What's new
Fantasy Football - Footballguys Forums

Welcome to Our Forums. Once you've registered and logged in, you're primed to talk football, among other topics, with the sharpest and most experienced fantasy players on the internet.

A statistical analysis of Mangini's decision (1 Viewer)

Chase Stuart

Footballguy
Should Mangini have gone for it on 4th and goal from the 2? The chance of getting a TD on 4th-and-goal from the 2 is about 43%. The likelihood of getting a FG from there is about 94%. Let's break it down.

Kicking the FG gives the Jets two likely scenarios: a miss or a hit. A field goal is worth three real points, but then the Colts would get the ball at around the 30 yard line following the kickoff (the Colts next kick return gave them the ball at the 32, IIRC). A first and 10 from the 30 is worth roughly 0.7 points, so the value of the made FG is 2.3 points. A miss is worth zero points, and would give Indianapolis the ball at the 9 yard line -- which is worth about -0.3 points for the Colts, or +0.3 points for the Jets. So kicking a FG gives the Jets an expected value of 2.27 points (.94 * 2.4 <plus> .06 * 0.3).

If the Jets score a TD, that's worth 6.24 points (6.00 <plus> .94 * 1 <plus> -0.7). If the Jets don't score a TD, the Colts take over at the 2 yard line, a value of about -1.4 to Indianapolis, or +1.4 to the Jets. For the choice of kicking a FG to be equivalent to going for the score, the odds of converting the fourth down would have to be just 18%. Considering the chance of converting is about 40-45%, there's a pretty safe margin of error that Mangini made the correct choice.

There are a couple of other issues to address. Going for the highest expected value play isn't always the best move; if the Jets were down by 2 with 0:01 to go, then kicking the FG is the obvious choice. But in the third quarter in a tie game, the +EV play is almost always going to be the correct decision. Further, considering the Colts have the highest scoring offense in the NFL, an argument could be made that a riskier decision with a higher reward -- expected value be damned -- should be the correct decision. That's not the case here since the +EV dictated going for it, but it would provide some support for going for it from the five yard line instead of the two. The Colts ended up scoring 2 TDs in 5 minutes late in the game, which makes me give a bit more credibility to the idea that you've got to put 7 on the board against Indianapolis when you can. The Jets have also had a fourth quarter defensive breakdown in every game this year, something that certainly was (and should have been) on Mangini's mind.

I did ignore one possibility when you go for it -- the one that happenned. But the odds of an interception in the end zone probably isn't much higher than the odds of a blocked kick returned for a TD, which would be much worse. If you added both possibilities in, the EV would probably tilt even more towards going for it. It was a good decision that had a poor result. In retrospect it hurt the Jets, but only because it failed. Vermeil's decision to go for the TD last year with just a few seconds on the clock when down by 3 was also the right decision; but if it had failed, it would have hurt the Chiefs. I view this on the same level.

(P.S. Before jwvdcw comes in and mentions the Cedric Houston injury, Kevan Barlow is the Jets goal-line RB (4 TDs this year), so that shouldn't factor into the analysis.)

 
He was outgunned, and had a shot at a TD versus a better team. Thumbs up to Mangini for going for it.
That's in line with my idea that perhaps going with the option that produces greater variance but less expected return is a good idea. I think there's some support for this, but still more on a straight EV analysis.
 
Chase, one more thing you didn't mention.... Nugent. He barely got his first PAT inside the right upright. His second was down the middle, and that was before the Mangini decision to go for it on 4th down. Nugent is struggling, and I don't think Mangini has much confidence in him at this point, and for good reason. Imagine if Nugent had missed there? I think that also played a role in Mangini's decision.... a decision I never second guessed, I liked the call.

 
I'm with ya - I've been debating this all week on a Jets board....

The Parcells halfback option call VS Detroit in 97 was a comparison brought up and what Parcells said was that when he feels he's outmanned and NEEDS help he calls on these plays - In 1997 the Jets OL was decimated and the offense had trouble getting out of it's own way - he called a highly risky play and it failed - The guy also turned a 1-15 team into a 9-7 team by making Ballsy calls all year and making more than not - You take the whole package.... I'm on record as being totally fine with that Call just like I am her today....

