In football, it's easier to determine the value of an offensive player than a defensive player from the box score. Chances are the guys with the most yards and touchdowns are the best players, but its more difficult to determine the value of a defensive player. Who’s more valuable, a defensive end with a few sacks, or a linebacker who makes tons of tackles all over the field? The problem is that the kinds of stats defensive players accumulate are impacted by the position they play. So, I’ve been working on a mathematical formula for calculating a defensive player’s value to a team regardless of position. I figured I could do this by applying a number value to every stat an individual player makes. I’m working from the box scores, so I’m not able to account for when and under what conditions a stat is recorded.
The first thing I wanted to do was select one stat to be the base stat against which all other stats would be measured. The goal of every defense is to get the offense off the field. There is only one play an individual defender can make to get the offense off the field on any down. That play is an interception. So I made an interception worth 10 points. The other statistics get values in relation to how they add to the defenses abilities to get stops. Those Values are as follows, with explanations..
Interception - 10 points
Forced Fumble – 6 points
Recovered Fumble – 4 points
A fumble recovery is worth as much as an interception, but it has to be divided between two players. Recovering a fumble is pure luck (per football prospectus) while forcing a fumble is a skill. So I’ve weighed the value of a fumble recovery to give more points to the player who created the fumble.
Pass Break Up - 3.5 points
The importance of sacks, tackles for loss, and incomplete passes is that they force the offense into situations that have a below average chance of producing a first down, and thus a higher chance of getting the offense of the field. So I calculated by the value of a pass break-up by looking at how an incomplete pass affects an offense’s ability to get a first down. I figures a pass defended is an incompletion forced by one player, and it doesn’t matter if the quarterback threw the ball out of bounds or if a cornerback knocked a pass away. I looked at every time Georgia Tech forced an incomplete pass and the offense failed to get a first down afterwards and divided by the number of times an incomplete pass was forced and the offense did get a first down latter. I then multiplied that number by 10, the points assigned for an interception, to get something relevant.
Quarter Back Hurries – 3.5 points
Same reasoning as above, only this time I’m attributing the incompletion to the quarterback being rushed by a defender.
Sacks – 7.9 points
I used the same process to calculate the value of sacks, as I did for pass break-ups. I divided the number of stops (failures to get a first down) after a sack, by the number of first downs recorded after a sack, and multiplied by 10.
Tackles for Loss – 8.5 points
I used the same process as I did for sacks. I’m not sure why this number is higher than the one for sacks. Maybe it’s because there are more opportunities for TFL’s on second and third down, and the offense has fewer opportunities to recover from a loss on those downs.
Tackles – 1.9
I divided the total number of Tech tackles by the total number of stops Tech made. For stops, I considered anything that wasn’t a touchdown a stop. Even if the offense kicked a field goal, Tech’s defense stopped them from achieving their ultimate goal. Counting field goal attempts as stops also doesn’t penalize the defense for playing well when the offense or special teams put them in bad fireld position.
Two things I’d like to be able to do, but can’t do from the box score are track the impact of a player drawing double-teams, and the impact of a cornerback shutting down his side of the field.
To track the impact of double teams, you would need to note each time a defensive lineman or line backer was blocked by two players, and then note the number of times a teammate recorded a stat other than a simple tackle (as a tackle is the result of almost every play). Then divided the later by the former and multiply by 10 to get a point total.
To track the impact of safeties and corner backs on opposing offenses you’d have to chart the percentage of times a team throws the ball into their area, find the inverse of that percentage and multiply by 3.5. This would increase the value of each pass defended for players offenses avoid and decrease the value of each pass defended for player’s offenses attack.
It’s not perfect, but check out the results and tell me if you think they are valuable. To be eligible a player has to have recorded a stat in at least three games.
The Black Watch
1 Derrick Morgan 152.3
2 Michael Johnson 115.8
3 Morgan Burnett 112.8
4 Vance Walker 103.2
5 Sedric Griffin 91.3
6 Darryl Richard 76.9
Honor Guard
7 Jahai Word-Daniels 76.2
Chosen Men
8 Mario Butler 59.8
9 Tony Clark 57.4
Enlisted Men
10 Kyle Jackson 55.1
11 Jason Peters 48.3
12 Rashaad Reid 47.9
13 Dominique Reese 47.5
14 Cooper Taylor 44.4
15 Travis Chambers 27.0
16 Jake Blackwood 23.4
17 Oshon Tango 19.3
Active Reserves
19 Brad Jefferson 15.2
20 Steve Sylvester 14.3
21 James Liipfert 12.4
22 Albert Rocker 12.3
National Guard
23 Anthony Barnes 9.5
24 Mario Edwards 8.8
25 Michael Peterson 7.6
Home Defense
26 Shane Bowen 6.7
27 Elris Anyaibe 6.7
No comments:
Post a Comment