PDA

View Full Version : Passing Efficiency: 07 vs 08 (and Hall vs Other BYU QBs)



pelagius
11-19-2008, 10:28 PM
There has been some chatter about Max Hall's performance as a passer. I took a look at both 07 and 08 seasons. In each case, what I did was compare Max Hall's passing efficiency to a predicted average efficiency of BYU quarterbacks from 1984-2006. If a typical BYU quaterback played the same opponent quality (based on my computer model), with the same experience in terms of games played, and in the same location, what passing efficiency would we expect (I will spare you the details of the regression specification). I compare Hall to that prediction. So this analysis insn't asking whether Hall's numbers are better than last year. Its asking how Hall's two season stack up to past BYU quarterbacks after taking into account his experience level and quality of opponent played.

http://fisher.osu.edu/~diether_1/b722/qb_07.png

Hall looks pretty good relative to history. In 07 Hall had 6 games that were better than the BYU predictive value given his experience level and oponent quality. One noticeable thing is the variance in his performance.

http://fisher.osu.edu/~diether_1/b722/qb_08.png

Hall looks slightly better this season even after controlling for experience and opponent quality. He has already been above the past BYU QB predictive over 6 times (with 2 games left). Also, he seems to be more consistent this year. Variance is noticeable less than the 07 season.

In both seasons he had one substantially sub-par game in terms of efficiency relative to predictive value (07: E Washington and 08: TCU). Nice season for Hall ... The biggest improvement historically for a BYU quarterback is from the 1st to the second year. Hall has improved at least as much as a typicallBYU qb and maybe a little better.

RockyBalboa
11-20-2008, 07:15 AM
There has been some chatter about Max Hall's performance as a passer. I took a look at both 07 and 08 seasons. In each case, what I did was compare Max Hall's passing efficiency to a predicted average efficiency of BYU quarterbacks from 1984-2006. If a typical BYU quaterback played the same opponent quality (based on my computer model), with the same experience in terms of games played, and in the same location, what passing efficiency would we expect (I will spare you the details of the regression specification). I compare Hall to that prediction. So this analysis insn't asking whether Hall's numbers are better than last year. Its asking how Hall's two season stack up to past BYU quarterbacks after taking into account his experience level and quality of opponent played.

http://fisher.osu.edu/~diether_1/b722/qb_07.png

Hall looks pretty good relative to history. In 07 Hall had 6 games that were better than the BYU predictive value given his experience level and oponent quality. One noticeable thing is the variance in his performance.

http://fisher.osu.edu/~diether_1/b722/qb_08.png

Hall looks slightly better this season even after controlling for experience and opponent quality. He has already been above the past BYU QB predictive over 6 times (with 2 games left). Also, he seems to be more consistent this year. Variance is noticeable less than the 07 season.

In both seasons he had one substantially sub-par game in terms of efficiency relative to predictive value (07: E Washington and 08: TCU). Nice season for Hall ... The biggest improvement historically for a BYU quarterback is from the 1st to the second year. Hall has improved at least as much as a typicall BYU qb and maybe a little better.

Sorry but this is bullshit.

According to Tooblue Hall has clearly regressed this year. ;)

Jeff Lebowski
11-20-2008, 07:52 AM
That's very interesting. It looks like they tend to peak around games 7-8 and then drop off. I wonder why.

pelagius
11-20-2008, 09:23 AM
That's very interesting. It looks like they tend to peak around games 7-8 and then drop off. I wonder why.

Are you referring to the control sample (the blue line)? It's pattern or Hall's pattern?

pelagius
11-20-2008, 09:31 AM
Are you referring to the control sample (the blue line)? It's pattern or Hall's pattern?

It think you must be referring to the blue line (the control sample). Its peaks in the middle and comes down because I am predicting the efficieny of that goup if they had played against the same opponent quality as Hall. It just so happens that over the last two years BYU has played better opponents towards the end of the year and our weakest have general been towards the middle.

Mormon Red Death
11-20-2008, 09:54 AM
That's very interesting. It looks like they tend to peak around games 7-8 and then drop off. I wonder why.

They start playing Utah...;)

Jeff Lebowski
11-20-2008, 09:58 AM
They start playing Utah...;)

That's game 12. Notice the spike. ...;)

pelagius
11-20-2008, 10:12 AM
I forgot to mention: A big shout out to the lost in transfer Indy Coug for very kindly sharing his historical passing efficiency data with me.

Jeff Lebowski
11-20-2008, 10:14 AM
I forgot to mention: A big shout out to the lost in transfer Indy Coug for very kindly sharing his historical passing efficiency data with me.

Yeah, where is Indy? Why hasn't he created an account yet?

BlueHair
11-20-2008, 10:28 AM
The thing that sucks about pass efficiency is that it doesn't count sacks as a passing play. Quarterbacks will take sacks rather than throw it away and lower their completion percentage.

pelagius
11-20-2008, 12:19 PM
The thing that sucks about pass efficiency is that it doesn't count sacks as a passing play. Quarterbacks will take sacks rather than throw it away and lower their completion percentage.

I think an easy adjustment would just be to reduce the passing yards by the sack yardage. Unfortunately, I don't have sack data. But its a nice idea. It would probably overpenalize but it would be reasonable to look at the difference.

BlueHair
11-20-2008, 03:26 PM
I think an easy adjustment would just be to reduce the passing yards by the sack yardage. Unfortunately, I don't have sack data. But its a nice idea. It would probably overpenalize but it would be reasonable to look at the difference.

If I'm not mistaken, the NFL counts sacks as negative passing yards. I know that's the way it's counted in game stats, but I'm not sure about pass efficiency.

pelagius
11-20-2008, 04:03 PM
A update to the comparison sample. Instead of using all quarterbacks from 1984-2006. I only use "good" quarterbacks and then once again predict how they woud perform with the same experience level, opponent quality, and home field advantage.

The good quarterback list is following: Bosco, Detmer, Walsh, Sarkinsian, Feterik, Doman, Beck

http://fisher.osu.edu/~diether_1/b722/qb_good_07.png

http://fisher.osu.edu/~diether_1/b722/qb_good_08.png

The good quarterback benchmark is higher in 07 than the higher benchmarket but in 08 in makes almost no difference. That's because "non good" quarterbacks usually didn't start more than 12 games. So Hall still looks good relative to "tougher" benchmark.