You take that onsides kick and the fact that Mangini and Co. PLANNED for situations like this and PLANNED to go for it in situations like this over the last group of Bozos that were here that would have run around the sidelines with no clue to what to do, probably call a TO and find a Magic 8 ball to answer the riddle.

If you want the Herm Edwards offense that the defense and every fan in the stands know what's coming, then fine.

Good coaches know when to deviate from the book and take a chance..... I think there's a great chance Parcells makes that same exact call.

WOULD I GO FOR IT - Probably not, but, I totally see where Mangini is coming from and it's just a relief that he actually THINKS and PLANS for this stuff and has the balls to do it.

I doubt Mangini makes that call again - And I just know some morons are going to be like "Hey Mangini 'learned'" when he doesn't but, the truth is The Jets don't face Manning and the Colts again this year and that's the key to that decision.

 
I'm not a fan of taking points off the board in the third quarter - you don't know how th rest of the game will play out. But, a 4th and goal at the 2 is a bit different to me. Even if you fail, you should get the benefit of great field position. That's no small benefit - field position is KEY in the NFL. Its not uncommon to force a 3 and out and then get the back in great field position again on the next series.

The fact that there was an interception, touchback, and then about an 80 yard drive by the Colts was all about poor execution by the players. The coach did the right thing, imo.

 
Nice analysis Chase.

Not really the same arena, but Madden made a simple, but good point on Sunday Night Football. He said something to the effect of "The same people who wouldn't have gone for it on 4th and goal and kicked the field goal, probably wouldn't have attempted the onside kick in the first half that lead to a touchdown. So, by my score, Mangini is ahead over those guys 7-3."

 
Nice analysis Chase. Not really the same arena, but Madden made a simple, but good point on Sunday Night Football. He said something to the effect of "The same people who wouldn't have gone for it on 4th and goal and kicked the field goal, probably wouldn't have attempted the onside kick in the first half that lead to a touchdown. So, by my score, Mangini is ahead over those guys 7-3."
Yeah, Except "those Guy's" don't want to accept that and they pretend the onside kick had a lot less risk...Thing is, if the Kick fails and Manning tosses the next ball from his own 35 for a TD and continues to beat on the Jets, we're all here talking about how Mangini KILLED the Jets with a stupid play call.There was also another 4th down play they went for And MADE!!!
 
Chase Stuart said:
massraider said:
He was outgunned, and had a shot at a TD versus a better team. Thumbs up to Mangini for going for it.
That's in line with my idea that perhaps going with the option that produces greater variance but less expected return is a good idea. I think there's some support for this, but still more on a straight EV analysis.
It's also worth noting that Pennington had never thrown an INT in 119 career attempts inside the opponent's 20 yard line. During the game, the announcers made the point that Pennington had to loft it up there and give his TE a chance at the ball. I disagree. James Dearth was the intended receiver - the long snapper who doesn't actually play TE in normal game situations. The Jets were specifically trying to catch the Colts unaware (Mangini compared the play to when the PAts used Vrabel to catch TDs). Dearth is obviously not an athletic, skilled receiver who was going to go up and battle two or three defenders for the ball. Pennington would have been wiser to either attempt to run it in or throw it away, since the INT freed the Colts from being pinnned against the goaline.
 
Chase Stuart said:
Reaper said:
WOULD I GO FOR IT - Probably not
Can you explain why in light of the original post? Curious to see your reasoning.
Chad Pennington has played for (?) years in the NFL and holds an impressive stat. He has never thrown an INT in the redzone. He threw an INT. That's never happenned before so amidst the gameplanning this had to be considerred an extremely unlikely result.They ran the previous 3 plays IIRC and seemed to run into a wall with no hole developping and really no hope for the running game at that instance. IF they were to go for it, they "had to" do something different like pass.As I mentioned a few times, I'd have put Brad Smith in as they did earlier in the game
 
PahtyTom said:
Nice analysis Chase. Not really the same arena, but Madden made a simple, but good point on Sunday Night Football. He said something to the effect of "The same people who wouldn't have gone for it on 4th and goal and kicked the field goal, probably wouldn't have attempted the onside kick in the first half that lead to a touchdown. So, by my score, Mangini is ahead over those guys 7-3."
Yes, and didn't the Jets also convert on fourth and one earlier in the game? "Those same people" wouldn't have done that, either.There's a common habit of viewing successful gambles as smart decisions and failed gambles as stupid decisions, but that's faulty thinking based on hindsight.IMO, all of those gambles were good ones even though only two out of three worked.
 
