She got game: Investigating how reputation can be leveraged to improve recruiting effectiveness in National Collegiate Athletic Association women’s basketball

Abstract

Although recruiting processes and outcomes in National Collegiate Athletic Association sports is an incredibly important facet of collegiate athletics, it is underdeveloped in several areas. Gaps in knowledge exist when it comes to better understanding actual recruits obtained, the role of reputation, and what factors may influence the school-choice decision of elite, female student-athletes. Probit analyses examining data from 500 National Collegiate Athletic Association Division I female basketball players, approximately 100 universities, and 20 National Collegiate Athletic Association conferences yield that recruits’ decisions are primarily influenced by the total number of Elite 8 teams and national championships from a team’s affiliated conference, geographic distance between recruits’ hometowns and the university, average arena attendance, and the percentage comparing the basketball arena capacity and game attendance. The results make both theoretical and practical contributions by demonstrating the predictive power of reputation, while also offering recruiters actionable information that potential recruits likely are considering each recruiting cycle.

Keywords

Gender intercollegiate athletics student-athletes

Introduction

Recruitment effectiveness in the context of National Collegiate Athletic Association (NCAA) sports represents the ability of individuals (coaches) to successfully attract and sign desirable recruits (student-athletes).¹ Understanding how to improve the effectiveness of athletic department recruitment efforts is a salient issue for sport scientists to explore because few endeavors will contribute more to the success of intercollegiate athletic departments than their capability to recruit and retain highly talented individuals. Numerous requests^2,3 have been made to better comprehend recruiting, but insufficient consideration has been given to several key areas. For instance, information tends to be at the core of recruits’ levels of attraction to an organization.^2,4 Yet, how reputational features (e.g. head coach win percentage and school’s academic ranking)—key information elements—influence recruits’ school-selection decisions have largely been overlooked. Limited research has been conducted on actual recruits obtained. Further, sport research on successfully recruited student-athletes has, with limited exception,⁵ been focused on men’s sports instead of women’s sports, which restricts opportunities to compare recruiting research across men’s and women’s sport contexts.

Overall, researchers have dedicated few resources towards examining the factors influencing the decisions of actual recruits obtained, how reputational features influence schools’ recruiting effectiveness, and the recruitment of NCAA Division I female athletes generally. In particular, although the NCAA reports that women’s basketball is the most sponsored women’s sport, and that over 300 schools in the US have a Division I women’s basketball team, inadequate attention has been given to the study of recruiting in this context. The purpose of the present investigation is to test an empirical reputational resource leveraging model of the factors that Entertainment and Sports Programming Networks (ESPN) Top 100 female NCAA Division I basketball recruits (From 2010 to 2014) may have considered when selecting a university. We accomplish this through the use of an extensive database of predictors that combines school-specific attributes with recruit-specific information to provide novel insight into how reputational factors affect schools’ ability to recruit female basketball players. Discussed next is how reputational features may influence sport recruiting outcomes.

Reputation and recruiting

Reputation constitutes⁶ the “ … perceptual identity formed from the collective perceptions of others, which is reflective of the complex combination of salient entity characteristics and accomplishments, demonstrated behavior, and intended images presented over some period of time as observed directly and/or reported from secondary sources, which reduces ambiguity about expected future behavior” (p.165). Operationally, although a variety of situationally determined reputation dimensions may exist, the reputation construct appears to have at its core two higher-order factors: performance/results and character/integrity.^6,7 The former dimension, performance/results, pertains to the collective record of past performance for an individual, team, or organization. The latter dimension, character/integrity, pertains to the variety of past behaviors that, collectively, convey of a sense of consistent, ethical behavior, and concern for other entities.⁶

In the context of sport, a support dimension of reputation also has been proposed.^1,4 This category captures support elements such as geographic proximity and fan attendance, which may be reassuring and attractive to recruits. Geographic proximity, for example, may be comforting to recruits because they know family and friends are not too far away.^1,4,8 Crowds, which may signal fan support of a sport team at a school, could sway the selection decision of recruits as well.^1,9,10

Reputation is a multifaceted construct, and the position taken in this study is that reputational features present within the aforesaid dimensions of reputation may influence recruiting via signals. According to Spence,¹¹ who formulated signaling theory in the context of economics, circumstances exist where there is incomplete information in the labor marketplace, and calculated efforts are made to persuade others via signals sent in this marketplace. Said differently, signaling stimulates attentional concentration by increasing the relevance and importance of stimuli, or reputational signals, in a particular environment, which then differentiates reputation in observers’ eyes. Accordingly, in this study, several different types of reputation information that is present in the sport marketplace is identified and then explored to see how various reputational features may influence recruiting outcomes for NCAA Division I women’s basketball programs.

Table 1.

Definitions of the three categories of reputation variables.

Category	Variable name	Definition	Source
Performance-based reputation	Drafting performance	Total number of players drafted into the WNBA from the school chosen in past five years	WNBA website
	Conference reputation	Total number of Elite 8 teams from a conference where the school chosen is affiliated in past five years	NCAA.com
	Conference national championship reputation	Total number of national championships from a conference where the school chosen is affiliate in past 15 years	NCAA.com
	Team Top 10 reputation	Total number of Top 10 finishes in past five years	AP final Top 25 Polls
	Head coach winning percentage	Head coach’s winning percentage based on the year the recruit signed	University websites
Support-based reputation	Geographic proximity	Distance from recruits’ hometowns to the university in miles	ESPN.com & Google maps
	Home game attendance	Total number of appearance on the Top 10 attendance rankings in past five years	NCAA.com
	Arena attendance per 2013	Average arena attendance per 2013^a
	Game atmosphere	Percentage comparing the capacity of the basketball arena and the average game attendance per 2013	University websites and NCAA.com
	Facility age	Basketball facility age in years	University websites
Character- and integrity-based reputation	Academic reputation	Is the university in the U.S. News and World Report Top 100	U.S. News & World Report
	NCAA sanctions	Is the team currently experiencing, or will face, NCAA sanctions	NCAA Major Infractions Database
	University type	Is the university a state or private school	U.S. News & World Report

AP: Associated Press; NCAA: National Collegiate Athletic Association; WNBA: Women’s National Basketball Association.

Average attendance was measured per 2013 because the NCAA only lists attendance numbers for women’s basketball teams back to 2011. Prior to 2011, the NCAA only lists the Top 50 in attendance. Hence, a standardized metric was used for this variable.

Methodology

The present study identifies and examines several potentially important pieces of reputation information that is present in the sport marketplace. To empirically examine reputation factors possibly contributing to recruiting effectiveness, a statistical model was adapted from a recruiting formula created by Dumond et al.,⁸ which is shown in equation (1)

P_{jz} = Φ (X_{j} β + Y_{z} δ + Z_{jz} θ)

(1)

where P_jz is the probability that recruit j selects school z and Φ is the normal cumulative distribution function. The X_j includes performance-based reputation (e.g. overall team performance, postseason performance, head coach winning %, etc.), and vector Yz consists of factors that are related to support-based reputation (e.g. geographic proximity, home game attendance, and age of the basketball arena). The vector Z_jz contains character and integrity-based reputation such as NCAA sanctions (i.e. if a team is currently experiencing or will experience NCAA sanctions in some form while a recruit plays for the team).

Data and participants

Archival data for this study were obtained through the use of numerous online and in-print resources, including but not limited to the Associated Press (AP), USA Today, the NCAA, the Women’s National Basketball Association (WNBA), ESPN.com, Google, the United States Office of Postsecondary Education, U.S. News & World Report, and the NCAA Major Infractions Database. The data for the team performance variable of total number of Top 10 finishes (over a five-year period of time), for example, were compiled from the AP final Top 25 Polls for the years 2010 through 2014. Table 1 lists and defines the three categories of reputation variables.

Recruits are only able to select a single school to attend. Thus, a dummy variable was created to indicate recruits’ decisions whether to attend a university. Additionally, a reference category was created which included universities, the recruits strongly considered attending but ultimately did not select. Reciprocal interest between a recruit and school was determined by including up to three schools, the recruit visited officially or unofficially and/or they noted to be a finalist for her school decision. This information was collected using ESPN.com, Google, Bing, Collegiate Girls Basketball Report, and LexisNexis® Academic.

Data were collected on the ESPN HoopGurlz Top 100 female high school athletes recruited to NCAA Division I women’s basketball for each year from 2010 to 2014 (n = 500). ESPN is one of the most successful sport networks in the world, and the ESPN HoopGurlz Top 100 is a focused effort by the media entity to provide a comprehensive ranking of the top high school women’s basketball talent. The annual list is compiled in part by Dan Olson, an experienced coach and recruiter who operates The Collegiate Girls Basketball Report (www.girlsbasketballreport.com).

Each basketball player in this study selected on average from three school options, resulting in a total of 1344 observations throughout the five-year window (2010–2014). Across the five-year span, approximately 100 universities across 20 different NCAA conferences attempted to recruit at least one of the top recruits. Of the Top 100 recruits, 77% were a four-star recruiting rank (M = 4.23, SD = 0.42), and more than 50 recruits lived within 400 miles from the institutions they visited (M = 594.48, SD = 651.51).

Measures

Three reputation-based categories were used to examine factors affecting female basketball players’ decision to select a university. The reputation-based variables were created by ensuring all the variable coding was such that greater values should be associated with increased reputation. An exception to this tenet of analysis was geographic proximity, which represents a support-based reputation variable. Both Dumond et al.⁸ and Magnusen et al.¹ reported that as the distance between recruits’ hometowns and a target university increases, recruits’ chances of selecting that university decrease. Thus, the inverse of geographic proximity was created during this phase.

Additionally, a five-season time frame was selected for most of the reputation measures for several reasons. First, although complete data were available for a five-year span, complete data were not available for a 10-year span. Second, Millennial student-athletes may only care about school and athletic department information within a five-year (i.e. season) window.⁹ As an exception to this point, the number of national championships won over the previous 15 seasons was used in this study. We chose to use a longer duration for this variable because only one team from a single conference wins a national championship per year. Methodologically, it was important not to limit national championships to five seasons as the data available would be insufficient to provide interpretable results.

The NCAA places boundaries on the number of universities an athlete can visit, and as such, athletes tend to visit only those universities in which they are most attracted. When an official visit could not be found, online and in-print news articles were used to identify unofficial visits and the schools of most interest to athletes (i.e. the finalists). Athletes only can sign with one university, and the final variable was the university selected by the athletes.

Model specification

The estimation used to model a given recruit’s university selection was a probit analysis assessed by maximum likelihood techniques. The specification of the research model shown in equation (2) assumes that all variables not included in the proposed model are constant from one potential selection to another. The quality of the official visit from one school to another for each recruit is considered the same. In line with this idea, the amount and quality of illegal recruiting tactics (unknown to the NCAA) are assumed to be equal from one school to another

\begin{matrix} \Pr (Y = 1) = 1 - \Pr [u < - (β_{0} + β_{1} Draft \\ + β_{2} Conference Reputation \\ + β_{3} Conference National Champ \\ + β_{4} Team Top 10 + β_{5} Coach Winning \\ + β_{6} Geo Proximity + β_{7} Home Attendance \\ + β_{8} Attendance per 2013 + β_{9} Atmosphere \\ + β_{10} Facility Age + β_{11} Academic Ranking \\ + β_{12} NCAA Sanctions + β_{13} University Type)] \end{matrix}

(2)

with Pr(Y = 1) denoting our dependent variable (i.e. probability of school selection). For the reputation dimensions, the empirical specifications include the number of players drafted into the WNBA (Draft), number of Elite 8 teams from the affiliated conference (Conference Reputation), number of national championships from the affiliated conference (Conference National Champ), number of Top 10 finishes (Team Top 10), head coach’s winning percentage (Coach Winning), distance between the recruit’s hometown and the school (Geo Proximity), number of Top 10 attendance ranking (Home Attendance), fan attendance in 2013 (Attendance per 2013), comparison of the arena capacity and average fan attendance per 2013 (Atmosphere), age of the basketball complex (Facility Age), Top 100 listed in the U.S. News and World Report (Academic Ranking), experiencing NCAA sanctions in some form while a recruit would play for the team (NCAA Sanctions), and whether a school is a state-run or a private school (University Type).

Results

A completely random selection method would correctly predict, on average, the college choices for nearly 40% of the recruits (see Table 2). Included in Table 3 is the overall predicted probability for school selection, along with the coefficients from the probit model and the marginal effects of each variable.

Table 2.

NCAA Division I women’s basketball recruiting descriptive statistics.

	Recruiting year
	2010	2011	2012	2013	2014
Sample size
Total number of records	271	269	266	243	295
Total number of recruits	100	100	100	99	100
Average choices/recruit	2.71	2.69	2.66	2.45	2.95
Performance-based reputation for the schools chosen
Average number of players drafted to the WNBA from the school chosen^a	3.08	2.84	2.93	2.46	2.44
Average number of Elite 8 teams from the conference^b	6.25	5.99	6.71	5.99	5.59
Average number of national championships from the conference^c	2.70	2.66	2.80	1.82	1.25
Average number of Top 10 finishes^d	1.37	1.22	1.21	1.18	1.23
Head coach winning percentage^e	65%	67%	65%	67%	66%
Support-based reputation for the schools chosen
Geographic proximity (distance between the school and the recruit’s hometown)^f	530	443	487	544	570
Average number of appearance on the Top 10 attendance ranking in past five years	1.23	0.74	0.84	0.74	0.98
Average arena attendance per 2013	3867	3676	3819	3976	4006
Game atmosphere^g (%)	33	34	34	37	33
Facility age	31	35	36	40	34
Character and integrity-based reputation for the schools chosen
Academic reputation (U.S. News and World Report)^h	60%	65%	74%	70%	68%
Percentage of choices with NCAA sanctions	5%	2%	0%	4%	1%
University type (percentage for the state university)	70%	70%	67%	63%	76%
Conference breakdown (n) for all records
American athletic conference	0	0	0	0	34
ACC	63	45	52	60	75
Atlantic 10	10	3	0	5	0
Big 10	17	28	12	22	56
Big 12	31	36	32	25	26
Big East	55	52	51	33	5
Conference USA	2	3	2	4	1
Ivy League	0	2	1	2	0
Mountain West Conference	6	2	0	0	1
Pac 12	38	40	39	36	40
SEC	47	47	71	54	53
West Coast Conference	1	2	1	0	0
Otherⁱ	1	9	5	2	4
Total	271	269	266	243	295

ACC: Atlantic Coast Conference; NCAA: National Collegiate Athletic Association; WNBA: Women’s National Basketball Association; SEC: Southeastern Conference.

In the five-year period prior to a recruit signing with a school.

For the five-year span prior to a recruit signing with a school.

Over the 15-year span prior to a recruit signing with a school.

In the five-year period prior to a recruit signing with a school.

Based on the year the recruit signed.

Miles.

Comparison of the arena capacity and the home game attendance.

The data are interpreted as percentage of schools chosen, which appeared on the Top 100.

Includes Colonial Athletic Association, Mid-American Conference, Missouri Valley Conference, Patriot League, Summit League, Sun Belt Conference, and Southwestern Athletic Conference.

Table 3.

Probit model results^a.

Variable		Coefficient	z	dy/dx	z
Performance-based reputation	Drafting performance	0.010 (0.025)	0.40	0.004 (0.009)	0.45
	Conference reputation	−0.034^b (0.015)	−2.24	−0.014^b (0.006)	−2.30
	Conference national championship reputation	0.037^b (0.019)	1.96	0.014^b (0.007)	1.98
	Team Top 10 reputation	0.030 (0.038)	0.79	0.013 (0.015)	0.89
	Head coach winning percentage	0.509 (0.341)	1.49	0.140 (0.162)	0.86
Support-based reputation	Geographic proximity	−0.001^c (0.001)	−3.31	−0.001^c (0.001)	−3.33
	Home game attendance (# of Top 10 attendance rankings)	−0.138 (0.143)	−0.96	−0.048 (0.053)	−0.88
	Arena attendance per 2013	−0.001^c (0.001)	−2.77	−0.001^c (0.001)	−2.81
	Game atmosphere	0.006^b (0.003)	1.93	0.002^b (0.001)	1.96
	Facility age	−0.003^d (0.002)	−1.65	−0.001^d (0.001)	−1.71
Character- and integrity-based reputation	Academic reputation	0.159^d (0.088)	1.81	0.056^d (0.035)	1.63
	NCAA sanctions	−0.263 (0.196)	−1.34	−0.138 (0.104)	−1.35
	University type	0.008 (0.094)	0.08	0.001 (0.036)	0.01
Log likelihood	−847.745
Likelihood ratio chi-square test	39.38
Prob > chi-square	0.0002
Pseudo R²	0.0277
Predicted probability for school selection	0.3683

NCAA: National Collegiate Athletic Association.

Note: Standard errors in parentheses.

Dependent variable: Probability of a recruit’s school selection.

b,c,d

Significance at 10, 5, and 1% levels, respectively. The dy/dx indicates the marginal effects for each variable.

The proposed recruiting model includes three different reputation categories of variable that may influence the school-choice decisions of NCAA Division 1 women’s basketball recruits. In terms of the performance reputation category, the results of the study show that the coefficient on drafting performance is positive, but not statistically significant. As expected, greater on-court performance leads to a greater probability that a recruit will select the university. Winning the national championship and team’s Top 10 appearance increases the predicted probability of school selection. Specifically, a conference national championship from the prior season increases the probability of school selection by approximately 1.5%. The team’s Top 10 finishes in NCAA Division I enhance the probability an athlete will select the school by 1.3%. A head coach’s winning percentage also had a positive impact on school selection; namely, when a coach’s winning percentage increases 1% on the year the recruit signs, recruits are 14% more likely to select that school.

Next, in terms of the character- and integrity-based reputation category, the coefficient for the academic reputation was positive and marginally significant. If a school is one of the Top 100 ranked universities, a recruit will be 5.6% more likely to select the school. University type, whether a school is a state or private university, did not have any impact on school selection. Additionally, NCAA sanctions had a non-significant negative impact on the school selection.

Regarding the support-based reputation, geographic proximity, the distance between the recruit and the university, had a significant positive impact on school selection. The further a school is from the athlete’s hometown, the less likely she is to pick that school. A team’s Top 10 attendance ranking in prior seasons does not statistically affect the likelihood of a recruit selecting that school. However, recruits are more likely to select universities with higher game atmosphere levels—the comparison of the basketball arena capacity and game attendance in 2013. Although the absolute number of attendance is not correlated with the probability a recruit will select the school, high school players appear to be attracted to the schools in which the basketball arena is at or near full capacity. Recruits also are more likely to select schools with a newer basketball facility than select a school with a longstanding basketball facility.

Discussion

The reputation of organizations is a complex variable, and different dimensions of reputation may play very different roles in attracting recruits.^1–3 In the case of NCAA Division I men’s basketball, for example, head coach performance, on-court success, stadium size, NCAA sanctions, Top 10 rankings, final four appearances, academic rankings, geographic proximity, and institutional categorization (private or public) are variables that appear to influence recruits’ selection decisions.^1,12 Such results are largely analogous to the findings of this study, which focused on NCAA Division I women’s basketball recruits. Coach record, historical conference success (i.e. winning national championships), Top 10 rankings, academic reputation, geographic proximity, game atmosphere levels, and facility age were all shown to influence the selection decisions of women’s basketball recruits. These results are also similar to studies on NCAA Division I football recruits.³ Geographic proximity, on-field success, conference affiliation, rankings, stadium capacity, and media exposure have been reported as factors that may be of great importance to football recruits as they make their school-choice decisions.^8,13

The results of the current study bring attention to the apparent shared aims and values of elite recruits, regardless of gender or sport affiliation. Many of the top recruits in both football and basketball are attracted to successful programs that can position them for success on the court (or field) as well as success upon graduation. These recruits also want support, be it in the form of updated facilities, supportive fan bases, and/or being relatively close to family and friends. Collectively, although the ways in which athletes are recruited in different sports are expected to be quite varied,^2,4,5 there does appear to be some level of commonality in the information that is important to the selection decisions of recruits for football, men’s basketball, and women’s basketball.

In the future, researchers should consider the inclusion of mixed-method approaches to recruiting research. Geographic proximity is important to recruits,^1,8 but such information, in and of itself, does not make clear the reason or reasons geographic proximity matters to recruits or, in some cases, does not matter to recruits. Consider how Taryn Griffey, the daughter of baseball legend, Ken Griffey Jr. and a four-star basketball recruit in the class of 2014, decided on which program to select. She committed to the University of Arizona, a school whose last NCAA tournament appearance was in 2005, despite living over 2000 miles away in the city of Orlando, Florida. She noted¹⁴ several reasons for her decision. Her brother was, at the time, playing football for Arizona. She also wanted to create her own identity, commenting: “One thing I really wanted ever since I was younger was make a name for myself and for a program, like how Skylar Diggins made her name at Notre Dame and Brittany Griner made her name at Baylor” (para. 6). Hence, following up quantitative data about recruiting with qualitative data could prove especially helpful to individuals seeking a more nuanced understanding of how and why certain factors are important to recruits.

Practically, hundreds of potential variables may influence recruits’ school-choice decisions. Researchers cannot adequately measure all possible variables when studying NCAA recruiting. Nonetheless, our results underscore the value of quantitative data to understanding recruiting processes. Thus, in addition to data provided herein, coaching staffs are recommended to create simple, customized surveys so that they can analyze recruit data using descriptive statistics. Gathering data, such as the extent to which former and current players were influenced by various predictors (e.g. academic reputation and facility age), will, in time, generate data patterns. These patterns should help coaches become more informed and be better positioned to develop effective short- and long-term recruiting strategies.

Footnotes

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

References

Magnusen

McAllister

Kim

et al.

The reputation playbook: exploring how reputation can be leveraged to improve recruiting effectiveness in NCAA men’s basketball. J App Sport Manage 2017; 9: 11–24.

Magnusen

Mondello

Kim

et al.

Political skill, influence strategy, and organization reputation in recruitment effectiveness in college sports. Thunderbird Inter Bus Rev 2011; 53: 687–700.

Treadway

Adams

Hanes

et al.

The roles of recruiter political skill and performance resource leveraging in NCAA football recruitment effectiveness. J Manage 2014; 40: 1607–1626.

Magnusen

Kim

Perrewé

et al.

A critical review and synthesis of student-athlete college choice factors: recruiting effectiveness in NCAA sports. Int J Sports Sci Coach 2014; 9: 1265–1286.

Magnusen

Kim

Perrewé

. Gaining a competitive edge when recruiting student-athletes: the role of political skill. Int J Sports Sci Coach 2014; 9: 1290–1310.

Zinko

Ferris

Blass

et al.

Toward a theory of reputation in organizations. In: Martocchio

(ed). Research in personnel and human resources management 2007; Vol. 26, Oxford: JAI Press/Elsevier Science Ltd, pp. 169–209.

Ferris

Harris

Russell

et al.

The role of reputation in the organizational sciences: a multi-level review, construct assessment, and research directions. In: Buckley

Wheeler

Halbesleben

JRB

(eds). Research in personnel and human resources management 2014; Vol. 32, Bingley: Emerald Group Publishing Ltd, pp. 241–303.

DuMond

Lynch

Platania

. An economic model of the college football recruiting process. J Sports Econ 2008; 9: 67–87.

Jessop A. FSU utilizes new football facilities to attract top recruits. Forbes, http://forbes.com/sites/aliciajessop/2014/09/20/fsu-utilizes-new-football-facilities-to-attract-top-recruits/ (2014, accessed 2 February 2017).

10.

Wood R. Recruit reaction: crowd impresses as Florida tops Kentucky. Rivals, https://florida.rivals.com/news/recruit-reaction-florida-stops-kentucky (2017, accessed 2 February 2017).

11.

Spence

. Job market signaling. Quart J Econ 1973; 87: 845–856.

12.

Evans

Pitts

. The determinants of NCAA basketball recruiting outcomes. App Econ Letters 2017; 25: 351–354.

13.

Mirabile

Witte

. A discrete-choice model of a college football recruit’s program selection decision. J Sports Econ 2015; 18: 211–238.

14.

Stieber Z. Ken Griffey Jr. daughter will play for University of Arizona. Epoch Times, http://theepochtimes.com/n3/61899-ken-griffey-jr-daughter-will-play-for-university-of-arizona/ (2013, accessed 2 February 2017).