Friday, October 5, 2012

Why I Hated the Debate Scoreboard


There has been plenty written about the debate in terms of the relative performaces of President Obama and Governer Romney. There has been far less written about the green and yellow lines beneath them that responded as they talked. According to CNN, these lines were instant polls - they were created by a studio audence of undecided voters watching the debate.


Those two lines bothered me more than anything that was said by either candidate. I found myself constantly having to turn my eyes away from the instant polls to try and figure out what the candidates were actually saying. There were a couple of times during the debate where I actually realized that I had been watching the lines and completely missed a sentence or phrase that had just been stated. So from a personal level, the lines bothered me as a viewer.

But as an American citizen, the two lines bothered me even more. I think that they represent everything that is wrong with our political system - an inability to think broadly about the problems facing the country, a focus on soundbites and gaffes over specific policy positions, and a short-term view that is always cast in terms of winners and losers rather than any hope for progress.

Why do we need to know what other people think about what each candidate is saying in order to decide what we think? Why do we need second-by-second updates from a room of people whose choices are no more valid than ours? And how do we know that this group of randomly selected people is actually representative of anything whatsoever. It is a truth proven in academic literature that an experiment that relies on a smalle sample of people can actually sometimes be worse than no experiment at all. Simply put, we are bad at understanding the differences in certainty between a small sample size and a large one. Without any information about the sample size of the instant poll, viewers likely overestimated the validity of the results that were presented on the graph below the video feeds of the two candidates.

CNN should be embarassed for trying to turn an important democratic event into a sports game. It was a decision that may have been in the best interest fo their ratings, but that certainly was not in the best interests of our republic.

Thursday, August 30, 2012

Some thoughts on admissions


In many ways, college admissions can be reduced to a fairly complicated multivariate resource allocation problem. Colleges select their incoming classes from a cohort of very talented students, trying to identify the high school seniors who are most likely to impact society in a meaningful way. While not every college states this mission explicitly, the goal of identifying and educating an extremely talented cohort of people is what drives universities to take admissions so seriously.

Past a certain size, this problem of selecting each subsequent class of students becomes prohibitively complicated, and so almost every university has a dedicated staff of admissions officials. Their role is to evaluate the pool of applicants and make the impossibly hard decisions about who to admit. While admissions officers are obviously still subject to the politics of any large organization, they are traditionally given the distance to operate with minimal external pressure. Essentially, admissions offices are structured to act as a sort of black box, taking a set of applicants as input and returning a set of admitted students as output. This process depends on the role of individual admissions officers being able to see the broad set of applicants and make decisions informed by this bulk population.

Currently, the system works as intended for the vast majority of applicants. However, when it comes to athletic recruiting, college admissions is broken. The preferential treatment of athletic recruits is unfair to other applicants, corrupts the overall system, and is not necessarily ideal for the athletes themselves. I understand that any discussion of this topic fundamentally challenges the legitimacy of athletes who currently attend top universities, and I also appreciate the defensiveness of many commentators when the subject is broached. However, in analyzing the current situation, I find very little justification for athletic recruiting in its current form. Furthermore, I think that majority of arguments supporting athletic recruiting are derived from a number of misconceptions about the zero-sumness of college admissions. This post, therefore, will take the form of a rebuttal to a number of common arguments given by defenders of the status quo. The responses will be centered around the Ivy League given my experience, although the trends driving recruiting are relevant to almost every university.

1) Questioning the legitimacy of the current admissions process is unfair to the many hardworking student-athletes who already attend the university.

While it may be true that questioning this policy is unfair to a number of students, this argument has been made historically in defense of a number of university policies. Similar arguments were made prior to the admission of Jewish students, black students, and women. Furthermore, this argument presumes that the current policy is fair to student athletes. It is not. Because of the existence of recruiting policies, even the most academically qualified student-athletes are viewed as less competent than their peers. In essence, belonging to an athletic team acts as a negative signal for academic talent. As a result, far from being unfair to current athletes, these policies have failed the cohort of very academically qualified athletes as well as the university in aggregate.

2) Athletics provides financial benefits to the university.