Chase Stuart said:
Reaper said:
WOULD I GO FOR IT - Probably not
Can you explain why in light of the original post? Curious to see your reasoning.
Because "By the Book" you probably get the points.. and I could go either way and I'm guessing that under pressure on the field, and the fact that the momentum just stopped as they carted my RB off the field I regroup and kick the FG......Would have been nice to lineup and just rip off a play. BUT, It's a borderline call - ONLY because the opponent is Manning. So, I probably don't have the balls Mangini did but, either way, if he saw something on tape and it sounds like they did and they actually planned for such situations then more power to them.I just see no reason to kill the coach and why it's a big deal that he went for it and I think it signals a great change in NY Jets and is a Wonderful departure from what we're used to.... In the big picture I love the call and the fact that we now have a coach that can't necessarily be out-guessed again.Do I make that Half-Back option call Parcells made in 1997 - NO! But, again, I recognized what we had in a coach and an attitude that players appreciate that The NY Jets ARE GOING FOR IT.....And again, I doubt this call is made against any other QB.
 
The only flaw that I see in your analysis is that the calculations of expected value become more of a decision tree than an isolated calculation of expected values. Furthermore, the coach is focused on winning/losing rather than necessarily maximizing the expected value on a particular play.

Meanining--what are the probabilities that a team will win/lose the game, from that point in the game, assuming they were either 1) tied, 2) +3 points, or 3) +7 points. You could obviously break this down into finer increments (+6 if they missed the PAT for instance), but for simplicity I chose these 3 options.

If there is 0:00 left in the game the probability of winning, if you have a 3-point lead, is 100%. The probability of winning with a 3-point lead will decline the further in time that you get away from the end of the game. At some point in game time the probability of a win based on kicking the FG will fall below the probability of a win based on going for the TD. That is, you kick a field goal with 0:01 on the clock, not for the expected value, but because the probability of a win is 94% versus going for the TD, in your example, which would give you a probability of winning of only 43%. With 15:00 on the clock, my guess is that the probabilities favor going for the TD, but I don't know for sure.

The problem that you get into with the expected value analyses is that it doesn't factor in game situations as well as the probability of winning analysis. That is, a team with a 3-point lead can react to that lead (by running the ball and killing the clock, for instance) that would further increase their probability of winning. That factor isn't captured by the expected values of plays based on down, distance and field position.

 
Chase Stuart said:
Reaper said:
WOULD I GO FOR IT - Probably not
Can you explain why in light of the original post? Curious to see your reasoning.
Because most fans and apparently GM's dont know a thing about EV. The safe play is the one you call. No chance of being the goat, because you did "what you should have."I am not saying going for it isn't the right call but much like a CEO, the safest thing to do is going to keep you around the longest.If he had kicked the FG most fans would say "I would have liked to see him go for it, but the FG was the 'right' call there." Even if they lost.Great analysis, just adding that popular opinion, no matter how ignorant, is part of the decision making process. I know it shouldn't be, but when your job is on the line...
 
The only flaw that I see in your analysis is that the calculations of expected value become more of a decision tree than an isolated calculation of expected values. Furthermore, the coach is focused on winning/losing rather than necessarily maximizing the expected value on a particular play.

Meanining--what are the probabilities that a team will win/lose the game, from that point in the game, assuming they were either 1) tied, 2) +3 points, or 3) +7 points. You could obviously break this down into finer increments (+6 if they missed the PAT for instance), but for simplicity I chose these 3 options.

If there is 0:00 left in the game the probability of winning, if you have a 3-point lead, is 100%. The probability of winning with a 3-point lead will decline the further in time that you get away from the end of the game. At some point in game time the probability of a win based on kicking the FG will fall below the probability of a win based on going for the TD. That is, you kick a field goal with 0:01 on the clock, not for the expected value, but because the probability of a win is 94% versus going for the TD, in your example, which would give you a probability of winning of only 43%. With 15:00 on the clock, my guess is that the probabilities favor going for the TD, but I don't know for sure.

