Recruiting Rankings vs. Actual Production, Part 1

Share on your favorite social network:
Share on FacebookTweet about this on TwitterEmail to someoneGoogle+share on TumblrShare on Reddit

“Despite their popularity among fans, rankings of college football
programs based on the perceived quality of their recruiting classes sometimes
has little relation to the success of those programs on the field in later
years.” — Ray Grasshoff, author of Beyond Friday Nights: College Football Recruiting for Players and Parents.

Before I started researching the correlation between college football
recruiting rankings and actual on-field production, I must admit that I was
firmly in Mr. Grasshoff’s camp, one that puts little stock in recruiting
rankings. However, as is usually the case, the journey provided many surprises
and lessons learned. Fortunately, there’s lots of ground to cover before any
conclusions are drawn.

The study I am presenting for your review stemmed from my observations on
many college message-boards. Independent of one specific site, knowledgeable
fans of college football everywhere cannot seem to agree on how much weight
recruiting rankings hold. A good-natured (usually) debate rages daily, to which
any regular reader of TSL’s recruiting board can attest.

Today’s article is the first in a four-part series that will examine a good
chunk of the raw data re: recruiting rankings and how they correlate to
player-performance, winning seasons, national championships and the efficiency
with which college staffs capitalize on “top” and “lesser” talent .
Since Virginia Tech is a member of the ACC, I have concentrated my research on
the current 12 members of our conference, focusing on the recruiting classes
spanning 2002-2008 (reliable rankings simply do not exist readily before 2002
and I wanted at least 3 years to examine the recent 2010 results). The
breakdown:

Part I: Sure to be the longest of the four articles, Part I will
focus on comparing recruiting rankings to player performance, specifically how
many All-ACC, All-American and NFL athletes were products of the classes inked
between 2002 and 2008. In addition, before the article tackles the real meat, I
am going to take care of some one-time “house-cleaning”, specifically
reviewing my criteria used in gathering and analyzing the data. Finally, as with
Parts I-III, I will draw a few conclusions and point out some interesting
tidbits.

Part II: I expect this article will create the most interest and
spark the most debates. Part II will focus on the correlation between recruiting
rankings and WINS (regular season, conference championships and bowl games).
Does that one 5-star WR push a program to greater heights? In addition, I have
compiled ranking statistics for the last 6 BCS champions; herein lays the
ultimate comparisons to VT and the rest of the ACC.

Part III: The most difficult article for me to research, Part III
will concentrate on how efficient the coaching staffs across the ACC are with
all 3 levels of recruits. For example, which staff makes the most of their
3-star recruits and which ones “fail” most often with 4 and 5-star athletes?
I will try to keep the data presented as objective as possible, but this piece
by nature will be qualitative and therefore somewhat subjective.

Part IV: At first I planned on making this final segment all about
conclusions, specifically factual ones, with lots of interesting observations
spanning the first three parts. However, while I still plan on drawing and
stating my personal conclusions, the main part of this final article will be
conclusions and observations written by four readers (two TSL/VT subscribers,
one TSL/UVA subscriber and one regular subscriber from www.tomahawknation.com
).

References: Simply put, Rivals is
the best national recruiting service when you include history, quantity/quality
of data and a willingness to adjust their rankings. I originally planned on
using a combination of Rivals and Scout.com but Rivals is much more reliable
overall, so all rankings are from Rivals. Other sites I used extensively: www.nfl.com,
espn.go.com, www.theacc.com,
www.sports-reference.com,
www.google.com
and Wikipedia. In an attempt to be fair and unbiased, I did not use any
information from TSL.

Criteria utilized: In order to save time on preparation (research
still took over 50 hours), I created some guidelines that provided me with
objective data while compromising statistical accuracy only minimally:

1) All-Big East teams were not included for 2002-04, due to lack of
accurate information and the relatively small (if any) sample size

2) All-Conference is only for players on the first or second team

3) A player only counts once towards All-Conference/American or
NFL, even if said player made the team multiple years

4) An All-American must be part of the CONSENSUS ALL-AMERICAN TEAM per
the following two reference links: http://www.sports-reference.com/cfb/awards/all-america-2010-2019.html
and http://www.sports-reference.com/cfb/awards/all-america-2000-2009.html

5) Kickers and punters are included in the total number of *s per class
and All-Conference/American or NFL IF they are listed by Rivals as signed
recruits

6) A recruit qualifies as “Pro” if the athlete was all-conference,
then drafted or active at least one NFL game. This rule creates a very
slight disadvantage to BC, Miami and VT. Obviously, the “Pro” number is
accurate only within the scope of this study. Miami is one
school that has placed many players into the NFL who did not meet my “qualifications”.

7) In all tables and graphs, “Athlete” (or ATH) encompasses RBs, WRs
and TEs. This was necessary due to time constraints; in addition, these
positions are often considered inter-changeable

(Part I is the only article that will include the References and Criteria
section, as well as an in-depth introduction)

So, without further delay, let’s take a look at some recruiting numbers,
see how they impact player performance (the first two data sets are in the form
of line graphs, without any discussion to follow).

Table 1:


School

Ave *s (2002-2008)

All-ACC totals

% Recruits All-ACC

All-ACC * Ave

All-Am Totals

NFL Totals

% Recruits in NFL
FSU
3.67
19 11.00
4.00
1
17
9.20
Miami 3.62 20 12.00 3.85 0 13 7.90
Clemson 3.30 23 14.00 3.74
3
13 8.00
UVA 3.26 17 11.00 3.71 1 13 8.50
Maryland 3.23 17 11.00 3.35 0 12 7.90
VT 3.22
27
16.00 3.63 1
17

10.20
NC State 3.19 14 8.50 3.14 0 8 5.00
UNC 3.19 15 9.00 3.47 0 10 6.00
BC 3.14 17
18.00
3.18 1 8 5.50
GT 3.12 17 12.50 3.24 1 7 5.20
Duke 3.02 5 3.50 3.00 0 4 2.70
Wake Forest 3.01 13 10.00 3.00 2 7 5.30

When looking at recruiting rankings and how they can most effectively be
compared, I believe there are two schools of thought. The first (average stars
per class or per period of several years) has an inherent flaw that is obvious
when looking at Table 1: certain schools have a major advantage due to
geographical location. In the ACC, FSU and Miami are located in a state that
produces tremendous amounts of D1 players, which in turn means there are many
more 4-5-star athletes available, enough to support 4 major D1 programs. South
Carolina produces the 4th-highest average amount of D1 players in the ACC
footprint (Florida, North Carolina, Georgia and South Carolina). However,
Clemson only has USC-E as serious in-state competition. Remember, ALL colleges
recruit inside-out (not just VT) … only a handful can also add annually to
their in-state haul from locations across the country. Georgia Tech is a
confusing anomaly to me, but being located in SEC country must offset the
Atlanta location.

The second school of thought is to use percentage of recruits that become
significant contributors, or in this case, All-ACC then professionals. There are
factors that reduce this method’s effectiveness, as well, but overall it
appears to be the more solid approach.

Five schools from Table 1 really stand out, four for good numbers and one for bad.
Florida State, even with its built-in advantage, is solid in almost every
category, as is Clemson. VT is middle-of-the-pack using average stars, but
easily leads everyone when using percentage of All-ACC and NFL. BC is the final positive,
landing the 4th-worst average stars per class but developing the highest percentage
All-ACC athletes. Once again, Georgia Tech stands out as a school that is
seriously underachieving … there is no excuse (in my opinion) for Wake Forest producing
a higher % NFL total.

UVA is the one school that is straight solid across the board, but not
spectacular enough in any area to crack my top 4 (top 5, yes).

In the seven recruiting years studied, the ACC only produced 10 consensus
All-Americans, a measly 8% of total All-Americans during that period. Clemson
alone accounted for 2%, which is an impressive number when one considers there
are 120 D1 colleges.

Finally, I was surprised to see Miami send only 7.9% of their recruits to the
NFL; they probably have more undrafted free-agents or pros that didn’t make
All-ACC by a wide margin. However, all schools are playing by my rules for this
study, so everyone has the potential to lose % points.

 

Table 2:


School

4/5* by OL

4/5* by QB

4/5* by Athlete

4/5* by DL

4/5* by LB

4/5* by DB

All-ACC >=4*

% Recruits >= 4*
BC
8
0 5 2 1 3 3 13.10
Clemson 7 2
14
7 5
11

11

28.22
Duke 0 0 2 1 0 0 0 2.04
FSU
11

4

27

23

21

16

11

58.29
GT 2 2 7 2 0 2 4 11.11
Maryland 6 2
14
5 1 4 6 21.05
Miami
11

5

27

17

11

18

11

53.94
UNC 2 1 12 8 4 3 5 17.75
NC State 5 2 11 5 2 3 2 17.39
UVA 5 3 8
10
4 6 6
23.53
VT 5
4
9 7 5 3
11
19.76
Wake Forest 0 0 0 0 1 0 0 0.75

Before we take a closer look at Table 2, a reminder that “Athlete”
encompasses RBs, WRs and TEs, at least in the system I used.

The objective with this table was to present an idea of how 4/5-star recruits
in the ACC break down by position while determining the percentage of 4/5-star recruits
per school from 2002-2008. If one looks only at the total number of
All-ACC per school 4/5-star recruits (which Part I is supposed to do), the usual four
suspects (Clemson, FSU, Miami and VT) are tied at the top, almost doubling their
next closest competitor. However, the overwhelming numbers that jump off the
page are 54% and 58% respectively for Miami and FSU. REMEMBER those numbers when
reading Part II, where wins/losses will correlate to rankings.

After reviewing those astounding numbers again, I honestly believe that a
chicken-and-the-egg phenomenon is in effect re: recruiting rankings in Florida.
The belief that excellent past high school classes “guarantee” future
Florida recruits gaining additional stars must be a real and uncontrollable
cycle.

