Pass defense in '08

[AdamJT13]

Do you agree, then, that defensive points allowed per play is a better measure of a defense than defensive points allowed per possession?

No, I don't. In that statistic, you'd essentially be penalized for forcing a three-and-out instead of giving up several first downs (which would boost your number of plays faced) before getting the ball back.

Say, for instance, Team A and Team B each face 10 possessions by their opponent, with every possession starting at the opponent's 25-yard line. Each of them allow one 10-play, 75-yard touchdown drive to start the game. Team A then forces a three-and-out the nine other times its opponent has the ball. Team B allows its opponents to drive downfield the nine other times, ending with three field goals, a missed field goal and five punts. Team A winds up allowing 120 yards and seven points on 37 plays. Team B allows 395 yards and 16 points on 88 plays.

All other factors being equal, Team A clearly played better on defense, but it would have a higher "points allowed per play." Had it allowed a couple of first downs and a few more plays instead of forcing three-and-outs, it would have "improved" in that statistic. And that's too big of a flaw.

 

[AdamJT13]

That's why Welker ate our lunch... Nate Jones was in coverage on him all day long.

Except when Greg Ellis was coveri... er ... trying to keep up with him.

 

[burmafrd]

3rd down conversion percentage is a very good stat. Thats what hurt us last year- we would give up the first on 3rd down.
Really need a new way to rate defenses. I have never thought the total yds measurement really told the tale (we were #1 in that in 2003 and we were FAR from the best D in the league).
To me you first look at the offenses that you faced and rate them. Then you look at everything else as against that stat. ST plays and things like INT returns for TDs against your offense should be removed. Look at field position when the offense starts; look at yds per rush and yds per pass; completion %; # of 20+ yd plays allowed; red zone percentage, 3rd down conversion %. Then give pts for each stat as regards how in the league it rates. Add it all up and like that and you would know whose D was really red meat and whose D was pasty bologna.

 

[AdamJT13]

3rd down conversion percentage is a very good stat.

Which defense would you have rather had last season -- the Colts' or the Chiefs'?

The Steelers' or the Falcons'?

The Buccaneers' or the Browns'?

 

[percyhoward]

No, I don't. In that statistic, you'd essentially be penalized for forcing a three-and-out instead of giving up several first downs (which would boost your number of plays faced) before getting the ball back.

I'm sure you (or somebody) will correct me if I'm wrong, but it seems like if you were facing more plays, that would boost your number of pass plays faced, artificially lowering your net yards per pass play as well.

Also, wouldn't factors unrelated to your defense also affect the per-possession number? Let's say your offense can't sustain long drives. That boosts number of possessions faced for your defense.

I'm dubious when it comes to the per possession stats, as opposed to per game stats. For example, if points per possession really is the best measure, then Seattle's defense is 1st, and Dallas' is 14th. Seattle's got a good D, but is it really that much better than ours?

 

[AdamJT13]

I'm sure you (or somebody) will correct me if I'm wrong, but it seems like if you were facing more plays, that would boost your number of pass plays faced, artificially lowering your net yards per pass play as well.

There's no indication that facing more pass plays naturally raises or lowers your average yards allowed per pass. More attempts merely move the average closer to a true indicator. For example, if you face five passes, your average might by 0.00 or 16.00, if one of those five passes included a rare long completion or five straight incompletions (which isn't that uncommon but certainly won't continue for long). After 400, 500 or 600 attempts, you'll get more of a true average -- not necessarily lower or higher than it should be.

Also, wouldn't factors unrelated to your defense also affect the per-possession number? Let's say your offense can't sustain long drives. That boosts number of possessions faced for your defense.

The number of possessions faced isn't a measure of a defense. A great defense could face a lot of possessions or a low number of possessions. The only thing that matters is what the defense does with those possessions.

I'm dubious when it comes to the per possession stats, as opposed to per game stats. For example, if points per possession really is the best measure, then Seattle's defense is 1st, and Dallas' is 14th. Seattle's got a good D, but is it really that much better than ours?

According to Football Outsiders, the Seahawks' defense faced 26 more possessions than ours did and gave up eight fewer points than ours did. Unless the starting field position was heavily against us, their defense fared better than ours did. If you could adjust for field position and strength of opponents, I'm pretty sure they'd still come out ahead.