The problem that you get into with the expected value analyses is that it doesn't factor in game situations as well as the probability of winning analysis. That is, a team with a 3-point lead can react to that lead (by running the ball and killing the clock, for instance) that would further increase their probability of winning. That factor isn't captured by the expected values of plays based on down, distance and field position.
I think you're right, but I addressed this in the OP. Anytime before the 4th quarter, the EV play is usually the right play. That applies a fortiori when you have had a defensive breakdown every 4th quarter to date, and when you play the highest scoring team in the NFL.
 
Interesting (to me) aside:

How interesting would the NFL be if you could only punt, say, 7 times a game... and every 4th down after the 7th punt you had to go for it?

 
Chase Stuart said:
Reaper said:
WOULD I GO FOR IT - Probably not
Can you explain why in light of the original post? Curious to see your reasoning.
Because most fans and apparently GM's dont know a thing about EV. The safe play is the one you call. No chance of being the goat, because you did "what you should have."I am not saying going for it isn't the right call but much like a CEO, the safest thing to do is going to keep you around the longest.If he had kicked the FG most fans would say "I would have liked to see him go for it, but the FG was the 'right' call there." Even if they lost.Great analysis, just adding that popular opinion, no matter how ignorant, is part of the decision making process. I know it shouldn't be, but when your job is on the line...
No, you're right on -- and that's what made me happy to see Mangini go against the grain. I think he's pretty insulated from fan criticism right now, and his job won't be on the line for awhile. I hope he continues to make the intelligent, if unpopular, decisions. That in the end will lead to the one thing that makes fans the happiest.
 
Chase Stuart said:
Reaper said:
WOULD I GO FOR IT - Probably not
Can you explain why in light of the original post? Curious to see your reasoning.
Because "By the Book" you probably get the points.. and I could go either way and I'm guessing that under pressure on the field, and the fact that the momentum just stopped as they carted my RB off the field I regroup and kick the FG......Would have been nice to lineup and just rip off a play. BUT, It's a borderline call - ONLY because the opponent is Manning. So, I probably don't have the balls Mangini did but, either way, if he saw something on tape and it sounds like they did and they actually planned for such situations then more power to them.I just see no reason to kill the coach and why it's a big deal that he went for it and I think it signals a great change in NY Jets and is a Wonderful departure from what we're used to.... In the big picture I love the call and the fact that we now have a coach that can't necessarily be out-guessed again.Do I make that Half-Back option call Parcells made in 1997 - NO! But, again, I recognized what we had in a coach and an attitude that players appreciate that The NY Jets ARE GOING FOR IT.....And again, I doubt this call is made against any other QB.
I'm just saying that "by the book", it's clearly better to go for it. I don't see what the argument is against that, unless you think the Jets have a <18% chance of converting from there. There's very little risk and a lot of reward. :shrug:
 
Chase Stuart said:
Reaper said:
WOULD I GO FOR IT - Probably not
Can you explain why in light of the original post? Curious to see your reasoning.
Because most fans and apparently GM's dont know a thing about EV. The safe play is the one you call. No chance of being the goat, because you did "what you should have."I am not saying going for it isn't the right call but much like a CEO, the safest thing to do is going to keep you around the longest.If he had kicked the FG most fans would say "I would have liked to see him go for it, but the FG was the 'right' call there." Even if they lost.Great analysis, just adding that popular opinion, no matter how ignorant, is part of the decision making process. I know it shouldn't be, but when your job is on the line...
No, you're right on -- and that's what made me happy to see Mangini go against the grain. I think he's pretty insulated from fan criticism right now, and his job won't be on the line for awhile. I hope he continues to make the intelligent, if unpopular, decisions. That in the end will lead to the one thing that makes fans the happiest.
It is ironic that the instability of the position in many cases prevents a coach from making the right call, thus limiting his ability to win games, which is what everyone really wants.
 
I loved the call. And the guts to do it. Against the best O in the league on the 2 yard line. Loved it.

I didn't love the actual play tho...

But I am glad he had balls to try it. We've been missing that since parcells left.

