A look at some historical QB stats and performances through Week 5

Submitted by Prince_of_Nachos on October 5th, 2021 at 6:14 PM

Was looking through a few QB stats that may be of interest.

ESPN has its own QB rating called QBR that it uses - the concept is familiar to most. It's a rating of player efficiency.

There is another tool they use called PAA which is supposed to measure the total points that a player contributes to a team. It is a cumulative stat, so a player that plays in more games, and makes more plays will accrue a higher PAA. Here is a summary of the top PAA seasons in CFB since 2004:

Based on these results, ranking by PAA seems to be a good proxy for elite QB play in CFB. Yes, there are a few system QBs in there (Colt Brennan, Case Keenum, Chase Clement), but in general these are Heisman trophy candidates and a few G5 players that were probably underrated given their subsequent NFL performance (Kaepernick & Carr). 

The two most dominant college seasons I can think of were Cam Newton and Jameis Winston, who show up at #25 and #26 on this list, so PAA may not adequately capture the impact of a QB run game. It's also possible that PAA is correctly capturing inefficiencies in those particular QBs' passing games.

Trivia question - Which B1G QB had the highest single game PAA performance since 2004, and against what opponent? (answer at end)

--------------------------------------------------------------------------------------------------------------------------------------

If we trust PAA as a decent proxy for QB performance, here are Big Ten QB rankings so far this year:

A couple things stand out here:

  • Adrian Martinez is performing on a whole different level than the rest of the league. While his efficiency (as measured by QBR) is essentially tied with CJ Stroud, he has put up double the PAA, ranking #3 nationally. This is with far inferior talent around him to Stroud (yes, Stroud did miss one game, but the point stands).
  • Graham Mertz is now the lowest ranked QB in the country by PAA, meaning he has done more statistically to hurt his team than anyone else
  • Cade has been very middling from a cumulative performance perspective (#69 PAA in the country), but he does rank respectably in terms of efficiency (#4 BIG / #18 nationally in QBR)

For those pushing for Harbaugh to use his QB and the passing game more, I wouldn't get your hopes up. There is a long history of Harbaugh QBs generating high efficiency seasons while making relatively fewer total plays:

 

In general, most QBs have a high correlation between their QBR and PAA rankings. Harbaugh QBs, especially at Michigan, have always had higher efficiency rankings than cumulative play rankings.

Cade's performance is currently stacking up very similar to Speight in 2016 in terms of his efficiency and his utilization level, for better or worse. 

Also, Andrew Luck was incredible.

------------------------------------------------------------------------------------------------------------------------------------

Trivia answer - The highest single game PAA performance in Big Ten history since 2004 was Jake Rudock defeating IU in 20T in 2015.

He scored a 19.5 in that game, or 40% of his cumulative PAA for the year. So when we as fans are asking for Harbaugh QBs that can go through a four-receiver progression and hit their target in stride with NFL throws that leave before the WR even breaks, ala Rudock to Chesson in that game, all we're asking for is the greatest QB performance in league history, every week :)

 

Comments

Mich1993

October 5th, 2021 at 7:10 PM ^

Sounds like PAA is volume based.  Not surprising a QB that hands off so backs can run for 300+ yards and doesn't play the second half of multiple games because his team is so far ahead would not do well. 

I'd be interested in seeing how Cade did in PAA against Wisconsin when he threw a little more. 

I'll take a QB leading an offense putting up 40 points per game. 

JHumich

October 5th, 2021 at 7:57 PM ^

So you're suggesting that we don't ask for a QB who can go through his progressions and make his throws?

You can fancy-stat-diary that all you want, but it sounds asinine to me.

Prince_of_Nachos

October 6th, 2021 at 12:44 AM ^

No, I'm suggesting that having a QB that can make a third or fourth read is a luxury in college football that few teams have. Rudock was able to do it in the latter half of his 5th year in college. Speight was able to (inconsistently) do it in 2016, his fourth 4th year in school. Both came in with scouting profiles that emphasized their football IQ.

I believe Ian Boyd's article on Monday emphasized this exact issue. Finding a highly intelligent QB who also has the raw throwing talent, and putting together a cohesive pass pro unit around him, is very difficult in this day and age. Most college teams emphasize quick passing attacks and pre-snap reads, and taking away the first read from a QB is often sufficient to stifle passing attacks. QBs that can make additional reads are a valuable rarity in today's game.

I hope Cade and JJ can get to the point where they're coming off receivers and making third/fourth reads. Cade right now seems comfortable with one read and then a checkdown. As a third year player with only six starts, it seems reasonable that he may still develop this skill.

The image below is a still of the throw where Rudock releases it as Chesson breaks. It a few moments, two Indiana DBs will careen into each other and Jehu will have a walk-in TD. This is the aspiration, but it's not easy to get there. This game was a statistical outlier, as were Andrew Luck's entire 2009 and 2010 seasons.

Having a QB that can do this is awesome, but it doesn't happen very often. That's my point.

grumbler

October 5th, 2021 at 8:17 PM ^

So, quarterbacks in QB-centric schemes contribute more to team success than QBs in run-centric schemes.  Whodathunkit?

We should not "trust PAA as a decent proxy for QB performance," but we can probably trust it as a measure of the importance of a QB to team success, if we combine it with QBR.

Ecky Pting

October 6th, 2021 at 9:54 AM ^

It seems like a tempo-adjusted PAA would be a more useful metric than the raw, cumulative PAA. Many of those QB's at the top of the list had the benefit of playing longer seasons by way of conference championship games plus a game or two in the CFB playoff. To that extent, comparing year-to-year or even within a season seems like apples and oranges?

bighouseinmate

October 6th, 2021 at 10:32 AM ^

An average of PAA per game, or better yet, per offensive snap, would probably yield a more accurate reflection on how well a qb has performed in relation to other qb’s. Then again, the effectiveness of a qb is largely dependent upon the team around him. As an example I’d say look at Oklahoma qb’s on the first list. 4 in the top ten, 3 of which played under the current coach and his offensive scheme, including the players (wr, te, ol, rb) recruited and their own skill levels. Does anyone believe that Murray, or Mayfield, for instance, would have done as well on the PAA scoring if they’d played for Kansas or KState? Having said this, though, I’d say that makes Joe Burrows’ season at LSU all the more remarkable. Yes, he did have very highly skilled players surrounding him (just as LSU does now that surround their current qb), but as we have seen at LSU recently, the offensive scheme and play calling does not lend itself to a “plug and play” situation like at Oklahoma or OSU. 

S5R48S10

October 6th, 2021 at 11:49 AM ^

Nebraska has played six games, while everyone on that list not named Illinois has played five.  That's why Martinez is "head and shoulders" above the conference in this cumulative stat.  

befuggled

October 6th, 2021 at 2:33 PM ^

Which is true, but if you look at a per-game average he still comes out well ahead of everybody else. Martinez has a PAA/game of 6.75 compared to 5.1 for Stroud (who didn't play against Akron) and 4.62 for Clifford.

Still hard to say if this is telling us anything useful.

UMForLife

October 6th, 2021 at 11:39 PM ^

The stats are very comparable. Mind you, Cade has less game experience than Speight at this time in their career. So, there is a chance Cade might improve. Wilton also was good at taking care of the football, expect for the last game of the regular season (argh.......) Also, deep balls seem to be Cade's plus. Wilton seems to have a better mid range game but hoping Cade will develop it as the season goes. Of course, we had better WRs with experience back then. I have the same thought as you after looking closely at their stats. Encouraged....