Per-game stats aren't as relevant because they're affected by things other than the defense itself. Let's say one defense allows 14 points on 15 possessions in a game, while another defense allows 13 points on seven possessions. The first faced 15 possessions because its offense had short possessions -- either quick scores, quick turnovers or three-and-outs. The second faced seven because its offense held the ball for a long time every possession. All other things being equal, the defense that allowed 14 points did much better than the defense that allowed 13 points on fewer than half as many possessions.

 

[percyhoward]

There's no indication that facing more pass plays naturally raises or lowers your average yards allowed per pass. After 400, 500 or 600 attempts, you'll get more of a true average -- not necessarily lower or higher than it should be.

Couldn't the same be said of plays in general then--whether run or pass--that the more plays you face, you get more of a true average? And it seems like the logic you're applying to plays should apply to possessions too.

Per-game stats aren't as relevant because they're affected by things other than the defense itself. Let's say one defense allows 14 points on 15 possessions in a game, while another defense allows 13 points on seven possessions. The first faced 15 possessions because its offense had short possessions -- either quick scores, quick turnovers or three-and-outs. The second faced seven because its offense held the ball for a long time every possession. All other things being equal, the defense that allowed 14 points did much better than the defense that allowed 13 points on fewer than half as many possessions.

Even if both defenses had allowed the same number of points, it sure sounds like the deciding factor in determining the better defense in that scenario is nothing more than what its own offense did. It's like saying, "We've got a great defense, the only thing missing that would put us over the top would be if our offense would stop all the long, sustained drives."

 

[AbeBeta]

There's no indication that facing more pass plays naturally raises or lowers your average yards allowed per pass. More attempts merely move the average closer to a true indicator. For example, if you face five passes, your average might by 0.00 or 16.00, if one of those five passes included a rare long completion or five straight incompletions (which isn't that uncommon but certainly won't continue for long). After 400, 500 or 600 attempts, you'll get more of a true average -- not necessarily lower or higher than it should be.

Hmm. The claim of a "true" indicator would be valid if the situations that lead to more attempts were the same as those that led to fewer attempts. True indicators are going to come from random events (e.g., no difference between the situations underlying attempt 1 and attempt 40). Facing more passes isn't always random. In fact, it usually isn't. For example, when you face 45 attempts in a game, it is often because the opponent is behind -- depending on how far behind, you might even allow short to mid range passes to be completed so long as you could keep the clock rolling. Those passes represent a true indicator of how you play in a soft zone or a prevent but not how your pass D performs overall.

 

[percyhoward]

*Seattle
TB
Pitt
NYG
Tenn
Wash
Balt
NE
*KC
*Chi
*SD
*Minny

All those defenses are ahead of ours in yards allowed per possession.

Those marked with * all rank behind the Cowboys in yards per game. (The traditional way of ranking defenses, and the one by which we had the #1 defense in 2003.)

 

[AdamJT13]

Couldn't the same be said of plays in general then--whether run or pass--that the more plays you face, you get more of a true average?

Yes, that's true. Facing more plays doesn't "artificially" lower or raise your average.

And it seems like the logic you're applying to plays should apply to possessions too.

That's true, as well -- facing more possessions doesn't "artificially" lower or raise your average.

Even if both defenses had allowed the same number of points, it sure sounds like the deciding factor in determining the better defense in that scenario is nothing more than what its own offense did.

No, using points per possession means it's based only on what the defense did when it was on the field. If you just go by the total points allowed, then the determining factor could be the team's own offense, because that could determine whether the defense faces more or fewer possessions.

 

[AdamJT13]

Hmm. The claim of a "true" indicator would be valid if the situations that lead to more attempts were the same as those that led to fewer attempts. True indicators are going to come from random events (e.g., no difference between the situations underlying attempt 1 and attempt 40). Facing more passes isn't always random. In fact, it usually isn't. For example, when you face 45 attempts in a game, it is often because the opponent is behind -- depending on how far behind, you might even allow short to mid range passes to be completed so long as you could keep the clock rolling. Those passes represent a true indicator of how you play in a soft zone or a prevent but not how your pass D performs overall.

I'm not sure the number of passes against a prevent defense is enough over the course of a season to make much of a difference overall (or even that defenses fare worse overall when in a prevent defense). But yes, situational statistics usually are more indicative than an overall average.