 
As an aside, for those of you who haven't already seen it, this article is pretty interesting: It's Fourth Down and What Does the Bellman Equation Say? A Dynamic Programming Analysis of Football Strategy, by David Romer.
The cliff note of the cliff note version of this link says "coaches are too conservative compared to what the numbers say they should do to maximize their teams odds of winning".Not going for TD's inside the 5 yard line and punting when facing a 4th and less than 10 anywhere inside the opponents territory are two prime examples. A 3rd would be Martyball. :bye:

 
As a Jet fan trapped in LA, who didn't see the game I am PLEASED that it happened. It may be the riskier choice, but sometimes you have to risk a letdown or failure if you wanna play with the big boys.

Go Mangini, GO!

 
As a Jet fan trapped in LA, who didn't see the game I am PLEASED that it happened. It may be the riskier choice, but sometimes you have to risk a letdown or failure if you wanna play with the big boys.Go Mangini, GO!
That's my point though; it shouldn't be lauded as the riskier choice: it should be praised as the right choice.Choosing a 40% chance at $1000 is a better option than a free $200. I wouldn't want my decision to go for the $1000 to be labeled as the risky choice, although I could see how it might be.
 
Choosing a 40% chance at $1000 is a better option than a free $200. I wouldn't want my decision to go for the $1000 to be labeled as the risky choice, although I could see how it might be.
I think that's the definition of a risky choice. I agree that it's the correct choice, but obviously you're taking on additional risk to increase your overall expectation of return.
 
I just think it was hilarious when they decided to go for it, that the idiot announcer... forget who it was now... said: "well this shows that Mangini has trust in his defense and that they can shut 'em down even if they don't convert"

yeah... that's it. ...Or that the Colts are likely to score 14+ more points and that the Jets won't help themselves with only 3 points.

 
I just think it was hilarious when they decided to go for it, that the idiot announcer... forget who it was now... said: "well this shows that Mangini has trust in his defense and that they can shut 'em down even if they don't convert"yeah... that's it. ...Or that the Colts are likely to score 14+ more points and that the Jets won't help themselves with only 3 points.
that reminds me... did anyone hear Thiessman on MNF try to explain why the Eagles WR (might have been Moats) was considered "not down" when Ahmad Carroll went to touch him down near the sideline?The player on the Eagles caught a ball short, near the sideline, and he fell down in the process... Carroll then argued to the ref while going to touch him "down" ....meanwhile the offensive player got up and ran another couple steps before being tackled.Thiesman, before seing the replay said that the offensive player was not considered down at first because "you can't just touch a man down... it's not like two-hand touch"(really he just didn't get whistled down because the guy got up before being touched by Carroll)I was in shock that a man who started as many NFL games as Theismann did, didn't actually know the rules.sorry, just thought this was utterly ridiculous at the time :bag:
 
Choosing a 40% chance at $1000 is a better option than a free $200. I wouldn't want my decision to go for the $1000 to be labeled as the risky choice, although I could see how it might be.
I think that's the definition of a risky choice. I agree that it's the correct choice, but obviously you're taking on additional risk to increase your overall expectation of return.
It's just semantics, but I would say the definition of a risky choice is choosing a 40% chance at $1000 over a free $400. Equivalent in the long run, but for one roll you're choosing to play for the risk. Playing it safe would be taking the $400, but you're playing for the variance.But when you choose the +EV play, risky isn't what comes to mind.Semantics, though. :shrug:
 
Choosing a 40% chance at $1000 is a better option than a free $200. I wouldn't want my decision to go for the $1000 to be labeled as the risky choice, although I could see how it might be.
I think that's the definition of a risky choice. I agree that it's the correct choice, but obviously you're taking on additional risk to increase your overall expectation of return.
It's just semantics, but I would say the definition of a risky choice is choosing a 40% chance at $1000 over a free $400. Equivalent in the long run, but for one roll you're choosing to play for the risk. Playing it safe would be taking the $400, but you're playing for the variance.But when you choose the +EV play, risky isn't what comes to mind.Semantics, though. :shrug:
We should probably talk about "variance" rather than "risk" to make things clearer. Choosing a 40% chance of $1000 has a higher variance than a sure $200, but a higher expected value as well."Risk" is kind of unclear in this context because the risk in both cases (going for it or kicking a field goal) is the same: the risk is that you'll lose the game. The question is which course of action will result in a lesser chance of losing the game. It may be the course of action whose variance is higher or it may be the course of action whose variance is lower. So I don't think "risk" is a good synonym for "variance."
 
