QB Accuracy Rankings

[Fan Boy]

27 minutes ago, TrueBlue4ever said:

 

Here is my issue with it - he is using a subjective "eye test" to decide personally what "looks like" a good pass to him, and giving it a grade. Now, he may have no skin in the game and will rank QBs evenly based on the throw and not the thrower, but if he doesn't prefer a certain style of pass, it will hinder the scoring for that QB. Using an efficiency rating (where Nichols is at 120.7 and #1 in the league for QBs with more than 50 pass attempts, and only behind Arbuckle for QBs with more than 6 attempts) incorporates raw data like completion percentage, TD/interception ratio, yards per pass and yards per attempt to determine a more objective number that more properly ranks QBs equally for the metric it tries to define. 

Streveler rates higher, and Nichols should rate highly because they use a short passing game with lots of short passes to the running backs, and those should be easier to throw than than crossing route into a seam between multiple defenders. But looking at some of the examples of what Ferguson grades his passes on, he seems to like the deeper downfield throws.

As an example, Mike Reilly likes to throw up the jump ball passes and let his tall receivers out-duel the shorter defensive backs on what we call "50/50 balls". But it's easy to say that the pass is accurate, because it is not behind there receiver or low, which he likes to pick on. And how many times do we see a receiver change their running pace to "slow down" or adjust to the ball in the air, to be able to speed up and the last moment to gain separation form the defender or catch the ball in stride. That may translate into an accurate pass, but how much of that is on the QB hitting the target perfectly and how much is on the receiver adjusting to the ball in flight to "move the target" and meet the ball? YAC yards seems to be a big thing based on Ferguson's "eye test" metric (because to him the perfect pass is hitting the receiver in stride so they don't have to change their route). 

I think of 3 passes in particular Nichols threw in the last game, two I believe came on the drive where Hardrick took the roughing penalty, that would be subject to a low grade but in my mind were put exactly where Nichols wanted them to be. The first pass was a sideline throw to Darvin Adams for about a 25 yard gain on 2nd and 10, where Adams made a great leaping catch. That would rank low on the rating scale because Adams had to adjust to the high wide ball and made no YAC yards, might score a 1 or 2. But Nichols put the ball where no other player was going to be able to make a play on it, so if Adams doesn't make the catch it falls incomplete out of bounds, and maybe Nichols trusts that Adams can make the high catch, so puts it up there for him rather than risking hitting him in stride down the sideline where a defender might step in front of it for an INT. Much safer than a Mike Reilly rainbow toss 40 yards down the middle of the field into double coverage where Duke Williams out jumps the corner and the safety, but the throw could be more "accurate" since it hits the target zone, but could be picked off more easily if not for the athleticism of his receiver. That to me is a flaw in the judgment system Ferguson employs.

The second pass was to Drew Wolitarsky. Nichols threw a dart over the middle that Wolitarsky went down to the turf to catch. Will score maybe at best a 2 because the receiver got zero YAC yards and had to go low to scoop the ball off the turf. But looked at another way, Nichols again puts the ball where only his receiver can make a play on it, threading a needle with defenders close by. And it went for 9 yards when the Bombers needed 8 for a first down. To me that is a perfectly placed ball, but according to Ferguson's video examples, that would be a bad pass. Wolitarsky makes a very nice catch, but was it a bad throw he had to go down for, or was it meant to be a low pass designed for Drew to go low and secure the first down with zero risk of interception if he can't squeeze it?

The last throw was the called back TD to Adams. Could score high because Nichols led the receiver and hit him in stride, or could score low because it was "too far in front" of Adams, and only the brilliant one-handed catch saves the off-target throw. Who, other than Ferguson, can say? Or does it even count because the penalty wiped out the pass altogether?

So it might be a nice idea, but unless Ferguson wants to justify his scoring on every pass publicly, all he is doing is assigning his own subjective ranking system of what in his mind makes a good pass, with no real consistent real raw data to validate his opinions. 

wow. I am not sure which is more amazing. that I read the whole thing or that I agree completely. Not just because "our" QB won't rank high but that any QB trying to avoid turnovers will rank low. I totally agree that Mike Reilly throws balls up in the air and if it isn't caught who knows if it bounces in the air for a pick. 

