The Michael McCorkle "Mac" Jones Thread

Eddie Jurak

Go Leafs Go
Lifetime Member
SoSH Member
Dec 12, 2002
32,327
Melrose, MA
No it's really not. Comparing the best pocket QB in the history of the league in his 18th year to a rookie while behind completely different lines with almost no teammates in common, has no value and even Bedard as clueless as he is knows that. Like the rest of the article, it's cheap and playing to the crowd while having less than zero actual value, it would be a shitty use of data as a tweet, in what is allegedly an article of analysis it's even worse. There are a lot of people writing positive articles about Mac that have substance (see Lazar's actual useful analysis, which has been pretty pro-Mac throughout). This is not one of them.
If the point were to ask "which one is better," then I would agree with you. But that's both a stupid question and not the point.
 

Eddie Jurak

Go Leafs Go
Lifetime Member
SoSH Member
Dec 12, 2002
32,327
Melrose, MA
Evan Lazar has tweeted out some Mac info today.
  • He was pressured on 25.8% of his dropbacks, lower than all other games besides the Jets.
  • According to PFF, Mac had 2 turnover-worthy plays against Houston, and he has 7 of them on the year. On the season, per throw, that works out to 3.9%, which rates him 19th out of 31 QBs.
  • Over the Pats 2 games, Mac has been markedly better against the blitz.
View: https://twitter.com/ezlazar/status/1447558129391214595?s=20

Evan Lazar: By far the biggest improvement in the numbers for Mac Jones over the last two games is against the blitz. If Mac continues to shred blitzes, teams will hopefully back off which will give him more time to throw downfield.

Mac has been blitzed on 35.8% of his drops, the third-highest rate in the NFL. His completion percentage (88.5% vs 70%), yards per attempt (8.6 vs 5.7), and passer rating (99.5 vs 76.8) all much better over the past 2 weeks than the first 3 against the blitz.
 

rodderick

Member
SoSH Member
Apr 24, 2009
8,299
Belo Horizonte - Brazil
Evan Lazar has tweeted out some Mac info today.
  • He was pressured on 25.8% of his dropbacks, lower than all other games besides the Jets.
  • According to PFF, Mac had 2 turnover-worthy plays against Houston, and he has 7 of them on the year. On the season, per throw, that works out to 3.9%, which rates him 19th out of 31 QBs.
  • Over the Pats 2 games, Mac has been markedly better against the blitz.
View: https://twitter.com/ezlazar/status/1447558129391214595?s=20

Evan Lazar: By far the biggest improvement in the numbers for Mac Jones over the last two games is against the blitz. If Mac continues to shred blitzes, teams will hopefully back off which will give him more time to throw downfield.

Mac has been blitzed on 35.8% of his drops, the third-highest rate in the NFL. His completion percentage (88.5% vs 70%), yards per attempt (8.6 vs 5.7), and passer rating (99.5 vs 76.8) all much better over the past 2 weeks than the first 3 against the blitz.
I'm kind of surprised with the 2 turnover worthy plays, I thought I could recall at least 3. But his improved performance against the blitz is awesome and shows he can adjust to how teams are choosing to defend against him. I always thought he'd be able to handle that approach because it suits his strengths (gets ball out quickly, makes good decisions). I can't see Mac being a QB that teams choose to blitz a lot going forward.
 

Eddie Jurak

Go Leafs Go
Lifetime Member
SoSH Member
Dec 12, 2002
32,327
Melrose, MA
I'm kind of surprised with the 2 turnover worthy plays, I thought I could recall at least 3. But his improved performance against the blitz is awesome and shows he can adjust to how teams are choosing to defend against him. I always thought he'd be able to handle that approach because it suits his strengths (gets ball out quickly, makes good decisions). I can't see Mac being a QB that teams choose to blitz a lot going forward.
Turnover-worthy plays might be an "eye of the beholder" thing. If PFF is using consistent methdology (I don;t know what they do), the 19 of 31 ranking may be reasonably accurate.

The good thing about the blitz numbers is that it shows some growth in his game.
 

rodderick

Member
SoSH Member
Apr 24, 2009
8,299
Belo Horizonte - Brazil
Turnover-worthy plays might be an "eye of the beholder" thing. If PFF is using consistent methdology (I don;t know what they do), the 19 of 31 ranking may be reasonably accurate.

The good thing about the blitz numbers is that it shows some growth in his game.
Oh, it's 100% an "eye of the beholder" thing, I was just thinking there were more than 2 passes in which defenders either had two hands on the ball or one defensive player knocked the ball away before another guy had a clear shot at an INT. But maybe one of those I'm recalling happened on the roughing the passer.
 

Big McCorkle

Member
SoSH Member
May 9, 2021
208
Turnover-worthy plays might be an "eye of the beholder" thing. If PFF is using consistent methdology (I don;t know what they do), the 19 of 31 ranking may be reasonably accurate.

The good thing about the blitz numbers is that it shows some growth in his game.
Lazar might have been looking at the numbers before regrading or something, as Mac was definitely credited with 3 TWPs yesterday. Which to a degree makes his passing grade (71.9)... not impressive per se, but that's rather high given the amount of TWPs, meaning that outside of those he was really good. (Which, yeah, is something that anyone with eyes could tell you yesterday: he was lights-out yesterday outside of those handful of complete headscratchers)
 

Cotillion

lurker
Jun 11, 2019
1,357
After Meyer's dropped that wide open pass that hit him in the chest Mac started to press pretty hard and threw at least 2 or 3 balls that were in the hands of defenders or into super tight coverage that could have been picked off...
 

Gash Prex

Member
SoSH Member
Apr 18, 2002
5,055
Mac had a couple of terrible series from the 2 min warning into the 3rd quarter - forcing the ball and some uncharacteristic bad decisions. Before and after he was on target all day
 

cournoyer

Member
SoSH Member
Dec 11, 2012
444
Enfield, Connecticut
Mac had a couple of terrible series from the 2 min warning into the 3rd quarter - forcing the ball and some uncharacteristic bad decisions. Before and after he was on target all day
I agree.. which also is fine. There is no such thing as the perfect quarterback. Even TB12 who comes as close as anybody to being perfect will make a couple headscratchers a game. I wanted Cam to begin the year as QB but I'll happily admit that Mac has really exceeded my expectations. I'm looking forward to seeing this offense progress as we start getting some continuity on the offensive line, while opening up the tight ends a bit in the passing game.
 

