The fact that in the middle of the release of this series we got Jared Goff's 18/18, 292 Yds, 2 TD, 0 Int game which recieved less (155.8) than a perfect passer rating (158.3) is in itself perfection.
Even funnier is that since it was only 18 passing attempts, it wouldn't have qualified to make this list. Much less funnier was that it was my Seahawks surrendering that pristine game. :,)
If Goff threw one more TD he would have gotten a perfect passer rating. TD percentage (i.e. the ratio of completions for TDs relative to the total number of attempts) factors a lot in passer rating, and rightfully so because it's directly resulting in points.
Removing the arbitrary limits in the passer rating formula, Jared Goff's "true" passer rating from that game would be just a smidge over 190. If you include Amon-Ra's 7-yard passing TD to Goff, the Lions as a whole were 19/19 for 299 yards, 3 TD, 0 INT (a "perfect" 158.3 using the NFL's formula) and the team's "true" passer rating would shoot up to 203.6, although that game still doesn't meet the 20 pass attempt minimum which they used for most of the videos in this series.
Alex’s legacy was already one of the greatest sports creators on the internet. But this? This might be the magnum opus to this point. And we don’t even know if he’s in his prime yet. Incredible work Alex. Jon has Scorigami, but Alex has true passer rating. You’re changing sports data with this one
We always knew (in no small part because Jon is awesome about saying it and reminding us) what an absolute Hall of Fame-caliber research guy Alex is. Now we're seeing that Alex can more than hold his own as a storyteller. Welcome to the S tier, Mr. Rubenstein.
tinyurl.com/37hbe38f for now as a primitive method, you can at least use that as a template and replace the numbers with those from whichever game your heart desires
Scorigami has already been made mainstream by ESPN. I believe that this mini series could cause a new and improved passer rating system. Great job Alex!
Another major weakness of passer rating is that it doesn't take sacks into account. Most modern research seems to suggest that sacks are more of a quarterback stat than an offensive line stat, and passer rating doesn't punish quarterbacks who hold on to the ball for too long and take the sack, but it does punish those who thrown the ball away. Looking at the three NFL quarterbacks with career passer ratings over 100, two of them (Aaron Rogers and Russell Wilson) have reputations for holding onto the ball for too long and taking a lot of sacks.
Im not sure what research told you sacks are more on the qb than the oline, but they might need to do some more of it Lol I do agree about qbs who hold the ball too long though. But for the most part, a lot of sacks are unavoidable
@@postverifiedlifestyle complete speculation on my part but it seems the research OP alluded to found a strong positive correlation between sacks taken and quarterbacks holding onto the ball. the longer the QB holds the ball, the more likely he is to get sacked.
Why does NO ONE criticize Rogers for this? All people talk about are his low interceptions. There's a reason he hasn't been to the SB in almost 15 years. Peyton, Montana and Mahomes got to 4, Elway got to 5, Brady got to 10 and I don't understand why Rogers is considered in that elite territory. He's probably a top 10 QB but he's not better than Unitas, Favre or Marino.
It also punishes QB's who have receivers drop the ball when it hit's them in the hands and it pops up for a in INT. There need to be errors like in baseball that don't account for those picks in the stats.
I created an 'inverse' QBR for QB's awhile back to sort of rank QB's by awful-ness per week. It assumes that 6 turnovers is the "worst" based on Jake Delhomme's terrible birthday game. There is a bunch more to it of course, but your formula allowed me to adjust mine and get even finer results. Thanks for your series on this!!
In my opinion, any metric that means to assess a quarterback's performance has to include sacks, fumbles, and rushing. It doesn't necessarily need to be situational because most situational statistics aren't available before 1994, but a 20-yard completion from one's own 15 on a 3rd and 30 in the waning seconds of the second quarter is fundamentally different from a 20-yard completion on a 4th and 17 from the opponent's 25 with less than a minute remaining in the 4th quarter.
That's the normal pronunciation for that name; blame mildly bastardised Italian spelling never updated to reflect changes in the English pronunciation over time
@@kourii Or an Italian American getting their name Anglicized. Not sure if it was a conscious decision by them to pronounce it that way or it's just how everyone around them found it easier to pronounce and they just rolled with it.
@@lordgemini2376 Italian-Englishman, actually. The name predates the American colonies. For the latter claim, contrast how Taliaferro has found itself pronounced versus more mangling of Italianate surnames in the US. There's a reason this one raises eyebrows among English speakers thitherto unfamiliar with it
Completion percentage has some some value since Anthony Richardson leads the league in Y/A with a 50% completion, I'd replace the completion % with first down % to avoid rewarding passes that were completed but didn't result in a 1st down or even negative yardage
Yeah, the Foles/Smith example given actually shows why it should be in there separately. Foles' yards per attempt is only just over 1% better, while Smith's completion percentage is over 8% better, so a 2% higher passer rating seems more than fair. It doesn't help that Alex thinks that 14.5 yards is 1.5 yards more than 14.35, when it's actually less than 6 inches.
@@tokinsloff312 The 1.5 yard difference was quotes for "yards per completion", the 14.5 and 14.35 numbers are "yards per attempt". The point is that Foles had a significantly better yards per completion, which with a worse completion percentage puts him about the same on yards per attempt - but he loses out on passer rating because the completion percentage is effectively double-penalised.
@@DumbMuscle Ok, my mistake, but doesn't that mean that Foles was missing a lot of short throws and compensating with a few big completions? Does that really make it a better passing performance? Surely it's better to be more efficient and put the ball at risk less often. Either way, a 2% higher passer rating for Smith is hardly a travesty. As the OP pointed out, if you don't account for completion percentage separately, then hitting a couple of big bombs skews the rating in a small sample size.
@@1998_MINEven if it did, Goff's uncapped passer rating would "only" be 190.0. Good enough for top 1% all time, but not like a top-10 performance or anything.
@@bendubz9000 Adding different passers together would break the experiment and all of a sudden you're not even analyzing anything. That's like Stacey King's "Together, Jordan and I combined to score 70 points" thing.
I always felt, going back to the '90s, that yards-per-attempt should be replaced by yards-per-completion if you wanted four elements, for exactly the reason of removing double-counting completion percentage. I think just removing completion percentage works, but there's something satisfying too (IMO) in having both accuracy and production being separate elements represented in the formula, rather than having accuracy baked into one of the elements.