[Floyd]

The rabid fan response to the fact that Nichols is not the most accurate passer in the league actually makes me more worried than Ferguson's ranking system.  Usually a sign that its a bit too close to the truth... ha

So far the response is generally 'Oh yeah? Well Ferguson is dumb' and then the poster proceeds to explain how 'their' ranking system would be better... because we're fans and he's just an ex-CIS QB and CFL analyst haha

Nichols has not been super accurate this year... there's nothing in Ferguson's ranking system that doesn't match up with the eye-test

 

[Firekid]

And to conclude - Mike Reilly 1-3 with 5 TD and 4 INT (Losers gonna lose Fergy)

Matt Nichols 3-0 with 7 TD and 1 INT (Winners gonna win Fergy)

 

[Floyd]

Just now, Firekid said:

And to conclude - Mike Reilly 1-3 with 5 TD and 4 INT (Losers gonna lose Fergy)

Matt Nichols 3-0 with 7 TD and 1 INT (Winners gonna win Fergy)

 

Ha why are you taking this so personally...  You think Ferguson has a vendetta against the bombers...? 

[Firekid]

No I didn't even read the stats lol! I just thnk folks are being too hard on Nichols when all he is doing is throwing TD passes and winning. 

[TrueBlue4ever]

30 minutes ago, TBURGESS said:

Yah, it's subjective, but so are folks around here's views. It's not results oriented. It's accuracy oriented. Efficiency rating means nothing in this analysis. You're right that short balls are easier to be accurate on, which is why Streveler ranked so highly this week. That doesn't seem to validate your 'likes deep balls' comment tho.

You seem to want to give Nichols points for results, not accuracy. Nichols throws balls in the dirt and the receivers dig them out. Accurate throws? Not unless that's the only place he could throw the ball and mostly it's not (Subjectively speaking). Good result? Sure. Nichols overthrows, but the receiver makes a great 1 handed grab. Accurate throw? Maybe? Good result? Yup.

It really doesn't matter anyway. It's just one mans opinion of how accurately QB's are throwing the ball. Assuming he doesn't have a bias for or against specific QB's it should be a reasonable way to look at QB accuracy. I can't think of a better way to do it... can you?

I guess my comments are in response the original post comment that this method would "settle arguments" or that Nichols is "having a bad year". I'm not sure how this system objectively or consistently measures good vs bad, so I don't see it as a valuable tool. Maybe a throw in the dirt is accurate because the QB wanted to keep it low to avoid the knockdown or INT, and was simply trying for a first down and not a 35 YAC yard type of play. Only the QB will truly know where they intended for the pass to go compared to where it ended up. Ricky Ray was a master at lofting the high sideline ball where only his receiver had a play on it, incredibly accurate to my mind, but who can say how Ferguson would rank it? (receiver had to go and get a ball in a tough spot, bad throw? - maybe it's meant to be in a tough spot so the defender also has a tough time getting to it). My "deep balls preference" theory is based solely on the limited examples Ferguson uses to outline his criteria, and he mentions low or behind the receivers throws as bad, and uses video of YAC yards catches as good, so I am only guessing based on what he provided as his measuring stick. Again, his lack of a clear consistent metric beyond "this looks good to me" diminishes the objective credibility of his rankings. 

A better way? Well, incorporating completion percentage would be a great start. I think we can all agree that a QB who threw at an 80% completion rate was more accurate than one who threw 50% on an objective basis of measurement. Maybe award higher points for longer completions. A QB who throws 70% overall, but is 13/13 in throws between 0-10 yards, 6/10 in throws from 10-20 yards, and 2/7 in throws over 20 yards may not be as highly ranked for accuracy as one who with a 66.6% completion rate but is 3/3 from 0-5 yards, 5/7 from 10-20 yards, and 12/20 from 20+ yards. Just as an example.

Hey, it's his thing, and kudos for thinking outside the box in finding a way to crunch numbers to evaluate QBs, but it seems too subjective and impossible to verify through consistent data to be taken as a way to rank QB performance fairly. Just my opinion.

Edited July 11, 2019 by TrueBlue4ever

[Brandon]

43 minutes ago, Firekid said:

And to conclude - Mike Reilly 1-3 with 5 TD and 4 INT (Losers gonna lose Fergy)

Matt Nichols 3-0 with 7 TD and 1 INT (Winners gonna win Fergy)

 

Going back to Bishop being undefeated with Toronto....  

