Study of Draft Prospects with Stats

[PonyExpress]

Unlocking the Secrets of the Draft
Part I, Defense: Linebackers from the BCS Conferences
Statistical Analysis of 20 BCS Conference Linebackers, 2004-2005 seasons.

I have used college statistics to study this years draft prospects. I felt a two year sample was a better indicator of a potential quality pro than a one year sample for some obvious reasons. Statistics over two years put a premium on Production Consistency Durability and Motivation. I think a player who has had the motivation to produce consistently over two consecutive years in college while remaining healthy has met certain bottom line criteria for success in pro ball. A two year analysis may be biased against a meteoric talent, but it errs in favor of safety and a high floor, and weeds out the one year wonders who could just as easily be smelling NFL green as coming into their own. This study is not suggesting that all highly productive players in college will become quality pros; far from it. More to the point, it is asking the question, if a player has not been productive, consistent, durable and motivated compared to his peers in college, why should an NFL team expect him to become so in the pros?
The first group I’m going to look at is Linebackers. Before beginning I’m making a few of assumptions: One, that all the major conferences in Division 1A have roughly the same talent level. I don’t think that’s a stretch after seeing Wisconsin manhandle Auburn and West Virginia’s rookies light up Georgia.

I. Linebacker Rankings: Impact in the Passing Game, 2004-2005 Seasons
Abbreviation Key:
PR: Position Rank
TFL: Tackles For Loss
S: Sacks
INTS: Interceptions
PBU: Passes Broken Up

Statistical Notes:
(a) In this study, for a LB a sack has roughly half the value of an INT.
(b) for a LB, a PBU has roughly a quarter the value of an INT.
(c) The tie breaker for positional rank is S + ((INTS +PBU/4) x 2). The second tiebreaker is the turnover column.

Pos. Rank, Player Sacks(PR) INTS + PBU/4(PR) PRaggregate
1. DQ. Jackson, MD 8(5) 7.5 (1) (6)
2. Hawk, Ohio St 11(2) 5 (4) (6)
3. Ingram, OK 7(6) 6.75(3) (9)
4. Alston, Stanf*** 17(1) 2.25(9) (10)
5. Greenway, Iowa 5(8) 7.25(2) (10)
6. Ellison, Oreg.St 9(4) 4.25(6) (10)
7. Havner, UCLA 3(10) 7.5 (1) (11)
8. Carpenter, OSU 10(3) 3.75(8) (11)
9. Wilkinson, GT 9(4) 4 (7) (11)
10. Sims, FSU 7(6) 4.25(6) (12)
11. Anderson, VT 6(7) 4.25(6) (13)
12. Nicholson, FSU 5(8) 4 (7) (15)
13. Gaither, Tenn 6(7) 3.25(8) (15)
14. McGarigle, Nw 3(10) 4.5 (5) (15)
15. McIntosh, UMia 10(3) .75(13) (16)
16. Parham, UV 11(2) .25(14) (16)
17. Robinson, ASU 9(4) 1 (12) (16)
18. Tulloch, NCSt. 6(7) 1.75(11) (18)
19. Ryans, Alabama 5(8) 2 (10) (18)
20. Hodge, Iowa 4(9) 1 (12) (21)

***By its nature a ranking system views the large number by which Alston outdistanced his peers in sacks as an anomaly, and reduced the significance of that number.

II. Linebacker Rankings: Impact in the Running Game, 2004-2005
Abbreviations:
PR: Position Rank
TFL-S: Tackles For Loss minus Sacks
TT-S: Total Tackles minus Sacks

Statistical Notes:
(a) For a LB, a TFL in the run game has roughly the value of ten conventional tackles.
(a) Total Tackles are ranked by decades. Thus 251 and 258 would be considered equals.
(c) Tie-breaker for pos. rank is ((TFL-S) x 10) + (TT-S). Second tiebreaker is TFL-S.

Pos. Rank, Player TFL – Sacks(PR) TT-S (PR) PR aggregate
1. Tulloch, NCSt. 25 (1) 214 (6) (7)
2. Havner, UCLA 21 (2) 221 (5) (7)
3. Robinson, ASU 20 (3) 200 (7) (10)
4. McGarigle, Nw 10 (10) 304 (1) (11)
5. Greenway, Iowa 13 (8) 264 (3) (11)
6. Hawk, OSU 14 (7) 251 (4) (11)
7. Hodge, Iowa 12 (9) 270 (2) (11)
8. Wilkinson, GT 18 (4) 196 (8) (12)
9. Nicholson, FSU 17 (5) 183 (9) (14)
10. McIntosh, Mia 15 (6) 190 (8) (14)
11. Parham, UVA 15 (6) 172 (10) (16)
12. Ellison, OregSt. 17 (5) 157 (12) (17)
13. Jackson, MD 6 (14) 252 (4) (18)
14. Gaither, Tenn 14 (7) 164 (11) (18)
15. Ryans, Ala 14 (7) 149 (14) (21)
16. Sims, FSU 12 (9) 151 (13) (22)
17. Anderson, VT 10 (10) 124 (16) (26)
18. Ingram, OK 9 (11) 104 (17) (28)
19. Carpenter,OS 7 (13) 132 (15) (28)
20. Alston, Stanf. 8 (12) 98 (18) (30)




III. Overall Linebacker Rankings, 2004-2005. Passing Game and Running Game Combined.
Abbreviations:
PR@Pass: Position Rank against the Pass
PR@Run: Position Rank against the run

Statistical Note:
(a) Tiebreaker for LB is PR@Pass.

Pos. Rank, Player PR@Pass PR@Run PR aggregate
1. Hawk, OSU 6 11 17
2. Havner, UCLA 11 7 18
3. Greenway, Iowa 10 11 21
4. Wilkinson, GT 11 12 23
5. DQ. Jackson, MD 6 18 24
6. Tulloch, NCSt. 18 7 25
7. Robinson, ASU 16 10 26
8. McGarigle, Nw 15 11 26
9. Ellison, OregSt. 10 17 27
10. Nicholson, FSU 15 14 29
11. McIntosh, Mia 16 14 30
12. Parham, UVA 16 16 32
13. Hodge, Iowa 21 11 32
14. Gaither, Tenn 15 18 33
15. Sims, FSU 12 22 34
16. Ingram, OK 9 28 37
17. Carpenter, OSU 1 28 39
18. Anderson, VT 13 26 39
19. Ryans, Ala 18 21 39
20. Alston, Stanf. 10 30 40


Observations:
This ranking system is not about athletic talent. It is about Production, Durability, Consistency, and Motivation over a two year period. This system is blind to every measurement of a player that is not reflected in on-field production. It is useful because it helps us identify possible risks and possible sleepers. For instance:

Demeco Ryans: Ryans was not productive at an elite level either in the running game or in the passing game compared to his fellow draft prospects over this two year period. He was captain and vocal leader of a very talented defense, and this may have masked his limitations as an individual player. A team that selects Ryans early in the draft expecting him to be a difference maker at the pro level would be taking a considerable risk.

Bobby Carpenter: Carpenter was second to last in productivity against the run among all 20 prospects studied. He had surprisingly few tackles for loss against the run, only 7 over a two year period while a starting outside LBer in the Big Ten. He also had over 120 fewer tackles over this two year period than AJ Hawk, and it is safe to assume that teams were not running away from Carpenter and at Hawk during that time. Carpenter’s marks against the pass are high, but his impact against the pass came primarily as a designated pass rusher, a non-traditional 4-3 linebacker role, and not in coverage. Half of his 8 sacks in 2005 came in a single game against MSU. His coverage production is middle of the road. All in all, Carpenter’s actual production does not justify his elite prospect status. A team drafting him high is taking a considerable risk.

Ernie Sims: Sims was a below average performer when compared to his draftable peers. He was middle of the road against the pass and sub-par against the run. We can guess as to the reasons for his elite prospect status: He’s from an elite program, he runs very fast and he hits very hard. However there are several negatives about him that are expressed in his production numbers. He has durability issues due to concussion problems. He is undersized and gets washed out of run plays. His instincts can be lacking as he tends to take poor angles and over-pursue. In other words, he is a better athlete than a football player, and the team drafting him very high will be taking a considerable risk.

Abdul Hodge: Hodge has been discussed in some places as a 2nd round prospect. In this study, Hodge finished dead last among all prospects against the pass. If his impact against the pass is so limited at the college level, is it likely that he will improve on it in the NFL? In other words, Hodge is a two down LBer in the NFL. It could be argued that an NFL team drafting a two down LBer in the 2nd rd would be reaching considerably.

Spencer Havner: Havner is an intriguing subject. Often dismissed by draft experts, he was very nearly the most productive LBer in college football over the last two years. He had a remarkable 20 TFLs against the run game during that time, while also exhibiting excellent production in coverage. Pro Football Weekly dismisses his coverage abilities by saying “… his interceptions were gifts thrown right at himâ€, and perhaps his numbers in that area are inflated by playing in the pass happy Pac-10. But conversely, you could argue that he is more pro ready due to experience in that facet of the game. As for his “accidental†interceptions and PBUs, over his 4 year career Havner had a remarkable 11 interceptions and 22 PBUs. He showed durability, consistency and motivation to play with pain, and also demonstrated above average athletic measurables at the combine, comparable with Chad Greenway, to whom his production most closely compares. My study suggests he could be a sleeper in the draft.

Gerris Wilkinson: Wilkinson seems to be an overlooked talent, ranking as the fourth most productive LBer over the last two years. He was Carpenter’s equal in the passing game, but showed much greater productivity against the run. Combined with his size (6’3â€, 233 lbs), sub 4.7 speed, and on-field leadership abilities, this study also suggests that Wilkinson could be a sleeper in this draft.

 

[PonyExpress]

Part II

Part II, Defense: 4-3 Defensive Ends from the BCS Conferences
Statistical Analysis of 18 BCS Conference Defensive Ends, 2004 and 2005 seasons.


I. 4-3 Defensive End Rankings: Impact in the Passing Game, 2004-2005 Seasons
Abbreviation Key:
PR: Position Rank
S: Sacks
FFP: Forced fumbles in the passing game
PBU: Passes Broken Up

DE Statistical Notes:
For a 4-3 DE in this study,
(a) a Forced Fumble is worth roughly half a turnover.
(b) a PBU is worth roughly ¼ of an interception.
(c) a PBU is worth roughly ½ of a forced fumble.
(d) a Sack has roughly the same value as a PBU.
(e) I have reduced the stats from Elvis Dumerville’s 2004 season by 1/3 because he was playing in a mid major conference at the time.
(f) The First Tie breaker in Position ranking for DEs is S + ((FFP + PBU/2) x 2). The second tie breaker is (FFP + PBU/2).


Pos. Rank, Player S(PR) FFP + PBU/2 (PR) PR aggregate
1. Dumerville, Lvl 25(1) 8(1) (2)
2. Lawson, NCSt 18(4) 5.5(3) (7)
3. Kiwanuka, BC 21(2) 3.5(6) (8)
4. Tapp, VT 19(3) 4 (5) (8)
5. McClover, Aub 16(5) 5 (4) (9)
6. M. Williams,NCS 21(2) 2.5(8) (10)
7. Keith, Kans 12(8) 6.5(2) (10)
8. Ninkovich, Pur 16(5) 4 (5) (10)
9. Lacasse, Syr 14(6) 4 (5) (11)
10. Haralson, Tenn 16(5) 3 (7) (12)
11. Edwards, Pur 14(6) 3.5(6) (12)
12. Adeyanju, Ind 11(9) 5 (4) (13)
13. Hali, PSU 13(7) 3.5(6) (13)
14. Wimbley, FSU 10(10) 3 (7) (17)
15. Jenkins, Stan 13(7) 1 (10) (17)
16. Rucker, USC 9(11) 3 (7) (18)
17. Anderson, Al 9(11) 2.5 (8) (19)
18. Oliver, LSU 10(10) 1.5 (9) (19)



II. 4-3 Defensive End Rankings: Impact in the Running Game, 2004-2005
Abbreviations:
PR: Position Rank
TFL-S: Tackles For Loss minus Sacks
TT-S: Tackles minus Sacks

Statistical Notes.
For a DE:
(a) a tackle for loss in the run game has roughly the same value as 10 conventional tackles.
(b) a decade of tackles is considered the same value (72 or 79 is statistically negligible)
(c) the tie-breaker for pos. rank for DEs is (TFL minus sacks) x 10 + (TT– S). The second tie breaker is TFL-S.



Pos. Rank, Player TFL-S (PR) TT-Sacks (PR) PR aggregate
1. Kiwanuka 21 (1) 97 (2) (3)
2. M. Williams 19 (2) 98 (2) (4)
3. Hali*** 16 (3) 101 (1) (4)
4. Anderson 16 (3) 72 (4) (7)
5. Lawson 14 (5) 90 (2) (7)
6. Haralson 15 (4) 73 (4) (8)
7. Tapp 13 (6) 90 (2) (8)
8. Keith 16 (3) 56 (6) (9)
9. Rucker 13 (6) 76 (4) (10)
10. Adeyanju 11 (6) 77 (4) (11)
11. Ninkovich 11 (7) 55 (6) (13)
12. Oliver 6 (11) 95 (2) (13)
13. Edwards 9 (8) 60 (5) (13)
14. Jenkins 5 (12) 85 (3) (15)
15. McClover 8 (9) 51 (7) (16)
16. Lacasse 5 (12) 70 (4) (16)
17. Dumerville 4 (13) 74 (4) (17)
18. Wimbley 7 (10) 44 (8) (18)

*** A shortcoming of the ranking model is exhibited here, where M. Williams has three fewer total tackles but 3 more sacks than Hali, but the two end up with same aggregate ranking (4). This is compensated for by placing Williams at #2 overall and Hali at #3.


III. Overall 4-3 Defensive End Rankings, 2004 and 2005. Passing Game and Running Game Combined.
Abbreviation Key
PR: Position Rank
PR@Pass: Pos Rank against the Pass
PR@Run: Pos Rank against the Run

Statistical Note:
(a) Tie breaker for pos rank among DEs is impact against the pass.

Pos. Rank, Player PR@Pass PR@Run PR aggregate
1. Kiwanuka 8 3 (11)
2. Lawson 7 7 (14)
3. M. Williams 10 4 (14)
4. Tapp 8 8 (16)
5. Hali 13 4 (17)
6. Dumerville 2 17 (19)
7. Keith 10 9 (19)
8. Haralson 12 8 (20)
9. Ninkovich 10 13 (23)
10. Adeyanju 13 11 (24)
11. McClover 9 16 (25)
12. Edwards 12 13 (25)
13. Anderson 19 7 (26)
14. Lacasse 11 16 (27)
15. Rucker 18 10 (28)
16. Jenkins 17 15 (32)
17. Oliver 19 13 (32)
18. Wimbley 17 18 (35)


Observations:

Kamerion Wimbley: Wimbley ranked last of all 18 DEs studied in on field production over the last two years. This study does not measure innate athletic talent, only how that talent makes an impact on the football field. By any measure, Wimbley’s impact was marginal compared to his peers. He played sparingly as a junior, showed durability problems by missing four games as a senior, and had minimal impact in the run game throughout. Wimbley is said to have a high character and be a late-bloomer; However, there is no on-field evidence to suggest that Wimbley is durable enough to play in the NFL, can stand up against the run for game after game, or is motivated to play the game for its own sake (he only turned it on as a senior). He could very well be that special person who can do all those things. But a team drafting him in the first rd based on one week of practice at the Senior Bowl and his athletic performance at the combine, while diminishing his track record in college, will be taking a considerable risk.


Mario Williams: Williams is ranked highly, the third most productive of all DE prospects in this study, but not as highly as his amazing athletic talent would lead us to believe. The reason is that he had only one great year. The question with Williams has always been his motor and consistency. We know he was laboring through an average junior season until his coach gave him a tongue lashing. We have to ask ourselves, is he motivated to get the pay day or to play football and maximize his potential? That inconsistent motor and immaturity is reflected in his #3 ranking in terms of production. It is still high, but not as high as you would like the potential #2 pick in the draft to be.

Manny Lawson: Lawson was tabbed as the compound freak who would jump up draft boards based on his athleticism. But studying his last two years of college play, he was much more than that. He was the #2 most productive DE in college football. He was active in every phase of the game: 2nd among his peers in PBUs with 9, showing his leaping ability (that’s more PBUs than Donte Whitner had over the same period playing safety for Ohio State); he showed durability, and consistently made plays behind the line of scrimmage both against the run and the pass. He was also special teams Captain for the Wolfpack and Defensive Team Captain. His production, combined with the upside his athleticism suggests, indicates Lawson could be one of the few prospects in the draft with both a high floor and a high ceiling.

Charlton Keith: this pass rush specialist came out of nowhere to finish 7th in production, leading all DEs with 10 PBUs over the last two years, as befits his basketball background. His cousins are NBA guard Eric Snow and former NFL LBer Percy Snow. A developmental player (only 12 strength lifts), but a possible sleeper to keep an eye on with his height (6’4.3â€), natural athleticism and college production.

Mathias Kiwanuka: Reportedly, Kiwanuka has been slipping on draft boards for the same reasons Wimbley has been rising: his offseason workouts. What is interesting is the remarkable consistency Kiwanuka has demonstrated over the last two years (actually his whole 4 year career). Unlike Wimbley, he played hurt and still produced. He is not a one year wonder, but stayed in school as the Captain of his team as it entered a new conference. He was equally productive against the pass and the run, surprisingly, registering the most TFLs against the run of any DE studied, and nearly led all DEs in TT minus sacks. He was a hallmark of production, consistency, durability and motivation over the last two years, and if this study can be believed, should have one of the highest floors in the draft. Pro Football Weekly chides his pass rush production as against “inferior competitonâ€. However, Maryland and NCState are not inferior, and he was stout against the run against Florida State. It will be interesting to follow the course of Kiwanuka’s career and see whether he is the boom or bust people are buzzing about. This study suggests that he is more likely to boom, and Wimbley more likely to bust, the opposite of conventional wisdom.

 

[PonyExpress]

Part III

Part III, Defense: Cornerbacks from the BCS Conferences
Statistical Analysis of 17 BCS Conference Cornerbacks, 2004-2005 seasons.


I. Cornerback Rankings: Impact in the Passing Game, 2004-2005 Seasons
Abbreviation Key:
PR: Position Rank
PBU: Passes Broken Up
TFL: Tackles For Loss
S: Sacks
INTS: Interceptions

Statistical Notes.
(a) for a CB, a PBU is worth roughly half the value of an INT. For a LB, a PBU is worth roughly a quarter the value of an INT. This difference in relative value of an INT is a combination of several factors: (i) responsibilities of each position; (ii) degree of difficulty of a statistic for each position.
(b) the tiebreaker in Pos Rank goes to the INT column.
(c) I have used the 2003 and 2004 seasons for Blackmon, since he played WR in 2005

Pos. Rank, Player PBU/2(PR) INTS (PR) PR aggregate
1. Youboty, OSU 12 (2) 5(5) 7
2. Zemaitis, PSU 8 (6) 8(2) 8
3. J. Williams, VT 10 (4) 6 (4) 8
4. D.Webb, UF 12 (2) 4(6) 8
5. Huff , UT 14 (1) 3(7) 8
6. Hill, Clem 14 (1) 3(7) 8
7. Gordon, Kans 6 (8) 9(1) 9
8. A. Philips, PSU 10(4) 5(5) 9
9. R. Marshall, FS*** 8 (6) 6(4) 10
10. K. Jennings, Mia 9 (5) 5(5) 10
11. Blackmon, BC 6 (8) 7(3) 11
12. T. Jennings, Geo 7(7) 6(4) 11
13. C. Griffin, UT 11(3) 2(8) 11
14. Minter, Geo 10(4) 2(8) 12
15. McPhearson, UM 9(5) 0(9) 14
16. Maxey, Mia 3 (9) 2(8) 17
17. Walker, USC 2 (10) 2(8) 18

***Marshall is not in a BCS conference. His stats are not truly comparable.
II. Cornerback: Impact in the Running Game, 2004-2005
Abbreviations:
PR: Position Rank
TT: Total Tackles
FFR: Forced Fumbles in the Running Game

Statistical Notes:
(a) for a CB, a FF against the run has a value of roughly 20 conventional tackles.
(b) Tackles are ranked by decades. Thus 51 and 58 would be considered equals.
(c) Tie-breaker is (FFR x 20) + T

Pos. Rank, Player TT (PR) FFR (PR) PR aggregate
1. Huff 182 (1) 5 (1) 2
2. C. Griffin 154 (2) 3 (3) 5
3. R. Marshall 139 (3) 1 (5) 8
4. Blackmon 110 (4) 2 (4) 8
5. K. Jennings 72 (7) 4 (2) 9
6. Webb 107 (5) 2 (4) 9
7. Zemaitis 98 (6) 2 (4) 10
8. Hill 106 (5) 1 (5) 10
9. Youboty 117 (4) 0 (6) 10
10. Minter 99 (6) 1 (5) 11
11. Philips 99 (6) 1 (5) 11
12. Gordon 75 (7) 2 (4) 11
13. T. Jennings 93 (6) 1 (5) 11
14. J. Williams 104 (5) 0 (6) 11
15. McPhearson 75 (7) 1 (5) 12
16. Maxey 48 (8) 0 (6) 14
17. Walker 34 (9) 0 (6) 15


III. Overall Cornerback Rankings, 2004-2005. Passing Game and Running Game Combined.

Statistical Notes.
(a) For a CB, in this study production in the passing game is considered twice as valuable as production in run support.
(b) Tie breaker is production in the Passing game


Pos. Rank, Player PR@Passx2 PR@Run PR aggregate
1. Huff 16 2 18
2. Youboty 14 10 24
3. D.Webb 16 9 25
4. Hill 16 10 26
Zemaitis 16 10 26
6. J. Williams 16 11 27
7. C. Griffin 22 5 27
8. R. Marshall 20 8 28
9. Gordon 18 11 29
A. Philips 18 11 29
11. K. Jennings 20 9 29
12. Blackmon 22 8 30
13. T. Jennings 22 11 33
14. Minter 24 11 35
15. McPhearson 28 12 40
16. Maxey 34 14 48
17. Walker 36 15 51
Observations:

Alan Zemaitis: Zemaitis is currently being downgraded due to his postseason workouts. In contrast, this study has him ranked as the 4th most productive CB over the last two years in college football, and also the best playmaker, posting 8 INTs and 2 Forced fumbles. His speed is questioned, but is comparable to that of Chris Gamble when he came out two years ago. Zemaitis’ production, consistency, durability and motivation during his college career compensate for some of his lower than expected measurables at PSU’s pro day and suggest he will be a solid Day One draft pick.

Dee Webb: has a lot of bad press for breaking Prothro’s leg against Alabama, for his ****y attitude, and for his gun ownership. However, Webb had an exceptional 23 PBUs over the last two years, more INTs than Tye Hill, and was solid in run support. Combined with his excellent speed in workouts and decent size, Webb has the potential to be a very solid pro, and this study suggests that he should be drafted higher compared to his peers than is being suggested elsewhere.

 

[PonyExpress]

Part 4

Part IV, Defense: Defensive Tackles from the BCS Conferences
Statistical Analysis of 17 BCS Conference Defensive Tackles, 2004 and 2005 seasons.


I. Defensive Tackles: Impact in the Passing Game, 2004-2005 Seasons.
Abbreviation Key:
PR: Position Rank
S: Sacks
BB: Batted balls

Statistical Notes:
(a) for DTs, a batted ball has roughly the same value as a sack.
(b) the tie breaker in position ranking is #sacks.

PR, Name S(PR) BB(PR) PRaggregate
1. K. Williams, LSU 11(2) 7(2) 4
2. Wroten, LSU 12(1) 3(5) 6
3. Lewis, VT 10(3) 4(4) 7
4. Cofield, NW 6(6) 8(1) 7
5. Oliver, LSU 10(3) 3(5) 8
6. Stanley, LV 10(3) 3(5) 8
7. Ngata, Oreg 7(5) 5(3) 8
8. Oshinowo, Stan 8(4) 3(5) 9
9. Wright, UT 6(6) 5(3) 9
10. Bunkley, FSU 10(3) 0(7) 10
11. Jolly, TAM 3(9) 7(2) 11
12. Mahelona,Ten 7(5) 0(7) 12
13. McDaniel,Ten 1(10) 5(3) 13
14. Harris, Mia 5(7) 0(7) 14
15. Dvoracek, OK 4(8) 1(6) 14
16. Watson, Mich 4(8) 0(7) 15
17. McCargo, NCSt (10) 0(7) 17


II. Defensive Tackle Rankings: Impact in the Running Game, 2004-2005
Abbreviations:
PR: Position Rank
TFL: Tackles For Loss minus Sacks
TT-S: Total Tackles minus Sacks

Statistical Notes:
(a) for a DT a FF has roughly twice the value as a Tackle For Loss.
(b) For a DT, a TFL in the run game has roughly the value of 5 conventional tackles.
(c) Total Tackles are ranked by decades. Thus 72 and 76 are considered equal.
(d) First Tie Breaker for Position rank is TT – S + 5(TFL + FFx2); second tiebreaker is #TFL+(FFx2)

Pos. Rank, Player TFL+ (FFx2) (PR) TT-S (PR) PR aggregate
1. Ngata 18 (3) 100 (1) 4
2. Mahelona 22 (1) 70 (4) 5
3. Harris 18 (3) 90 (2) 5
4. Bunkley 20 (2) 68 (5) 7
5. K. Williams 11 (6) 107 (1) 7
6. Wroten 15 (4) 81 (3) 7
7. Cofield 10 (7) 109 (1) 8
8. Jolly 10 (7) 106 (1) 8
9. Wright 15 (4) 76 (4) 8
10. McCargo 15 (4) 71 (4) 8
11. Watson 12 (5) 73 (4) 9
12. Oshinowo 10 (7) 87 (3) 10
13. Oliver 6 (8) 95 (2) 10
14. Dvoracek 15 (4) 45 (6) 10
15. Stanley 11 (6) 68 (5) 11
Lewis 11 (6) 68 (5) 11
17. McDaniel 4 (9) 25 (7) 16



III. Overall Defensive Tackle Rankings, 2004 and 2005. Passing Game and Running Game Combined.

Abbreviation Key
PR@Pass: position rank against the pass
PR@Run: position rank against the run

Note#1: For a DT, by the nature of his position his impact against the run is twice as valuable as his impact against the pass.
Note#2: Tiebreaker for DTs is production against the run.

Pos. Rank, Player PR@Pass PR@Run PR aggregate
1. Ngata 8 4 16
2. K. Williams 4 7 18
3. Wroten 6 7 20
4. Mahelona 12 5 22
5. Cofield 7 8 23
6. Harris 14 5 24
7. Bunkley 10 7 24
8. Wright 9 8 25
9. Jolly 11 8 27
10. Oliver 8 10 28
11. Oshinowo 9 10 29
12. Lewis 7 11 29
13. Stanley 8 11 30
14. McCargo 17 8 33
16. Watson 15 9 33
15. Dvoracek 14 10 34
17. McDaniel 13 16 45


Observations:

Kyle Williams: appears to be a sleeper in this draft. Impacts on the field in every way possible: sacks, batted balls, TFLs, FFs. His production indicates a well rounded game, a measure of consistency, motivation and durability that is underscored in reports about him from the Senior Bowl. This study indicates a high floor.

Broderick Bunkley: An interesting study. Like Kamerion Wimbley, a one year wonder. Had no batted passes in last two years. Convicted of shoplifting from WalMart as a Freshman. Has endured major durability issues, which partly accounts for his sub-par junior season. Had an amazing final year, like Mario Williams, but there are some questions that his great combine performance should not erase, especially with the track record FSU has of producing immature athletes.

Barry Cofield: An interesting candidate for the 3-4 DE at 6’4†304 lbs. A Solid contributor against both the run and the pass, used his height to bat an exceptional 8 passes over the last two years. Make sure to test him for steroids (Remember Castillo, Cofield’s teammate, whom SD drafted in Rd 1 last year)!

Gabe Watson: Stats show that Watson was not productive, consistent or motivated over this two year period against the run or the pass. The fact he put out effort during the Senior Bowl should not remove doubt that any team thinking of drafting him on the first day is taking a big risk.

 

[PonyExpress]

Part V

Part V, Defense: Safeties from BCS Conferences
Statistical Analysis of 20 BCS Conference Safeties, 2004-2005 seasons.

I. Safety Rankings: Impact in the Passing Game, 2004-2005 Seasons
Abbreviation Key:
PR: Position Rank
PBU: Passes Broken Up
INTS: Interceptions

Statistical Notes.
(a) for a DB, in this study a PBU is worth roughly half the value of an INT. For a LB, a PBU is worth roughly a quarter the value of an INT. This difference in relative value of an INT is a combination of several factors: (a) responsibilities of each position; (b) degree of difficulty of a statistic for each position.
(b) the tiebreaker goes to the turnover column.
.

Pos. Rank, Player PBU/2 (PR) INTS (PR) PR aggregate
1. Bullocks, Neb 10 (1) 6(4) 5
2. Smith, Syr 6 (5) 9(1) 6
3. Simpson, SC 8 (3) 7(3) 6
4. Watkins, FSU 8 (3) 7(3) 6
5. Hudson, NCSt**9 (2) 4(6) 8
6. Bing, USC 7 (4) 6(4) 8
7. Page, UCLA 7 (4) 4(6) 10
8. Lowry, PSU 2 (9) 8(2) 11
9. Landry, GT 4 (7) 5(5) 12
10. Pollard, Purd 4 (7) 4(6) 13
11. Salley, OSU 5 (6) 3(7) 13
12. Harper, Ala 3 (8) 4(6) 14
13. Whitner, OSU 4 (7) 3(7) 14
14. Blue, UGA 4 (7) 2(8) 15
15. Allen, Tenn 4 (7) 2(8) 15
16. Peprah, Ala 4 (7) 1(9) 16
17. Slay, TT 1 (10) 3(7) 17
18. Harrell, PSU 2 (9) 1(9) 18
19. Iwur, Col 2 (9) 1(9) 18
20. Ware, USC 2 (9) 1(9) 18

** Hudson tore a tendon in his knee at the Senior Bowl


II. Safety Rankings: Impact in the Run Support, 2004-2005
Abbreviations:
PR: Position Rank
TT: Total Tackles
FFR: Forced Fumbles in the Run Game

Statistical Notes:
(a) for a Safety, an FFR is worth roughly 10 conventional tackles, while for a CB, a FFR is worth roughly 20 conventional tackles. This difference in relative value of a FF is a combination of several factors: (a) responsibilities of each position; (b) degree of difficulty for each position.
(b) Total Tackles are ranked by decades. Thus 51 and 58 would be considered equals.
(c) First Tie-breaker goes to TT + (FFR x 10). Second tie breaker goes to #Ints.

Pos. Rank, Player TT (PR) FFR (PR) PR aggregate
1. Blue 176 (2) 4 (2) 4
2. Pollard 188 (1) 2 (4) 5
3. Allen 158 (4) 4 (2) 6
4. Page 151 (4) 4 (2) 6
5. Slay*** 137 (6) 8 (1) 7
6. Smith 156 (4) 3 (3) 7
7. Harper 146 (5) 3 (3) 8
8. Bullocks 141 (5) 3 (3) 8
9. Iwur 171 (2) 0 (6) 8
10. Watkins 133 (6) 3 (3) 9
11. Simpson 164 (3) 0 (6) 9
12. Landry 157 (4) 1 (5) 9
13. Whitner 142 (5) 1 (5) 10
14. Hudson 126 (7) 2 (4) 11
15. Bing 113 (8) 3 (3) 11
16. Lowry 129 (7) 1 (5) 12
17. Harrell 137 (6) 0 (6) 12
18. Peprah 94 (9) 2 (4) 13
19. Ware 67 (10) 3 (3) 13
20. Salley 98 (9) 1 (5) 14

***In the case of Slay, the ranking system by its nature views his amazing edge over his peers in number of FFs as an anomaly, and reduced its significance.


III. Overall Safety Rankings, 2004-2005. Passing Game and Running Game Combined.

Statistical Notes.
(a) For a Safety, production in the passing game is of equal value with production in run support.
(b) Tie breaker: Production against the pass.


Pos. Rank, Player PR@Pass PR@Run PR aggregate
1. Bullocks 5 8 13
2. Smith 6 7 13
3. Simpson 6 9 15
4. Watkins 6 9 15
5. Page 10 6 16
6. Pollard 13 5 18
7. Bing 8 11 19
Hudson 8 11 19
9. Blue 15 4 19
10. Landry 12 9 21
11. Allen 15 6 21
12. Harper 14 8 22
13. Lowry 11 12 23
14. Whitner 14 10 24
15. Slay 17 7 24
16. Iwur 18 8 26
17. Salley 13 14 27
18. Peprah 16 13 29
19. Harrell 18 12 30
20. Ware 18 13 31


Observations:

Anthony Smith, Syracuse: Smith could be a major sleeper. This study ranks him as the second most productive safety in college football over the last two years. His production indicates tremendous ball skills and willing support of the run. Seems very much under the radar compared with Whitner, Simpson, Bullocks and Watson.

Donte Whitner: Whitner has all the athletic measurables, but is a one-year wonder at the position leaving early. In that one year he managed 6 PBUs and 2 INTs, respectable but not elite production (Smith had 8 PBUs and 6 INTs), but has had durability concerns (knee surgery before last year) and another year would have demonstrated his small stature is durable enough to play safety. An enticing prospect considering his combine speed and intensity on the field, but durability and consistency are issues that make drafting him in rd 1, as some have suggested, a risk.

 

[drpatriot]

Excellent post that gives insight beyond what a 40 time and vertical leap will tell you.

I think you're onto something, but I would be interested to see how well your system would rank prospects from previous years.

 

[PonyExpress]

Excellent post that gives insight beyond what a 40 time and vertical leap will tell you.

I think you're onto something, but I would be interested to see how well your system would rank prospects from previous years.

Sorry about the spacing in the stat lines. It looks awful. I tried to fix it but for some reason it didn't work out on screen. i think the information would be much friendlier to the eyes if the columns were in line.

 

[bosfan]

Thanks for that PE, will definitely be referring to that as we near the draft. Certainly one of the keys to evaluating talent is to eliminate the hype - this helps a lot.

 

[Flying Fungi]

thanks buddy--I'm a fan of Kiwi...

Mathias Kiwanuka: Reportedly, Kiwanuka has been slipping on draft boards for the same reasons Wimbley has been rising: his offseason workouts. What is interesting is the remarkable consistency Kiwanuka has demonstrated over the last two years (actually his whole 4 year career). Unlike Wimbley, he played hurt and still produced. He is not a one year wonder, but stayed in school as the Captain of his team as it entered a new conference. He was equally productive against the pass and the run, surprisingly, registering the most TFLs against the run of any DE studied, and nearly led all DEs in TT minus sacks. He was a hallmark of production, consistency, durability and motivation over the last two years, and if this study can be believed, should have one of the highest floors in the draft. Pro Football Weekly chides his pass rush production as against “inferior competitonâ€. However, Maryland and NCState are not inferior, and he was stout against the run against Florida State. It will be interesting to follow the course of Kiwanuka’s career and see whether he is the boom or bust people are buzzing about. This study suggests that he is more likely to boom, and Wimbley more likely to bust, the opposite of conventional wisdom.

 

[chowder]

interesting analysis. But do these stats take into account injuries? For example, since Carpenter was injured for part of the season, could that be the reason for his low ranking in stats if he wasn't on the field?

 

[Ochmed Jones]

Based on these stats, if Tulloch was 5 inches taller, he would be ideal for the Pats. Great post.

 

[Box_O_Rocks]

Bump!

Interesting read PE, many thanks!

 

[Feep_FLA]

Outstanding post.... PonyExpress Really interesting reading.

 

[c_nice_37]

This is really great work. It gives an uncanny insight into the 'total player' and not just a pass rush freak for example.

Good stuff, I appreciate the analysis.

 

[big mike]

It seems rather odd to use the same numbers system for rating all LBs, whether ILB, MLB, or OLB. For example, OLBs generally aren't going to have as many INTs as MLBs, as OLBs are more often playing in the backfield. Incorporating INTs into a "rating against the pass" for all LBs equally just generally skews the results.

The analysis also completely ignores the impact of injuries and other issues. For example, Carpenter shows up with only 132 tackles. That ignores the fact that in his junior year he had 93 tackles (http://www.nflfans.com/x/2006/showplayer.php?key=Bobby Carpenter ), and his number of tackles dropped drastically his senior year due to injury and other issues.

My point being - just analyzing based on stats doesn't give you much of a whole picture. If you just look at stats, you'd think Richard Seymour isn't a top DLman.

Add to that that it seems rather silly to equate a TFL to ten conventional tackles. So, if you tackle a guy for a 0.5 yard loss, that's 10 times as good as tackling him for a 0.1 yard gain? And a 0.5 yard loss counts just as much as tackling someone for a 10 yard loss? That's the problem with these pure statistical analyses - they ignore far too much, make too many assumptions, etc..

 

[PonyExpress]

It seems rather odd to use the same numbers system for rating all LBs, whether ILB, MLB, or OLB. For example, OLBs generally aren't going to have as many INTs as MLBs, as OLBs are more often playing in the backfield. Incorporating INTs into a "rating against the pass" for all LBs equally just generally skews the results.

The analysis also completely ignores the impact of injuries and other issues. For example, Carpenter shows up with only 132 tackles. That ignores the fact that in his junior year he had 93 tackles (http://www.nflfans.com/x/2006/showplayer.php?key=Bobby Carpenter ), and his number of tackles dropped drastically his senior year due to injury and other issues.

My point being - just analyzing based on stats doesn't give you much of a whole picture. If you just look at stats, you'd think Richard Seymour isn't a top DLman.

Add to that that it seems rather silly to equate a TFL to ten conventional tackles. So, if you tackle a guy for a 0.5 yard loss, that's 10 times as good as tackling him for a 0.1 yard gain? And a 0.5 yard loss counts just as much as tackling someone for a 10 yard loss? That's the problem with these pure statistical analyses - they ignore far too much, make too many assumptions, etc..

It sounds like the statistics aren't telling you what you want them to. You say Carpenter's production was limited due to injury and "other issues". What are those "other issues"? This is why a study of production is useful, forcing you to ask these questions and separate fact from fiction. A statistical study is just a tool, helping identify questions that need to be asked and answered. If Carpenter is so talented, and I'm not saying he isn't, why did he only make 7 tackles for loss in two years against the run while playing in the Big 10? That question needs to be answered by carefully watching ALL his game tape. Why did he accumulate so few tackles? My methodology is certainly open to criticism and I can be accused of making some arbitrary decisions. But attacking my methodology does not absolve Carpenter from having to answer some questions about his play. That is why these studies can be useful if their limitations are understood.

 

[bucky]

PonyExpress,

I applaud your hard work and the results you have attained. My doubt regarding the validity of the counclusions implied by your stats is similar to my doubts of the "football scientist" on ESPN. Football is a game where teams have entirely different defensive systems with entirely different responsibilities for players who seemingly play the same position. That's why comparing 2 guys who are considered ILBs may seem like an apples-to-apples comparison, but when you dig down under the covers, it really isn't.

Certainly, a player's statistics are an extremely important part of his "resume" when considering where he should be drafted. But those statistics should be considered in the context of the team's defensive scheme, surrounding players, and even opponents. For example, a great CB will not have great stats if the CB on his other side stinks, or if his DL stinks against the run and can't rush the passer, or if they play a third of their games against option teams.

 

[PonyExpress]

PonyExpress,

I applaud your hard work and the results you have attained. My doubt regarding the validity of the counclusions implied by your stats is similar to my doubts of the "football scientist" on ESPN. Football is a game where teams have entirely different defensive systems with entirely different responsibilities for players who seemingly play the same position. That's why comparing 2 guys who are considered ILBs may seem like an apples-to-apples comparison, but when you dig down under the covers, it really isn't.

Certainly, a player's statistics are an extremely important part of his "resume" when considering where he should be drafted. But those statistics should be considered in the context of the team's defensive scheme, surrounding players, and even opponents. For example, a great CB will not have great stats if the CB on his other side stinks, or if his DL stinks against the run and can't rush the passer, or if they play a third of their games against option teams.

Your arguments are sound. However, I think they apply more to the pro than to the college game. I think great players, big-time NFL prospects, distinguish themselves in college football by making plays, because their talent is so much more than their peers. Lack of production in the NFL can sometimes be explained when it applies to truly great players, although the NFL mantra is "great players make big plays in big games". Lack of production in college football is a real question mark. As far as my "conclusions", they are really just personal observations. The rankings are not ranking TALENT, but production. You have to draw your own conclusions. If you feel comfortable drafting a player who was not as productive as his peers in college, then you must have watched the tape that justifies your conclusions.

 

[Handel]

Manny Lawson FTW!

 

[DaBruinz]

PE -
I also want to applaud you on your work. It really is very well done and thought out.

As you say, this is just one tool to use to help evaluate players. If you develop any more tools and would like to share them with us, please don't hesitate because it will help broaden our perspective.