I looked up the play-by-play. His first 20 dropbacks: 4 completions for 37 yards, 5 sacks for -44 yards, a net of -7 yards, plus two fumbles on sacks, one run back for a TD by the Lions. But his *intangibles* were tremendous!
(At that point the Lions were up 38-3, went into a consume-time prevent defense, and Tim was able to run up his passing stats all the way to a 56 rating. That's more clutch 4th-Q play by Tebow, saving his best for last, striving against all odds for the comeback!)
Jim Glass i truly love your post. I know we are on the same page almost always... All this intangibles, clutch, QBR and EPA/WPA-Madness is worth nothing. You need eleven 11 guys on offense and one of them needs to have the talent to throw a football. No more, no less, either 1st Qtr or 2nd, 3rd or 4th.
In GB for example they have it, in other cities they just don´t...
I just look at Y/PP and can see who´s good and who´s not. Simple as that, as it always was.
I don't think I understand the offensive line stats very well. San Francisco's line has -5.4 EPA this week. Compared to CLE, SF had fewer sacks, QB hits, tackles for a loss... and better run success rate. Why is their o-line score worse?
this is not a criticism, more a comment on how hard it is to use statistics to define an individuals player value in a game.
Hussein Abdullah is listed as the 2nd most valuable defensive player in the league. I presume that's a result of 2 pass defences, a late game sack and 7 total tackles.
But if you watched the game you would see a guy playing a terrible game for the most part. He was torched in man coverage for a 39yd TD. He was out of position several other times. The sack was only the result of another player missing the sack and having it fall into his lap. I honestly don't recall the two pass defences, but I suspect the coaches aren't going to be praising him during film session.
Fair question. O-line -WPA and -EPA are situation dependent, -WPA much more so. A line can have fewer sacks or hits, but worse -EPA or -WPA. Think of these as opponent front-seven WPA or EPA. If an opposing DL causes a fumble or tips a ball that results in an interception, it can have a big effect on WPA and EPA that wouldn't be reflected in other stats.
One team could have had a critical stuff on 3rd and 1 or 4th and inches. It might only appear as a single "stuff" or a tackle for a loss, but the leverage in the situation can cause a big loss of EPA or WPA.
This isn't an "advanced" stat, but if you don't mind some trivial stats: after the Chiefs' win tonight, 19 of the 32 teams have winning records. I'm not sure that's a historic high, but it strikes me as unusually high, even conditional on there being no teams with even records.
Also, a majority of the teams in the Canadian Football League are currently tied for the best record in the league, which is always the case before the season starts, but is unusual a week before the end of the season -- five of the eight teams are 10-7.
Brian, if memory serves, the niners were stuffed on the goal line several times, including on 4th. I bet that's where the negative EPA is coming from. Thanks for the tip.
Enhancement idea: for the top players by week pages where there are more than 50 players in the set, refresh the page each time the user sorts by a new column.
When I sort this week's top defensive players grid to see which players had the highest EPA, I'm missing Detroit's Cliff Avril because he was not in the top 50 players by WPA.
Leave a Reply
Note: Only a member of this blog may post a comment.
Welcome to the Advanced Football Analytics archive. This is where you can find over 1,500 ground-breaking articles full of the most innovative research and analysis of modern NFL football.
Interested in publishing your own football research, analysis, or stat-based commentary? Advanced NFL Stats Community is the site to share your thoughts and ideas. There's plenty of data available to get started. All submissions will be accepted and published. Check it out!
Support Military Families
If you enjoy Advanced Football Analytics, please consider a small donation to The Fisher House, a place where families of injured servicemen can stay while visiting their hospitalized heroes.
Tebow! Wow, -0.46 EPA/p.
I looked up the play-by-play. His first 20 dropbacks: 4 completions for 37 yards, 5 sacks for -44 yards, a net of -7 yards, plus two fumbles on sacks, one run back for a TD by the Lions. But his *intangibles* were tremendous!
(At that point the Lions were up 38-3, went into a consume-time prevent defense, and Tim was able to run up his passing stats all the way to a 56 rating. That's more clutch 4th-Q play by Tebow, saving his best for last, striving against all odds for the comeback!)
Jim Glass i truly love your post. I know we are on the same page almost always... All this intangibles, clutch, QBR and EPA/WPA-Madness is worth nothing. You need eleven 11 guys on offense and one of them needs to have the talent to throw a football. No more, no less, either 1st Qtr or 2nd, 3rd or 4th.
In GB for example they have it, in other cities they just don´t...
I just look at Y/PP and can see who´s good and who´s not. Simple as that, as it always was.
Karl, Germany
Would be nice to see win probabilities for next weeks games by now :-)
I don't think I understand the offensive line stats very well. San Francisco's line has -5.4 EPA this week. Compared to CLE, SF had fewer sacks, QB hits, tackles for a loss... and better run success rate. Why is their o-line score worse?
Here's the link:
http://live.advancednflstats.com/index.php?gameid1=2011103009
this is not a criticism, more a comment on how hard it is to use statistics to define an individuals player value in a game.
Hussein Abdullah is listed as the 2nd most valuable defensive player in the league. I presume that's a result of 2 pass defences, a late game sack and 7 total tackles.
But if you watched the game you would see a guy playing a terrible game for the most part. He was torched in man coverage for a 39yd TD. He was out of position several other times. The sack was only the result of another player missing the sack and having it fall into his lap. I honestly don't recall the two pass defences, but I suspect the coaches aren't going to be praising him during film session.
Ian B-
Fair question. O-line -WPA and -EPA are situation dependent, -WPA much more so. A line can have fewer sacks or hits, but worse -EPA or -WPA. Think of these as opponent front-seven WPA or EPA. If an opposing DL causes a fumble or tips a ball that results in an interception, it can have a big effect on WPA and EPA that wouldn't be reflected in other stats.
One team could have had a critical stuff on 3rd and 1 or 4th and inches. It might only appear as a single "stuff" or a tackle for a loss, but the leverage in the situation can cause a big loss of EPA or WPA.
I agree with the comment above about the difficulties. Stats are only telling a part of the story.
Karl, I appreciate the love, thanks. I need it more than I will ever tell you. :-)
This isn't an "advanced" stat, but if you don't mind some trivial stats: after the Chiefs' win tonight, 19 of the 32 teams have winning records. I'm not sure that's a historic high, but it strikes me as unusually high, even conditional on there being no teams with even records.
Also, a majority of the teams in the Canadian Football League are currently tied for the best record in the league, which is always the case before the season starts, but is unusual a week before the end of the season -- five of the eight teams are 10-7.
Brian, if memory serves, the niners were stuffed on the goal line several times, including on 4th. I bet that's where the negative EPA is coming from. Thanks for the tip.
Enhancement idea: for the top players by week pages where there are more than 50 players in the set, refresh the page each time the user sorts by a new column.
When I sort this week's top defensive players grid to see which players had the highest EPA, I'm missing Detroit's Cliff Avril because he was not in the top 50 players by WPA.