Choosing a 40% chance at $1000 is a better option than a free $200. I wouldn't want my decision to go for the $1000 to be labeled as the risky choice, although I could see how it might be.
I think that's the definition of a risky choice. I agree that it's the correct choice, but obviously you're taking on additional risk to increase your overall expectation of return.
It's just semantics, but I would say the definition of a risky choice is choosing a 40% chance at $1000 over a free $400. Equivalent in the long run, but for one roll you're choosing to play for the risk. Playing it safe would be taking the $400, but you're playing for the variance.But when you choose the +EV play, risky isn't what comes to mind.Semantics, though. :shrug:
We should probably talk about "variance" rather than "risk" to make things clearer. Choosing a 40% chance of $1000 has a higher variance than a sure $200, but a higher expected value as well."Risk" is kind of unclear in this context because the risk in both cases (going for it or kicking a field goal) is the same: the risk is that you'll lose the game. The question is which course of action will result in a lesser chance of losing the game. It may be the course of action whose variance is higher or it may be the course of action whose variance is lower. So I don't think "risk" is a good synonym for "variance."
Doesn't this make the riskier decision always the negative EV decision? By that definition, I would think any decision that has a negative EV relative to another -- in this case, kicking the FG instead of going for the TD -- would be the course of action that will result in a greater chance of losing the game. So by one definition, kicking the FG is riskier, because you increase your chances that you're going to lose, since you're making a stupid decision.Is that a fair interpretation of what you said?
 
Doesn't this make the riskier decision always the negative EV decision? By that definition, I would think any decision that has a negative EV relative to another -- in this case, kicking the FG instead of going for the TD -- would be the course of action that will result in a greater chance of losing the game. So by one definition, kicking the FG is riskier, because you increase your chances that you're going to lose, since you're making a stupid decision.
The dictionary definition of "risk" is "exposure to loss." In the gambling scenario, a 100% chance at $200 exposes you to monetary loss less than a 40% chance at $1000 does. The 100% chance does expose you to loss of opportunity, and reduced EV, but I think by any reasonable definition, the 100% chance at $200 is less risky. Similarly, investing in a municipal bond is less risky than investing in the stock market, even though the expected return on stock market investments is probably higher in most scenarios. That's a fairly common way to talk about risk; a risky portfolio is one that has a high exposure to loss, in exchange for higher EV (maybe).In the football scenario, I'm not sure going for it on fourth down is the riskier choice, except in the sense that coaches who expose themselves to criticism (by making unconventional calls) are more likely to get fired quickly.
 
Choosing a 40% chance at $1000 is a better option than a free $200. I wouldn't want my decision to go for the $1000 to be labeled as the risky choice, although I could see how it might be.
I think that's the definition of a risky choice. I agree that it's the correct choice, but obviously you're taking on additional risk to increase your overall expectation of return.
It's just semantics, but I would say the definition of a risky choice is choosing a 40% chance at $1000 over a free $400. Equivalent in the long run, but for one roll you're choosing to play for the risk. Playing it safe would be taking the $400, but you're playing for the variance.But when you choose the +EV play, risky isn't what comes to mind.Semantics, though. :shrug:
We should probably talk about "variance" rather than "risk" to make things clearer. Choosing a 40% chance of $1000 has a higher variance than a sure $200, but a higher expected value as well."Risk" is kind of unclear in this context because the risk in both cases (going for it or kicking a field goal) is the same: the risk is that you'll lose the game. The question is which course of action will result in a lesser chance of losing the game. It may be the course of action whose variance is higher or it may be the course of action whose variance is lower. So I don't think "risk" is a good synonym for "variance."
Doesn't this make the riskier decision always the negative EV decision? By that definition, I would think any decision that has a negative EV relative to another -- in this case, kicking the FG instead of going for the TD -- would be the course of action that will result in a greater chance of losing the game. So by one definition, kicking the FG is riskier, because you increase your chances that you're going to lose, since you're making a stupid decision.Is that a fair interpretation of what you said?
What I was trying to say was that the word "risk" isn't very useful in this context. Going for it on fourth down is either smart or stupid, but it's not risky in any meaningful sense, IMO. The risk you take by going for it -- namely, that it will cause you to lose the game -- is the same risk you take by attempting a field goal instead. Going for it may produce a range of results that has a higher variance (using a Romer-type scoring system based on actual points as well as ball possession and field position), but variance and risk are not synonymous in this context, IMO.
 
