• Please be aware we've switched the forums to their own URL. (again) You'll find the new website address to be www.steelernationforum.com Thanks
  • Please clear your private messages. Your inbox is close to being full.

DEL's Quarterback Rating Thread

deljzc

deljzc

Regular Member
Member
Joined
Apr 13, 2014
Messages
5,307
Reaction score
4,795
Points
113
I plugged the performances of Roethlisberger and our defense into my quarterback formula (a cleaned up improved version of the NFL's quarterback rating system).

This week, both Roethlisberger and Kiser performed worse using my criteria than that of the NFL's outdated system. That is primarily because Roethlisberger's offense produced only 14 points on 9 meaningful possessions. And for Kizer, his 7 sacks and poor scrambling performance (4 scrambles for 16 yards) means 11 dropbacks of mostly bad outcomes that the traditional QB rating system ignores.

I will start to include all the meaningful stats that go into the formula (these are all the stats I use to get my rating):

Ben Roethlisberger - 87.7
24 of 36
0 rushes, 0 yards
1 sack
263 passing yards
1 turnover
9 drives, 14 points produced
2 each, 20+ yard plays

D. Kizer (Defense) - 66.2
20 of 30
4 rushes, 16 yards
7 sacks
222 passing yards
1 turnover
9 drives, 17 points produced (note I ignore missed XP's and 2-pt conversions, all TD's are 7 points)
4 each, 20+ yard plays


Note in the old rating system Roethlisberger has a 95.0 rating and Kizer an 85.7
 
can you add Vikings QB stats as well ?
 
You mean you want to know how well Bradford did last week using my system?

Sam Bradford, Week 1 - 188.8 DQR
27 of 32
346 passing yards
0 runs, 0 rushing yards
1 time sacked
0 turnovers
8 drives, 30 points
8 plays 20+ yards (could be more but I didn't have time to check)

That's about as close to perfect as you can get.
 
THANK. YOU. Deljzc !!!!!!! That is a WAY better QBR system than the ******* ESPN pukes attempt at glamorizing their t*mmy b*y. I'd take your system any day to use as to how well my QB was actually doing in real life game. The NFL version is to glamorize fantasy football to get more fans hooked and conned. AGIAN, thank you




Salute the nation
 
deljzc said:
You mean you want to know how well Bradford did last week using my system?

Sam Bradford, Week 1 - 188.8 DQR
27 of 32
346 passing yards
0 runs, 0 rushing yards
1 time sacked
0 turnovers
8 drives, 30 points
8 plays 20+ yards (could be more but I didn't have time to check)

That's about as close to perfect as you can get.
Click to expand...


The Steelers will have to give him some different looks - pressure.As he is very accurate. Can't afford for him to be too comfy in that pocket.
 
deljzc said:
Like Roethlisberger, Bradford is a much different quarterback on the road. If he was as good as he looked last week he wouldn't be on his 3rd team. Just saying....
Click to expand...

I hope you're right, at least this week. We cannot afford to let him pick us apart. Their D ain't too shabby either.
 
I kind of agree that Minnesota is a pretty solid team right now. There line and running game are much improved and they are pretty healthy. I think Mike Zimmer is a better coach than Marvin Lewis and the team he's building is similar in style to Cincinnati (without the prisoners running the show). Zimmer has a lot of experience defending Roethlisberger as well (maybe not great, but he knows him).

The current point spread/money line seems to indicate we are about a 70.5% favorite to win the game but I think it could be close.

Certainly not a "gimmee" type game in my opinion. We better bring our work hats.
 
Last edited:
deljzc said:
I kind of agree that Minnesota is a pretty solid team right now. There line and running game are much improved and they are pretty healthy. I think Mike Zimmer is a better coach than Marvin Lewis and the team he's building is similar in style to Cincinnati (without the prisoners running the show). Zimmer has a lot of experience defending Roethlisberger as well (maybe not great, but he knows him).