UVA comes in at a solid 4th spot, based largely on their excellent job
recruiting defensive linemen. I must admit that between 2002 and 2005, UVA
landed some very strong classes, by stars at least.

BC landing eight 4/5-star offensive linemen is not a surprise, since that is considered their
specialty area. However, VT garnering only three 4/5-star DBs was shocking, as was
the four top QBs they signed in only seven years.

Clemson is the most impressive to this point, with FSU/VT close behind then
Miami/UVA about even. But, I am not drawing conclusions in this article, right?
I’m just presenting the data.


Table 3
:


School

Ave *s (2002-2008)

Recruits 3-Stars or Less Per Class

Total All-ACC 3-Stars or Less

Total NFL 3-Stars or Less

5* Recruits Per Class

Total All-ACC = 5*

Total NFL = 5*
Clemson 3.30 16.71 11.00 4.00 0.43 3.00 2.00
FSU
3.67

10.43
12.00 2.00
2.14

4.00

4.00
Miami 3.62 10.71 14.00 5.00 1.71 3.00 3.00
UNC 3.19 19.86 11.00
8.00

0.29
1.00 1.00
UVA 3.26 16.71 11.00 6.00 0.57 3.00 2.00
VT 3.22
19.14

16.00
6.00 0.43 3.00 3.00

Now we take a look at the six schools most VT and ACC fans believe are the
top recruiters, or best at taking solid players and making them great. Another
reminder; using NFL numbers in this study is effective because any recruit that
is considered a professional must first have made All-ACC.

In the opinion of many college football fans and experts, 3-star or less recruits are the true key to building and maintaining a consistent, winning
program. At the same time, 5-star recruits usually turn out special and are the
difference-makers in winning a conference or the BCS title. Think of the 3-stars
as a Warthog and the 5-star a Stealth fighter.

Obviously, with all of the 4/5-star recruits they sign, FSU and Miami
average, by far, the lowest number of 3-stars or less per class. FSU gets a C for
their 3-star All-ACC totals and an F for 3-star NFL recruits, but Miami seems
pretty solid in getting the most from their “lesser” talents. VT is known
for developing underrated players, from walk-ons to 3-stars and they have earned
the highest 3-star or less All-ACC count. However, it does appear that 3-stars and
lower have a much harder time making the leap from an excellent college player
to one who plays on Sundays (less than 50%). Only UNC and UVA have had success pushing
their 3-stars into a football career after college (by %). Clemson finally fails
at producing, albeit with “lesser” talent.

The most positive % in this chart is 88, the percentage All-ACC 5-stars that
play on Sundays. I truly believe that this statistic best illustrates how
successful most ACC programs are with their top-ranked recruits. It should be
noted that VT is the only ACC team in the 2002-2008 period to have all of their
5-star recruits make both All-ACC and the NFL.

One piece of my research I would have augmented if I had the time was
breaking down the 4-stars to 3s, 2s and walk-ons. Sounds like an excellent
dataset for determining staff efficiency.

Conclusions and comments:

After reviewing my research and finally writing Part I of this series, I have
reached a few conclusions, specific to this first section:

1) Virginia Tech was the best program in the ACC between 2002 and 2008 at
making players from all ranking levels into All-ACC and NFL athletes. A
close second was FSU, followed by Clemson, Miami and UVA. Duke and Wake
Forest are just too handcuffed by an almost complete lack of 4/5-star
recruits to consistently produce All-ACC and NFL young men.

2) Five-star recruits have a higher success rate than I believed going
into this project, while there is just too fine a line between 3s and 4s to
definitively state that a 4-star will usually become the more productive
player.

3) I am convinced that recruiting services (including Rivals, which I
respect) are grading Florida recruits too high, especially if they attend
one of the big 3 Florida schools or an SEC powerhouse. There is obviously
more pure talent in Florida than, say, Virginia, but at some point the cycle
needs to be broken and a future class of Florida recruits be evaluated as if
they resided in Alaska

4) So far, my position on recruiting rankings as they relate to actual
production (in this case player development) has only slightly opened.
However, since I already know the data I will be using for Part II, I am
pretty sure that’s about to change

5) UVA and Maryland both pleasantly surprised me, especially during the
first 3-4 years of my 7-year recruiting period. However, UVA definitely
should have won more games with the amount of All-ACC and future pros on
their rosters.

Part II should be out in a few days and I firmly believe the comparison
between ACC teams and the programs that won the last 6 BCS title games will
really make some people think twice about the importance of landing and PLAYING
the very-highly ranked recruits. However, don’t worry … the focus will
remain on the 12 ACC schools, overall wins, conference championships and bowl
games.

Please feel free to point out any mistakes that I missed or make suggestions
for inclusion in the next three parts. If you’re interested, the spreadsheet
containing all the data used in Part I can be accessed via the link at the
bottom of this page.

Share on your favorite social network:
Share on FacebookTweet about this on TwitterEmail to someoneGoogle+share on TumblrShare on Reddit