Jimbodandy

Well-Known Member
Gold Supporter
SoSH Member
Jan 31, 2006
5,895
around the way
Lazar might have been looking at the numbers before regrading or something, as Mac was definitely credited with 3 TWPs yesterday. Which to a degree makes his passing grade (71.9)... not impressive per se, but that's rather high given the amount of TWPs, meaning that outside of those he was really good. (Which, yeah, is something that anyone with eyes could tell you yesterday: he was lights-out yesterday outside of those handful of complete headscratchers)
I maintain that the headscratchers are the natural by-product of taking off the training wheels and looking downfield more. It's the next step in the progression. He will have games where the bad guys catch those balls too, instead of dropping them. But for him to become a top 10 QB or better, he needs to learn what he can/can't do downfield. This is what happens. He and the receiver will read the zig - zag thing differently, for example. One of those bad throws, it looked like he expected someone to sit down in the zone, but the receiver kept up the seam. And of course he needs to test his arm throwing into windows. Sometimes he'll have the arm for those windows, sometimes he won't. Part of the process imo.
 

Cellar-Door

Member
SoSH Member
Aug 1, 2006
22,821
Oh, it's 100% an "eye of the beholder" thing, I was just thinking there were more than 2 passes in which defenders either had two hands on the ball or one defensive player knocked the ball away before another guy had a clear shot at an INT. But maybe one of those I'm recalling happened on the roughing the passer.
Honestly from what I can tell turnover worthy plays by QBs is one of the PFF stats that holds the least value. Following their employees on twitter, I'm always amazed how often they declare a throw not the QBs fault. Like even the slightest issue with a WR they declare the throw good. I think of the Eagles pick 6 that Smith fell down on, they declared it not Hurts' fault, but the throw was WAY behind where Smith was when he fell, if he doesn't fall the ball is too far behind him to even reach back for... but because he fell... not TO-worthy.
 

Eddie Jurak

Go Leafs Go
Lifetime Member
SoSH Member
Dec 12, 2002
32,327
Melrose, MA
Evan Lazar film review shows a bunch of good and a few bad plays.

https://www.clnsmedia.com/lazars-film-review-patriots-qb-mac-jones-is-way-ahead-of-the-curve-relative-to-other-rookies/



This graph of all the QBs, completion percentage above expected vs. EPA per play highlights where Mac is compared with the other rookies.

He's in the top right quadrant, meaning that he is above average on both metrics (though just barely so for EPA). Fields, Wilson, Lance, Lawrence , and Mills are all in the lower left quadrant, meaning below average on both metrics.

So if you want to be a contrarian, then keep ignoring everything telling you that Jones is operating well in an NFL offense five games into his career. Five. Not 305 like Brady. Five.

Every rookie makes mistakes, and Jones is making rookie mistakes just like the rest of them (more on that later). But he’s also making reads that other guys simply aren’t at this stage.
In other words, Jones knows where the opening will be in the defense and the best way to get the ball there.

As we highlighted in the few plays he’d like to have back, the next step for the Pats’ rookie is understanding his physical limitations and the speed in which NFL passing windows shrink.

But at this stage, he’s operating well ahead of the curve and is in a great place with his development.
 

Big McCorkle

Member
SoSH Member
May 9, 2021
208
What's the argument for CPOE as a good metric to evaluate QB play? I'm asking this in all honesty.
Well, I know the specific reason for an EPA+CPOE composite is that the combination is more predictive of future EPA than just EPA alone.

But more generally it's valid because completed passes tend to be "good" and incomplete passes tend to be "bad," and by adjusting for depth of target you can generally judge how good a QB is at completing passes relative to other QBs while factoring in the difficulty level of the throws being made. Of course, it's a noisy stat with WR quality in particular going to play a role, but that's true of most stats out there, and it's more or less measuring a real thing (as opposed to a box score composite like passer rating, which is dumb and arbitrary.)
 

rodderick

Member
SoSH Member
Apr 24, 2009
8,299
Belo Horizonte - Brazil
Well, I know the specific reason for an EPA+CPOE composite is that the combination is more predictive of future EPA than just EPA alone.

But more generally it's valid because completed passes tend to be "good" and incomplete passes tend to be "bad," and by adjusting for depth of target you can generally judge how good a QB is at completing passes relative to other QBs while factoring in the difficulty level of the throws being made. Of course, it's a noisy stat with WR quality in particular going to play a role, but that's true of most stats out there, and it's more or less measuring a real thing (as opposed to a box score composite like passer rating, which is dumb and arbitrary.)
I just want to know what EPA/Play doesn't tell you that CPOE does. It might be more predictive, but when accounting for past performance, if a guy has higher EPA/Play hasn't he just played better/had better results? In that graph for instance, the composite would tell you Kyler has been about as good or better than Brady, even though Brady has higher EPA/Play.
 

Super Nomario

Member
SoSH Member
Nov 5, 2000
13,410
Mansfield MA
Lazar might have been looking at the numbers before regrading or something, as Mac was definitely credited with 3 TWPs yesterday. Which to a degree makes his passing grade (71.9)... not impressive per se, but that's rather high given the amount of TWPs, meaning that outside of those he was really good. (Which, yeah, is something that anyone with eyes could tell you yesterday: he was lights-out yesterday outside of those handful of complete headscratchers)
Yeah, I think Mac had kind of an odd performance with probably his most good plays on the season but also maybe his most bad plays. He was lucky to just throw the one pick, but he made a lot of nice plays and the offense moved the ball consistently.

I just want to know what EPA/Play doesn't tell you that CPOE does. It might be more predictive, but when accounting for past performance, if a guy has higher EPA/Play hasn't he just played better/had better results? In that graph for instance, the composite would tell you Kyler has been about as good or better than Brady, even though Brady has higher EPA/Play.
EPA / play is probably less predictive, but it's a very good descriptive statistic. CPOE is intended to be more predictive; I'm not sure if it actually is or not.
 

Cellar-Door

Member
SoSH Member
Aug 1, 2006
22,821
Yeah that chart feels misleading, because it assume CPOE is equally important as EPA, and I can't think of any reason someone would grade that way. I mean, Herbert has below average CPOE.... it doesn't matter at all, his EPA is really high and he's a really good QB. It's putting way too much weight on completions. In particular... completions are not always good. In many cases an incomplete pass 12 yards down field is a better play than a complete screen. Putting together a chart like this is going to overvalue players who attempt and complete a lot of short passes, regardless of the value of the passes, and undervalue players who attempt more downfield passes, you can attempt to adjust for likelihood of completion, but as far as I know they don't adjust for quality of attempt. So a 3rd and 8 completion behind the line is a positive CPOE play.

