Week Twelve: Detroit Lions (9-1) @ Indianapolis Colts (5-6)

[RedRamage]

16 hours ago, chasfh said:

I have a Football 101 question here: just how legitimate is the idea of Goff, or any other quarterback, having a "perfect game" based on QB rating?

I honestly have no idea what the calculation is used in passer rating and I think it's insane odd that a "perfect" score is 158.3.

Having said that, saying the QB was "perfect" is probably akin to saying that a pitcher who allowed no base runners pitched a perfect game. A perfect game for a pitcher might include many pitches that were less than ideal but the hitter was just bad/unlucky or the defense came up with a great play.

There's obviously going to be instances where the pass could have been better, but I guess how ever the calculation happens it's determined that the final outcome was optimal, even if there was a mistake here or there.

[Motown Bombers]

Smith looked good in his debut. 

[Motown Bombers]

8 minutes ago, RedRamage said:

I honestly have no idea what the calculation is used in passer rating and I think it's insane odd that a "perfect" score is 158.3.

Having said that, saying the QB was "perfect" is probably akin to saying that a pitcher who allowed no base runners pitched a perfect game. A perfect game for a pitcher might include many pitches that were less than ideal but the hitter was just bad/unlucky or the defense came up with a great play.

There's obviously going to be instances where the pass could have been better, but I guess how ever the calculation happens it's determined that the final outcome was optimal, even if there was a mistake here or there.

I’m not sure how it’s calculated, but Goff’s name literally is perfect. 

Edited November 20, 2024 by Motown Bombers

[Motown Bombers]

St Brown is second in the league in ASS. 

 

[chasfh]

1 hour ago, RedRamage said:

Having said that, saying the QB was "perfect" is probably akin to saying that a pitcher who allowed no base runners pitched a perfect game. A perfect game for a pitcher might include many pitches that were less than ideal but the hitter was just bad/unlucky or the defense came up with a great play.

I believe the difference is that in baseball, batter outcome as it relates to change in base-out state is germane to the disposition of the game in a way individual pitches are not. If a pitcher throws a ball outside the zone, that does not change the win probability of the game in any way. Win probability changes only when base-out state changes. In football, throwing an incomplete pass does negatively change win probability.

IOW, in a baseball perfect game, every batter disposed of raises the win probability for the pitcher's team, because it never goes down. But even in a quarterback's "perfect game", when he throws an incomplete game, win probability drops. That is, practically by definition, imperfect.

[gehringer_2]

6 minutes ago, chasfh said:

I believe the difference is that in baseball, batter outcome as it relates to change in base-out state is germane to the disposition of the game in a way individual pitches are not. If a pitcher throws a ball outside the zone, that does not change the win probability of the game in any way. Win probability changes only when base-out state changes. In football, throwing an incomplete pass does negatively change win probability.

IOW, in a baseball perfect game, every batter disposed of raises the win probability for the pitcher's team, because it never goes down. But even in a quarterback's "perfect game", when he throws an incomplete game, win probability drops. That is, practically by definition, imperfect.

yes - there is a lot of imprecision in tying the concept of a baseball perfect game to the pitcher individually. Or another way to put it is that in baseball a perfect game is a net outcome, but it does not require or necessarily reflect anything like perfect performance by the pitcher at a one to one level. He can give up any number of line-drive at'em balls, get behind in the count, throw the ball to the backstop, etc. While on the other hand he could give up nothing but Ks and soft contact and a fielder could boot away his perfecto.

So in FB, in one sense the standard is higher for a QB because every throw does have a specific outcome. On the other hand for a QB his receiver could drop a perfect throw, and the QB also is rated on a much smaller number of 'chances' per game. The probability of being perfect in 25 throws is going to be better than in a 100.

Then if you really want to push into the weeds, QBs are not rated directly on mistakes they do make that are just as bad to the game outcome as bad throws, and that would be mis-reads. If the QB checks into the wrong running play and hands the ball off for a loss, nothing is tallied against his performance at all (other than by the HC who happens to be watching!).