I think TD% is an unnecessary metric. Why does a receiver who gets tackled at the 1 count for less than somebody who managed to stretch out and break the plane? Meanwhile, sacks and fumbles are hugely important, but not counted at all.
The problem with that is that you can make the same argument for all the stats. Completion percentage doesn't account for having a contested catch monster as your receiver, or receivers dropping easy catches. Yards per attempt doesn't account for YAC or defenders whiffing easy tackles. Interceptions doesn't account for defenders dropping easy catches or receivers making great "defensive" plays.
If we are going to rewrite the formula a bit, then I think we should include sacks as a negative somehow. Sometimes it is poor line play, but that can be true of interceptions as well. A sack is a failure to attempt to pass. You could maybe add sacks to the attempts for yards per attempt.
I wouldn't mind swapping out yards per attempt with adjusted net yards per attempt, that takes into account QBs who are good in the pocket/mobile - only the end results of throws are currently taken into account, rather than all passing plays. We can rework the weights to keep the ranges the same for aesthetics if we need to.
My suggestion: Referring to the passing rating that's displayed as "Adjusted Passer Rating" or APR, and their true passer rating, without constraints, as "Uncapped Passer Rating" or UPR. APR would be more of a relative sample throughout a season, and give a rough idea of how well they played on average, while UPR would more accurately show how amazing or awful they were on a game-to-game basis. Plus, it'd be easier to distinguish the two that way!
It would be fascinating to dig deeper into Luck's numbers and see why the passer rating undervalues him. Perhaps this would give a clue about a better way to do the calculation.
The trouble is that completion percentage is a complementary to yards per attempt. If you complete a pass behind the line of scrimmage and the receiver gets tackled immediately, you take a YPA hit but you still did something fundamentally "passer" that you should arguably get credit for. But you also made a bad decision that you deserve a penalty for. Having both in the mix allows you to balance them in your rating.
such a great set of data; have you considered using the data to collect all of the new ratings of each player and re-rank the players according to the new rankings based on their total output?
I mean you must have completion % AND YPC. You can’t have one without the other. A really really good example is Lamar vs Joe last week. I was at that game. Lamar was hitting passes at an incredible clip it seemed bc they were short slants the bengals linebacker couldn’t cover. He had two balls I counted only an All Pro could make. He struggled to throw to the right side entirely. If you’re hitting a majority of short passes on broken coverage 6 yards in the air, I want to know. Joe was airing the ball out and his arm was the weapon. It’s crazy how much better a pure passer Joe looked by a wide margin but the data wouldn’t immediately show that. Thats why in my algorithm I also include both YAC and sacks. YAC bc for the example I just used, a high majority of Lamar’s yards were his receivers doing WORK post catch. That’s not a QB stat. Even more so when the bengals notice they struggle to cover shifty receivers so they back off and it only gives Lamar more of an option to hit short slants. Im not saying it should be punished but a YAC heavy game in no scenario is a quarterback friendly environment - it should be isolated. Sacks as well should be considered in tandem with pressure and a signal callers choice to hold or throw. Back to Lamar, he’s fantastic at avoiding sacks. Multiple times he saved 5 or 6 yards which they kept crucially converting on 3rd that saved that game. Joe on the other hand has a tendency to not give a play up and sometimes wants to stay in the pocket too much . This is tough to distinguish but it’s why this is so hard.
This a pretty gross mischaracterization of Lamar’s game last week. Also a gross mischaracterization of Burrow’s game they torched the Ravens on slants and hitches on the boundary all game and one of his touchdowns was thrown behind the LOS that Ja’Marr Chase took like 70 yards.
Lamar hit a deep seam to Charlie Kolar, a layered shot to Mark Andrews on a deep crosser against tight man coverage, and couple of deep outs to Zay Flowers that were thrown with anticipation. Also hit Mark Andrews on a wheel route up the left sideline, and several times hit throws over the middle of the field that were backside reads. None of them had huge YAC gains and were downfield throws.
as i've been watching these videos, i also had the idea of reconfiguring how passer rating is measured, and then adjusting the ratio to make 100 a "good day at the office"--better than average but nothing outstanding--and making 0 something where you'd go "rough one out there today" but not end up with a wikipedia page because of it. take off the top and bottom limits and just leave it from there. then i realized that's kind of how they came up with it to begin with, right? they picked a contemporary bad game and set that at 0, a contemporary good game and set that to 100 (or 90 or 110 or whatever), and just rolled from there. which really prompts the question: why did they put the limits on? was it solely an ego thing where the historically awful games are rated the same as the generally bad ones?
Could we get an episode where Alex shows how his version of passer rating changes what the currently defined passer rating is? I think it would be fun to see who gets an even bigger uptick and who digresses from what their current game passer rating is.
tinyurl.com/3685fzyj sheet 1 there sorts best to worst by actual passer rating without the completion percentage element (9/8/19 lamar remains best, however 10/19/75 joe namath replaces 10/9/72 pastorini at bottom), sheet 2 sorts it by who got screwed the most from it to who benefited the most from it
Love this series, but it's genuinely INFURIATING that they haven't once shown what the big graph would look like compared to the big graph with the NFL's artificial constraints. I guess it would be similar but with a ton of names bunched at 0 and 158, but it would just be awesome to see a comparison of the two. I keep waiting for him to say "here's what the NFL's chart looks like, and here's what mine looks like," but it never comes. Pleaaaaaase!!
quick question, do you think total attempted passes should be accounted for in passer rating? I'm imagining scenarios in which QB performances are either underrated or overrated thanks to the other parts of the game, especially when the ground game is either really good or really bad. Like when QBs are stuck without a solid ground game to fall back on, they're probably forced to throw worse passes. Or when a defense is really scared of a ground game that is popping off, it's entirely possible for that to create "easier" passes and likely less downs in which the team even bothers to try a passing play. More importantly, a really good running game probably means the QB has more realistic attempts at touchdowns, and at least on paper, there's nothing in passer rating that would account for that
I think passer rating could use a fourth component: Pass attempt percentage, the percentage of a team's offensive plays that were pass attempts. Maybe exclude spiking the ball to stop the clock and kneeling to run out the clock from the denominator. But the idea is very simple: Even with a fixed minimum of pass attempts, It's a lot harder to put up good numbers passing it say, 20 times out of 30 plays (against an opponent like the NCAA service academies running down the clock with 10-minute drives) than 20 times out of 60 plays (thanks to spending most of the game handing off to Adrian Peterson).
the 14.5 to 14.35 is yards per attempt, foles did indeed average over 1.5 more yards per completion (thus producing a slightly higher yards per attempt despite the inferior completion %)
Personally, I'd prefer some kind of measure that doesn't punish the QB for having bad receivers. I guess it would be best described as 'Accuracy'? Obviously it would be a stupidly hard stat to track and measure, especially for older games, but if a QB puts the ball right in his receivers chest and the guy drops it, or even bumbles it to the other team, it's weird to me that the QB is the one who's statistically punished for the mistake.
