Four years after Sports Illustrated hailed the debut of Value Add Basketball, the first Value Add Football database appears today.

Michigan State quarterback Connor Cook ranks as the only player worth more than a touchdown to his team (7.01), while the top field goal kicker and punter (Florida State’s Roberto Aguayo and Alabama’s JK Scott) give about three points a game. Notre Dame’s Ronnie Smith plays as one of four offensive lineman worth more than five points.

Here are the 20 most valuable and top player at each position:

Rnk Player Team PosRnk Position Value Add Conference
1 Cook, Connor Michigan State Spartans 1 QB 7.01 Big 10
2 Boykin, Trevone TCU Horned Frogs 2 QB 6.83 Big 12
3 Kessler, Cody USC Trojans 3 QB 6.75 PAC-12
4 Jones, Cardale Ohio State Buckeyes 4 QB 6.12 Big 10
5 Hackenberg, Christian Penn State Nittany Lions 5 QB 6.06 Big 10
6 Striker, Eric Oklahoma Sooners 1 LB 5.92 Big 12
7 Buckner, DeForest Oregon Ducks 1 DE 5.87 PAC-12
8 Chubb, Nick Georgia Bulldogs 1 RB 5.70 SEC
9 Allen, Jack Michigan State Spartans 1 OLine 5.61 Big 10
10 Bosa, Joey Ohio State Buckeyes 2 DE 5.58 Big 10
11 Stanley, Ronnie Notre Dame Fighting Irish 2 OLine 5.55 Ind
12 Elliott, Ezekiel Ohio State Buckeyes 2 RB 5.47 Big 10
13 Williams, Marquise North Carolina Tar Heels 6 QB 5.38 ACC
14 Conklin, Jack Michigan State Spartans 3 OLine 5.38 Big 10
15 Drango, Spencer Baylor Bears 4 OLine 5.29 Big 12
16 Cash, Jeremy Duke Blue Devils 1 DBack 5.26 ACC
17 Oakman, Shawn Baylor Bears 3 DE 5.20 Big 12
18 Robinson, A’Shawn Alabama Crimson Tide 4 DE 5.07 SEC
19 Goff, Jared California Golden Bears 7 QB 5.06 PAC-12
20 Thomas, Justin Georgia Tech Yellow Jackets 8 QB 5.05 ACC
24 Zettel, Anthony Penn State Nittany Lions 1 DLine 4.90 Big 10
30 North, Marquez Tennessee Volunteers 1 Receiver 4.72 SEC
79 Fournette, Leonard LSU Tigers 1 Return 4.00 SEC
168 Prescott, Dak Mississippi State Bulldogs 1 PuntTeam 3.50 SEC
417 Aguayo, Roberto Florida State Seminoles 1 Kicker 3.00 ACC
640 Scott, JK Alabama Crimson Tide 1 Punter 2.72 SEC

SEC fans may grumble at having only two of the top 20, but the conference dominates the top 100:

SEC: 29 of top 100

Big 12: 18

ACC: 17

Big Ten: 13

Pac-12: 11

American Athletic: 4

Independents: 4

MAC: 2

Sun Belt: 1

CUSA: 1

The two biggest obstacles to setting up a Value Add Football ranking has been the difference in the value of different positions, and the lack of common stats. While a guard, forward, and center are relatively equal on the basketball court and all record the same stats, it is very difficult for even a dominant offensive guard to equal the value of an above average quarterback.

But the system calculates that JK Scott’s incredible punting last year improved the Tide by 2.72 points a game, though he was even more valuable when he averaged 55 yards a punt against Ohio State in the National Semifinal.

To that end, eight of the 20 most valuable players are quarterbacks including one non-NFL prospect in Justin Thomas, who runs the complicated option so well that the team shattered the record by converting 58% of all third down conversions. They are the football example of the “tempo free” basketball team, taking their time to score and keep the ball away from the opponent as the most efficient offense in the country. Thomas’ stunning upset at Georgia, near upset of Florida State in the ACC title game, and crushing win over former national contender Mississippi State has him as one of the most valuable returning players. He will have to do even more this year with most of his skilled teammates gone, but the line should still be strong.

Field Goal Kickers Ratings the only Precise Measurement

The only spot that is precise is the ratings for field goal kickers, where the basketball analytics term “Effective Field Goal Percentage” can be run based on the distance of each field goal attempt. The average kicker hits 44% of his attempts from over 50 yards out, and very high percentages from every other 10-yard range down. Aguayo scored 30 more points for Florida State than an average kicker would have given the same field goals and extra points he attempted, and once you adjust to see how many points above a below0average “replacement” kicker would have made (as in baseball WAR or Value Add Basketball) it comes to 3.0 extra points per game.

Value Add Football 1.0 an Art Not a Science (non-technical stop reading here)

While Value Add Basketball is a precise science, Value Add football will always be a more subjective art. Without getting into formulas, the basic attempt to figure out the actual points each player improves his team are as follows:

1. Offensive, Defensive and Special Teams Efficiency Ratings and Positions:

The operating assumption is that a team that is very strong at every position could beat a team of replacement players (borderline D1 players) by 55 points. That would entail the six groups of players charged with stopping points (first six listed below) being about 27.5 points better than a replacement team at stopping points, and the six offensive groups of players being 27.5 points better at scoring points than the replacement players.

Group Position starters per player Group
D Back 4 1.8 7.2
D End 2 3.3 6.6
D Line 2 2.3 4.6
LB 3 2.2 6.6
Punt Team 2 0.5 1
Punter 1.0 1.5 1.5
Total Defense 27.5
Back 1.5 2.3 3.5
Kicker 1 1.5 1.5
Line(C/LT+) 5 1.7 10.5
QB 1 4.7 4.7
Receiver 3.5 1.8 6.3
Return 2 0.5 1
Total Offense 27.5

This table breaks down the estimated value of each position, based on NFL salaries with adjustments for the college game (e.g., running backs are more valuable in college than in the pros).

2. Each of the 12 groups above is rated numerically as well, so the running backs and lineman other than the left tackle get more credit if the team is stronger at running, whereas the left tackle, receivers, and quarterback get more credit for teams are more successful throwing (however running quarterbacks can get credit in an option).

3. Obviously players expected to start (1st string) rank higher than their backups, and all else being even seniors hold an edge over juniors, etc. A big flag is that all returning starters, or players that are expected to leap frog another starter rate higher compared to a starter that simply moved into the spot because the better player graduated.

4. A player evaluated as elite, preseason All-Conference players or projected draft picks, further calculate as a bigger part of his unit’s success.

5. Sum of the parts cannot be greater than the whole. Once all these calculations are made, the total value of all 1st string players on each team is compared to the actual preseason projections for the team. To take an extreme case, if a team somehow had three offensive All-Americans and three defensive All-Americans and yet projected as only a .500 team, the calculations would estimate that the other players on the team were not very good.

Obviously these five steps are a very simplified explanation of the math.

Suffice it to say that database of the points that each of the 6,186 players would improve their team is an “ALPHA” test, not even a BETA test yet, which is why the google doc with each players’ value is public and you can go there and type in the column to the right to note when players are missing from a team or if one players is clearly better and another player clearly worse than their rating.

You can click on the database and sort all players  can be sorted by team, position, value etc., but for me to see notes and update the list you need to type in Column J by typing over the words that say “type missing players, notes here.” The Value Add Basketball database can also be edited here.

And yes, in a first attempt at ranking 6,000+ plus players there will be players missing, and other times where you know from watching your team that another offensive lineman should be rated higher than they one that initially shows up first, etc. The nice thing is you can actually make the note by the player in the database and this list can continue to grow more accurate with input.