Last edited by a moderator:
A few thoughts:

1) I didn't realize that play was in the 3rd quarter. I guess I hadn't been paying attention but the way it was replayed as a critical decision implied that it was later in the game. This is important because with so much time left the decision not just to take 3 points (as if that would decide the game) is a lot less questionable;

2) that play was at the end of a 18-play, 8.5 minute drive that virtually defined the third quarter of play and included 3 3rd down conversions already. That defense had to have been tired.

3) Wouldn't everyone agree that in all likelihood the Colts offense would fare better on average (i.e. more points) from the 30-yard line?

4) What's the risk of an INT being run back for an TD in that situation? For some reason it seems like more get run back from the defensive team's red zone than from the middle of the field.

5) How often do plays that are in "and-goal" situations result in defensive penalties that give the offense a first down?

Overall, I too like what Mangini did there. It was pretty clear that a 17-14 lead wasn't going to hold up against Manning and the Colts with around 20 minutes of game left.

 
Last edited by a moderator:
Should Mangini have gone for it on 4th and goal from the 2? The chance of getting a TD on 4th-and-goal from the 2 is about 43%. The likelihood of getting a FG from there is about 94%. Let's break it down.

Kicking the FG gives the Jets two likely scenarios: a miss or a hit. A field goal is worth three real points, but then the Colts would get the ball at around the 30 yard line following the kickoff (the Colts next kick return gave them the ball at the 32, IIRC). A first and 10 from the 30 is worth roughly 0.7 points, so the value of the made FG is 2.3 points. A miss is worth zero points, and would give Indianapolis the ball at the 9 yard line -- which is worth about -0.3 points for the Colts, or +0.3 points for the Jets. So kicking a FG gives the Jets an expected value of 2.27 points (.94 * 2.4 <plus> .06 * 0.3).

If the Jets score a TD, that's worth 6.24 points (6.00 <plus> .94 * 1 <plus> -0.7). If the Jets don't score a TD, the Colts take over at the 2 yard line, a value of about -1.4 to Indianapolis, or +1.4 to the Jets. For the choice of kicking a FG to be equivalent to going for the score, the odds of converting the fourth down would have to be just 18%. Considering the chance of converting is about 40-45%, there's a pretty safe margin of error that Mangini made the correct choice.

There are a couple of other issues to address. Going for the highest expected value play isn't always the best move; if the Jets were down by 2 with 0:01 to go, then kicking the FG is the obvious choice. But in the third quarter in a tie game, the +EV play is almost always going to be the correct decision. Further, considering the Colts have the highest scoring offense in the NFL, an argument could be made that a riskier decision with a higher reward -- expected value be damned -- should be the correct decision. That's not the case here since the +EV dictated going for it, but it would provide some support for going for it from the five yard line instead of the two. The Colts ended up scoring 2 TDs in 5 minutes late in the game, which makes me give a bit more credibility to the idea that you've got to put 7 on the board against Indianapolis when you can. The Jets have also had a fourth quarter defensive breakdown in every game this year, something that certainly was (and should have been) on Mangini's mind.

I did ignore one possibility when you go for it -- the one that happenned. But the odds of an interception in the end zone probably isn't much higher than the odds of a blocked kick returned for a TD, which would be much worse. If you added both possibilities in, the EV would probably tilt even more towards going for it. It was a good decision that had a poor result. In retrospect it hurt the Jets, but only because it failed. Vermeil's decision to go for the TD last year with just a few seconds on the clock when down by 3 was also the right decision; but if it had failed, it would have hurt the Chiefs. I view this on the same level.