[chasfh]

Just now, gehringer_2 said:

yes - there is a lot of imprecision in tying the concept of a baseball perfect game to the pitcher individually. Or another way to put it is that in baseball a perfect game is a net outcome, but it does not require or necessarily reflect anything like perfect performance by the pitcher at a one to one level. He can give up any number of line-drive at'em balls, get behind in the count, throw the ball to the backstop, etc. While on the other hand he could give up nothing but Ks and soft contact and a fielder could boot away his perfecto.

So in FB, in one sense the standard is higher for a QB because every throw does have a specific outcome. On the other hand for a QB his receiver could drop a perfect throw, and the QB also is rated on a much smaller number of 'chances' per game. The probability of being perfect in 25 throws is going to be better than in a 100.

Then if you really want to push into the weeds, QBs are not rated directly on mistakes they do make that are just as bad to the game outcome as bad throws, and that would be mis-reads. If the QB checks into the wrong running play and hands the ball off for a loss, nothing is tallied against his performance at all (other than by the HC who happens to be watching!).

Flurry of what-ifs and whatabouts notwithstanding, my conclusion is that a perfect game as applied to a baseball pitcher makes a hell of a lot more sense than a perfect game as applied to a football quarterback.

[gehringer_2]

4 minutes ago, chasfh said:

Flurry of what-ifs and whatabouts notwithstanding, my conclusion is that a perfect game as applied to a baseball pitcher makes a hell of a lot more sense than a perfect game as applied to a football quarterback.

I've always been interested in the limits of measurement. Measurement is what has allowed us to do science, which is what has given us the modern world, but it's always good to explore the limits and understand what your measurement do and don't mean. Can keep you out of a lot of trouble in the "lab".

Edited November 20, 2024 by gehringer_2

[chasfh]

9 minutes ago, gehringer_2 said:

I've always been interested in the limits of measurement. Measurement is what has allowed us to do science, which is what has given us the modern world, but it's always good to explore the limits and understand what your measurement do and don't mean. Can keep you out of a lot of trouble in the "lab".

Good thing baseball doesn't exist in a lab. I don't know enough about football to hazard a guess on that one. 😏

I'm still going to refer to 27 up 27 down as a perfect game, even if multiple pitchers take part in one.

 

[KnoxP]

43 minutes ago, gehringer_2 said:

I've always been interested in the limits of measurement. Measurement is what has allowed us to do science, which is what has given us the modern world, but it's always good to explore the limits and understand what your measurement do and don't mean. Can keep you out of a lot of trouble in the "lab".

Played with this just a bit.  Seems to place a great deal of value on TD passes and and some ratio of yards and attempts/completions?  Its wild some of the #'s you can plug in and still not get a perfect rating.  For example: 40/50, 500y and 6TD, not perfect.

https://www.nfeloapp.com/tools/qb-passer-rating-calculator/ 

Edited November 20, 2024 by KnoxP
add

[Sports_Freak]

49 minutes ago, chasfh said:

Good thing baseball doesn't exist in a lab. I don't know enough about football to hazard a guess on that one. 😏

I'm still going to refer to 27 up 27 down as a perfect game, even if multiple pitchers take part in one.

 

I was thinking of this...using the logic of not hitting a wide open receiver for a long TD pass instead of a receiver for a shorter gain. In baseball, it shouldn't be a perfect game if a pitcher misses his spot? Even if he retires the hitter, he wasn't trying to pitch the ball in that spot. 

[RedRamage]

1 hour ago, chasfh said:

I believe the difference is that in baseball, batter outcome as it relates to change in base-out state is germane to the disposition of the game in a way individual pitches are not. If a pitcher throws a ball outside the zone, that does not change the win probability of the game in any way. Win probability changes only when base-out state changes. In football, throwing an incomplete pass does negatively change win probability.