The current point spread/money line seems to indicate we are about a 70.5% favorite to win the game but I think it could be close.

Certainly not a "gimmee" type game in my opinion. We better bring our work hats.
Click to expand...
I wonder how many times a team wins when their QB is 20 points lower than the opponent? Sure it happens, but how often?

The Vikings game will tell me where the Steelers stand. The talent on the team is there, always has, but sloppy penalties and one example of poor coaching were on display vs Cleveland.
 
So you're rating system has Ben having a much better game (33%) than Kiser compared to the old system (11%)
 
Just a couple of questions about your rating:

-Do sacks indiscriminately effect the score? In other words- is there a difference between the QB holding the ball too long vs. overall poor O-Line play?
-How do you evaluate the scrambling criteria? Because I find it hard to believe that 4ypc is considered a poor performance.
 
Bigappleyinzer said:
Just a couple of questions about your rating:

-Do sacks indiscriminately effect the score? In other words- is there a difference between the QB holding the ball too long vs. overall poor O-Line play?
-How do you evaluate the scrambling criteria? Because I find it hard to believe that 4ypc is considered a poor performance.
Click to expand...

Sacks are used in a criteria I call "Big Plays/Risk". But they are also counted as a "drop back" or pass attempt for most of the statistical analysis.

The old QB rating system used four statistics: YPA, Completion %, TD% and INT%. And again, all these percentages only use official pass attempts as the denominator in the formula.

First, I get rid of official pass attempts as the denominator in my formulas. I use total "drop backs". That means sacks and scrambles (not QB sneaks or kneel downs) are counted as "pass attempts" in my book because to me the play call was a pass play and I want to know what happens in those calls.

So instead of yards per attempt (old system), I used yards per pass play. I add up yards passing + scrambling/rush yards per pass play. Note I do not use sack yardage (I strongly considered it) but you'll see why in a little bit. That is partly why 16 yards on 4 rush attempts is bad. Because these aren't run calls. They are pass calls. And a good pass play is 7-8 yards, not 4. When a QB scrambles, he normally (not always) gets a bad result. I mean Kizer scrambled 4 times and 3 of those times were bad results/unsuccessful plays (1 yard, 1 yard and 3 yards). Now I know the 3-yard run resulted in a personal foul penalty for 15 yards, but I'm not considering that (really it's hard to imagine any QB rating system that does).

In the same idea, I use turnovers per pass play and include fumbles. This just makes a lot more sense to me to include fumbles lost in QB evaluation.

I do keep completion percentage the same. It's a simple idea, simple to explain. No reason I thought to change it.

The most controversial change I made was I ditched the TD passes per attempt completely and I decide to rate QB's on the general effectiveness at scoring points per possession. Call this my "field general/game manager" criteria. I just think it's important to rate QB's on the overall effectiveness of the offense, their ability to check into the right plays, keep drives going, etc. There are just too many drives to me that the QB does a lot of the work only to get a 1-yard TD run by the RB to not be rewards something for that. Yes, that means QB's are rewarded if they hand off one time and the RB runs 80 yards for a touchdown. To me the goal of good quarterbacking is the TD, not whether or not HE threw the TD pass.

So while those four things replace the original four, I added a fifth category (which isn't weighted as much as the others). I call this the sack/big play/risk category. I wanted to penalize sacks (since I think the QB really is over 50% responsible for most sacks) but I understood the logic that sacks are acceptable in the attempt to "go deep". In other words, I didn't want to reward "dink and dunk" too much. So I created a +/- criteria that compares 20+ yard passing plays vs. sacks then divides that total by pass attempts (i.e. pass plays). So to me it's okay if you take 30 sacks but get 60 big plays out of it as compared to a QB that takes 20 sacks but only gets 40 big plays. Big plays matter towards winning and I wanted to reward that.