I'd just ignore the x axis altogether. It's still a good result for Mac, his EPA is solid for a rookie.
 

Big McCorkle

Member
SoSH Member
May 9, 2021
208
I just want to know what EPA/Play doesn't tell you that CPOE does. It might be more predictive, but when accounting for past performance, if a guy has higher EPA/Play hasn't he just played better/had better results? In that graph for instance, the composite would tell you Kyler has been about as good or better than Brady, even though Brady has higher EPA/Play.
The purpose of incorporating something like CPOE into a predictive measure would be to account for those plays where a QB executes a screen pass to a runningback behind the line of scrimmage who then takes it eighty yards for a touchdown. EPA/Play is, more or less, a very, very accurate measure of true outcome, while CPOE is an attempt to factor in the quality of the process that led to that true outcome. PFF passing grade appears to actually be the best for that, though; it's easily the most correlated measure with future EPA. (Which is partially why I'm generally rather happy to cite their more subjective stuff.)

Ultimately, though, a graph using PFF grades along the X instead of CPOE is going to tell you just about the same thing about Mac Jones: through five weeks he's been slightly below average, as an NFL quarterback. Not as a rookie quarterback, as an NFL quarterback, and much better than any of the other rookies in his class so far.

Yeah that chart feels misleading, because it assume CPOE is equally important as EPA, and I can't think of any reason someone would grade that way. I mean, Herbert has below average CPOE.... it doesn't matter at all, his EPA is really high and he's a really good QB. It's putting way too much weight on completions. In particular... completions are not always good. In many cases an incomplete pass 12 yards down field is a better play than a complete screen. Putting together a chart like this is going to overvalue players who attempt and complete a lot of short passes, regardless of the value of the passes, and undervalue players who attempt more downfield passes, you can attempt to adjust for likelihood of completion, but as far as I know they don't adjust for quality of attempt. So a 3rd and 8 completion behind the line is a positive CPOE play.
So, it's funny, CPOE kind of does the opposite of what you're saying in terms of overvaluing and undervaluing players. Most QBs tend to regress towards the trendline, but there are outliers who have consistently stayed above and below it, in my memory, over the last decade or so. The two prime examples are Brady and Wilson; Brady has always had a substantially higher EPA/play than his CPOE would predict (him being really good at not taking sacks, executing the short-to-intermediate passing game to perfection, just never really outside of a few years having a credible deep threat), and Wilson has always had a worse EPA/play than his CPOE would predict, in part because of him taking bad sacks, and in part because he's ridiculously accurate on deep balls (and has had great targets to go to downfield, in general.) CPOE really favors passers who are good at completing deep shots. Which makes sense, given that the whole point of CPOE is to not give as much credit to the QBs who are completing the easy stuff and give more credit to QBs who are completing the hard stuff.
 

Cellar-Door

Member
SoSH Member
Aug 1, 2006
22,821
So, it's funny, CPOE kind of does the opposite of what you're saying in terms of overvaluing and undervaluing players. Most QBs tend to regress towards the trendline, but there are outliers who have consistently stayed above and below it, in my memory, over the last decade or so. The two prime examples are Brady and Wilson; Brady has always had a substantially higher EPA/play than his CPOE would predict (him being really good at not taking sacks, executing the short-to-intermediate passing game to perfection, just never really outside of a few years having a credible deep threat), and Wilson has always had a worse EPA/play than his CPOE would predict, in part because of him taking bad sacks, and in part because he's ridiculously accurate on deep balls (and has had great targets to go to downfield, in general.) CPOE really favors passers who are good at completing deep shots. Which makes sense, given that the whole point of CPOE is to not give as much credit to the QBs who are completing the easy stuff and give more credit to QBs who are completing the hard stuff.
I get the thought process, but I don't think it really does. It rewards guys with high completions on deep shots, but it also rewards guys who don't take chances, and who take thier checkdowns (that can be good or bad). If you usually take the easy throw you are going to have higher completion rates than average in those distances because the nature of NFL defense in the modern era is they're willing to give you short completions on long to-go plays

I think the historical data back that up.
In 2016 the top 7 included these 3 players: Sam Bradford, COdy Kessler, Brian Hoyer. Those 3 were also all at the bottom of the league in AYTS.
In 2017 the top 4 included Brees and Keenum both near the bottom (or in Brees's case dead last by a lot) in AYTS
In 2018 Brees, Foles, COusins and Cam where high in CPOE, near the bottom in AYTS
In 2019 didn't have any real scrubs to mix in (Brees, Cousins, Carr, Jimmy G) but the pattern continued,
In 2020 it did buck the trend, the top guys in CPOE were throwing further downfield.

Not sure if they tweaked it in 2020, but otherwise it looks like it rewards a particular kind of QB, the one who makes good throws well short of the sticks. Now sometimes that means Drew Brees, but a lot of times it means "game manager" backup types like Kessler, Hoyer, Keenum.

I don't think it's useless, but I also don't think comparing it with EPA as an equal measure makes much sense.
 

Eddie Jurak

Go Leafs Go
Lifetime Member
SoSH Member
Dec 12, 2002
32,327
Melrose, MA
I get the thought process, but I don't think it really does. It rewards guys with high completions on deep shots, but it also rewards guys who don't take chances, and who take thier checkdowns (that can be good or bad). If you usually take the easy throw you are going to have higher completion rates than average in those distances because the nature of NFL defense in the modern era is they're willing to give you short completions on long to-go plays
The bolded seems wrong. If NFL defenses are strategically giving up certain types of passes, then the expected compleition percentage on those throws will be higher. I don't know for sure but I'd assume they are using the same sample to set expected completion percentages for different types of throws as to assess whether QBs are above or below that.
I don't think it's useless, but I also don't think comparing it with EPA as an equal measure makes much sense.
I would argue that it is a misconception of the purpose of graphs to assume that putting one measure on the x-axis and another on the y is somehow intended to claim that they are equal measures. (The point of it is really to see the relationship between the two measures.).

What stands out in this graph is that no one with above average CPOE is bad (using the y-axis as the measure of bad). Indeed, Mac is the worst of the above-average CPOE QBs, and he's still a touch above average. There can be productive QBs who are a little below average at hitting their throws (Jameis this year), but there are also obviously many unproductive ones.

Brady's spot on that graph is also fascinating. In a way he's like the opposite of Mac.
 

