Getting Ready for the Combine; Must Read Statistics from @DraftMetrics

Getting Ready for the Combine

By Tony Villiotti of Draftmetrics

The NFL Combine has evolved from a virtually private workout into a major “made for television” event. There is constant chatter about how meaningless the Combine is, but it continues to gain popularity among fans. This year’s on-field workouts begin on March 3 with the bench press generally occurring the day prior to each position’s on-field workouts. The schedule for on-field activities is as follows:

March 3	Offensive line
March 4	Quarterbacks, wide receivers, tight ends
March 5	Defense line, linebackers
March 6	Defensive backs

There are two purposes for this article. The first is to present the historical standards for the Combine. This gives the television viewer a sense of what makes a good Combine performance for each drill. A fan can then make a judgment regarding how well a player performed and enhances the viewing experience. The second purpose is to review whether Combine performance is predictive of draft position.

Following are the seven drills for which I have accumulated historical information, principally from public information:

10 yard split in the 40 yard dash (10)	Broad Jump (BJ)
40 Yard Dash (40)	20 yard shuttle (20)
Bench Press (BP)	3 – cone drill (3C)
Vertical Jump (VJ)

The following table shows the average by drill over the past 10 Combines (2007-2016) for each playing position. This table shows averages for (1) all combine participants and (2) for only those participants that are drafted.

	All Participants								Drafted Participants
	10	40	BP	VJ	BJ	20	3C	10		40	BP	VJ	BJ	20	3C
QB	1.66	4.83	NA	21	111	4.35	7.11	1.66		4.81	NA	31	111	4.32	7.12
SRB	1.56	4.52	19	35	119	4.24	7.02	1.56		4.48	19	35	120	4.21	6.98
LRB	1.58	4.57	21	35	118	4.31	7.00	1.57		4.55	21	35	118	4.29	7.02
WR	1.56	4.49	NA	35	121	4.24	6.93	1.55		4.47	NA	35	121	4.23	6.90
TE	1.62	4.70	22	33	116	4.39	7.13	1.63		4.71	22	33	116	4.38	7.12
C	1.80	5.24	27	27	101	4.67	7.72	1.79		5.22	27	27	101	4.65	7.65
G	1.82	5.30	26	27	100	4.81	7.86	1.80		5.25	28	27	101	4.75	7.77
T	1.80	5.23	25	28	103	4.77	7.81	1.79		5.20	25	28	104	4.74	7.76
SDE	1.62	4.75	23	34	117	4.40	7.10	1.61		4.71	24	34	118	4.36	7.16
LDE	1.67	4.86	25	32	113	4.49	7.32	1.66		4.82	25	32	114	4.41	7.30
DT	1.74	5.09	28	29	105	4.66	7.67	1.73		5.07	29	29	105	4.64	7.66
ILB	1.62	4.70	22	33	115	4.33	7.14	1.60		4.70	23	33	116	4.30	7.11
OLB	1.60	4.69	23	34	117	4.33	7.14	1.60		4.68	23	34	118	4.33	7.12
CB	1.55	4.49	NA	36	122	4.18	6.93	1.54		4.47	NA	36	122	4.17	6.92
S	1.56	4.56	17	35	120	4.24	7.00	1.55		4.53	17	35	121	4.21	6.97

• SRB denotes running backs that weigh less than 215 pounds
• LRB denotes running backs that weigh 215 pounds or more
• SDE denotes defensive ends that weight 260 pounds or less
• LDE denotes defensive ends that weigh more than 260 pounds

The bench press is ignored for quarterbacks, wide receivers and corners as very few participants at those playing positions perform the bench. I have also excluded fullbacks since there are so few results in all drills.

The next table is essentially the same as the preceding one but compares the average results of early round draftees (rounds 1-3) to later round draftees (4-7).

	Rounds 1-3								Rounds 4-7
	10	40	BP	VJ	BJ	20	3C	10		40	BP	VJ	BJ	20	3C
QB	1.62	4.74	–	33	115	4.34	7.14	1.67		4.85	–	30	109	4.35	7.14
SRB	1.53	4.44	20	36	122	4.18	6.93	1.58		4.62	19	34	119	4.24	7.02
LRB	1.56	4.50	22	36	121	4.27	6.94	1.57		4.58	21	34	117	4.31	7.07
WR	1.55	4.45	–	36	123	4.24	6.90	1.55		4.48	–	35	120	4.23	6.90
TE	1.63	4.69	22	34	117	4.39	7.15	1.63		4.73	21	33	115	4.38	7.10
C	1.78	5.21	35	27	100	4.68	7.68	1.80		5.23	28	27	102	4.63	7.64
G	1.80	5.23	28	28	102	4.70	7.77	1.81		5.27	27	27	100	4.78	7.78
T	1.79	5.17	26	29	105	4.71	7.74	1.79		5.23	24	28	103	4.76	7.78
SDE	1.60	4.68	23	34	119	4.34	7.18	1.62		4.76	24	35	116	4.38	7.14
LDE	1.64	4.79	26	32	114	4.44	7.25	1.67		4.85	25	32	113	4.51	7.38
DT	1.73	5.07	29	29	105	4.65	7.66	1.73		5.08	28	29	104	4.65	7.67
ILB	1.59	4.67	23	34	117	4.28	7.11	1.61		4.72	22	33	116	4.32	7.10
OLB	1.59	4.66	23	35	120	4.33	7.10	1.61		4.69	23	34	116	4.32	7.13
CB	1.53	4.44	–	36	124	4.15	6.89	1.55		4.49	–	35	121	4.20	6.96
S	1.53	4.50	16	36	122	4.18	6.92	1.57		4.58	17	34	119	4.25	7.02

These tables set the historical standard for the Combine. Using that information as the base, can it be shown that Combine results are predictive of draft position? A large difference between all participants and players selected in rounds 1-3 indicates a drill that might be significant for a particular playing position.

An analysis of all draftees versus those drafted in the first three rounds shows that, of 102 data points, 91 had better results among players drafted in the first three rounds versus all draftees. Despite that overwhelming disparity, the difference between the two groups of players was small.

Here is the distribution of the 91 data points when comparing draftees from round 1-3 and round 4-7:

0.0-0.99%	29	3.0-3.99%	8	6.0-6.99%	2
1.0-1.99%	37	4.0-4.99%	2	7.0-7.99%	2
2.0-2.99%	10	5.0-5.99%	1	Total	91

Over 70% of the “positive” data points showed a 2% or lower difference. So there is very little difference among all participants and those drafted in the first three rounds. A few observations:

• In general, the Combine tells us very little at least when it comes to averages
• The 10 yard split and 40 yard dash differences were inconsequential
• The bench press probably showed the largest percentage difference but this often came down to a bench press or two
• Aside from the bench press, a favorable vertical jump difference was the next highest, but those differences were small
• The same could be said about the broad jump, but the differences were a shade smaller than even the vertical jump
• Differences in both the 20 yard shuttle and the three-cone drill were the most inconsequential of all

Maybe it is more meaningful to look at the predictive aspect of a drill by measuring the results of the top Combine performers to see whether those participants are drafted early more frequently than others. This analysis does show significantly different results than the previous averages.

In this analysis, the percentage of participants who finished in the top 10% (plus ties) of each drill and drafted in rounds 1-3 was compared to the percentage of all other participants drafted in those rounds. To facilitate the identification of significant differences I developed and used an index that is reflected in the following table. As an example of the index, let us say that 40% of top 10 performers are drafted in the first three rounds and 20% of the rest of the participants are drafted in those rounds. This results in an index of 2.00, or 40% divided by 20%.

Here are the indices for each drill and each playing position.

	10 yd Split	40 yd dash	Bench	Vertical Jump	Broad Jump	20 yd Shuttle	3-Cone Drill
QB	1.73	2.34	NA	1.69	2.02	1.46	1.66
SRB	NA	NA	NA	NA	NA	NA	NA
LRB	1.71	2.03	1.25	1.52	2.02	1.46	1.66
WR	1.62	2.09	NA	1.66	1.67	1.24	0.84
TE	1.10	1.64	1.41	0.87	1.03	1.09	0.72
C	1.38	1.38	0.38	0.40	0.40	0.69	0.62
G	2.42	2.55	1.37	0.70	1.40	2.74	1.74
T	1.49	1.55	1.19	1.47	1.54	1.81	1.79
SDE	2.30	2.56	1.50	1.12	1.49	2.65	1.43
LDE	1.24	1.80	1.39	1.08	1.44	1.83	1.51
DT	1.07	1.49	1.50	1.47	1.34	0.74	1.43
ILB	3.41	1.93	0.78	2.31	1.93	1.50	1.01
OLB	1.20	1.74	0.98	2.10	1.85	0.52	1.54
CB	1.69	1.95	NA	1.58	1.82	1.24	1.20
S	2.90	2.28	0.64	1.95	1.17	1.53	1.87

Before proceeding with further explanation, it should be noted that no small running backs were drafted during the 10-year period. That is why SRB (or small running backs as described earlier) shows “NA” for each drill.

To get away from bombarding the reader with numbers, I assigned ratings (expressed as plus or minus) that might give a clearer picture of the frequency of top 10% performers being drafted in the first three rounds. Here is the scale I used:

Index Range	Rating	Index Range	Rating
Less than .250	—	1.150-1.450	+
.250-.549	—	1.450-1.750	++
.550-.849	–	More than 1.750	+++
.850-1.150	0

I then applied the scale to the indices calculated earlier with the following results:

	10 yd Split	40 yd dash	Bench	Vertical Jump	Broad Jump	20 yd Shuttle	3-Cone Drill	Net
QB	++	+++	NA	++	+++	++	++	+14
SRB	NA	NA	NA	NA	NA	NA	NA	NA
LRB	++	+++	+	++	+++	+	++	+14
WR	++	+++	NA	++	++	+	+	+11
TE	0	++	+	0	0	0	–	+2
C	+	+	—	—	—	–	–	-6
G	+++	+++	+	–	+	+++	++	+12
T	++	++	+	++	++	+++	+++	+15
SDE	+++	+++	++	0	++	+++	+	+14
LDE	+	+++	+	0	+	+++	++	+11
DT	0	++	++	++	+	–	+	+7
ILB	+++	+++	–	+++	+++	++	0	+14
OLB	+	++	0	+++	+++	—	++	+9
CB	++	+++	NA	++	+++	+	+	+12
S	+++	+++	–	+++	+	++	+++	+14
All	+25	+36	+6	+18	+23	+17	+18	+143

The net number is the net pluses (expressed as a positive) and minuses. This measures the likelihood, by drill, of top 10% performers being selected in the first three rounds as compared to all participants selected in those rounds.

A few observations regarding the top performers and their draft position:

• Centers are the only playing position with a net negative result, indicating that Combine results are not important at the position
• Tight ends showed a very minor positive correlation between Combine performance and draft results, principally due to the 40 yard dash
• Defensive tackles and outside linebackers all showed positive results, but not much of one

• Offensive tackles had the highest positive rating, followed closely by quarterbacks, large running backs (215 pounds or more), small defensive ends (260 pounds and lower), inside linebackers and safeties
• All playing positions were positive in the “speed drills” (40 yard dash and the 10-yard split)
• The bench press has the smallest correlation across all positions
• The broad jump trailed only the speed drills in positive correlation