I think that's a different thing you're trying to measure. Someone like Mike Vick or Eagles-era Randall Cunningham isn't necessarily _better_ than Drew Brees or Joe Montana because they get a big chunk of their yards on the ground, but you could certainly have something like a QB Total Offense Rating. And in the case of someone like Fran Tarkenton, I almost think you'd have to weight the rushing yards he got from scrambling more heavily than you'd weight the same yards from a QB in 2024 because it was so much rarer for a quarterback to do that in 1964 or 1974. Tarkenton would come off as even more transcendent than his already Hall of Fame career made him look as a pure passer. Passer rating is exactly that...one's effectiveness throwing passes, and on some level, the mobile quarterback still has an advantage because the NFL counts sacks as negative passing yards (college treats it as a run statistically speaking.)
@@badgoogle4509 because the shorthand is ‘QB Rating’ and is used to evaluate QBs against one another. If he’s putting half the RBs yard on the field and it goes into the void …
Passer rating is fine. The problem is people take it out of context to make it mean more than it actually does. Passer rating, which is short for passer efficiency rating, only measures, as its name suggests, passing efficiency. It isn't an all-encompassing measurement of quarterback play. Fumbling the ball or taking a sack before throwing a pass is obviously not going to negatively affect passer rating, because passer rating only measures passes thrown. Heck, a wide receiver can have a passer rating like Mohamed Sanu, who has a career perfect passer rating. A defensive tackle, like Dontari Poe, who threw a TD pass as a member of the Chiefs, has a passer rating. Anybody who throws the ball in a game that isn't negated by a penalty has a passer rating--QB, WR, kicker, DE, whoever. The more people understand this, the less controversial this metric would be. If you want a more in-depth measurement of QB play, use ESPN's QBR or PFF's grading system. People who complain about or overrate passer efficiency rating never seem to use it in its proper context.
Two more metrics I'd like to see factored in somehow: YAC vs. yards in the air, and INTs that were the receiver's fault. On the one hand, I don't think QBs should be automatically penalized for yardage totals that rely heavily on YAC (a lot of that is scheme and skill at reading defenses), but on the other hand, some receivers do absolutely crazy shit to bail out their QBs. And with INTs, to me there's a huge difference between throwing a terrible pick because the QB is inept, and throwing an on the money pass that the receiver bobbles right into a DB's hands
Not all YAC is created equal. There's a big difference between a screen pass where the receiver breaks two tackles to free himself and, say, getting the deep safeties to bite on play action from your own goal line, airing it out halfway down the field to an open receiver, and having said receiver take it the remaining 45 yards for the 99-yard touchdown. Which in turn creates a problem with trying to factor in YAC and Kadarius Toney-induced pick-sixes; it's always going to be a judgment call and the NFL would have to employ something like MLB's "official scorer"...at least with current metrics, where the ball is spotted on the field is where the ball is spotted on the field, guaranteed statistical uniformity. (eta: yes, I read your comment before I made mine, it was only after the fact that I, in my Friday night 41-hour-workweek-induced brain fog, realized I pretty much repeated back what you said in my first paragraph. I stand behind the point about the "official scorer" however.)
@@SimuLord Agreed! Would be cool to have an official scorer in the NFL. I was also thinking today, I would love to know which of the better than 158.3 games Alex identifies here came against the hardest pass defenses (we could use DVOA for a start). It's like a pitcher throwing a no-hitter against a team with a stacked lineup
There needs to be an overall rating that considers rushes, fumbles, sacks, delivery time etc. I guess QBR makes an attempt at it, but it’s very difficult because you have to consider every play and circumstance
Your issue with the double inclusion of completion percentage is my biggest problem with OPS and why I think it’s a terrible stat. 1. It adds two stats together which is already bonkers for a statistic to do and leads to… 2. It values singles twice as much as just getting on base. Slugging only considers at-bats and not total plate appearances, so due to that and how OPS is calculated by just adding two stats together, it means hitting to get on base is valued more than just getting on base, which doesn’t make sense since a single or a walk are the same result. An improved OPS would be Total Bases / Plate Appearances, as that quantifies how many bases you expect a player to net just by going into the batters box. It’s simple, it’s elegant, it’s not adding two things together and double weighting something that shouldn’t be.
a single and a walk are often not the same result when baserunners are aboard...you cant ever advance a baserunner more than 1 base on a walk, nor can you ever advance a baserunner even 1 base on a walk if 1st base is open
I completely disagree about completion percentage being removed. You can't tell me 1/5 for 80 yards and a TD is just as good as 5/5 for 80 yards and a TD. 5/5 eats more clock and likely doesn't result in a punt. 1/5 could easily result in a punt since 4 plays are not productive at all.
I understand your point and agree to an extent, but also have a counter. 1/5 and 5/5 are not equally good, but when both have 80 yards and a touchdown attached. 1/5 clearly didn't result in a punt because 80 yards and a touchdown was accrued. In the aggregate, going 5 for 5 with 80 yards is a lot more likely than going 1 for 5 with 80 yards. However, in this argument we've determined that 80 yards and a td are the actual results that occurred and it would be wrong to say that one deserves a higher passer rating than the other because it's more likely to happen again. Passer rating is like WAR in baseball. It does not care about circumstance or likelihoods.
Definitely agree. Completion percentage is one of the best simple measurements we have for accuracy, and really the only one that extends far enough back in history to be useful for broad comparisons. Sure, it has its flaws, but for the most part a QB who completes 65% of his passes is going to be better than a QB who completes 55% of his passes.
Also if we're adjusting the formula fundamentally, why not also normalize the final value to be on a scale from 0 to 100? Negative values never sit right with me, and the ceiling shouldn't be some arbitrary number.
Think another good addition is incorporating volume. So the number gets higher with volume if you're above average and lower with volume if you're below average.