Morgan's Magic Snowplow

Member
SoSH Member
Jul 2, 2006
19,591
Philadelphia
I get the thought process, but I don't think it really does. It rewards guys with high completions on deep shots, but it also rewards guys who don't take chances, and who take thier checkdowns (that can be good or bad). If you usually take the easy throw you are going to have higher completion rates than average in those distances because the nature of NFL defense in the modern era is they're willing to give you short completions on long to-go plays
The bolded seems wrong. If NFL defenses are strategically giving up certain types of passes, then the expected compleition percentage on those throws will be higher. I don't know for sure but I'd assume they are using the same sample to set expected completion percentages for different types of throws as to assess whether QBs are above or below that.
I think what CD is arguing is that guys who take checkdowns will end up with an easier body of short throws even when compared to the body of short throws made by the rest of the league. I think there is some merit there. Imagine two kinds of play situations, one a 3rd and 14 and one a 3rd and 4. On the first play its easy to hit a receiver for five yards, on the second its much more difficult because the defense is trying to prevent that. Guys who regularly take the checkdown on 3rd and 14 will end up with a better short pass completion rate. How much this affects the calculations I don't know. 3rd and longs aren't actually very common and if a guy is checking down like 10% more often on those plays vis-a-vis a more adventurous QB it may not actually make a big difference.

All that said, I think there is another plausible reason for CPOE being correlated with AYTS, which is simply that CPOE is essentially a measure of accuracy and guys who don't have the arm/mindset to throw downfield need to be very accurate in order to stay in the NFL. Checkdown Artists who aren't highly accurate with the football become real estate agents, not NFL quarterbacks.

To me CPOE is a useful stat as long as you consider it a measure of an important QB trait (accuracy), which is of course only one important trait among many, not a comprehensive measure of QB value. It also might be the kind of stat that is more useful at the lower end than the higher end. Having a very high CPOE may be only weakly predictive of future excellent QB performance, as you get a mix of guys whose elite accuracy is added to a robust all around skillset and guys who are only keeping jobs in the league because of their accuracy. But having a CPOE lower than about -2 for multiple years is pretty much a death sentence. Some guys have slightly below average CPOE regularly probably because they take so many shots and throw into so many tough windows (Aaron Rodgers is one) and some guys have a freak year where everything falls apart around them (Tom Brady in 2019). Some rookies like Allen and Lamar have also struggled mightily and then improved on this stat, so a low score in your rookie year isn't necessarily a killer (which is good news for fans of the non-Macs, because they all rate awfully so far). But you really don't see good QBs repeatedly put up CPOE worse than -2. If that happens its a pretty good sign that you just lack the minimum level of accuracy that is necessary to succeed in the league, no matter what your other traits might be.
 
Last edited:

Eddie Jurak

Go Leafs Go
Lifetime Member
SoSH Member
Dec 12, 2002
32,327
Melrose, MA
I think what CD is arguing is that guys who take checkdowns will end up with an easier body of short throws even when compared to the body of short throws made by the rest of the league. I think there is some merit there.
This seems absurd on its face to me. Defenses that expect a player to always throw short (or think he doesn't have the arm to complete deep shots) aren't going to respond by conceding the short stuff.
Imagine two kinds of play situations, one a 3rd and 14 and one a 3rd and 4. On the first play its easy to hit a receiver for five yards, on the second its much more difficult because the defense is trying to prevent that.
CPOE considers air yards, down and distance, and spot on the field in determing what the expected completion percentage is.

To me CPOE is a useful stat as long as you consider it a measure of an important QB trait (accuracy), which is of course only one important trait among many, not a comprehensive measure of QB value.
Accuracy on that QBs mix of throws. The QB who throws a lot of deep balls (Russell Wilson) is going to have a lower completion percentage than the one who throws mosts short passes (Drew Brees). CPOE attempts to control for that - Wilson and Brees are both good by CPOE.

Having a very high CPOE may be only weakly predictive of future excellent QB performance, as you get a mix of guys whose elite accuracy is added to a robust all around skillset and guys who are only keeping jobs in the league because of their accuracy.
Judging from the data I've seen, high CPOE puts a floor under a QB. A high CPOE can be roughly average (Mac) to elite (Mahomes) in by better measures of productivity. An average (or slightly above/below average) QB can be anywhere from bad to good. A poor CPOE QB is going to be poor overall.[/quote][/QUOTE]
 

SMU_Sox

queer eye for the next pats guy
SoSH Member
Jul 20, 2009
7,429
Dallas
All football stats have a lot of noise in them. I like ANY/A which is adjusted net yards per attempt. That’s my measuring stick. Mac Jones is currently 27th at 5.12 yards. All the other rookies who qualify (so no Lance) are below that. Lawrence is 28th. Mills is 30th. Wilson is 31st and Fields is 32nd. All of the rookies have atrocious ANY/A’s. Life as a rookie QB especially on teams that have serious roster issues is going to be tough.
 

Morgan's Magic Snowplow

Member
SoSH Member
Jul 2, 2006
19,591
Philadelphia
This seems absurd on its face to me. Defenses that expect a player to always throw short (or think he doesn't have the arm to complete deep shots) aren't going to respond by conceding the short stuff.
CPOE considers air yards, down and distance, and spot on the field in determing what the expected completion percentage is.

Accuracy on that QBs mix of throws. The QB who throws a lot of deep balls (Russell Wilson) is going to have a lower completion percentage than the one who throws mosts short passes (Drew Brees). CPOE attempts to control for that - Wilson and Brees are both good by CPOE.

Judging from the data I've seen, high CPOE puts a floor under a QB. A high CPOE can be roughly average (Mac) to elite (Mahomes) in by better measures of productivity. An average (or slightly above/below average) QB can be anywhere from bad to good. A poor CPOE QB is going to be poor overall.

I didn't realize CPOE also considers down and distance. In that case, I agree with you that the checkdown issue should be controlled for.

I'm not sure about the floor point, though, unless the floor is Sam Bradford, which is kind of a low one. I haven't done any kind of sophisticated analysis but just eyeballing the data I think a bad CPOE tells you a lot more than a good CPOE.
 

Cellar-Door

Member
SoSH Member
Aug 1, 2006
22,821
I didn't realize CPOE also considers down and distance. In that case, I agree with you that the checkdown issue should be controlled for.

I'm not sure about the floor point, though, unless the floor is Sam Bradford, which is kind of a low one. I haven't done any kind of sophisticated analysis but just eyeballing the data I think a bad CPOE tells you a lot more than a good CPOE.
I missed that too, so my explanation probably doesn't make sense, but the question then becomes... why does CPOE love bad QBs who throw short?
Honestly the more I look through scores from past years the less I can figure out what value it has, it doesn't seem to tell us much about QB quality either in descriptive or predictive ways.