 

[percyhoward]

In (points per play), you'd essentially be penalized for forcing a three-and-out instead of giving up several first downs (which would boost your number of plays faced) before getting the ball back.

Using points per possession, isn't a team being "penalized" by facing fewer possessions and allowing fewer points, in the case of a team that faced 10 possessions and gave up 2 TD's compared to another that faced 15 possessions and gave up three? Both result in 1.40 ppp, but I wouldn't say those two defenses performed at the same level.

Even if the defense that gave up 21 points on 15 drives forced three-and-outs on all 12 of the other drives, that only adds up to 36 plays, as opposed to the 24 or so other plays when they were getting their butts handed to them. Their job was not to make good plays in clusters, it was to keep the opponent from scoring; they failed three times. I don't cut them slack because of how the plays were divided up into possessions--how many times they managed to make 3 good plays in a row.

The defense that gave up the 2 TD's ***made more plays*** spread out over the 60 minutes, and gave up fewer points. It played better. Period. Per-possession stats don't show that.

Using points per possession means it's based only on what the defense did when it was on the field.

I'd say per-possession stats are based on how a defense performed and how many times the defense went onto and came off the field. For me, the flaw lies in that second part.

Can you give a (realistic, please) example of a small sample size (one game) where per-play stats are misleading, and per-possession stats tell the real story?

 

[AdamJT13]

Using points per possession, isn't a team being "penalized" by facing fewer possessions and allowing fewer points, in the case of a team that faced 10 possessions and gave up 2 TD's compared to another that faced 15 possessions and gave up three?

How is either one being penalized? They're being graded for what they did on the field. Facing only 10 possessions wasn't any sort of accomplishment by the first defense -- it faced fewer possessions either because it failed more or because its own offense sustained drives better.

Both result in 1.40 ppp, but I wouldn't say those two defenses performed at the same level.

Why not? You measure the defense by what it does on the field, not by how many times it is put on the field. In fact, the better the defense, the more likely it is to be on the field more.

Even if the defense that gave up 21 points on 15 drives forced three-and-outs on all 12 of the other drives, that only adds up to 36 plays, as opposed to the 24 or so other plays when they were getting their butts handed to them. Their job was not to make good plays in clusters, it was to keep the opponent from scoring; they failed three times. I don't cut them slack because of how the plays were divided up into possessions--how many times they managed to make 3 good plays in a row.

I have no idea what you're trying to say. Put the defense that allowed two touchdowns in 10 possessions on the field five more times, and it's likely to allow another touchdown -- it already allowed one touchdown every five possessions. If it plays exactly the same, it will allow one in the next five, too. It would have to perform better than it already has in order to not allow one.

The defense that gave up the 2 TD's ***made more plays*** spread out over the 60 minutes, and gave up fewer points. It played better.

There's nothing in those stats that says it made more plays. This defense stopped the opponent eight times. The other defense stopped the opponent 12 times. Couldn't the other defense have made more plays?

I'd say per-possession stats are based on how a defense performed and how many times the defense went onto and came off the field. For me, the flaw lies in that second part.

That's NOT a flaw. The flaw would be to NOT consider how many times the defense was put on the field. The more times your defense gets put on the field, the more likely it is to allow yards and points, period. A defense that is put on the field only eight times has an inherent advantage over a defense that is put on the field 16 times.

Let's say Defense A allows a touchdown on every other possession -- one possession is scoreless, the next is a touchdown, the next is scoreless, the next is a touchdown, on and on for as many possessions as it gets put on the field.

Now let's say Defense B allowed one touchdown every three possessions -- one possession is scoreless, the next is scoreless, the next is a touchdown, the next is scoreless, another scoreless, then another touchdown, on and on.

Is there ANY scenario in which you would say Defense A is better? Of course not. Half of the time it's on the field, it allows a touchdown. The other defense allows a touchdown only once every three possessions. Defense B is succeeding 67 percent of the time, compared to 50 percent for Defense A.

Unless you can say that Defense A -- the one that fails half of the time -- is better than Defense B, there's no way you can say that points per possession isn't more relevant than points per game or points per play.

Can you give a (realistic, please) example of a small sample size (one game) where per-play stats are misleading, and per-possession stats tell the real story?