Rewinder: 2006 Orange Bowl 3OTs Beef History: Eric Lindros vs Philadelphia Flyers Organization Overlap: Greatest College Football Coach (Saban) and the Greatest NFL Coach (Belichick) coaching in the same division Beef History: Michael Jordan and Isiah Thomas Beef History: Phil Jackson and Pat Riley Untitled: Reggie Miller Untitled: Barry Sanders Untitled: Carl Yastrzemski Untitled: Ted Williams Untitled: Jim Kelly Collapse: Early 90s Buffalo Bills Collapse: 1987-2000 Florida State Seminoles Collapse: Moneyball (2002) Oakland Athletics Rewinder: 2012 Olympic 100m Men’s Final Rewinder: The Catch By Willie Rewinder: Wide Right I, Wide Right II Rewinder: Hank Aaron and Barry Bonds breaking the respective HR record(s) Untitled: Don Nelson (Coach)
I agree with no artificial boundaries when purely looking at data, but this data is still created by ppl whos livelihood depends on it, so having an artifical floor of somwhere between 0 and -10 seems reasonable to me and maybe release the actual data after retirement, but them having a bad performance must not be turned into public mockery, no one deserves that, especially not for playing a game badly
We have play-by-play data stretching back 30 years, which ain't nothing. The super advanced stats like air yards and throwaways have only been available for a decade or so.
@@dfp_01One angle you could go back generations with first downs passing (and rushing for that matter) is a team perspective though. I.e. The total number of passing and rushing first downs allowed and produced for seasons are known going back to 1941.
This was a good series but it also feels like it came... 5 years too late? like these days we have QBR that is a far more informative metric than passer rating, and it's not even a way off to the side only the analytics crowd likes it (like EPA/play) type of stat, its even on espn, cbs etc this isn't to say QBR is perfect either for the record
Shouldn’t fumbles be part of this since in a way they are worse than interceptions because fumbles usually happen not so far down feel like a deep past might
Interesting to see just how much of a passing league the NFL has become. When graphed by number of attempts, the graph is so much denser. It also seems like the rating goes up over time. Is that because interceptions get you pulled out of the game earlier?
Yeah and because there are too many rules protecting the precious quarterbacks Today's top passers would never develop if they had to go up against 70s defenses
@@1998_MINThat and quarterbacks have gradually been getting more accurate with the ball over time. It used to be that they would just sling the ball all over the yard with reckless abandon and hope their receiver would come down with it
I think SecA is closest to what you’re looking for. It’s secondary average. It includes base on balls, total bases, hits, stolen bases, caught stealing, and at bats.
It shouldn’t be. A double could drive in a run from first and certainly will from second. A single might score a run from second and won’t score a run from first.
The fact that in the middle of the release of this series we got Jared Goff's 18/18, 292 Yds, 2 TD, 0 Int game which recieved less (155.8) than a perfect passer rating (158.3) is in itself perfection.
Even funnier is that since it was only 18 passing attempts, it wouldn't have qualified to make this list. Much less funnier was that it was my Seahawks surrendering that pristine game. :,)
Baffling, though maybe someday SB will bring that game up including the Goff TD reception.
@@jojodelacroixSeahawks won three games against teams that weren’t that great, and have fallen off since
If Goff threw one more TD he would have gotten a perfect passer rating. TD percentage (i.e. the ratio of completions for TDs relative to the total number of attempts) factors a lot in passer rating, and rightfully so because it's directly resulting in points.
Removing the arbitrary limits in the passer rating formula, Jared Goff's "true" passer rating from that game would be just a smidge over 190.
If you include Amon-Ra's 7-yard passing TD to Goff, the Lions as a whole were 19/19 for 299 yards, 3 TD, 0 INT (a "perfect" 158.3 using the NFL's formula) and the team's "true" passer rating would shoot up to 203.6, although that game still doesn't meet the 20 pass attempt minimum which they used for most of the videos in this series.
Alex’s legacy was already one of the greatest sports creators on the internet. But this? This might be the magnum opus to this point. And we don’t even know if he’s in his prime yet.
Incredible work Alex. Jon has Scorigami, but Alex has true passer rating. You’re changing sports data with this one
We always knew (in no small part because Jon is awesome about saying it and reminding us) what an absolute Hall of Fame-caliber research guy Alex is.
Now we're seeing that Alex can more than hold his own as a storyteller.
Welcome to the S tier, Mr. Rubenstein.
But this doesn’t really fix passer rating and this metric isn’t supposed to be applied to single games anyway.
@@SometimesCompitent If it's not supposed to be applied to single games why does every broadcast show it on every game?
@@doqit's not super uncommon in sports to see a stat used in a non intended way as it sticks around
Now I'm wondering which game had the biggest discrepancy in passer ratings between opposing quarterbacks.
Same. Someone with like a 186.3 versus someone with -32.9 would be a massive gap, and I want to know who to make fun of, lol.
Just look at the list of biggest blowouts
Yall gotta make a data portal or something so we can play with the passer ratings from our favorite QB's and games.
tinyurl.com/37hbe38f
for now as a primitive method, you can at least use that as a template and replace the numbers with those from whichever game your heart desires
Yes please
@@SecretBaseSBN you guys are the actual literal unequivocal best omg
Scorigami has already been made mainstream by ESPN. I believe that this mini series could cause a new and improved passer rating system. Great job Alex!
Seeing Jon on NFL Network talking about Scorigami was a huge there's-our-boy moment.
@@SimuLord that was proud moment for a lot of people
loving these videos so far, hope to see more in depth stats based videos like this in the future
They have 100’s of these videos, you obviously are new to the channel. Welcome.
Another major weakness of passer rating is that it doesn't take sacks into account. Most modern research seems to suggest that sacks are more of a quarterback stat than an offensive line stat, and passer rating doesn't punish quarterbacks who hold on to the ball for too long and take the sack, but it does punish those who thrown the ball away. Looking at the three NFL quarterbacks with career passer ratings over 100, two of them (Aaron Rogers and Russell Wilson) have reputations for holding onto the ball for too long and taking a lot of sacks.
Im not sure what research told you sacks are more on the qb than the oline, but they might need to do some more of it Lol
I do agree about qbs who hold the ball too long though. But for the most part, a lot of sacks are unavoidable
@@postverifiedlifestyle complete speculation on my part but it seems the research OP alluded to found a strong positive correlation between sacks taken and quarterbacks holding onto the ball. the longer the QB holds the ball, the more likely he is to get sacked.