I mean, okay, so if a QB has a high CPOE, he has a floor of Cody Kessler/Brian Hoyer.... maybe?
On the other hand some of the really bad QBs do very poorly in CPOE... however, guys with negative CPOE over the last 5 years also include... Brady, Rodgers, Mahomes, Allen, Rivers, and guys can go from very high to very low from year to year, and rookies with really poor CPOE in particular can be good later (Allen, Lamar) and ones with high CPOE (Kessler) can not be NFL players.
 

Big McCorkle

Member
SoSH Member
May 9, 2021
208
I wish I could find it, but there's this really neat graph depicting the N+1 predictive power of various measures in relation to EPA... okay I found it. Mind, and this is noted in the follow-up tweet here, the EPA+CPOE composite is meant to be predictive for the next season, not the current season. Of course, it's worth noting that an r^2 of 0.2 to 0.25 is absolutely not that high in absolute terms-- but is incredibly high for something like football, which is just hilariously noisy in general.

View: https://twitter.com/reinhurdler/status/1440707964243902464/photo/1
 

rodderick

Member
SoSH Member
Apr 24, 2009
8,299
Belo Horizonte - Brazil
I wish I could find it, but there's this really neat graph depicting the N+1 predictive power of various measures in relation to EPA... okay I found it. Mind, and this is noted in the follow-up tweet here, the EPA+CPOE composite is meant to be predictive for the next season, not the current season. Of course, it's worth noting that an r^2 of 0.2 to 0.25 is absolutely not that high in absolute terms-- but is incredibly high for something like football, which is just hilariously noisy in general.

View: https://twitter.com/reinhurdler/status/1440707964243902464/photo/1
Yeah, all I'm seeing is that CPOE by itself is kinda useless for that and that there's very little meaningful difference between EPA/Play and the composite.
 

Big McCorkle

Member
SoSH Member
May 9, 2021
208
Yeah, all I'm seeing is that CPOE by itself is kinda useless for that and that there's very little meaningful difference between EPA/Play and the composite.
The composite is a bit better as the season goes on, but yeah, in-season it's not much better. Do remember that the difference you see as the season goes on isn't nothing, though; the composite is a bit more predictive than EPA/play alone. The big conclusion to be drawn is that PFF would actually appear to be pretty darn good at grading quarterback play.
 

Jimbodandy

Well-Known Member
Gold Supporter
SoSH Member
Jan 31, 2006
5,895
around the way
All football stats have a lot of noise in them. I like ANY/A which is adjusted net yards per attempt. That’s my measuring stick. Mac Jones is currently 27th at 5.12 yards. All the other rookies who qualify (so no Lance) are below that. Lawrence is 28th. Mills is 30th. Wilson is 31st and Fields is 32nd. All of the rookies have atrocious ANY/A’s. Life as a rookie QB especially on teams that have serious roster issues is going to be tough.
This post should be pinned in the forum.

Being a rookie in the hardest position in pro sports is hard. You're gonna struggle.
 

BusRaker

Member
SoSH Member
Aug 11, 2006
1,042
All football stats have a lot of noise in them. I like ANY/A which is adjusted net yards per attempt. That’s my measuring stick. Mac Jones is currently 27th at 5.12 yards. All the other rookies who qualify (so no Lance) are below that. Lawrence is 28th. Mills is 30th. Wilson is 31st and Fields is 32nd. All of the rookies have atrocious ANY/A’s. Life as a rookie QB especially on teams that have serious roster issues is going to be tough.
Made we wonder how the 32 day 1 (to filter out injuries) fared in the first year as the primary QB. I would arrive at conclusion that the QB's that weren't straight from college to starter obviously helped their team out more than the day one people and the ones that came in part way through kicked ass in their second year

Team Day 1 Starter First Year Team Year # ANY/A Notes
Kansas City Chiefs Patrick Mahomes
2018​
Chiefs
1​
8.89​
Dallas Cowboys Dak Prescott
2016​
Cowboys
1​
7.86​
San Francisco 49ers Jimmy Garoppolo
2019​
49ers
6​
7.22​
Minnesota Vikings Kirk Cousins
2015​
FT
4​
7.14​
Houston Texans Tyrod Taylor
2015​
Texans
5​
7.1​
Atlanta Falcons Matt Ryan
2008​
Falcons
1​
7.01​
Seattle Seahawks Russell Wilson
2012​
Seahawks
1​
7.01​
Pittsburgh Steelers Ben Roethlisberger
2004​
Steelers
1​
6.93​
Los Angeles Chargers Justin Herbert
2020​
Chargers
1​
6.84​
Cleveland Browns Baker Mayfield
2018​
Browns
1​
6.77​
Green Bay Packers Aaron Rodgers
2008​
Packers
4​
6.64​
New Orleans Saints Jameis Winston
2015​
Bucs
1​
6.44​
Baltimore Ravens Lamar Jackson 2018 * Ravens
1​
5.99​
Part-time, 8.19 in year two
Cincinnati Bengals Joe Burrow
2020​
Bengals
1​
5.72​
Chicago Bears Andy Dalton
2011​
Bengals
1​
5.65​
Arizona Cardinals Kyler Murray
2019​
Cardinals
1​
5.55​
Denver Broncos Teddy Bridgewater
2014​
Vikings
1​
5.46​
Tampa Bay Buccaneers Tom Brady
2001​
Patriots
2​
5.39​
New York Giants Daniel Jones 2019 * Giants
1​
5.38​
Carolina Panthers Sam Darnold
2018​
Jets
1​
5.24​
Tennessee Titans Ryan Tannehill
2012​
Dolphins
1​
5.23​
Indianapolis Colts Carson Wentz
2016​
Eagles
1​
5.09​
Las Vegas Raiders Derek Carr
2014​
Raiders
1​
4.82​
Buffalo Bills Josh Allen 2018 * Bills
1​
4.37​
Part-time, 5.72 in year two
Los Angeles Rams Matthew Stafford 2009 * Lions
1​
3.64​
Part-time, 6.98 in year three
Washington Football Team Ryan Fitzpatrick
2008​
Bengals
4​
3.58​
Detroit Lions Jared Goff 2016 * Rams
1​
2.82​
Part-time, 7.72 in year two
Jacksonville Jaguars NA
Miami Dolphins NA
New England Patriots NA
New York Jets NA
Philadelphia Eagles NA
 

Jimbodandy

Well-Known Member
Gold Supporter
SoSH Member
Jan 31, 2006
5,895
around the way
I'm not sure that "in their first year as primary QB" tells much in the context of a rookie QB, but it is interesting data nonetheless.