Without trying to sound like I'm defending QBR "perfect game" status, I disagree a bit here. The count that a batter is in changes the probability of getting on base and getting on base changes the probability of scoring a run and the probability of scoring a run changes the win probability. So I think yeah... a ball outside the zone (assuming the batter doesn't swing at it) does chance the win probability.

If you want to argue: "As long as the batter is out, what does it matter if the pitcher threw 3 balls or 0 balls?" then I guess I'd counter with: "As long as the team scored, or even just took a bunch of time off the clock so the opposing team isn't able to make a come back, what does it matter if the QB threw an incomplete pass a few times?"

If you want to argue: "A ball outside of the zone could be a strategic move to setup the batter, so even a 'bad pitch' might be a good strategy." Then I'd argue the same could be said in football. A deep pass with a lower success rate might be done to setup a defense for a run play or a screen or whatever. A ball thrown away before taking a sack is better move than tacking the sack or throwing a very low completion percentage pass that could be intercepted.

[RedRamage]

1 hour ago, KnoxP said:

Played with this just a bit.  Seems to place a great deal of value on TD passes and and some ratio of yards and attempts/completions?  Its wild some of the #'s you can plug in and still not get a perfect rating.  For example: 40/50, 500y and 6TD, not perfect.

https://www.nfeloapp.com/tools/qb-passer-rating-calculator/ 

Anyone know the actual formula? I played around a bit with it too.

30 for 30, 500 yards, 0 TDs: 118.7... not even close to perfect. You could be zipping the ball all over the place, hitting received on a dime, but the receivers all get tackled just shy of the End Zone so your team runs the ball in. Suddenly your rating drop like a rock. If I increase the TDs to 3 I get to 152...still not a perfect game.

Here's another stupid one: My defense keeps getting turn overs giving me great field position. I only need to throw 10 passes to get the win and each pass results in TD, but because we had such amazing field position I only racked up 124 yards pass. 10 for 10, 10 TDs, 124 yards = 157.9. Not a perfect game. What more could I possibly have done as a QB??

So I want to re-iterate @chasfh, while I disagree that individual pitches don't change win probability, I do NOT want to appear like I'm defending the QBR metric or how it defines a "perfect" QB game.

[RedRamage]

Okay... now that y'all got me going down this rabbit hole I decided to go all out...

https://en.wikipedia.org/wiki/Passer_rating

Go read that page for all the nitty gritty details, but short version is that there are 4 calculations made. Each calculation is weighted to give a result between 0 and 2.375. When the formulas were created a result of 1.0 would be league average based upon data from 1960-1970. I don't know why 2.375 is considered the "cap" but that's what they picked. If the result of any one of those four calculations is negative, it's set to 0. If it's more than 2.375, it's set to 2.375.

Then you do this: 

It seems the inventors of the formula wanted 66.7 to equal average while 100+ was excellent and presumably 33.3 was awful. Still seem super dumb to me, but whatever. Obviously as passing has become a bigger and bigger part of the league the numbers have shifted up. In 2020 League average of 93.6.

I will say that I don't think the designers ever intended that someone who hit the highest possible score had a "perfect game." I think they were just trying to set a cap to prevent a QB having one aspect of their game be super good and skew things because of that. So we shouldn't (imho) be using it to say that a QB had a perfect game. It would be more accurate to say his game was so good it topped the rating measurement system. But that doesn't really roll off the tongue. 

Here's the mark you need to hit for each equation to meet that 2.375 cap:

• 77.5% completion percentage
• 12.5 yards per attempt
• 11.875% TD/ATT (1 TD/8.421ATT)
• No interceptions

It's also worth noting that fumbles, sacks, and rushing yards/TDs by the QB are NOT included in the calculations at all. A QB who had 18/25, 250 yards, 3 TDs and 1 INT gets a ratting of 126.6 while a QB who had 18/25, 250 yards, 3 TDs and 0 INTs, but 3 lost fumbles get a rating of 143.3. Why? Because this formula was very specifically designed to be PASSER RATING. Even though most of us talk about it like QBR or Quaterback Rating, it's not. QBR is an ESPN stat and it's totally different than this.