Why does NO ONE criticize Rogers for this? All people talk about are his low interceptions. There's a reason he hasn't been to the SB in almost 15 years. Peyton, Montana and Mahomes got to 4, Elway got to 5, Brady got to 10 and I don't understand why Rogers is considered in that elite territory. He's probably a top 10 QB but he's not better than Unitas, Favre or Marino.
It also punishes QB's who have receivers drop the ball when it hit's them in the hands and it pops up for a in INT.
There need to be errors like in baseball that don't account for those picks in the stats.
@@robertdascoli949 And it also rewards a QB for throwing a screen and their receiver breaking 5 tackles on the way to an 80 yard TD
That graph of rating versus attempts is beautiful.
I created an 'inverse' QBR for QB's awhile back to sort of rank QB's by awful-ness per week. It assumes that 6 turnovers is the "worst" based on Jake Delhomme's terrible birthday game. There is a bunch more to it of course, but your formula allowed me to adjust mine and get even finer results. Thanks for your series on this!!
In my opinion, any metric that means to assess a quarterback's performance has to include sacks, fumbles, and rushing. It doesn't necessarily need to be situational because most situational statistics aren't available before 1994, but a 20-yard completion from one's own 15 on a 3rd and 30 in the waning seconds of the second quarter is fundamentally different from a 20-yard completion on a 4th and 17 from the opponent's 25 with less than a minute remaining in the 4th quarter.
I don’t know which language/alphabet to blame for “Taliaferro” being pronounced “Tolliver”
Anti-italian sentiment at the time
That's the normal pronunciation for that name; blame mildly bastardised Italian spelling never updated to reflect changes in the English pronunciation over time
@@kourii Or an Italian American getting their name Anglicized. Not sure if it was a conscious decision by them to pronounce it that way or it's just how everyone around them found it easier to pronounce and they just rolled with it.
@@lordgemini2376 Italian-Englishman, actually. The name predates the American colonies. For the latter claim, contrast how Taliaferro has found itself pronounced versus more mangling of Italianate surnames in the US. There's a reason this one raises eyebrows among English speakers thitherto unfamiliar with it
as a Bills fan as soon as you started talking about negative rating i was like "well here comes ol' Nate probably" and sure enough there he is
Completion percentage has some some value since Anthony Richardson leads the league in Y/A with a 50% completion,
I'd replace the completion % with first down % to avoid rewarding passes that were completed but didn't result in a 1st down or even negative yardage
Yeah, the Foles/Smith example given actually shows why it should be in there separately. Foles' yards per attempt is only just over 1% better, while Smith's completion percentage is over 8% better, so a 2% higher passer rating seems more than fair. It doesn't help that Alex thinks that 14.5 yards is 1.5 yards more than 14.35, when it's actually less than 6 inches.
@@tokinsloff312 The 1.5 yard difference was quotes for "yards per completion", the 14.5 and 14.35 numbers are "yards per attempt". The point is that Foles had a significantly better yards per completion, which with a worse completion percentage puts him about the same on yards per attempt - but he loses out on passer rating because the completion percentage is effectively double-penalised.
@@DumbMuscle Ok, my mistake, but doesn't that mean that Foles was missing a lot of short throws and compensating with a few big completions? Does that really make it a better passing performance? Surely it's better to be more efficient and put the ball at risk less often. Either way, a 2% higher passer rating for Smith is hardly a travesty. As the OP pointed out, if you don't account for completion percentage separately, then hitting a couple of big bombs skews the rating in a small sample size.
Are we going to get a video that includes Goffs "perfect game!? Really wonder how that stacks up!
Sadly doesn't meet the minimum threshold of passes thrown for this project
@@1998_MINEven if it did, Goff's uncapped passer rating would "only" be 190.0. Good enough for top 1% all time, but not like a top-10 performance or anything.
@@dfp_01 however, add in ARSB's one successful attempt to Goff, and that rating balloons to 203.6
@@1998_MIN I wonder how many other performances we'd see if the scope of this experiment had a threshold minimum of 15 passes.
@@bendubz9000 Adding different passers together would break the experiment and all of a sudden you're not even analyzing anything. That's like Stacey King's "Together, Jordan and I combined to score 70 points" thing.
imma be honest this could’ve been one vid lol, i respect the hustle
Probably because most people have instagram reel / tiktok / yt short brainrot attention spans
Nah you don't see the vision
Really been enjoying the passer rating series!
This looks like a solid series, great job Phil.
I always felt, going back to the '90s, that yards-per-attempt should be replaced by yards-per-completion if you wanted four elements, for exactly the reason of removing double-counting completion percentage. I think just removing completion percentage works, but there's something satisfying too (IMO) in having both accuracy and production being separate elements represented in the formula, rather than having accuracy baked into one of the elements.
I sort of thought tangents were the main point 😂 love them.
I think TD% is an unnecessary metric. Why does a receiver who gets tackled at the 1 count for less than somebody who managed to stretch out and break the plane? Meanwhile, sacks and fumbles are hugely important, but not counted at all.
I think it's silly too. If you have a goal line back, or just dont throw deep much, it's not fair.
The problem with that is that you can make the same argument for all the stats. Completion percentage doesn't account for having a contested catch monster as your receiver, or receivers dropping easy catches. Yards per attempt doesn't account for YAC or defenders whiffing easy tackles. Interceptions doesn't account for defenders dropping easy catches or receivers making great "defensive" plays.
@@tokinsloff312 not sure if YAC should even be credited to qbs but i see the general point
@@freddiesimmons1394 Yeah, as much as we love them, stats are never perfect. They do make for great debates and SB videos, though!
The odds of a goalline stall are low, but never zero
If we are going to rewrite the formula a bit, then I think we should include sacks as a negative somehow. Sometimes it is poor line play, but that can be true of interceptions as well. A sack is a failure to attempt to pass. You could maybe add sacks to the attempts for yards per attempt.
Agree, somehow tie it to holding on to the ball too long.
Thanks for making this video. I've had similar gripes against passer rating as soon as I learned the formula. The artificial caps are infuriating!
Definitely my favorite ongoing series 😂 fr I love the qbrA
9:20 Oh, the irony of talking about mathematics and using "exponentially", when it's not even multiplicatively but only additively.
I wouldn't mind swapping out yards per attempt with adjusted net yards per attempt, that takes into account QBs who are good in the pocket/mobile - only the end results of throws are currently taken into account, rather than all passing plays. We can rework the weights to keep the ranges the same for aesthetics if we need to.