Absolutely. Let's look at how Cleveland's defense did against Houston in Week 12, compared to how Jacksonville's defense did against Houston in Week 6.

Houston's average starting field position was almost exactly the same in both games, so that's not much of a factor. And let's forget about whatever injuries, field conditions, etc., might have been a factor and just compare the points each defense allowed.

Both defenses allowed Houston's offense to score 17 points. So, on a per-game basis, you'd have to say they performed the same -- 17 points allowed each.

The Texans ran 56 offensive plays against Cleveland's defense and 69 offensive plays against Jacksonville's defense. Using that "points per play" stat, you'd have to say Jacksonville's defense (.246 points per play) performed better than Cleveland's defense (.304 points per play).

The Texans had 13 offensive possessions against Cleveland and 10 offensive possessions against Jacksonville. On a per-possession basis, you'd have to say Cleveland's defense (1.31 points per possession) performed better than Jacksonville's defense (1.70 points per possession).

So, we have three different ways of looking at points allowed -- points per game, points per play and points per possession -- and each of them tells us something different. Points per game tells us Cleveland's defense and Jacksonville's defense performed equally well. Points per play tells us Jacksonville's defense performed better. And points per possession tells us Cleveland's defense performed better. Obviously, only one of those can be correct.

To me, there's no doubt that Cleveland's defense was better at stopping the Texans than Jacksonville's defense was. Cleveland's defense allowed shorter drives, both in yardage, plays and time of possession. The Jaguars' defense allowed several long, time-consuming drives -- four of them that were at least two plays and at least 29 seconds longer than ANY drive the Browns' defense allowed. The only reason the Texans scored as many points against Cleveland as they scored against Jacksonville was because they got three extra chances to have the ball against Cleveland.

In the game against Cleveland, the Texans' average possession lasted only 1 minute, 53.6 seconds, and the Browns' average possession lasted 2:43.3. But in the game against Jacksonville, the Texans' average possession lasted 3:04, and the Jaguars' average possession lasted 3:15.6. Because both teams had shorter possessions in the Cleveland-Houston game, they could squeeze more possessions into the 60-minute time limit. The longer possessions in the Jacksonville-Houston game resulted in fewer total possessions. So in large part because the Browns' defense got off the field more quickly than the Jaguars' defense, the Browns ended up facing three more possessions.

If you were to believe that "points per play" shows that Jacksonville's defense performed better, you'd be punishing the Browns' defense for not allowing more first downs, which would result in more plays being run against them, longer drives, more time of possession for the Texans and, as a result, fewer total possessions for the Texans.

And if you were to believe that "points per game" shows that both defenses performed the same, then you would have to believe that teams are likely to score the same number of points no matter how many times they get the ball.

To me, both of those ideas are ludicrous, as is the idea that a defense allowing a touchdown 50 percent of the time is somehow better than one that allows a touchdown 33 percent of the time (all other things being equal, of course).

 

[percyhoward]

How is either one being penalized? They're being graded for what they did on the field. Facing only 10 possessions wasn't any sort of accomplishment by the first defense -- it faced fewer possessions either because it failed more or because its own offense sustained drives better.

AND if its own offense sustained drives better, the defense that gave up the 2 TD's is still ranked no higher than the one that gave up 3. The success of its own offense is the only thing that kept it from ranking higher. Just because their offense could move the ball, they end up ranked no higher than a defense which gave up 50% more points than they did.

You say that the better the defense, the more likely it is to be on the field more. Don't forget that the worse its offense, the more likely it is to be on the field more too.

Put the defense that allowed two touchdowns in 10 possessions on the field five more times, and it's likely to allow another touchdown

When you say "that defense" you're just basing it on that 10-posession defensive performance. You couldn't put that same defensive performance on the field 5 more times. It would totally change the nature of each of the 10 possessions. It would mean they were shorter. Which plays do you then remove from each possession? You could remove the TD plays, for example, and you'd have a shut out. You could create more TD's against them out of thin air, I guess.

There's nothing in those stats that says it made more plays. This defense stopped the opponent eight times. The other defense stopped the opponent 12 times. Couldn't the other defense have made more plays?

If they did, I'd like to know how they gave up 7 more points. They sure didn't make plays when it counted. They must have been pretty bad in the red zone, for example.