Edit: also, going back to 2008 will swag the data somewhat, let alone 2004 and 2001.
 

rodderick

Member
SoSH Member
Apr 24, 2009
8,299
Belo Horizonte - Brazil
I'm not sure that "in their first year as primary QB" tells much in the context of a rookie QB, but it is interesting data nonetheless.

Edit: also, going back to 2008 will swag the data somewhat, let alone 2004 and 2001.
Yup. Just as a reference point, Brady's 5.39 ANY/A was the 12th best mark in the league in 2001, but would have ranked 29th last year.
 

rymflaherty

Member
SoSH Member
Jun 27, 2010
3,140
Norfolk
Sorry, Dolphins fan interjecting, but I saw the Fins left blank in the grid, which caused me to seek out Tua’s number last year.
If statmuse is to be believed, his Any/A was 5.79.

And I suppose this is the point where it is obligatory for me to moan about how the national media has labeled Tua a bust, people act as if he’s awful, yet his number was higher than any of this years rookies.
 

rodderick

Member
SoSH Member
Apr 24, 2009
8,299
Belo Horizonte - Brazil
Sorry, Dolphins fan interjecting, but I saw the Fins left blank in the grid, which caused me to seek out Tua’s number last year.
If statmuse is to be believed, his Any/A was 5.79.

And I suppose this is the point where it is obligatory for me to moan about how the national media has labeled Tua a bust, people act as if he’s awful, yet his number was higher than any of this years rookies.
Nope, his ANY/A last year was 5.40, per Pro Football Reference (the guys who invented the stat). Still higher than all of these year's rookies. And yeah, as a Dolphins fan I can see how it could bother you that basically no one was sold on Tua last year while you have people talking about how Mac would be the first QB taken this year in a redraft already when at this point they have incredibly similar rate stats (Tua slightly better).
 

Cellar-Door

Member
SoSH Member
Aug 1, 2006
22,821
Sorry, Dolphins fan interjecting, but I saw the Fins left blank in the grid, which caused me to seek out Tua’s number last year.
If statmuse is to be believed, his Any/A was 5.79.

And I suppose this is the point where it is obligatory for me to moan about how the national media has labeled Tua a bust, people act as if he’s awful, yet his number was higher than any of this years rookies.
It was 5.40, though I would say, you can't really compare until the year is over, Rookies often (but not always) put up better numbers as they year goes on, so final numbers will be different (Lawrence in particular has had much better numbers the last couple weeks).

I think Tua's problem is he was inserted into the lineup for Fitz who was a 6.85 last year and Herbert who he was taken before was really good. I don't think Tua's rookie year is a concern, this year is where you want to see improvement.
 

rodderick

Member
SoSH Member
Apr 24, 2009
8,299
Belo Horizonte - Brazil
It was 5.40, though I would say, you can't really compare until the year is over, Rookies often (but not always) put up better numbers as they year goes on, so final numbers will be different (Lawrence in particular has had much better numbers the last couple weeks).

I think Tua's problem is he was inserted into the lineup for Fitz who was a 6.85 last year and Herbert who he was taken before was really good. I don't think Tua's rookie year is a concern, this year is where you want to see improvement.
Tua was the opposite. He had a 100 passer rating (including 7 TDs and 0 INTs) in his first 5 starts, then played his worse games as the season wore on.
 

Cellar-Door

Member
SoSH Member
Aug 1, 2006
22,821
Tua was the opposite. He had a 100 passer rating (including 7 TDs and 0 INTs) in his first 5 starts, then played his worse games as the season wore on.
Yep, not all players do, I was just saying comping full seasons to the first 3-5 games for these guys won't give you the full picture. Like FIelds hasn't been good, but his number is getting obliterated by that first start where his coach tried to get him murdered to justify going back to Andy Dalton
 

rodderick

Member
SoSH Member
Apr 24, 2009
8,299
Belo Horizonte - Brazil
Yep, not all players do, I was just saying comping full seasons to the first 3-5 games for these guys won't give you the full picture. Like FIelds hasn't been good, but his number is getting obliterated by that first start where his coach tried to get him murdered to justify going back to Andy Dalton
Oh, you're completely right in general. Most guys improve and their first starts seem to be shakier.
 

SMU_Sox

queer eye for the next pats guy
SoSH Member
Jul 20, 2009
7,429
Dallas
Mac’s first game was also his best ANY/A.

Overall Mac had a good day. He is playing so well right now IMO for a rookie with mediocre pass pro and no real superstar weapons to throw to. There were some plays I think Mac would want to have back. In college the underneath defender doesn't have the range like underneath defenders do in the NFL - either Mac is missing the underneath defenders dropping back or he isn't respecting their range. He could have had a couple throws picked that way (though both would have been difficult picks). I didn't think it was a velocity problem either as both these throws were in the 10-20 yard range and he stepped into them. On one of them he hesitated too much and gave the underneath guy way too much time to recover and nearly pick it off.

He absolutely has to stop attempting long throws when he doesn't drive. If he isn't going to step into the throw downfield, well... DO NOT THROW IT. He has done this flick throw in at least 3 of the games so far and each time it is disastrous. The ball just dies and is easily picked off or batted down. He had the opportunity to step into it too (both times he did it this game). Or maybe he did on one and on the other he didn't - but, again, he just doesn't have the arm to make that throw unless he can step into it. Even if Mac Jones eventually improves his arm strength he will never be someone who can rifle it in off his back foot. That's ok - for him to win he has to get the ball out earlier, step into throws, and use the clean pocket to his advantage.

He also hesitates/double-clutches and had one or two plays where he locked onto a guy too early and the defender just read his eyes and made the break. Those are all things he can improve upon. I think he has the skillset to be a good game manager and as long as he is protected make the right reads and good throws. He can sustain drives in-structure.
 

SMU_Sox

queer eye for the next pats guy
SoSH Member
Jul 20, 2009
7,429
Dallas
It's no secret the Patriots have struggled to throw the ball more than 20 yards down the field. I decided to do a deep dive on it.

I thought coming out Mac's arm was the most similar to Burrow's (although Burrow's pinpoint accuracy was... just extraordinary).

Burrow notoriously struggled downfield last year too.