This was a really good series!
My suggestion: Referring to the passing rating that's displayed as "Adjusted Passer Rating" or APR, and their true passer rating, without constraints, as "Uncapped Passer Rating" or UPR. APR would be more of a relative sample throughout a season, and give a rough idea of how well they played on average, while UPR would more accurately show how amazing or awful they were on a game-to-game basis. Plus, it'd be easier to distinguish the two that way!
apr & upr are the same in big season/career samples anyway, so there's no real reason to just not roll with upr ubiquitously
It would be fascinating to dig deeper into Luck's numbers and see why the passer rating undervalues him. Perhaps this would give a clue about a better way to do the calculation.
Damn I’m in here early lol excited to watch this for the first time knowing that it’ll prolly become a frequent rewatch over the years
Best sports videos on UA-cam.
Love this series!!
Nice series - bring on the long cut.
Great game, SB!
The trouble is that completion percentage is a complementary to yards per attempt. If you complete a pass behind the line of scrimmage and the receiver gets tackled immediately, you take a YPA hit but you still did something fundamentally "passer" that you should arguably get credit for. But you also made a bad decision that you deserve a penalty for. Having both in the mix allows you to balance them in your rating.
such a great set of data; have you considered using the data to collect all of the new ratings of each player and re-rank the players according to the new rankings based on their total output?
I mean you must have completion % AND YPC. You can’t have one without the other. A really really good example is Lamar vs Joe last week. I was at that game. Lamar was hitting passes at an incredible clip it seemed bc they were short slants the bengals linebacker couldn’t cover. He had two balls I counted only an All Pro could make. He struggled to throw to the right side entirely. If you’re hitting a majority of short passes on broken coverage 6 yards in the air, I want to know. Joe was airing the ball out and his arm was the weapon. It’s crazy how much better a pure passer Joe looked by a wide margin but the data wouldn’t immediately show that. Thats why in my algorithm I also include both YAC and sacks. YAC bc for the example I just used, a high majority of Lamar’s yards were his receivers doing WORK post catch. That’s not a QB stat. Even more so when the bengals notice they struggle to cover shifty receivers so they back off and it only gives Lamar more of an option to hit short slants. Im not saying it should be punished but a YAC heavy game in no scenario is a quarterback friendly environment - it should be isolated. Sacks as well should be considered in tandem with pressure and a signal callers choice to hold or throw. Back to Lamar, he’s fantastic at avoiding sacks. Multiple times he saved 5 or 6 yards which they kept crucially converting on 3rd that saved that game. Joe on the other hand has a tendency to not give a play up and sometimes wants to stay in the pocket too much . This is tough to distinguish but it’s why this is so hard.
This a pretty gross mischaracterization of Lamar’s game last week. Also a gross mischaracterization of Burrow’s game they torched the Ravens on slants and hitches on the boundary all game and one of his touchdowns was thrown behind the LOS that Ja’Marr Chase took like 70 yards.
Lamar hit a deep seam to Charlie Kolar, a layered shot to Mark Andrews on a deep crosser against tight man coverage, and couple of deep outs to Zay Flowers that were thrown with anticipation. Also hit Mark Andrews on a wheel route up the left sideline, and several times hit throws over the middle of the field that were backside reads. None of them had huge YAC gains and were downfield throws.
We no longer rate quarterbacks solely on their arm. QB rating without rushing TDs or rushing yardage, positive or negative, is a disservice.
as i've been watching these videos, i also had the idea of reconfiguring how passer rating is measured, and then adjusting the ratio to make 100 a "good day at the office"--better than average but nothing outstanding--and making 0 something where you'd go "rough one out there today" but not end up with a wikipedia page because of it. take off the top and bottom limits and just leave it from there.
then i realized that's kind of how they came up with it to begin with, right? they picked a contemporary bad game and set that at 0, a contemporary good game and set that to 100 (or 90 or 110 or whatever), and just rolled from there. which really prompts the question: why did they put the limits on? was it solely an ego thing where the historically awful games are rated the same as the generally bad ones?
Could we get an episode where Alex shows how his version of passer rating changes what the currently defined passer rating is? I think it would be fun to see who gets an even bigger uptick and who digresses from what their current game passer rating is.
tinyurl.com/3685fzyj
sheet 1 there sorts best to worst by actual passer rating without the completion percentage element (9/8/19 lamar remains best, however 10/19/75 joe namath replaces 10/9/72 pastorini at bottom), sheet 2 sorts it by who got screwed the most from it to who benefited the most from it
Love this series, but it's genuinely INFURIATING that they haven't once shown what the big graph would look like compared to the big graph with the NFL's artificial constraints. I guess it would be similar but with a ton of names bunched at 0 and 158, but it would just be awesome to see a comparison of the two. I keep waiting for him to say "here's what the NFL's chart looks like, and here's what mine looks like," but it never comes. Pleaaaaaase!!
i.imgur.com/VaVfqZg.png
this is like the 4th video from you on the same subject? am i crazy?
quick question, do you think total attempted passes should be accounted for in passer rating? I'm imagining scenarios in which QB performances are either underrated or overrated thanks to the other parts of the game, especially when the ground game is either really good or really bad. Like when QBs are stuck without a solid ground game to fall back on, they're probably forced to throw worse passes. Or when a defense is really scared of a ground game that is popping off, it's entirely possible for that to create "easier" passes and likely less downs in which the team even bothers to try a passing play. More importantly, a really good running game probably means the QB has more realistic attempts at touchdowns, and at least on paper, there's nothing in passer rating that would account for that
babe wake up SB released made another chart party about the NFL
Oh baby here we go
Hearing "Phil Rivers" just sounded wrong.
I think passer rating could use a fourth component: Pass attempt percentage, the percentage of a team's offensive plays that were pass attempts. Maybe exclude spiking the ball to stop the clock and kneeling to run out the clock from the denominator. But the idea is very simple: Even with a fixed minimum of pass attempts, It's a lot harder to put up good numbers passing it say, 20 times out of 30 plays (against an opponent like the NCAA service academies running down the clock with 10-minute drives) than 20 times out of 60 plays (thanks to spending most of the game handing off to Adrian Peterson).
in case you guys wanted to correct it, the graph @9:48 does not show Nick Foles averaging over 1.5 more yards per completion (14.5 to 14.35)
the 14.5 to 14.35 is yards per attempt, foles did indeed average over 1.5 more yards per completion (thus producing a slightly higher yards per attempt despite the inferior completion %)
4:46 "And last but not least..." but in this case, yes, actually least
Here we go
Personally, I'd prefer some kind of measure that doesn't punish the QB for having bad receivers. I guess it would be best described as 'Accuracy'?
