Anything like WAR for football players?

[JameisLoseston]

Is there any advanced stat that measures a player's value to the team in a particular season, in terms of team wins, on the scale of a 16 game schedule? I think this would actually be more interesting for football players than baseball, because each position would also have different WAR values in addition to specific players; QBs would obviously be the highest, for example. But what is the second most valuable position to have an elite player at? I was thinking about this when writing my last topic, and considering how elite RBs seem to offer almost no difference in team wins over replacement RBs, whereas going from a top tier QB to a backup is usually season-ending.

I actually wonder if the QB has the highest difference from elite to replacement value of any position in any team sport. Taking data from instances in which elite QBs have gotten injured, sometimes they seem like they can swing up to 50% or more of the team's games; just remove Dak, and a playoff bound team like the Cowboys can suddenly become one of the worst teams in football. Baseball players, meanwhile, absolutely max out at about 10%.

[rhickok1109]

https://www.pff.com/news/nfl-what-is-pf ... is-the-mvp

[RyanChristiansen]

Pro Football Reference has its own Approximate Value rating, explained here:

https://www.pro-football-reference.com/ ... x37a8.html

[RichardBak]

As much as I enjoy research, I've always been far more interested in the personalities and social issues of sports than drilling too deep into stats. WAR, analytics, etc. just don't get my blood pumping. My feelings are best summed up with a quote by NYT columnist Russell Baker some 40 or 50 years ago. "The trouble with baseball," Baker wrote, "is that it grew up and became engineering."

[Bryan]

Pro Football Reference has its own Approximate Value rating, explained here:

https://www.pro-football-reference.com/ ... x37a8.html

PFR's Approximate Value rating is good when looking at a player's value to his team for that individual season, but its a little wonky when comparing players from different teams and/or across eras. In 2001, Favre puts up a 3921-32TD-15INT-94.1 rating with a 122 Rate+, along with 56 rushing yards and 1 TD. Kordell Stewart has a 3109-14TD-11INT-81.7 rating with a 104 Rate+, along with 537 rushing yards and 5 TDs. Both earned an AV score of 15.

[JohnR]

As much as I enjoy research, I've always been far more interested in the personalities and social issues of sports than drilling too deep into stats. WAR, analytics, etc. just don't get my blood pumping. My feelings are best summed up with a quote by NYT columnist Russell Baker some 40 or 50 years ago. "The trouble with baseball," Baker wrote, "is that it grew up and became engineering."

Nice quote. I have tremendous respect for the stats chasers working tirelessly to get it right. But it's the characters and the rivalries for me.

[ChrisBabcock]

I sort of take the middle road. I'm a bit of a numbers geek an I'll admit stats are interesting. However the "why" behind the stats sheds light on how the sport evolved. ....and can help give an informed reply to statements like... "Bobby Layne must have been terrible! Look at his stats!"

[RichardBak]

I sort of take the middle road. I'm a bit of a numbers geek an I'll admit stats are interesting. However the "why" behind the stats sheds light on how the sport evolved. ....and can help give an informed reply to statements like... "Bobby Layne must have been terrible! Look at his stats!"

I probably should have added that I've always loved numbers. In fact, as a kid I once won a Science Fair ribbon for my project on the laws of probability. I rolled a pair of dice 1,000 times and tallied up the results. (It wasn't until later that I realized I could've just made up the results and saved myself about 10 hours of work!) Anyway, I played APBA football and baseball for many years with a good buddy and we completed several full seasons and kept meticulous stats. And prior to that, back in the '60s, my two brothers and I played several seasons with a dice football game we devised ourselves. Have some fond memories of rolling those dice, sometimes into the wee hours of the morning. But as I look back on it, it was the camaraderie and good-natured competition that stands out best, not the actual numbers we were compiling and faithfully recording. So whether it's the real deal NFL or some imaginary league, it's always been the human element that's most appealing to me. Drilling too deep into the numbers just isn't as interesting to me, though I can certainly understand how others can find it fascinating and derive a good deal of satisfaction from it. To each their own!

I haven't played a game of APBA football in more than a decade. Still have the box up on a closet shelf. Hmmm....might be time to trot out the '62 season...though it's never as much playing solo as it is face-to-face.

[JameisLoseston]

https://www.pff.com/news/nfl-what-is-pf ... is-the-mvp

This is perfect; I wish there was more said about how to use the formula yourself, but it gives a good idea of what I was looking for in regard to value differing by position. AV, by comparison, does an atrocious job of this, and I really think it just doesn't work at all; I've actually found that its primary usefulness comes when it goes in the negatives, to reveal the worst seasons. It does seem that RBs are the least valuable major position, and the difference between, say, 2006 LT here vs. his AV, and compared to even a second tier QB, is stunning. Has the RB always been a worthless sham? WR being the second most valuable position also doesn't surprise me, but I wonder if that may not have been the case in prior eras. The one obvious anomaly here is 2017 Tom Brady, one of the most "default" modern MVPs, being over 5, when every other 4+ season is a traditionally all-time QB season. Something wrong there, but overall it looks good.

[Bryan]

https://www.pff.com/news/nfl-what-is-pf ... is-the-mvp

This is perfect; I wish there was more said about how to use the formula yourself, but it gives a good idea of what I was looking for in regard to value differing by position. AV, by comparison, does an atrocious job of this, and I really think it just doesn't work at all; I've actually found that its primary usefulness comes when it goes in the negatives, to reveal the worst seasons. It does seem that RBs are the least valuable major position, and the difference between, say, 2006 LT here vs. his AV, and compared to even a second tier QB, is stunning. Has the RB always been a worthless sham? WR being the second most valuable position also doesn't surprise me, but I wonder if that may not have been the case in prior eras. The one obvious anomaly here is 2017 Tom Brady, one of the most "default" modern MVPs, being over 5, when every other 4+ season is a traditionally all-time QB season. Something wrong there, but overall it looks good.

I tend to be skeptical about football stat equations where I don't know the exact calculations. PFF's model seems to be measuring performance variability by position...positions that have a large gap between great and terrible also have the highest WAR totals. I liked PFF's quote of "The average WAR value for a running back with a significant number of snaps in a season is 0.1"; if their equation always produces an insignificant result, then how can anyone derive meaning from it? It sounds like the math is screwy if every RB performance is put into a computer and then they all receive an 'average' rating. It would be like Mel Kiper grading the draft and giving every team a "B"...is there any conclusion to be drawn from that?