This argument was refuted categorically in the book The Game of Life. While it is true that certain sports, such as basketball, can provide enough revenue to offset operating expenses, this argument ignores the extraordinary costs that athletics require of the university. The operating expenses for athletics programs vastly outweigh any money they bring in from ticket sales, and when facilities costs are accounted for, the expenditures that are required to maintain a completive athletic program at any level outweigh the costs. Obviously this isn't a reason not to have athletics; the university at a fundamental level is structured around subsidizing unprofitable activity. This argument, however, is not unique to athletics, and the fact of the matter is that running a large athletics program is expensive.

3) People arguing against athletic recruiting want to end college athletics.

Very few people actually want to end college athletics. Speaking personally, my ideal outcome would be to restructure college admissions in order to promote the type of athletics that was common forty or fifty years ago. During this period, athletes were true amateurs and being a student-athlete meant that in general, athletes were more academically qualified than their peers and spent less time engaged in sport. Without reforming the current structure of recruiting, universities face an athletic arms race in which competitive coaches and schools continue recruiting more highly talented athletes and the American ideal of the multitalented student-athlete fades from reality. We already have schools which specialize in training artists and musicians; the athletic recruiting that happens at top Division I schools (obviously not the focus of this post) could be better served by a similar structure.

While I am very open to a rebuttal, I have not yet heard a first-principles argument which suggests that the university is the proper place to foster preprofessional athletic talent. Just as certain very talented musicians choose to go to college over conservatory, so too would certain athletes. Unlike now, such a system would allow those individuals whose sole goal was to eventually play at the professional level to forgo the burden of attending four years of college rather than being enabled to dedicate themselves wholly to their sport. College athletes should be a place where the participants are students first and athletes second, not the other way around.

4) Recruiting is necessary if we want to ensure that we are able to compete for top athletes.

This argument is based on two assumptions: first, that we want to recruit the best athletes regardless of other characteristics, and second, that we are interested in competing directly with those institutions that do. I have no objective rebuttal to this question; more than the others, it speaks to a fundamental disagreement in values that will not be reconciled via my short exposition on the subject.

However, I will take this time to explicitly state that I disagree strongly that the absolute athletic skill of a given team is what makes athletics valuable for the university. Rather, I think that it is the sense of camaraderie, loyalty to a common cause, dedication to self-improvement and commitment to others that make athletics such a positive aspect of the university experience. I believe that the quest for top athletes has become dominated by the unsupported assumption that it is absolute, rather than relative, athletic talent which allows these virtues to be exposed most clearly. I think that those who defend athletic recruiting by using this line of reasoning will have difficulty explaining why universities, rather than postsecondary athletic schools or semiprofessional athletic leagues, should have a responsibility for nurturing budding athletic talent. If the sole goal is to optimize for athletic skill, why maintain any impediments for athletes to reach their top potential? One could imagine a restructured post-secondary experience where the best athletes postpone attending college and instead focus entirely on sport for one or two years. This would allow for optimization along a single parameter, and would clearly yield a set of schools or leagues with the top athletes from a given year. It would also open the opportunity to experience college athletics to the many people who are not willing to commit as much time or don't have the raw skill to compete currently, providing real choices in the market for athletic talent.

5) Athletes are already academically qualified enough to get in without existing recruiting policies; they are merely a formality.

If this argument were true, changing the current policy would not have any measurable effect on admissions rates for athletes or non-athletes. In the Ivy League, each and every student-athlete must meet minimum academic standards in order to receive an offer of admission. Generally, defenders of recruiting argue that this standard is the academic bar that should guarantee academic competence. They envision a straw man - a bookish, nerdy, socially awkward student with a 2400 SAT score - and contrast that image with the well-rounded student athlete. What this argument fails to consider is that according to many admissions officers, roughly 50% of the applicants for admission to top universities are academically qualified. In the real world of college admissions, the brilliant athlete would probably be admitted regardless of his status as a recruited athlete, and the brilliant, one-dimensional student would get rejected. Most top schools could create between two and five statistically identical unique classes of students. Given this fact, how can one justify being above the bare minimum (a 1160 on the 1600 SAT) as good enough? The SAT is imperfect, but does remain a good predictor of future success according to numerous studies. The simple fact is that there are a subset of athletic recruits who would not get into a top school without recruiting, regardless of whether they met the bare minimum set of academic standards. Recruiting arbitrarily manipulates their potential for admission and artificially inflates their value in the market of admissions.

6) Athletics allow alumni to stay collected to the university.

