Statistical milestones for 2014

[percyhoward]

Some team stats are.

Not just team stats. Passer rating is right up there.

The big three are points per drive differential, pass rating differential, and touchdown percentage differential.

 

[Hawk]

Hayes definitely deserves a mention, but "over his first five seasons" was the operative phrase. He got his 50th TD catch in his 65th game, but that wasn't until his sixth season (1970). Dez has 40 in 59 games over his four seasons.

Interestingly, although each team played a 14-game season in Hayes' era, teams averaged 1.6 TD passes per game in 1965, and 1.6 per game in 2013 as well.

You are missing the TD in 1965 Hayes ran the ball...thus 50

Bob Hayes - Pro-Football-Reference.com

 

[percyhoward]

You are missing the TD in 1965 Hayes ran the ball...thus 50

Bob Hayes - Pro-Football-Reference.com

Hey it's 53 if you count rushing and punt return TD, but the Dez stat was TD receptions only. Like I said, Hayes deserves to be mentioned, because it's not his fault they played fewer games in the 60s.

 

[Hawk]

Hey it's 53 if you count rushing and punt return TD, but the Dez stat was TD receptions only. Like I said, Hayes deserves to be mentioned, because it's not his fault they played fewer games in the 60s.

I think Dez's stats include his Run back Td also. But Bobby Had more TD's in his 1st 4 season's than Dez. Hayes had a bad year in his 5th season.

Plus I have to say I missed the part were they were supposed to be just receptions. You have to understand Hayes started out lining up in the back field, that's why he wore #22

 

[percyhoward]

I think Dez's stats include his Run back Td also. But Bobby Had more TD's in his 1st 4 season's than Dez. Hayes had a bad year in his 5th season.

Plus I have to say I missed the part were they were supposed to be just receptions. You have to understand Hayes started out lining up in the back field, that's why he wore #22

Right. Dez's stats do not include his punt return TD. He has 40 TD receptions and 2 punt return TD.

 

[Hawk]

Right. Dez's stats do not include his punt return TD. He has 40 TD receptions and 2 punt return TD.

Thxs for the conversation, I enjoy talking about some of the legends once in awhile

 

[percyhoward]

Thxs for the conversation, I enjoy talking about some of the legends once in awhile

Same here. Now if we can just get Howley and Harris in the HOF with Hayes.

 

[erod]

Not just team stats. Passer rating is right up there.

The big three are points per drive differential, pass rating differential, and touchdown percentage differential.

I disagree. A quarterback can throw a hitch play and watch his wide receiver go 90 yards for a score, and that gets way more statistical value than a 3rd-down slant between two defenders for five yards and a first down.

Points per drive is a reflection of offensive lines and running backs. Passer rating is heavily impacted by how soft a defense with a big lead is playing. Touchdown percentage is overvalued for a quarterback with a poor short yardage running game.

Individual stats are hugely ambiguous and circumstantial.

 

[CowboyStar88]

How about some bikini contests, that's a good point in measuring....just saying...
I'll volunteer to measure the Cowboys Cheerleaders. What ya all non stat lovers think of stats now...

Those are my favorite kinda stats

 

[percyhoward]

I disagree. A quarterback can throw a hitch play and watch his wide receiver go 90 yards for a score, and that gets way more statistical value than a 3rd-down slant between two defenders for five yards and a first down.

Depends on which metric you use. If you use first down conversion percentage or drive success rate, their value is the same. Of course the hitch play will be worth more if you use a metric with a higher correlation to wins -- like points per drive, pass rating, or TD percentage. After all, that play scored points, and points lead to wins more than first downs do.

Points per drive is a reflection of offensive lines and running backs.

No, it isn't necessarily a reflection of that. Points per drive is simply points divided by number of drives. It tells us how many points an offense scores (or defense allows) without depending on how many possessions it has (or faces). That's why it's better than points per game.

Take two defenses that allow 20 points per game. The defenses seem equal. But Team A's defense faces 9 drives per game, while Team B's faces 11. Points per drive allows us to see that these defenses aren't equal after all. It could be that Team B has an offense that can't stay on the field.

Touchdown percentage is overvalued for a quarterback with a poor short yardage running game.

TD% is not overvalued at all. Unless you want to say winning is overvalued. I think what you mean is that teams that don't run well in the red zone will have have lower TD% than teams that do, regardless of QB play. Obviously that's true, but TD% is not a QB stat anyway.

