Personally I have never put much stock in the QB efficiency rating .... to me it just tells me how the team is doing generally, not so much the QB ... I like the TD to interception ratio much better to differentiate the QB .... after that I like total passing yardage divided by passing attempts

Recently someone suggested I look at total passing yardage on 2nd downs (perhaps divided bt passing attempts on  2nd down) .... I don't know if the cfl even keeps such stats or the wherewithal to calculate same ... I dunno

What is your favorite stat for QB evaluation in the CFL?

Conversions on second downs.

Wins.

Just now, Throw Long Bannatyne said:

Wins.

Too obvious. You lose a bonus point.

I would say TD:INT ratio is my favorite stat, but I also like completion % 

I would say chimp avatars are pretty important as well.

The QB efficiency rating is essentially the 'gold standard'  for QB effectiveness because it is the statistic that has the highest positive correlation with the number of wins said quarterback produces. To put it in a statistical nutshell, the correlation of variable x to variable y indicates its ability to predict the outcome (output) of a given variable (y - in this case, # of wins by a QB), based on a known quantity (input) of a known variable (x - which in this case, would be the QB efficiency rating). Correlation values range from -1 (a perfect negative 1:1 correlation) to +1 (a perfect positive 1:1 correlation) - a quick google search shows that the general consensus is that the correlation between QB rating and # of wins is about 0.50ish. To date, no other statistic has a higher correlation with the number of wins a QB has.

This means that, generally speaking, if you were to look at all QB efficiency ratings since the dawn of football to present day, the QB efficiency rating is the best statistic to determine the number of wins a quarterback will get. The higher their QB efficiency rating, the more wins they are likely to have. Analysis has also been done on other statistics such as total passing yards, completion % etc.

In terms of testing, you could hypothetically use anything as your x for testing your null hypothesis. For example, you could do analysis testing the correlation between the temperature the day of the game and # of wins (hint: you probably wouldn't see any correlation that is meaningful). The testing done to date shows that the number of wins a QB has is correlated most consistently with the QB efficiency rating.

Interesting breakdown. But since stats in football are so complex, QB efficiency would also be affected by the quality of the team around the QB. This in turn would affect the number of wins too. I think QB stats are partially to differentiate the ability of QB from the number of wins he receives. The only QB stat with a perfect correlation to team wins is QB wins and we admit this doesn't tell the whole story. We look to other stats as an attempt to see how much of an affect the QB had in contributing to team wins.

The only stat that matters for qbs are the ones that help you support your pre-existing beliefs. 

On 4/29/2017 at 9:57 AM, Eternal optimist said:

The QB efficiency rating is essentially the 'gold standard'  for QB effectiveness because it is the statistic that has the highest positive correlation with the number of wins said quarterback produces. To put it in a statistical nutshell, the correlation of variable x to variable y indicates its ability to predict the outcome (output) of a given variable (y - in this case, # of wins by a QB), based on a known quantity (input) of a known variable (x - which in this case, would be the QB efficiency rating). Correlation values range from -1 (a perfect negative 1:1 correlation) to +1 (a perfect positive 1:1 correlation) - a quick google search shows that the general consensus is that the correlation between QB rating and # of wins is about 0.50ish. To date, no other statistic has a higher correlation with the number of wins a QB has.

This means that, generally speaking, if you were to look at all QB efficiency ratings since the dawn of football to present day, the QB efficiency rating is the best statistic to determine the number of wins a quarterback will get. The higher their QB efficiency rating, the more wins they are likely to have. Analysis has also been done on other statistics such as total passing yards, completion % etc.

In terms of testing, you could hypothetically use anything as your x for testing your null hypothesis. For example, you could do analysis testing the correlation between the temperature the day of the game and # of wins (hint: you probably wouldn't see any correlation that is meaningful). The testing done to date shows that the number of wins a QB has is correlated most consistently with the QB efficiency rating.

 

I do beliebe you have the horse behind the cart: the number of wins predicts the efficiency ratio, not the other way round .... thats why I am looking for a better predictor

Never been a fan of QBR. 

Its not an advanced metric but offensive TDs per game, plays per drive, and red zone % are favorites of mine. 

Its impossible to have a metric show you purely the efficiency of a qb. Football isnt like baseball or basketball. There are too many factors out of the qbs control. You can throw a perfect pass and have it kholert'd to a defender and get a pick. You can throw the ball right to a defender and have it dropped, or have a wr make a crazy play. You can have an OL turnstile and get hit as your throw on a quick timing route etc. 

If the offense is healthy, especially in the cfl, the qb is being efficient. Time of possession can be skewed by running plays, hurry up offense, big plays, screen passes etc. Thats why I like plays per drive. Its not a qb stat, its a healthy offense stat. When you suffer frequent bouts of 2 and outs it drops that number like a rock. You of course have to temper it with big plays. If you constantly rely on big plays for your offense thats a big factor in few drives per play. But not being able to chunk out yards makes it hard to win games you dont blow out. 

Tds are tds. If you rush 70 yards on 5 plays then play action pass 1 yard for the major it doesnt matter. Or vice versa. The offense worked if you moved the ball and scored. 

TLDR; no single stat will give you a best view of qbs. Each qbs worst statistical are is probably the best point to compare and locate issues. 

Oh people can come up with statistics to prove anything Kent. Forty percent of all people know that.

The first thing I tend to look at is completion percentage. Its flawed and can be heavily dependent on quality of receivers, time given by blocking, system the QB is in, but its does a decent job telling how much a QB is able to do his job. Especially if you're talking about consistent high completion percentage over a career with multiple coordinators and systems.