This last category is also the reason why I didn't count sack yardage (negative) in total yards above. I think that would "double up" the damage sacks do to your QB score and I thought that was a bit much, at least in this iteration.

Note also, while the old formula made all 4 criteria almost exactly the same weight, I slightly message my five categories to be weighted slightly different from each other to reflect my bias and opinion as to which is most important (i.e. I lower completion % slightly in the weighting system).

That's it in a nutshell. It's not a perfect stat because it's completely based on numbers. SITUATIONS ARE NOT CONSIDERED. It also is a bit odd since you need at least one possession to be calculated. It works best over a game or season, not particularly a quarter or 2-3 passes.
 
Great read Del, like how well thought out the rating system is. Do you by chance have a list of how the QBs rated out in your system last year?
 
Fiji Mariner said:
Great read Del, like how well thought out the rating system is. Do you by chance have a list of how the QBs rated out in your system last year?
Click to expand...

Yes. I have data from the last 4-5 seasons actually. I will create a google doc and post the link next week.
 
SUPER good read and as one can tell you have put a lot of work into this. I thank you for your time and effort. I'm looking forward to seeing the past season's ratings and as I've said "this QBR of your's is by far superior to the homemade ESPN crap".


Salute the nation
 
Here are week 2 numbers. As always, DQR can be derived just from the gamebook here:

http://www.nfl.com/liveupdate/gamecenter/57257/PIT_Gamebook.pdf

Ben Roethlisberger - 103.9
23 of 35
0 rushes, 0 yards
2 sacks
243 passing yards
0 turnover
11 drives, 26 points produced
3 each, 20+ yard plays

C. Keenum (Defense) - 61.6
20 of 37
0 rushers, 0 yards
2 sacks
167 passing yards
0 turnovers
11 drives, 10 points produced
3 each, 20+ yard plays
 
Comparing the two one could argue that the Vikings have a better D in comparison to the clowns. With the differential between week one and two it has a look of week one being unprepared. Or rusty, order both. I think Tomlin needs to play players particularly Ben a little more in preseason.

Sent from my XT1635-01 using Steeler Nation mobile app
 
Last edited:
QB ratings don't mean ****!


The TEAM plays to win the game!
 
The lack of big plays right now is a bit concerning.

Roethlisberger and this offense had been on a good uptick with big plays (while reducing sacks btw). From 2013-2015 he averaged just about a 20+ yard play every 10 pass attempts. Last year he had 59 big plays in 509 pass attempts.

So far this year it's been only 5 big plays in 71 pass attempts. And the blame can't all go on Bryant because last year we didn't have him and Roethlisberger was great at generating big plays.

Once that happens (not sure why it hasn't yet), I think you'll see everything kind of break open for this offense. Points up, Bell better, offensive line look better, etc.
 
I am not blaming Roethlisberger for the lack of big plays in my opinion. Right now Bell has not got it going and in a large part because our offensive line has not go it going. I think Munchak needs to light a fire under those guys and get some holes opened for Bell and then when Bell gets a hole he has to do something with it. A few runs where he had room went for 7-8 yards. He needs to turn those into 20+ scampers. Once that happens the safeties start crashing down on play action a little and some one on one coverages open up down the field.

That is my two cents.
 
FordFairLane said:
I am not blaming Roethlisberger for the lack of big plays in my opinion. Right now Bell has not got it going and in a large part because our offensive line has not go it going. I think Munchak needs to light a fire under those guys and get some holes opened for Bell and then when Bell gets a hole he has to do something with it. A few runs where he had room went for 7-8 yards. He needs to turn those into 20+ scampers. Once that happens the safeties start crashing down on play action a little and some one on one coverages open up down the field.

That is my two cents.
Click to expand...

I agree it is too early to raise significant concerns right now. To me, we look at statistical issues at every quarter pole of the season. Give them two more games (Chicago and Baltimore) and let's see what this team looks like.
 
You must log in or register to reply here.
Top