Unless you can say that Defense A -- the one that fails half of the time -- is better than Defense B, there's no way you can say that points per possession isn't more relevant than points per game or points per play.

That scenario doesn't make points per game look bad at all, actually.

And I'm not trying to say that PPG is the best measure, or even that there is a single best measure. I don't believe there is. But if there is, it can't be PPP.

I'm assuming that when a team gets the ball with a lead and 2 minutes left, then gets a couple of first downs, that does NOT count as a possession, under per-possession stats?

 

[AdamJT13]

AND if its own offense sustained drives better, the defense that gave up the 2 TD's is still ranked no higher than the one that gave up 3. The success of its own offense is the only thing that kept it from ranking higher. Just because their offense could move the ball, they end up ranked no higher than a defense which gave up 50% more points than they did.

Why should they be higher when the defenses were equally successful? If the defense that faced 10 possessions had been on the field one more time, they might have held the opponent scoreless -- which would make them more successful than the other defense -- or they might have allowed a field goal or touchdown, which would have made them less successful.

You say that the better the defense, the more likely it is to be on the field more. Don't forget that the worse its offense, the more likely it is to be on the field more too.

That's why you HAVE to base a defense on how many possessions it faces. You can't punish or reward a defense for things unrelated to the defense, and you certainly can't punish a defense for being more successful.

When you say "that defense" you're just basing it on that 10-posession defensive performance. You couldn't put that same defensive performance on the field 5 more times. It would totally change the nature of each of the 10 possessions. It would mean they were shorter.

No, it would not. It could mean that team's offense scored more quickly or had more three-and-outs.

Consider a defense that allows the opponent to hold the ball for 3:00 on each possession. If that team's offense holds the ball for only 1:00 on each possession, the defense would face 16 possessions. If that team's offense holds the ball for 2:00 on each possession, its defense would face 12 possessions. If that team's offense holds the ball for 3:00 on each possession, its defense would face 10 possessions. And if that team's offense holds the ball for 4:00 on each possession, its defense would face nine possessions.

Nothing about the opponent's possessions changed at all. The team's own offense determined how many possessions the opponent had.

If they did, I'd like to know how they gave up 7 more points.

Because they faced more possessions. They made more plays (to stop possessions) AND gave up more scoring plays because they were put on the field five more times.

Look at it like two baseball pitchers. One pitches six innings, allowing two runs. Another pitches nine innings, allowing three runs. Who got more outs? Who allowed more runs? Obviously, the answer to both questions is the second pitcher, who pitched more innings. Likewise, a defense that allows three "runs" in 15 "innings" got more "outs" than a defense that allowed two "runs" in 10 "innings."

They must have been pretty bad in the red zone, for example.

There's no way to even deduce that they were ever IN the red zone.

That scenario doesn't make points per game look bad at all, actually.

Except that points per game ignores how often the defense gets put on the field. Most pitchers will allow fewer runs if they pitch one inning than if they pitch nine.

And I'm not trying to say that PPG is the best measure, or even that there is a single best measure. I don't believe there is. But if there is, it can't be PPP.

Certainly not if "PPP" means "points per play." But points per possession is by far the best raw measure out of the three options (game, possession and play). If it's adjusted for field position and the opponent, it's even better.

I'm assuming that when a team gets the ball with a lead and 2 minutes left, then gets a couple of first downs, that does NOT count as a possession, under per-possession stats?

Irrelevant possessions at the ends of halves should be thrown out, yes.

 

[percyhoward]

Why should they be higher when the defenses were equally successful? If the defense that faced 10 possessions had been on the field one more time, they might have held the opponent scoreless -- which would make them more successful than the other defense -- or they might have allowed a field goal or touchdown, which would have made them less successful.

The point is, they weren't equally successful. If the defense that faced 10 possessions had been on the field one more time, that extra possession would have had to come in overtime, or in the next game.

It's either that, or...

You couldn't put that same defensive performance on the field 5 more times. It would totally change the nature of each of the 10 possessions. It would mean they were shorter.

No, it would not. It could mean that team's offense scored more quickly or had more three-and-outs.

That's possible, but it's equally possible that the same thing happened with the opposing offense (just as it's possible for that same team's offense to have had even longer possessions, for that matter). There are endless scenarios, but for the sake of the argument, don't you have to apply the "all else being equal" and assume that both teams' possessions were shorter?

Look at it like two baseball pitchers. One pitches six innings, allowing two runs. Another pitches nine innings, allowing three runs. Who got more outs? Who allowed more runs? Obviously, the answer to both questions is the second pitcher, who pitched more innings. Likewise, a defense that allows three "runs" in 15 "innings" got more "outs" than a defense that allowed two "runs" in 10 "innings."

You forget that baseball isn't a timed game. There is a higher priority on retiring the side in order, because that reduces the number of innings left to be played. In football, when a defense forces three-and-outs, they aren't reducing the number of possessions left for them to face. In fact, all else being equal, isn't it just the opposite? Isn't that defense, despite a great three-play performance, giving itself more work to do later on?

Points per game ignores how often the defense gets put on the field. Most pitchers will allow fewer runs if they pitch one inning than if they pitch nine.

Baseball influences how some people look at possessions, I think. If baseball were a timed game, it wouldn't be nearly as important for a pitcher to retire the side in order, because it would add to the number of batters he'd have to face over the course of the game. He (and/or his relievers) would still have to fill the 60 minutes or whatever it was.

Sure a pitcher allows fewer runs in one inning than in nine. But that inning always represents one ninth of the game. In football, a possession could represent anything from 1/5 to 1/100 of the game.

I'm assuming that when a team gets the ball with a lead and 2 minutes left, then gets a couple of first downs, that does NOT count as a possession, under per-possession stats?

Irrelevant possessions at the ends of halves should be thrown out, yes.

So if the offense gets the first downs needed in order to be able to then run out the clock, that defensive failure is considered an "irrelevant" possession, and therefore not counted. If, however, the defense forces three-and-out on that very same type of possession, then it DOES count, doesn't it?

That's messed up, and more than a little bit.

 

[AdamJT13]

The point is, they weren't equally successful. If the defense that faced 10 possessions had been on the field one more time, that extra possession would have had to come in overtime, or in the next game.

Not if its offense had one (or more) shorter possessions.

That's possible, but it's equally possible that the same thing happened with the opposing offense (just as it's possible for that same team's offense to have had even longer possessions, for that matter). There are endless scenarios, but for the sake of the argument, don't you have to apply the "all else being equal" and assume that both teams' possessions were shorter?

No, because we're only talking about the defenses.

You forget that baseball isn't a timed game. There is a higher priority on retiring the side in order, because that reduces the number of innings left to be played. In football, when a defense forces three-and-outs, they aren't reducing the number of possessions left for them to face. In fact, all else being equal, isn't it just the opposite? Isn't that defense, despite a great three-play performance, giving itself more work to do later on?

That would depend on what its offense does.

Personally, I'd much rather have a defense that forces quick three-and-outs, because it can give that team's offense more possessions, but perhaps more importantly, it usually gives the offense better field position. If your defense forces a three-and-out from the 20, you'll probably get the ball back near midfield. If your defense lets the opponent get two or three first downs before punting, you'll probably get the ball back inside your own 20.

I can't think of a single scenario when I'd rather have my defense allow a first down than force the opponent to punt. And I certainly can't imagine anyone thinking the defense was better BECAUSE it allowed a first down.

Baseball influences how some people look at possessions, I think. If baseball were a timed game, it wouldn't be nearly as important for a pitcher to retire the side in order, because it would add to the number of batters he'd have to face over the course of the game. He (and/or his relievers) would still have to fill the 60 minutes or whatever it was.

Whether it's football, baseball or even basketball, no matter how many possessions/innings there are, both teams get essentially the same number (depending on what happens at the end of halves/games). So if baseball was timed and your pitcher walked the bases loaded on purpose to create longer innings and therefore fewer innings, it would also reduce the number of times his team came to bat. So, allowing one two runs in six innings wouldn't be any better than allowing three runs in nine or four runs in 12, because his offense would have to produce runs at the same rate in order to win in any of those scenarios.

So if the offense gets the first downs needed in order to be able to then run out the clock, that defensive failure is considered an "irrelevant" possession, and therefore not counted. If, however, the defense forces three-and-out on that very same type of possession, then it DOES count, doesn't it?

That's messed up, and more than a little bit.

If you're calculating points per possession, you would count any possesion when the offense tried to score and had time to score. It would be "messed up" to do anything otherwise.

If you're calculating drive success rating, then you would credit a success for preventing the offense from running out the clock and a failure for allowing the offense to run out the clock.

 

[percyhoward]

Not if its offense had one (or more) shorter possessions.

Again, I know that. As you know, that's just one scenario. I could easily say "not if the opposing offense had one or more shorter possessions." You assume, for some reason, that it would always play out in the way that best supports your argument.

No, because we're only talking about the defenses.

Then explain why you just wrote, "Not if its offense had one (or more) shorter possessions."

That would depend on what its offense does.

Did it again. Not "busting your chops," but you are including what the offenses do. You have to, because you can't divorce what the offense does from what the defense does.

Personally, I'd much rather have a defense that forces quick three-and-outs, because it can give that team's offense more possessions, but perhaps more importantly, it usually gives the offense better field position. If your defense forces a three-and-out from the 20, you'll probably get the ball back near midfield. If your defense lets the opponent get two or three first downs before punting, you'll probably get the ball back inside your own 20.

I can't think of a single scenario when I'd rather have my defense allow a first down than force the opponent to punt. And I certainly can't imagine anyone thinking the defense was better BECAUSE it allowed a first down.

Is it really hard to imagine anyone thinking it was better despite allowing a first down on a particular possession, but allowing fewer points in the game?

Personally, I also like watching our D force three-and-outs, and winning the field position battle. Just not to the point that it becomes more important than the fact that it allowed 2 TD's as opposed to 3.

Whether it's football, baseball or even basketball, no matter how many possessions/innings there are, both teams get essentially the same number.

A reminder here that we were comparing number of possessions across games, not within one game.

So if baseball was timed and your pitcher walked the bases loaded on purpose to create longer innings and therefore fewer innings, it would also reduce the number of times his team came to bat.

Maximum score on a possession is 8 points. What's the maximum score in an inning?

If you're calculating points per possession, you would count any possession when the offense tried to score and had time to score.

If they aren't trying to score, but only trying to get a couple of first downs, so that they can then just let the clock run out, but the defense stops them on a three-and-out, then you're saying that the possession is NOT counted in that defense's points per possession?

 

[AdamJT13]

Again, I know that. As you know, that's just one scenario. I could easily say "not if the opposing offense had one or more shorter possessions." You assume, for some reason, that it would always play out in the way that best supports your argument.

Either one would support my argument. This was YOUR scenario of a defense that faced 10 possessions. YOU said it would face 11 possessions only "in overtime, or the next game." If you to refute your own argument, be my guest.

Then explain why you just wrote, "Not if its offense had one (or more) shorter possessions."

Because you cant assume that more possessions in a game means both teams' possessions were shorter. It could be one, or the other, or both.

Is it really hard to imagine anyone thinking it was better despite allowing a first down on a particular possession, but allowing fewer points in the game?

Someone might think that, but they also might be wrong.

Personally, I also like watching our D force three-and-outs, and winning the field position battle. Just not to the point that it becomes more important than the fact that it allowed 2 TD's as opposed to 3.

Would you rather allow two touchdowns on three possessions or three touchdowns on 20 possessions?

A reminder here that we were comparing number of possessions across games, not within one game.

It doesn't matter. You get the ball essentially the same number of times that your opponent does, whether it's in a game or over the course of a season.

The problem with using per-game stats is that not all teams play at the same pace. The Colts' defense faced 158 possessions last season. The Bears' defense faced 210 -- a whopping 33 percent more, or 3.25 possessions per game! If the Colts' defense allowed 228 points (it did), and the Bears' defense had allowed 235 (it didn't), which defense would you say was better at preventing scoring? There's NO WAY I would say the Colts' defense was better at it, given the fact that it faced 52 fewer possesions and allowed only one fewer touchdown.

Maximum score on a possession is 8 points. What's the maximum score in an inning?

The same for both teams, which is all that matters.

If they aren't trying to score, but only trying to get a couple of first downs, so that they can then just let the clock run out, but the defense stops them on a three-and-out, then you're saying that the possession is NOT counted in that defense's points per possession?

Not if it's done correctly.

 

[dmq]

I won't sugar coat it. I think our pass defense was terrible last year. I think this may be a different story. I just remember Eli's comments at halftime of the first game where he said throwing to an open receiver in front of Reeves was gonna be open all day.