1 hour ago, WildPath said:

The first thing I tend to look at is completion percentage. Its flawed and can be heavily dependent on quality of receivers, time given by blocking, system the QB is in, but its does a decent job telling how much a QB is able to do his job. Especially if you're talking about consistent high completion percentage over a career with multiple coordinators and systems.

In that case you must LOVE Drew Willy.

59 minutes ago, gcn11 said:

In that case you must LOVE Drew Willy.

And Sam Bradford

2 hours ago, gcn11 said:

In that case you must LOVE Drew Willy.

lol, exactly what I was thinking of as a counter argument when I posted it. I thought it was time to replace Willy long before it happened, but I probably wasn't as hard on him as others. The worst part is Willy was solved, put pressure on him and he's useless. Couldn't beat the blitz.

However, when Ricky Ray was in his prime I think he did a great job moving the ball and being successful on short throws. Big difference is he could make you pay for putting pressure on him. I remember several games he just nickle and dimed us and it was so frustrating to be on defence the whole time.

On 5/2/2017 at 0:13 PM, WildPath said:

lol, exactly what I was thinking of as a counter argument when I posted it. I thought it was time to replace Willy long before it happened, but I probably wasn't as hard on him as others. The worst part is Willy was solved, put pressure on him and he's useless. Couldn't beat the blitz.

However, when Ricky Ray was in his prime I think he did a great job moving the ball and being successful on short throws. Big difference is he could make you pay for putting pressure on him. I remember several games he just nickle and dimed us and it was so frustrating to be on defence the whole time.

% is good, but you have to temper it with yards per completion. Thats the difference between a willy / nichols passing at a 65+ % and a khari passing at a 55-60% range. Willy was like 8-9 YPC nichols 9-11 and khari 14-16.  

From Derek Taylor on Twitter.....

 

Agreed wbbfan. But you can never get the full picture of course. Even looking at completion % and yards per completion can be altered by the system the QB is placed in and then the receivers as well. The receivers ability to get downfield and produce YAC yards. Khari had an amazing compliment of receivers.

On 5/2/2017 at 10:13 AM, WildPath said:

lol, exactly what I was thinking of as a counter argument when I posted it. I thought it was time to replace Willy long before it happened, but I probably wasn't as hard on him as others. The worst part is Willy was solved, put pressure on him and he's useless. Couldn't beat the blitz.

However, when Ricky Ray was in his prime I think he did a great job moving the ball and being successful on short throws. Big difference is he could make you pay for putting pressure on him. I remember several games he just nickle and dimed us and it was so frustrating to be on defence the whole time.

I can remember games were Ray was completing 10-15 passes in a row against the Bombers and running 23 out of 25 completions only because the receivers dropped the occasional ball.

5 minutes ago, Throw Long Bannatyne said:

I can remember games were Ray was completing 10-15 passes in a row against the Bombers and running 23 out of 25 completions only because the receivers dropped the occasional ball.

2008 East Semi Final..................

3 hours ago, WildPath said:

Agreed wbbfan. But you can never get the full picture of course. Even looking at completion % and yards per completion can be altered by the system the QB is placed in and then the receivers as well. The receivers ability to get downfield and produce YAC yards. Khari had an amazing compliment of receivers.

Yeah you look at every stat and you begin to get an idea. No stat will give you the full picture. All stats taken together can still be skewed heavily. 

Khari did early on. After bruce left he had milt and blink out of the back field. After bruce our WR core out side of milt was pretty weak.

On 5/1/2017 at 1:43 PM, BigBlue said:

 

I do beliebe you have the horse behind the cart: the number of wins predicts the efficiency ratio, not the other way round .... thats why I am looking for a better predictor

A quick wikipedia  search of the formula for determining the quarterback rating yields the following. As you can see, the number of wins a QB has is not a input variable for determining the quarterback efficiency ratio.

Straight from wikipedia; as mentioned above:

The four separate calculations can be expressed in the following equations:

a=(COMPATT−.3)×5{\displaystyle a=\left({{\text{COMP}} \over {\text{ATT}}}-.3\right)\times 5}

b=(YDSATT−3)×.25{\displaystyle b=\left({{\text{YDS}} \over {\text{ATT}}}-3\right)\times .25}

c=(TDATT)×20{\displaystyle c=\left({{\text{TD}} \over {\text{ATT}}}\right)\times 20}

d=2.375−(INTATT×25){\displaystyle d=2.375-\left({{\text{INT}} \over {\text{ATT}}}\times 25\right)}

where

ATT = Number of passing attempts

COMP = Number of completions

YDS = Passing yards

TD = Touchdown passes

INT = Interceptions

If the result of any calculation is greater than 2.375, it is set to 2.375. If the result is a negative number, it is set to zero.

Then, the above calculations are used to complete the passer rating:

Passer Rating=((a+b+c+d)6)×100{\displaystyle {\text{Passer Rating}}=\left({(a+b+c+d) \over 6}\right)\times 100}

AND THIS PROVES WHAT?

Does it change the position of the horse behind the cart?

15 hours ago, BigBlue said:

AND THIS PROVES WHAT?

Does it change the position of the horse behind the cart?

Wins is not part of the calculation, so how can wins be a driver in QB Efficiency?

On 2017-05-07 at 10:02 PM, BigBlue said:

AND THIS PROVES WHAT?

Does it change the position of the horse behind the cart?

That some sites copy and paste real well.