Obviously it would be a stupidly hard stat to track and measure, especially for older games, but if a QB puts the ball right in his receivers chest and the guy drops it, or even bumbles it to the other team, it's weird to me that the QB is the one who's statistically punished for the mistake.
Im shocked that a rushing yards isn’t a consideration for an updated metric.
I think that's a different thing you're trying to measure. Someone like Mike Vick or Eagles-era Randall Cunningham isn't necessarily _better_ than Drew Brees or Joe Montana because they get a big chunk of their yards on the ground, but you could certainly have something like a QB Total Offense Rating. And in the case of someone like Fran Tarkenton, I almost think you'd have to weight the rushing yards he got from scrambling more heavily than you'd weight the same yards from a QB in 2024 because it was so much rarer for a quarterback to do that in 1964 or 1974. Tarkenton would come off as even more transcendent than his already Hall of Fame career made him look as a pure passer.
Passer rating is exactly that...one's effectiveness throwing passes, and on some level, the mobile quarterback still has an advantage because the NFL counts sacks as negative passing yards (college treats it as a run statistically speaking.)
it's called the passer rating, why would rushing yards be considered for a passing metric.
@@badgoogle4509 because the shorthand is ‘QB Rating’ and is used to evaluate QBs against one another. If he’s putting half the RBs yard on the field and it goes into the void …
Passer rating is fine. The problem is people take it out of context to make it mean more than it actually does. Passer rating, which is short for passer efficiency rating, only measures, as its name suggests, passing efficiency. It isn't an all-encompassing measurement of quarterback play. Fumbling the ball or taking a sack before throwing a pass is obviously not going to negatively affect passer rating, because passer rating only measures passes thrown. Heck, a wide receiver can have a passer rating like Mohamed Sanu, who has a career perfect passer rating. A defensive tackle, like Dontari Poe, who threw a TD pass as a member of the Chiefs, has a passer rating. Anybody who throws the ball in a game that isn't negated by a penalty has a passer rating--QB, WR, kicker, DE, whoever. The more people understand this, the less controversial this metric would be.
If you want a more in-depth measurement of QB play, use ESPN's QBR or PFF's grading system.
People who complain about or overrate passer efficiency rating never seem to use it in its proper context.
Two more metrics I'd like to see factored in somehow: YAC vs. yards in the air, and INTs that were the receiver's fault. On the one hand, I don't think QBs should be automatically penalized for yardage totals that rely heavily on YAC (a lot of that is scheme and skill at reading defenses), but on the other hand, some receivers do absolutely crazy shit to bail out their QBs. And with INTs, to me there's a huge difference between throwing a terrible pick because the QB is inept, and throwing an on the money pass that the receiver bobbles right into a DB's hands
Not all YAC is created equal. There's a big difference between a screen pass where the receiver breaks two tackles to free himself and, say, getting the deep safeties to bite on play action from your own goal line, airing it out halfway down the field to an open receiver, and having said receiver take it the remaining 45 yards for the 99-yard touchdown.
Which in turn creates a problem with trying to factor in YAC and Kadarius Toney-induced pick-sixes; it's always going to be a judgment call and the NFL would have to employ something like MLB's "official scorer"...at least with current metrics, where the ball is spotted on the field is where the ball is spotted on the field, guaranteed statistical uniformity.
(eta: yes, I read your comment before I made mine, it was only after the fact that I, in my Friday night 41-hour-workweek-induced brain fog, realized I pretty much repeated back what you said in my first paragraph. I stand behind the point about the "official scorer" however.)
@@SimuLord Agreed! Would be cool to have an official scorer in the NFL. I was also thinking today, I would love to know which of the better than 158.3 games Alex identifies here came against the hardest pass defenses (we could use DVOA for a start). It's like a pitcher throwing a no-hitter against a team with a stacked lineup
There needs to be an overall rating that considers rushes, fumbles, sacks, delivery time etc. I guess QBR makes an attempt at it, but it’s very difficult because you have to consider every play and circumstance
Would kill to see the Secret Base Passer Rating
BIG BEN!!!!
id love to see the analysis of this data w/o completion percentage.
tinyurl.com/3685fzyj
@@SecretBaseSBN Holy shit my man came through!!
Having a game where complete, incomplete, and intercepted all have the same % is pretty wild.
Your issue with the double inclusion of completion percentage is my biggest problem with OPS and why I think it’s a terrible stat.
1. It adds two stats together which is already bonkers for a statistic to do and leads to…
2. It values singles twice as much as just getting on base.
Slugging only considers at-bats and not total plate appearances, so due to that and how OPS is calculated by just adding two stats together, it means hitting to get on base is valued more than just getting on base, which doesn’t make sense since a single or a walk are the same result.
An improved OPS would be Total Bases / Plate Appearances, as that quantifies how many bases you expect a player to net just by going into the batters box. It’s simple, it’s elegant, it’s not adding two things together and double weighting something that shouldn’t be.
a single and a walk are often not the same result when baserunners are aboard...you cant ever advance a baserunner more than 1 base on a walk, nor can you ever advance a baserunner even 1 base on a walk if 1st base is open
When’s the next Collapse episode coming
I completely disagree about completion percentage being removed. You can't tell me 1/5 for 80 yards and a TD is just as good as 5/5 for 80 yards and a TD. 5/5 eats more clock and likely doesn't result in a punt. 1/5 could easily result in a punt since 4 plays are not productive at all.
It's still in the yards per attempt formula. They just aren't counting it twice now.
is a completion for zero yards better than an incompletion?
@@SecretBaseSBNdepends on the scenario. 2 minute drill or 1st quarter.
I understand your point and agree to an extent, but also have a counter. 1/5 and 5/5 are not equally good, but when both have 80 yards and a touchdown attached. 1/5 clearly didn't result in a punt because 80 yards and a touchdown was accrued. In the aggregate, going 5 for 5 with 80 yards is a lot more likely than going 1 for 5 with 80 yards. However, in this argument we've determined that 80 yards and a td are the actual results that occurred and it would be wrong to say that one deserves a higher passer rating than the other because it's more likely to happen again.
Passer rating is like WAR in baseball. It does not care about circumstance or likelihoods.
Definitely agree. Completion percentage is one of the best simple measurements we have for accuracy, and really the only one that extends far enough back in history to be useful for broad comparisons. Sure, it has its flaws, but for the most part a QB who completes 65% of his passes is going to be better than a QB who completes 55% of his passes.
I think yards to score has to be considered. If their defense gives them a short field isn’t there an artificial boost to the touchdown percentage?
Can't wait for the one hour mega cut
I understand exactly what you're say'n... i agree! Tom Brady was a blithering incompetent.
I'm convinced!
3:47 Using the term “crank out” for a man with the last name beathard seems at least mildly intentional
Charts and catchy little tunes plz
Also if we're adjusting the formula fundamentally, why not also normalize the final value to be on a scale from 0 to 100? Negative values never sit right with me, and the ceiling shouldn't be some arbitrary number.
Think another good addition is incorporating volume. So the number gets higher with volume if you're above average and lower with volume if you're below average.
Can you add Jared Goffs recent perfect game? With an exception that it didn't reach a total of 20 passes so an asterisk yeah. But he caught a TD too!
tinyurl.com/msn4ca8y
190
I knew that Big Ben 12-TD stretch would make its way into this at some point. It would have been malpractice had it not. Thanks
"Is that relevant to this story? No. But, I find it interesting." There is no greater to pursue something IMO.
Could we please get a Carey Price or Roberto Luongo untitled video?
I never understood the limits they put on it. Does anyone know why? Was it just to protect players feelings?
Rewinder: 2006 Orange Bowl 3OTs
Beef History: Eric Lindros vs Philadelphia Flyers Organization
Overlap: Greatest College Football Coach (Saban) and the Greatest NFL Coach (Belichick) coaching in the same division
Beef History: Michael Jordan and Isiah Thomas
Beef History: Phil Jackson and Pat Riley
Untitled: Reggie Miller
Untitled: Barry Sanders
Untitled: Carl Yastrzemski
Untitled: Ted Williams
Untitled: Jim Kelly
Collapse: Early 90s Buffalo Bills
Collapse: 1987-2000 Florida State Seminoles
Collapse: Moneyball (2002) Oakland Athletics
Rewinder: 2012 Olympic 100m Men’s Final
Rewinder: The Catch By Willie
Rewinder: Wide Right I, Wide Right II
Rewinder: Hank Aaron and Barry Bonds breaking the respective HR record(s)
Untitled: Don Nelson (Coach)
10:32 Eh, Nature prefers the running game.
Nate Peterman 😂😂
Why is yards per attempt not total passes/total passing yards?
I hope someone with juice in the NFL sees this series and makes the changes.
I'm disappointed you didn't run the numbers with the new formula
tinyurl.com/3685fzyj
havent built a chart for it yet, but in case you wanna peruse it at all
I agree with no artificial boundaries when purely looking at data, but this data is still created by ppl whos livelihood depends on it, so having an artifical floor of somwhere between 0 and -10 seems reasonable to me and maybe release the actual data after retirement, but them having a bad performance must not be turned into public mockery, no one deserves that, especially not for playing a game badly
Imagine if there was a component for non touchdown passing first downs.
Admittedly we wouldn’t be able to use it very far back, but still.
We have play-by-play data stretching back 30 years, which ain't nothing. The super advanced stats like air yards and throwaways have only been available for a decade or so.
@@dfp_01Fair point.
But that still leaves a lot of quarterback performances where passing first down totals aren’t known.
@@dfp_01One angle you could go back generations with first downs passing (and rushing for that matter) is a team perspective though.
I.e. The total number of passing and rushing first downs allowed and produced for seasons are known going back to 1941.
I think they should use a normalized percentile for passer rating.
This was a good series but it also feels like it came... 5 years too late? like these days we have QBR that is a far more informative metric than passer rating, and it's not even a way off to the side only the analytics crowd likes it (like EPA/play) type of stat, its even on espn, cbs etc
this isn't to say QBR is perfect either for the record
It's an offense passing rating, not a passer rating.
What was Garo Yepremian's passer rating tho?
What even is the point of an artificial ceiling?
Why doesnt the NFL have an error stat, where a WR gets credit for an int if it touches them first?
Does yds/attempt include YAC?
yes
Lol. Me before clicking on the video: Bet they mention Nick Foles in the first like 90 seconds of this.
Shouldn’t fumbles be part of this since in a way they are worse than interceptions because fumbles usually happen not so far down feel like a deep past might
Fumbles are usually considered running errors rather than passing errors.
It's still gonna be lies until yards after completion are out of the equation.
Here we go. Chart time.
All this is why average net yards per attempt (any/a) is a better metric
@NFL your move...
Interesting to see just how much of a passing league the NFL has become. When graphed by number of attempts, the graph is so much denser. It also seems like the rating goes up over time. Is that because interceptions get you pulled out of the game earlier?
Yeah and because there are too many rules protecting the precious quarterbacks
Today's top passers would never develop if they had to go up against 70s defenses
@@1998_MINThat and quarterbacks have gradually been getting more accurate with the ball over time. It used to be that they would just sling the ball all over the yard with reckless abandon and hope their receiver would come down with it
@@dfp_01 I would love to return to that. A good run is way way more interesting to watch than airing it out
Is there something like slugging percentage that takes stolen bases into account? Is a double equal in value to a single plus a steal?
I think SecA is closest to what you’re looking for. It’s secondary average. It includes base on balls, total bases, hits, stolen bases, caught stealing, and at bats.
Personally I think total bases should count walks/HBP's/reaching on an error/stolen bases, as they are those technically earned by the player.
It shouldn’t be. A double could drive in a run from first and certainly will from second.
A single might score a run from second and won’t score a run from first.
ua-cam.com/video/eaTCyQMyOu8/v-deo.html
inspired by rickey henderson from the '89 alcs, can i interest you in bases conquered
@@SecretBaseSBN Awesome, thanks!
1:41 aaaaaaaaaa bee!!
Amazing how close GOFF was to making this series... Too bad he was 2 passes short of the 20 pass baseline. No less a perfect game tho!
well, that & the cutoff is after the '23 season lol, but he pulled off a 190:
tinyurl.com/msn4ca8y
11/17/24 jared goff needs to be added. Pretty sure it beats that nick foles stat.
tinyurl.com/yc3vacna
not quite, foles' td% that game was nearly twice as high