(P.S. Before jwvdcw comes in and mentions the Cedric Houston injury, Kevan Barlow is the Jets goal-line RB (4 TDs this year), so that shouldn't factor into the analysis.)
I think this is very thoughtful analysis, however, I think it is flawed. The reason is, football situations are just that, situational. In poker, or in any random event, you can create a true stastistical anaylsis. A fair die will roll a 6 one in six times in the long run. So perhaps in the long run, a team will convert a two point converserion about 43% of the time. However, that average is not necessarily an accurate indicaator of success in this situation. I would guess (and it is only a guess) 4th and 2 in a game 28-6 with 15 seconds remaining, may not be as contested by the defense as a 4th and 2 in a 28-20 game with 15 seconds remaining. In the stastical grouping, all of these situations are lumped together.To further complicate the matter, the 2006 Jets in week 4 do not have a 43% of making a first down against the 2006 Colts. I am not sure what their percentage is, but I would assert that it is different. My guess, and again it is a guess, is that it would be slightly lower given some of their difficulties running the ball. However, you could argue that it was higher given that situation. In any case, there is not concrete number. The coach has to go by feel.

Another factor in the decision that is well outside the statistical data is the psychological impact the action could have. To make another poker analogy, sometimes I would make s slighty negative expected value play if I thought it could tilt my opponens. Sometimes I would avoid a slightly positive expected play if I thought that a negative result would either make my opponents play better or might lead to me playing worse. This again goes to feel.

I think Mangini made the right decision. I support your use of statistical analysis to defend the decision. I just don't see it nearly as rigid as you do.

 
I think this is very thoughtful analysis, however, I think it is flawed. The reason is, football situations are just that, situational. In poker, or in any random event, you can create a true stastistical anaylsis. A fair die will roll a 6 one in six times in the long run. So perhaps in the long run, a team will convert a two point converserion about 43% of the time. However, that average is not necessarily an accurate indicaator of success in this situation. I would guess (and it is only a guess) 4th and 2 in a game 28-6 with 15 seconds remaining, may not be as contested by the defense as a 4th and 2 in a 28-20 game with 15 seconds remaining. In the stastical grouping, all of these situations are lumped together.To further complicate the matter, the 2006 Jets in week 4 do not have a 43% of making a first down against the 2006 Colts. I am not sure what their percentage is, but I would assert that it is different. My guess, and again it is a guess, is that it would be slightly lower given some of their difficulties running the ball. However, you could argue that it was higher given that situation. In any case, there is not concrete number. The coach has to go by feel.Another factor in the decision that is well outside the statistical data is the psychological impact the action could have. To make another poker analogy, sometimes I would make s slighty negative expected value play if I thought it could tilt my opponens. Sometimes I would avoid a slightly positive expected play if I thought that a negative result would either make my opponents play better or might lead to me playing worse. This again goes to feel.I think Mangini made the right decision. I support your use of statistical analysis to defend the decision. I just don't see it nearly as rigid as you do.
Statistics are tools and like every tool they're "flawed" to the extent that they are not perfect, good or even sometimes relevant at all for addressing a given problem. To think otherwise is like saying a screwdriver is flawed because you can't sand with it. In each case ther are limitations, however both belong in your toolbox. Your overall criticism would be valid were Chase asserting that his analysis was the end-all and completely reflected everything about that game situation. I think he was looking for something a bit more concrete than a "hunch" to use as an indicator of the wisdom of the decision, or the lack thereof.
 
A few thoughts:1) I didn't realize that play was in the 3rd quarter. I guess I hadn't been paying attention but the way it was replayed as a critical decision implied that it was later in the game. This is important because with so much time left the decision not just to take 3 points (as if that would decide the game) is a lot less questionable; 2) that play was at the end of a 18-play, 8.5 minute drive that virtually defined the third quarter of play and included 3 3rd down conversions already. That defense had to have been tired. 3) Wouldn't everyone agree that in all likelihood the Colts offense would fare better on average (i.e. more points) from the 30-yard line? 4) What's the risk of an INT being run back for an TD in that situation? For some reason it seems like more get run back from the defensive team's red zone than from the middle of the field. 5) How often do plays that are in "and-goal" situations result in defensive penalties that give the offense a first down? Overall, I too like what Mangini did there. It was pretty clear that a 17-14 lead wasn't going to hold up against Manning and the Colts with around 20 minutes of game left.
I agree..but on a drive like that you "have" to come away with some points and the lead. The D had done pretty good prior to that..only 14 points heading into the fourth quarter. If you get the lead and make a stop..you have a chance to take more time off the clock and make it a 2 score game as well.
 

Users who are viewing this thread

Top