Via PFF's charting stats.

Burrow in 2020 had the 37th rated adjusted deep completion percentage out of 39 qualifying QBs for deep passes (20 yards of air+): 20.8%. In 10 games he had 48 attempts, 9 completions, and 1 drop for an 18.8% completion rate. He attempted a deep throw in 11.9% of his dropbacks which was 20th overall. The high was Drew Lock at 16.0% (Flacco was #1 at 19.4% but he's one of the 7 part time starters). Brady was next at 14.9%. The median was Matt Ryan at 12.5%. Those 9 receptions went for 293 yards, 6.1 yards per attempt - also 37th worst in the league, with 1 TD and 1 int. He had 3 turnover-worthy-plays on those 48 attempts, and a TWP% of 5.6 - 11th best in the league. He took the 13th longest time to throw on those at 3.17 seconds.

This year Burrow is 23rd in adjusted deep completion percentage at 41.2%. He is 7 of 17 for 4 TDs and 1 INT with no drops and 16.8 yards per attempt - 6th best in the league. He attempted a deep throw in 11.7% of his dropbacks which is 19th overall. He has 2 turnover worthy plays, a 10.5% rate - 7th worst. He is 11th in time to throw at 3.36 seconds. Burrow is one of 7 guys this year who have been pressured 7 times on their dropbacks, tied for 15th highest, so he slightly overindexes on his under pressure rate for dropbacks. That's a lot of improvement so far in the year. He made great strides with his ADJ% and Y/A.

Jones in 2021 has the 33rd rated adjusted deep completion percentage out of 33 qualifying QBs for deep passes (20 yards of air+): 21.1%. In 5 games he has 19 attempts, 4 completions, and no drops* for a 21.1% completion rate. He attempted a deep throw in 10.0% of his dropbacks which is 26th overall. The high is Derek Carr at 17.1% (Fields is #1 at 18.1% but he has 2 starts). Teddy Bridgewater is next at 16.8%. The median is Davis Mills at 12.5% like in 2020. Those 4 receptions went for 97 yards, 5.1 yards per attempt - lowest in the league, with 0 TD and 2 ints. He has 4 turnover-worthy-plays on those 19 attempts, and a TWP% of 16.7%, 2nd worst in the league. He has the 9th fastest time to throw on those at 3.01 seconds. Jones is one of 5 guys this year who have been pressured 8 times on their dropbacks, tied for 10th highest, so he greatly overindexes on his under pressure rate for dropbacks.

*The Meyers drop last week against HOU doesn't count because the Texans committed and the Patriots accepted the penalty.

Mechanically I have had some serious issues with at least 5-6 of his deep passes. He won't fully step into the throw and drive it and it dies. I mentioned that earlier. He also has been late on a handful. It doesn't help that he has not always had a clean pocket and didn't get to practice much this summer with Agholor, his primary downfield threat. Agholor is not a WR1 downfield ace either... and Bourne and Meyers are 4.6 guys and not known for their deep receiving prowess. The Patriots offense is also historically more of a horizontal offense. What was Tom like back when he played for the Pats who also had issues in 2018 and 2019 with sub par receiving groups (don't wanna argue this). How about Tom on the Bucs? What about Tom in 2007?

2021: 14.2%, 12th. 12-32, 50.0% ADJ% (11th), 339 yards, 2 TDs 1 INTs, 10.6 Y/A (28th), 0 TWP 0.0% (1st).
2020: 14.9%, 3rd. 36-91, 42.9% ADJ% (23rd), 1,233 yards, 11 TDs 4 INTs, 13.5 Y/A (19th), 6 TWP 5.7% (12th).
2019: 10.1%, 27th. 25-62, 41.9% ADJ% (19th), 728 yards, 7 TDs 3 INTs, 11.7 Y/A (21st), 5 TWP 6.7% (15th).
2018: 11.4%, 19th. 23-65, 43.1% ADJ% (12th), 714 yards, 8 TDs 2 INTs, 11.0 Y/A (24th), 5 TWP 6.9% (16th).
2017: 13.8%, 8th. 33-80, 46.3% ADJ% (6th), 1,104 yards, 5 TDs 4 INTs, 13.8 Y/A (9th), 5 TWP 5.2% (11th).
2016: 11.3%, 22nd. 23-49, 51.0% ADJ% (8th), 834 yards, 8 TDs 1 INTs, 17.0 Y/A (4th), 2 TWP 3.6% (6th).
2015: 10.7%, 26th. 24-67, 44.8% ADJ% (11th), 833 yards, 3 TDs 2 INTs, 12.4 Y/A (14th), 4 TWP 4.9% (5th).
2014: 10.3%, 35th. 17-60, 33.3% ADJ% (32nd), 528 yards, 5 TDs 2 INTs, 8.8 Y/A (35th), 3 TWP 4.3% (10th).
2013: 10.5%, 31st. 20-66, 39.4% ADJ% (25th), 694 yards, 4 TDs 4 INTs, 10.5 Y/A (26th), 5 TWP 6.3% (16th).
2007: 12.6%, 19th. 31-73, 50.7% ADJ% (3rd), 1,192 yards, 16 TDs 5 INTs, 16.3 Y/A (6th), 6 TWP 7.5% (10th).

Mac
2021: 10.0, 26th. 4-19, 21.1% ADJ% (33rd), 97 yards, 0 TDs 2 INTs, 5.1 Y/A (33rd), 4 TWP 16.7% (33rd).

A lot to digest here. I think the data shows you scheme and what the play-calling was mattered and also that your roster (OL + WR + TE) matters. When Brady had good pass catchers his rates and raw stats look better. When he didn't they didn't. Brady always had a low TWP rate - so even when the team wasn't good at deep passing he wasn't turning the ball over. In an average year here he took about 4 shots a game and 5 in years when he had the supporting casts. So we're still talking about a small number of passes per game which are prone to some volatility based on the sample size. Aside from 2014 Brady gave the team a floor of around a 40% ADJ. In Tampa not only does he have a terrific supporting cast but he also is in a scheme that takes more shots. It wasn't that Brady couldn't throw deep but in most years the offense and the play-makers weren't deep threats. When they were Brady threw it deep. Part of what made Brady great was taking what the defense gave him and hitting on it - and if you look at his time in NE while he didn't throw deep as much in the years he didn't have the horses to do it when the defenses gave him those deep shots he usually executed well on them (adjusted completion percentage).
Mac hasn't looked good on deep passes. Now while these passes are only 10%-11% of the Pats offense in a typical year they can account for close to 15-20% of the passing yards. His turnover worthy play rate is simply unacceptable. His adjusted competition % is also woefully bad. It's been 5 games and there is a laundry list of reasons why the deep passing game isn't and hasn't worked well for this team and Mac shouldn't be shouldering the blame - it's a team problem though.