 

[Motown Bombers]

I was told there would be no math. 

[Motown Bombers]

 

[KnoxP]

1 hour ago, RedRamage said:

Okay... now that y'all got me going down this rabbit hole I decided to go all out...

https://en.wikipedia.org/wiki/Passer_rating

Go read that page for all the nitty gritty details, but short version is that there are 4 calculations made. Each calculation is weighted to give a result between 0 and 2.375. When the formulas were created a result of 1.0 would be league average based upon data from 1960-1970. I don't know why 2.375 is considered the "cap" but that's what they picked. If the result of any one of those four calculations is negative, it's set to 0. If it's more than 2.375, it's set to 2.375.

Then you do this: 

It seems the inventors of the formula wanted 66.7 to equal average while 100+ was excellent and presumably 33.3 was awful. Still seem super dumb to me, but whatever. Obviously as passing has become a bigger and bigger part of the league the numbers have shifted up. In 2020 League average of 93.6.

I will say that I don't think the designers ever intended that someone who hit the highest possible score had a "perfect game." I think they were just trying to set a cap to prevent a QB having one aspect of their game be super good and skew things because of that. So we shouldn't (imho) be using it to say that a QB had a perfect game. It would be more accurate to say his game was so good it topped the rating measurement system. But that doesn't really roll off the tongue. 

Here's the mark you need to hit for each equation to meet that 2.375 cap:

• 77.5% completion percentage
• 12.5 yards per attempt
• 11.875% TD/ATT (1 TD/8.421ATT)
• No interceptions

It's also worth noting that fumbles, sacks, and rushing yards/TDs by the QB are NOT included in the calculations at all. A QB who had 18/25, 250 yards, 3 TDs and 1 INT gets a ratting of 126.6 while a QB who had 18/25, 250 yards, 3 TDs and 0 INTs, but 3 lost fumbles get a rating of 143.3. Why? Because this formula was very specifically designed to be PASSER RATING. Even though most of us talk about it like QBR or Quaterback Rating, it's not. QBR is an ESPN stat and it's totally different than this.

 

I found a link, full stop.  You did some heavy lifting.  At least it now has some story behind it, albeit using 1960-70 as a benchmark!  

[Motor City Sonics]

When is Iffy coming back?   Secondary is already great

[Motown Bombers]

3 minutes ago, Motor City Sonics said:

When is Iffy coming back?   Secondary is already great

He was just placed on IR again, so it will be at least another 4 weeks. 

[RedRamage]

1 hour ago, Motor City Sonics said:

When is Iffy coming back?   Secondary is already great

Hey, this is the obscure trivia about stats thread not the -- um... wait, sorry, my mistake. Carry on.

[Motown Bombers]

 

[Motown Bombers]

Pics of Dan Campbell in high school with massive shoulder pads have surfaced, and Lions players are starting to use the pic as their profile pic on social media. 

[RedRamage]

Tigers huh?  Hmm... if Dan gets bored after winning the next 3 Super Bowls, maybe he'll want to coach the Tigers??

[Motor City Sonics]

2 hours ago, RedRamage said:

Tigers huh?  Hmm... if Dan gets bored after winning the next 3 Super Bowls, maybe he'll want to coach the Tigers??

Screw that,  The Red Wings are the ones that need him.

 

[lordstanley]

A couple of names I haven't thought of in a bit and have already written off, for this season at least: are Barnes and Davenport both definitely out for the season, including the playoffs? Both injuries were reported as "most likely" ending their seasons. Any chance either one attempts a January or February comeback like some hope Hutch will?