If BC had our defense and we had BC's defense.... would Nichols be 3- 0 and Reilly 1 - 3......

[TBURGESS]

12 minutes ago, TrueBlue4ever said:

I guess my comments are in response the original post comment that this method would "settle arguments" or that Nichols is "having a bad year". I'm not sure how this system objectively or consistently measures good vs bad, so I don't see it as a valuable tool. Maybe a throw in the dirt is accurate because the QB wanted to keep it low to avoid the knockdown or INT, and was simply trying for a first down and not a 35 YAC yard type of play. Only the QB will truly know where they intended for the pass to go compared to where it ended up. Ricky Ray was a master at lofting the high sideline ball where only his receiver had a play on it, incredibly accurate to my mind, but who can say how Ferguson would rank it? (receiver had to go and get a ball in a tough spot, bad throw? - maybe it's meant to be in a tough spot so the defender also has a tough time getting to it). My "deep balls preference" theory is based solely on the limited examples Ferguson uses to outline his criteria, and he mentions low or behind the receivers throws as bad, and uses video of YAC yards catches as good, so I am only guessing based on what he provided as his measuring stick. Again, his lack of a clear consistent metric beyond "this looks good to me" diminishes the objective credibility of his rankings. 

A better way? Well, incorporating completion percentage would be a great start. I think we can all agree that a QB who threw at an 80% completion rate was more accurate than one who threw 50% on an objective basis of measurement. Maybe award higher points for longer completions. A QB who throws 70% overall, but is 13/13 in throws between 0-10 yards, 6/10 in throws from 10-20 yards, and 2/7 in throws over 20 yards may not be as highly ranked for accuracy as one who with a 66.6% completion rate but is 3/3 from 0-5 yards, 5/7 from 10-20 yards, and 12/20 from 20+ yards. Just as an example.

Hey, it's his thing, and kudos for thinking outside the box in finding a way to crunch numbers to evaluate QBs, but it seems too subjective and impossible to verify through consistent data to be taken as a way to rank QB performance fairly. Just my opinion.

I agree that it won't settle any arguments or 'prove' anything. It does, however match my 'eye' test. For example I think a lot of Nichols in the dirt throws aren't cuz he had to and a lot of his deep balls aren't accurate (Although his deep balls have been much better this year). I'd guess that Ferguson would have give Ricky Ray big accuracy points because putting the ball where it needs to be is a very accurate throw.

Completion % isn't that great a stat for accuracy. Throw a ton of short passes and it goes up. Throw a ton of deep passes and it goes down. Can't establish the run, don't have enough time, or have to play from behind so the defence knows you have to throw it goes down. Receivers can make the QB look good or look bad and greatly affect the completion %. Have an OC who favours intermediate to deep balls and the completion % suffers. Have a PLAP style offence and it goes up. 

[do or die]

Does the ball hit the receiver in the hands?  Does the receiver have to slow down to catch it.  What kind of window (coverage) was the ball thrown through?  What kind of pressure, was the QB under?  How good are the hands of the guy you are throwing it to?

People always searching too hard for the statistical or other metric indicators, here.   For ole traditionalists like me..... my right and left eyeballs, will suffice.

[17to85]

1 hour ago, Floyd said:

The rabid fan response to the fact that Nichols is not the most accurate passer in the league actually makes me more worried than Ferguson's ranking system.  Usually a sign that its a bit too close to the truth... ha

So far the response is generally 'Oh yeah? Well Ferguson is dumb' and then the poster proceeds to explain how 'their' ranking system would be better... because we're fans and he's just an ex-CIS QB and CFL analyst haha

Nichols has not been super accurate this year... there's nothing in Ferguson's ranking system that doesn't match up with the eye-test

 

I would say Nichols had 1 game where he was inaccurate and 2 where he was pretty damn accurate. 

[TBURGESS]

8 minutes ago, do or die said:

Does the ball hit the receiver in the hands?  Does the receiver have to slow down to catch it.  What kind of window (coverage) was the ball thrown through?  What kind of pressure, was the QB under?  How good are the hands of the guy you are throwing it to?

People always searching too hard for the statistical or other metric indicators, here.   For ole traditionalists like me..... my right and left eyeballs, will suffice.

I trust my eyes too, but forum arguments of my eyes are better than your eyes quickly become no they aren't, yes they are, repeat ad nauseam.

[do or die]

4 minutes ago, TBURGESS said:

I trust my eyes too, but forum arguments of my eyes are better than your eyes quickly become no they aren't, yes they are, repeat ad nauseam.

Damn your eyes, sir.

[TBURGESS]

Just now, do or die said:

Damn your eyes, sir.

Damn your eyes too, sir! 😂

Back the the 'stat'.... If you put 6 football literate fans of neither team in a room with the game film and the catch radius tool, I'll bet you would quickly come to consensus on 90% of the throws. If 3 are one teams fans and 3 are the other teams fans, then that 90% would go down considerably.

 

[do or die]

Just now, TBURGESS said:

Damn your eyes too, sir! 😂

Back the the 'stat'.... If you put 6 football literate fans of neither team in a room with the game film and the catch radius tool, I'll bet you would quickly come to consensus on 90% of the throws. If 3 are one teams fans and 3 are the other teams fans, then that 90% would go down considerably.

 

Sorta like politics....

[Jpan85]

Nichols probably does not have a lot 5 out of 5 throws but I am sure he is very low on the 1 and 2 out of 5 throws. Not many throws wow you during the game but he gets the ball where it needs to be. I would have to agree with much of ranks. There needs to be a decision making ranking.

[JuranBoldenRules]

Each data set has pretty significant flaws.

Basic QB rating doesn’t factor in fumbles, sacks (whether QB’s fault or not) and gives a ton of weight to TD’s.  So if a QB was lucky enough to get the ball in the red zone a couple times a game off a turnover and throw a TD pass each time the number will go way up....if in the same situation the QB passes the ball to the 1 yard line and then it is run in from there his number will be way lower.

 

QB rating isn’t even worth looking at other than over a full season or career.

 

The QB accuracy thing is sort of like curling percentages, has some pretty concrete scoring but also pretty subjective by the person doing the scoring.  At least it’s the same guy doing it every week.

[JuranBoldenRules]

16 hours ago, Noeller said:

But I'm still not sure how he knows which play is called.. Seems to me that's a big part of it. 

It’s pretty easy to tell if you’re watching the film and able to reverse a couple times to watch what each receiver is doing and where the QB is looking.

He doesn’t need the play call he’s just looking at the routes.

[TrueBlue4ever]

Let me just interject here and say THIS is what a good football forum offers. Lots of differing opinions, disagreement between posters, but civil discussion and lots of intelligent thoughts and valid arguments on all sides, regardless of your stance. I have enjoyed the debate on this topic. You stay classy, San Diego!

[17to85]

2 hours ago, JuranBoldenRules said:

 

The QB accuracy thing is sort of like curling percentages, has some pretty concrete scoring but also pretty subjective by the person doing the scoring.  At least it’s the same guy doing it every week.

Good call on that. Highly subjective.

[Floyd]

Really surprised by this response - we’ve been wailing for years asking for more cfl stats - then someone busts his ass to do it - we don’t quite like it... and start crying that there’s too many stats now ha

[NorthernSkunk]

45 minutes ago, Floyd said:

Really surprised by this response - we’ve been wailing for years asking for more cfl stats - then someone busts his ass to do it - we don’t quite like it... and start crying that there’s too many stats now ha

The only stat that truly matters is that elusive cup.  

[17to85]

2 hours ago, Floyd said:

Really surprised by this response - we’ve been wailing for years asking for more cfl stats - then someone busts his ass to do it - we don’t quite like it... and start crying that there’s too many stats now ha

Stats are fine, but assigning too much weight to a subjective stat like this is fool hardy. 

[Floyd]

11 hours ago, 17to85 said:

Stats are fine, but assigning too much weight to a subjective stat like this is fool hardy. 

Except no one is doing that

[17to85]

You appear to be...

[USABomberfan]

On 2019-07-11 at 2:26 PM, 17to85 said:

I would say Nichols had 1 game where he was inaccurate and 2 where he was pretty damn accurate. 

Really almost seems like Mr Ferguson's accuracy system neglects completion percentages and doesn't factor in deep vs short range passes.  All of which Nichols had last week ... AND ... yes his completion percentage was higher than Streveler's short passes numbers.  Hence I say he's got a lot of bonkers in this.