I think you understood me to mean that "the big three" (points per drive differential, pass rating differential, and touchdown percentage differential) were all individual stats. That's not what I said. I was making two points. One, that those three are the best metrics. The other, that not all the best metrics are necessarily team stats. Like pass rating, for example.

Passer rating is heavily impacted by how soft a defense with a big lead is playing.

No, that's a different stat you're talking about there -- passing yards. Pass rating actually goes down about 10 points when teams are trailing big late (lots of INT in those situations).

This is a common confusion, though, because of all the passing yards that come against a prevent defense. The Browns, Texans, and Commanders (all bottom 10 teams in pass rating) ranked in the top half of the league in passing yards for this reason.

Just remember that passer rating and passing yardage are two completely different stats. Passer rating includes yards, attempts, completions, TD, and INT, so it gives you a much more complete picture. Yards is just yards. You can't just look at yards and conclude "well, stats are pointless because Jeff George had more yards than Aikman." Use pass rating. In their five best seasons, Aikman's average NFL rank was 4th and George's was 11th.

Since 2008, there have been 43 teams that finished the season ranked among the top 12 in pass rating differential, turnover differential, and points-per-drive differential. All 43 teams made the playoffs or won at least 10 games.

I could list win correlations to prove the importance of these statistics, but I think 43-for-43 speaks for itself.

 

[percyhoward]

Here are the rankings in the differentials from last season. Order is pass rating, turnovers, points per drive.

Seattle (1st, 1st, 2nd)
Philadelphia (4th, 4th, 9th)
San Francisco (5th, 4th, 6th)
Kansas City (7th, 2nd, 10th)
Carolina (11th, 6th, 4th)
New England (12th, 8th, 7th)

Dallas (18th, 9th, 17th)

The Cowboys' turnover differential was good enough, but we missed the post season because of differentials in pass rating and points per drive that were merely average. Thanks to the defense in both cases.

pass rating
offense: 95.7 (7th)
defense: 96.0 (26th)
differential -0.3 (18th)

points per drive
offense: 2.25 (4th)
defense 2.31 (30th)
differential -0.06 (17th)

 

[TheDude]

The problem isn't with measuring. It's with people measuring the wrong things and trying to use that data to derive information. For example, if my tailor wants to measure the distance from my bellybutton to my tailbone, that's creepy. But, whatever. As long as he's not using that data point to try to make a decent pair of pants, it's just a measurement. The fact that some measurements aren't useful doesn't mean we need to do away with the notion of inseams and waist measurements. Right?

stats and correlation rarely describe causation. The more factors input, the more difficult the isolation. Your waist measurements ch..errr.. may change over time, it doesnt mean your tailor made you a bad pair of pants because you got into cheesecake

 

[Idgit]

stats and correlation rarely describe causation. The more factors input, the more difficult the isolation. Your waist measurements ch..errr.. may change over time, it doesnt mean your tailor made you a bad pair of pants because you got into cheesecake

In the analogy, measuring the inseam doesn't cause the pants to fit. The pants either fit, or they don't. The measurements attempt to gauge conditions under which pants are going to be comfortable. If you measure the wrong thing, the measurements aren't effective. If you measure wrong, the measurements aren't effective. If you measure properly, and then do the wrong thing with the data, the pants aren't going to be comfortable. If I get fat after the measurements are made, that doesn't mean the measurement was wrong. It means the conditions changed, which is another situation in which you can't trust your measurements.

Either way, I want the guy making my pants to be measuring and not just going by what his eyes might tell him. Don't you?

 

[percyhoward]

Since 2000
4th qtr, trailing by >8 points
16,450 of 28,146 for 182,913 yd 1,217 td 1,130 int 75.6
all other situations
125,635 of 208,397 for 1,450,099 yd 8,565 td 6,062 int 82.9

4th qtr, leading by >8 points
3,999 of 6,468 for 45,623 yd 354 td 154 int 91.3
all other situations
138,086 of 230,075 for 1,587,389 yd 9,428 td 7,038 int 81.8

Being behind late in the game does not inflate pass rating. It inflates passing yards.

A team that's behind by more than 8 points in the 4th quarter passes four times as much as its opponent, for four times as many yards--and a rating that's almost 16 points worse.

 

[Hardline]

You can use stats to prove anything.
Three out of ten people know that.