It's like he's trying to have us lol suzzer

With two TOs and the 2-minute warning down by one score? Sure, depending on how I felt about my defense. If I was playing Marty Shottenheimer in the playoffs I'd do it 100% of the time. His record has to be perfect going 3-and-out in those spots.

If you kickoff you need a 3-and-out to get the ball back around midfield with ~2 min. and no TOs (unless they throw). Onside kick and you have a 90% chance of needing a 3-and-out to get the ball back around the 10 with ~2 min and no TOs (where they will be way less likely to throw). Maybe I'm way off but I don't think that's better. Maybe someone can do the math.

yeah i think with 2 TOs and the 2 minute warning that's a reasonable position - it has to at least be close either way, and probably situation dependent (i.e. if they have a good offense or you have a rotten defense, on-sides is the play, but if not, kicking it away is the play)

I agree with this. If you have enough clock stoppages to stop it after each down of a 3 and out it's usually better to kick (there are some variables including just how good an onside kick you have, how well you expect to cover a normal kick, how likely you are to get a 3 and out, etc but usually).

Yeah and in that spot, you have to assume you're going to make your 2pt. conversion (whenever you take it), since otherwise you lose. So it's better to take it later and keep more pressure on their offense.

Seems like a perfect time for a surprise onside, then. Expected onsides are 20% to succeed, but surprises are 60% to succeed. Line up to kick off, then surprise them (if they don't have the hands team out).

Surprise or not, I'd still onside most of the time. All the opponent needs is a single first down to seal the game. Here's a 3-and-out chart from last season. Even the worst team gets a first down on 70% of their possessions. However, it'd be less in this situation, because teams will be more conservative than usual.

http://espn.go.com/blog/afceast/post...he-percentages

Expected onsides are 10%. Not sure about surprise. But that would be a semi-surprise at best. They would certainly be prepared for both.

gus: an onside kick would not be a surprise there, they are most likely going to have a hands team out.

If we use the AFNFL calculator, kicking off gives us a WP of 0.02 (assuming a touchback, which it usually is with the hands team). If we onside kick and get it, which is ~20% of the time (albeit under the old rules), our win probability is .13. If they recover it, our win probability is 0.01. Therefore, the EV of onsiding is 0.2 * 0.13 + 0.8 + 0.01, or 0.034. That's an almost 70% improved chance of winning, although I'd imagine there's a fairly substantial margin of error with those calculations. Still though, it just doesn't make much intuitive sense to trade ~30 yards of field position for an almost 20% chance of getting the ball.

Edit: As I see there's some discrepancies between onsides data, ANFLS has it at 20% for non surprise onsides.

http://www.advancednflstats.com/2009...ide-kicks.html

Wow ok I just looked it up. I still have a hard time believing 20% success rate for expected onsides kick. Does that feel right to anyone?

(lol the third link for "what are the odds of recovering an onsides kick" is 2p2)

20% seems very high. If that's correct most teams should onside it every time vs some of the super premium offenses in the league.

Also, the 70% no 3-out number means next to nothing in that situation. It's not a slightly different situation than the typical drive it's a completely different situation.

i'm surprised looking at anfls that they don't particularly care about harbaugh not taking the 15 yards after the akers FG try

The calculator doesn't seem to take into account TOs. This is a gigantic factor in this decision.

Also it would be nice to get more precise than .01 and .02. The margin of error on those could almost flip the equation if it's actually .095 and .024.

how often do teams turn it over though in that situation? i remember one time McNabb i believe threw a pick 6 to Newman, but the majority of the time they are gonna go run run run punt and on a rare occasion run run pass but they dont throw a pick much there imo and they hardly fumble.

The pass saves a TO (or a big chunk of time) if it's incomplete. A lot of coaches like to take the bold move there to end the game, especially when the alternative is a conservative run and most likely punting to them at midfield with 2 min left.

ahhhh TOs = timeouts

i thought that was Turn Over

disregard

I agree w/ suzzer when down 9, but think its a slam dunk to go for 2 when you score to go down 8. Only question there is whether coaches know to do this and don't for job security reasons or if they are all just ignorant.

I still think people are underestimating the number of times the opponent will 3 and out.

Just 2 weeks ago, the Broncos were down by 3, had 2 timeouts with 3:43 left, and did a full kick off. The Raiders got 3 first downs. And the Raiders aren't exactly Tom Brady's Patriots.

http://www.nfl.com/gamecenter/201109...yze=playbyplay

That's just one game, though. Has anyone done a study of how often a team that only needs a first down to seal the game actually gets the first down?

joe girardi

why? jw the reasoning.

A lot of my stupid friends firmly believe in the "keep the players motivated" argument. Like if they fail to get the two, the team will be so devastated that they can't play to their full capability. And of course that outweighs the fact that it gives you a better chance to win the game.

edit: oops i was referring to a different situation, my bad

yeota,

conversion rate is 45 percent

make it first time, kick xp second time (win): .45 x 1 = 45%
miss it first time, make it second time (tie): .55 x .45 = 25%
miss it first time, miss it second time (lose): .55 x .55 = 30%

It is closer to 10% since the rule changed so that you can't stack one side of the ball. Or so says wikipedia. The 20% figure has a larger sample size but is over both rulesets.

thanks

definitely something i wouldnt have thought of.

The coach at the D3 school I go to actually did that late in the game on Saturday, but I assume it was more due to lack of confidence in the kicker or because he wanted to run another play on offense rather than any kind of math/logic thing.