This is just something to keep an eye on as we watch Mac going forward. It's also something that we should look at for the offense as a whole and each part of it: pass pro, skilled position players, and offensive play-calling.
 
Last edited:

rymflaherty

Member
SoSH Member
Jun 27, 2010
3,140
Norfolk
Appreciate the additional info @Cellar-Door
I’ll leave it at that, as I wasn’t trying to derail things. Interesting conversation here, and it definitely inspires me to go dig into some more numbers as it applies to my team.
 

Devizier

Member
SoSH Member
Jul 3, 2000
15,087
Somewhere

rodderick

Member
SoSH Member
Apr 24, 2009
8,299
Belo Horizonte - Brazil
Came across this and it made me laugh given how much we've discussed the strengths/weaknesses of PFF's grading:
View: https://twitter.com/BarryOnHere/status/1446677297524846592


Apparently drilling the CB directly in the chest isn't turnover worthy.
Barry is among the top 5 worst Twitter accounts. And yeah, a low throw away from the defender which the receiver gets a hand on is pretty borderline for "turnover worthy". I think PFF is way more charitable with that stat than most people think. If they weren't, Mac would have had about 5 of those against the Texans and they only counted 3.
 

Cellar-Door

Member
SoSH Member
Aug 1, 2006
22,821
That's literally a top 5 worst Twitter account. And yeah, a low throw away from the defender which the receiver gets a hand on is pretty borderline for "turnover worthy". I think PFF is way more charitable with that stat than most people think. If they weren't, Mac would have had about 5 of those against the Texans and they only counted 3.
the account it a troll account I get, I just want the video. If that's not a TO worthy throw the stat is useless. Sure it's lowish, but the WR doesn't "get a hand on it" he tackles the defender who is about to make a fairly easy pick. This is far from the first time we've seen this with PFF, they have really weird vague standards on these that don't make any sense, some of their stuff is good, but TWT is a spot that really emphasizes how much of their stuff is really shaky because it's opinion calls with shaky standards and a tendancy to give QBs the BotD.
 

rodderick

Member
SoSH Member
Apr 24, 2009
8,299
Belo Horizonte - Brazil
the account it a troll account I get, I just want the video. If that's not a TO worthy throw the stat is useless. Sure it's lowish, but the WR doesn't "get a hand on it" he tackles the defender who is about to make a fairly easy pick. This is far from the first time we've seen this with PFF, they have really weird vague standards on these that don't make any sense, some of their stuff is good, but TWT is a spot that really emphasizes how much of their stuff is really shaky because it's opinion calls with shaky standards and a tendancy to give QBs the BotD.
I don't think that pick is fairly easy, but I'd have no problem characterizing it as turnover worthy. It's just that PFF is conservative in general when charting those, every QB will have discrepancies in that department. It's impossible to make those judgements without subjectivity, but in a large enough sample size I think they end up being reflective of how a QB has played (just go back to the graph posted above that indicated PFF grade more accurately predicts future performance than EPA/Play). Of course if you go play by play any stat, subjective or objective, will look stupid. A QB will have positive EPA on a play in which the defender bats an easy pick up into the air and the receiver snags it. Won't mean the stat is dumb.

What racist troll Barry McCockiner is doing there is his usual schtick of "Brady only throw checkdowns to open receivers and PFF are biased towards him".
 

Sille Skrub

Dope
Dope
Mar 3, 2004
5,250
Massachusetts
@elias I'd like to address some of your post.

The Pats absolutely did not "do Cam dirty." Mac won the job fair and square in camp. Almost every report out of camp was that Mac was outperforming Cam. Now, I don't know if these are to be trusted, but they are all we had to go on without actually being there. If there was any unconscious bias or effort to elevate Mac for reasons other than what happened on the field, it would have had to be a collaborated effort among every single member of the football media covering the Patriots camp. Furthermore, even if their performances were equal, the Patriots invested a first round pick in Jones, the first time they used a first round pick on a QB since Bledsoe; why wouldn't they want to start the development of said asset as soon as they thought he was ready?

As for doing Cam dirty, in 2020, when no other team expressed any serious interest, the Patriots paid Cam $3.5mil to throw footballs above receivers' heads, in front of them and at their feet. It was an awful performance. Furthermore, even after that dumpster fire of a season, they brought him back and gave him every opportunity to win the job back. They have paid him $3.5mil this season to do whatever he pleases except throw more of those said passes at receivers' feet. If that is doing someone dirty, please sign me up every day of the week and twice on the proverbial NFL Sunday.

I will agree that some of the talk about Mac has been silly and that this board has taken pleasure in mocking other boards making similar posts about their players. However, this is what being a fan is all about. It shouldn't be a shock to anyone that on a Boston sports based sports message board that there are actually posts slanted towards athletes on Boston teams.
 

Cellar-Door

Member
SoSH Member
Aug 1, 2006
22,821
I don't think that pick is fairly easy, but I'd have no problem characterizing it as turnover worthy. It's just that PFF is conservative in general when charting those, every QB will have discrepancies in that department. It's impossible to make those judgements without subjectivity, but in a large enough sample size I think they end up being reflective of how a QB has played (just go back to the graph posted above that indicated PFF grade more accurately predicts future performance than EPA/Play). Of course if you go play by play any stat, subjective or objective, will look stupid. A QB will have positive EPA on a play in which the defender bats an easy pick up into the air and the receiver snags it. Won't mean the stat is dumb.
I think the QB grades are generally pretty good, but I think TWT is the worst element, and one where they have a lot of inconsistency internally. I also think they give the benefit of the doubt some places and not others, it's human nature probably, that a QB you know is good gets BotD on borderline, but it's a flaw. I think PFF needs a better more consistent standard on TWT if they want to use it, as is it feels like a free for all. I just grabbed that video because it came up in someone's feed, but it's a thing I've been annoyed at for a while, they have no real standard that I can tell, and seem to routinely call actual turnovers, and dropped picks not-turnover worthy for dubious reasons.

Edit- and on Mac, that he only got credited with 2 TWT against HOU is insane, he had several more pretty clear ones. My issue is... they seem to not want to give TWT to QBs, but if we want to use it as a measurement of what it says it is they need to actually do that.
 
Last edited: