Through several months spent exploring the role of numerous factors in college rowing performance, the most telling and significant variable examined by the Sparks team continues to be international recruiting, especially among IRA heavyweight programs.
International athletes generally entail a greater recruiting cost that their domestic teammates. We define recruiting cost as the time, effort, and financial expenditure required to recruit an athlete. The majority of varsity college rowers of international origin are recruited athletes, as opposed to walk-ons. Walk-on athletes carry a relatively low recruiting cost; they can generally be recruited through on-campus advertising, boathouse tours, and learn-to-row clinics. Recruited athletes, on the other hand, require several years’ worth of organizational and logistical effort. Coaches must scout, contact, and coordinate with potential recruits to organize official visits, review grades and test scores, and provide advice through the application process. This process is cumbersome enough with domestic athletes. Recruiting internationally makes for a more logistically intensive process.
Importantly, coaches have a limited number of official recruiting spots to offer athletes, which adds an economic wrinkle to the recruiting process. In economic terms, the opportunity cost of each recruiting spot is the next best athlete who could have been chosen to fill that spot. Due to the scarcity of recruiting spots and the large surplus of potential recruits, coaches must exercise due diligence throughout the recruiting process, taking care to allocate their recruiting spots in the best manner possible.
If coaches must make an allocation decision with their recruiting spots, and if it’s been demonstrated that international athletes contribute to improved team performance, then we can deduce that coaches should pursue international athletes, subject, to certain financial and logistical constraints. How should a coach go about choosing prospective international recruits?
Over the next two articles, our team will be exploring the differences between countries in terms of two outputs: retention and performance. Specifically, we will attempt to determine which countries produce athletes that are more likely to be retained year over year, and we will attempt to determine which countries produce athletes that contribute most to team performance. By examining the intersection between these two outputs, we hope to paint a more detailed and nuanced picture of international recruiting than what has been presented in the past. Through this process, we will include American athletes as a benchmark.
The Breadth
In 2016 and 2017, both the IRA National Championship and the NCAA Division I National Championship were contested by athletes from an impressive array of countries. In 2017, 33 nations were represented at the IRA National Championship, with 26 nations represented at the NCAA Division I National Championship. In 2016, 31 and 25 nations were represented, respectively.
For the purposes of our analysis, we placed a few restrictions on the set of schools we examine. Only schools that entered a varsity eight in their respective championship in 2015, 2016, and 2017 were included. As well, only schools with available roster data for these three years were included. To mitigate outliers, we only included countries that were represented by at least ten athletes across 2016 and 2017.
To calculate retention rates, the Sparks research team added up the total number of freshman, sophomore, and junior athletes from each country in 2016, and compared these totals to the total number of sophomore, junior, and senior athletes from each country in 2017. There were several situations in which retention rates exceeded 100%. This was largely driven by the presence of transfer athletes and walk-ons who began rowing after their freshman year.
The Numbers
Looking at both the IRA and NCAA data, we can see that the United States ranks among the bottom of qualified countries in terms of retention. This could be explained by several factors. First, it stands to reason that walk-on athletes would be disproportionately American. If we accept that to be the case, then there is a possibility that walk-on athletes would be more likely to stop rowing than their recruited teammates. Second, American athletes account for the vast majority of collegiate rowing athletes at American universities. This large number (1546 in the IRA and 1242 in the NCAA) means that outliers have relatively small influence.
Retention Rates for Qualified* Teams, IRA Heavyweight Men, 2016-2017
*Criteria for inclusion: Teams that appeared at NCAA Regatta in 2015, 2016, and 2017 with complete roster data
What does this mean in context? For example, let’s take a hypothetical team with 10 athletes from a certain country. Say there is a major organizational change, such as a coaching shakeup, and five of these athletes quit the team. Five out of 1546 or 1242 American athletes is just a drop in the bucket. But if these athletes are from a country like Germany, which only had 22 IRA and 13 NCAA athletes across two years, retention numbers will be strongly affected.*Criteria for inclusion: Teams that appeared at IRA Regatta in 2015, 2016, and 2017 with complete roster data.
Retention Rates for Qualified* Teams, NCAA Division I Women, 2016-2017
*Criteria for inclusion: Teams that appeared at NCAA Regatta in 2015, 2016, and 2017 with complete roster data
There are some other interesting data points here. Clearly, there was a shock in the Canadian athlete IRA retention rate in 2017. Digging into the numbers, it appears several Canadian athletes simply stopped rowing for their universities. However, this shock is largely influenced by the relatively low number of Canadian athletes in the IRA.
We can also see the difference in international recruiting between the IRA and NCAA. In a previous article, we explained that international athletes, on the aggregate, have a more significant effect on performance in the IRA than the NCAA. From the data here, we can see that IRA teams recruit from a wider range of countries, and in higher numbers. Outside of the wider breadth of countries recruited by the IRA, the names are roughly similar on each side, with Great Britain, Australia, and Canada featuring prominently.
Conclusions
While the jury is out on the role of outliers, we can generally conclude that international athletes are more likely to be retained than American athletes. If a coach is making an allocation decision based on estimated retention, then he or she is best served to recruit internationally, from countries like Germany, New Zealand, and Great Britain.
The calculation, however, is not this simple. When making the decision to expand international recruiting, coaches must weigh the increased expected future retention against the logistical and financial cost of recruiting internationally.
To more accurately make this determination, coaches need to know which countries are more likely to produce better-performing athletes. If a certain country produces highly retained and highly performing athletes, then they are probably worth the additional recruiting cost. However, if another country produces highly retained but relatively low performing athletes, they are probably not worth the additional recruiting cost.
Next time, we will attempt to tackle this question. Through a statistical analysis, we will seek to determine which countries produce relatively high and relatively low-performing athletes. Then, we will compare this against retention to produce a more definitive ranking of international recruiting locations.
