. As soon as a team starts intentionally missing PATs, then other teams are going to adjust, have real-time advanced stats involved in crunch-time decisions, etc. It'll be like micro trading on wall street, or whatever it's called. Where they trade using computers by the nanosecond.

And yet, with some teams, they can't even make seemingly basic time management decisions correctly. Compared with the pressure of the decision I present above, more common mistakes pale in comparison. Today, in the world we're living in, most coaches in the above situation who choose to go for it, and fail, would face extreme criticism and a likely firing. Maaaybe Belichick wouldn't be called in for a talk with the owner (though I doubt even he would do it). I think every other coach might have a significant portion of the fan base and media calling for their firing. What would save them? The advanced stat. But the advanced stat is still, as Rojo says, a few years away from being accepted by enough people to hold popular sway (another decision bias). Fascinating!!

I just can't believe that every team isn't fighting to hire rojo. Of course, rojo has calculated the odds that show he is better off just lounging around on his trust fund. There is zero probability that rojo would work at an NFL job or any other job.

Super Massive Genius wrote:

There is zero probability that rojo would work at an NFL job or any other job.

Rojo's last job was head coach of a D1 cross team. What was your last job?

rojo wrote:

Sure enough, I was right.

Using the win probability calculator here:

http://wp.advancedfootballanalytics.com/winprobcalc1.php

the Cowboys had a 74% chance of winning the game up 4 with 2:32 left. I f they'd been up just 3, their win percentage would have been 82%.

they don't say exactly how they arrive at those figures, so it's just hocus-pocus.

Let F be the probability that they score a field goal when down 3, T3 be the probability of a TD in that situation, and T4 the probability of a TD when down 4.

Assuming they can't score more than once, i.e. no overlap between the outcomes of TD and FG, the probability of victory down 3 is the probability of no score + half the probability of a score (i.e. ovetime win),

1 - (F + T3) + (F + T3)/2 = 1 - (F + T3)/2

and when down 4 is

1 - T4

Substituting the results from that magic calculator,

1 - T4 = .74

T4 = .26

1- (F + T3)/2 = .82

F + T3 = .36

Indicating a 36% chance of a score on the drive. It is safe to assume a TD is at most as likely as a FG in that scenario, and even at that maximum likelihood, the expected score per drive is only 7x.18 + 3x.18 = 1.80. But most NFL teams scored more than that this year even during 3-down-and-punt regulation, from an average field position near their own 30:

Detroit was not one of them, but would certainly have a higher points per drive in 4 down territory than their overall 1.76. More realistically it would be 26 percent FG and 10 percent TD for a 1.54. Whatever that website's secret formula is, they seem to greatly underestimate the likelihood of last minute FG's.

Rojo's overlooking the fact that Dallas went for it on a couple fourth downs on the final scoring drive when they had a chance to tie the score with a field goal. Maybe Detroit manipulated the whole scenario so they'd be up three instead of six so Dallas wouldn't try for the touchdown. Then look what happened, they got burned...damn you cybermetrics!!!!

Bad Wigins wrote:

Let F be the probability that they score a field goal when down 3, T3 be the probability of a TD in that situation, and T4 the probability of a TD when down 4.

Assuming they can't score more than once, i.e. no overlap between the outcomes of TD and FG, the probability of victory down 3 is the probability of no score + half the probability of a score (i.e. ovetime win),

1 - (F + T3) + (F + T3)/2 = 1 - (F + T3)/2

and when down 4 is

1 - T4

What you've done isn't correct.

Even keeping your assumption that overtime means a 50% probability of winning, the expression for the probability of winning when down 3 would be T3 + F/2. And when down 4, it's T4.

1-T4 is probability of LOSING. 1-(F+T3)/2 is not meaningful. For one thing, you've assumed that a touchdown could lead to overtime (since you divided T3 by 2), but it wouldn't. Touchdown = victory whether down 3 or 4. And then you subtracted from 1 which again shows the probability of NOT scoring, and therefore of losing.

But even beyond that, you've made an assumption that this is the last possession of the game. You can't assume that. There is a non-zero probability that a T3 or T4 could be scored quickly, then then the other team will get time to work back down the field. Or they'll run back a kick.

Not to mention that your assumption that overtime is a 50% chance of winning - I don't think it is. I would think the home team would have a greater chance the longer the game goes on.

Your crazy point would be stronger had the Lions scored a TD on their final drive.

Or maybe it would be stronger if you showed stats from actual results instead of from a calculator.

But let's say they did miss that extra point.

The Lions drove to the Cowboys 42 where they faced a 4th and 3.

Down 4 they had to go for it. And fumbled, ending the game.

Down 3 they would have had a decision to make.

Try for a 59 yard field goal or go for it.

Prater has kicked a 64 yarder before (longest in NFL history)

Maybe Prater would have made a playoff record and hit the field goal, tying the game. And the Lions may have gone on to win in OT.

Sure, miss that extra point on purpose.

Here's a question, why would anyone be a Cowboy fan?

It is widely known that the Dallas Cowboys are D-bags, and therefore, all Cowboy fans are D-bags by association.

“Worldly wisdom teaches that it is better for reputation to fail conventionally than to succeed unconventionally.” John Maynard Keynes

"In no other branch of mathematics is it so easy for experts to blunder as in probability theory" - Martin Gardner

Rojo isn't anywhere near an expert in probability.

Even if the results of the calculator that Rojo's got the numbers from were actually true, his conclusion was faulty. Here's why.

i presume the calculator that Rojo used analyzed hundreds of NFL late fourth-quarter games and compared the finish results of teams with either a 4 point or 3 point lead. If his statistics were accurate and repeatable, then a bookie (perhaps Rojo himself) could give odds and make a profit over the long run. The law of large numbers.

The problem here is that Rojo hasn't correctly framed the issue here. Should a coach in one game (not hundreds) forgo an extra point with a three point lead because it will improve the likelihood of winning for extremely likely (74% for a 4 point lead) to extremely likely (82% with a 3 point lead)?

The coach isn't playing in a theoretical world where he can push the button on the computer simulator hundreds of times. He is in the real world with 1 trial opportunity. If the team wins they go on to play another day. If it loses, they go home for the winter, He doesn't get to repeat games over and over until the variance is squeezed out of the results and he gets the result that he wants.

One shot. Extremely likely to win with a 4 point lead. Extremely likely to win with a 3 point lead.

Kick the extra point. Take the 4 points and force the opposing team to score a touchdown. They have one chance to beat you, not a hundred.

-------------------------------------------------------------

This is not the same as having a 2 point lead with 1:30 to go and your opponent is 1st and goal inside the 10. In that case you have almost zero chance of winning the game since your opponent can run down the clock and kick the sure FG. Your only chance is to give up a TD right away, preserving time on the clock, so that you have a chance at winning the game with a TD of your own. (see Superbowl XXII) That strategy gets foiled if the offense recognizes what's happening and the running back takes a knee rather then running into the endzone. Which is what DeMarcus Lawrence should have done immediately after his 4th quarter fumble recovery in yesterday's game.