There has been a decent amount of research done about whether this is true, and while it may be true in the context of a large public university, longitudinal studies covered in detail in The Game of Life indicate no quantifiable relationship between athletic spending, athletic success, and alumni opinion. While there is certainly a vocal minority of alumni who do care very much about athletic success, this minority is simply not representative of alumni on the whole. Additionally, many of these alumni were varsity athletes themselves, and so have a much more personal attachment to the institution of varsity athletics that is not generalizable to the overall alumni pool.

7) There are specific athletes who clearly deserve to be at the university based on their intellect, moral character, compassion, commitment to their team, and raw athletic talent.

This is absolutely true. This is also a sweeping generalization with no bearing on the discussion of recruiting. Firstly, it assumes that these student-athletes would not have been admitted without athletic recruiting, which I find to be a very convenient assumption. Even assuming that some of them would not have been admitted, and that a top school would have lost some exceptionally promising young students, the argument ignores the fact that they would have been replaced by students who, in the judgement of the admissions committee, would have in some way contributed more to the university community. I don't think that it is justifiable to assume that every exceptional athlete would have been rejected and that every student who hypothetically would have replaced them would have been categorically worse. 

8) Varsity athletics fosters a sense of community in the student body.

In my experience, limited as it was, athletics fostered a sense of community very similar to other activities which involved the same number of students. While it is valid to argue that many students attend their friend's sporting events, it was difficult to see how this sense of community was any different than the community formed by any other group on campus. While the validity of my statement can be challenged, I would press any defender of athletic recruiting solely on these grounds to consider the sense of community that athletics forms in high schools, small towns, and large cities. Even assuming that athletics fosters a sense of community that is unique, I simply cannot accept that that sense of community depends on whether or not the athletes themselves were recruited or not.

9) Getting rid of recruiting would make individual teams less competitive.

While this would be a likely outcome, the role of the university is not to play host to the most talented athletes. According to mission statements of many leading universities, the goal of the university is to primarily serve as a space for creating and sharing the cumulative collection of human knowledge. I think that athletics can help teach skills of leadership and of teamwork, but I worry that single-mindedly focusing on competing for the sake of maximizing the level of competition corrupts the goals of the institution. There are other places for the best athletes to train and push their bodies; the university is built to be a place where the primary mission is meant to be focused on training the mind.

10) Without recruiting, our athletic teams would be unbalanced, particularly in rare positions.

While this is a valid point, I fail to see how this is different than the analogous argument for a band; without actively recruiting musicians, we would never have enough oboes to have a functioning ensemble, much less a great one. This argument assumes that without recruiting, coaches would have no input in the admissions process. I think that a more realistic scenario would be similar to the current process for gifted artists and musicians, where coaches would provide a valuable, frank assessment of the applicant's abilities in a specific area, in this case athletics, which would be used by admissions counselors to accurately assess the relative talents of individual applicants and match them with the unique needs of each subgroup within the university.

11) Admissions is not a zero-sum game.

At the end of the day, every single student admitted to a school represents close to twenty students who were rejected. Even assuming that only half of those were academically qualified, the opportunity cost of an individual admitted student is extremely high. Many advocates focus on the negative effects that eliminating recruiting would have on individual applicants, but ignore the thousands of students who are rejected as a result of these policies. While I remain uncertain about whether the most elite schools are, in fact, the best places to go for college, given the widespread assumption that they are, I think that it is unfair to focus so heavily on individual athletic recruits getting rejected without thinking about those who would be admitted in their stead.

12) Recruited athletes commit early to schools, which means that they have to make a decision and probably are committed more to an individual school.

Many rational high school seniors would immediately commit to their top choice university by mid-fall. Arguing that a recruited athlete shows dedication to a school by accepting some sort of preliminary likely letter ignores the fact that almost every other applicant would happily take that opportunity to both commit to their top choice school early and skip the incredibly stressful period of college recruiting, which, without such a commitment, can last from October-May of their senior year.

Each of these defenses points to the implicit assumption that college athletics is fundamentally different from any other extracurricular activity. More generally, it rests on the normative claim that college athletics should be different. While I have addressed each of these issues individually, each defense relies on this assertion. Additionally, defending a policy of athletic recruiting assumes that these talents could not be priced into the admissions decision. Similar problems are faced by every other non-academic organization on campus. Outside of recruiting, we have developed processes which allow non-expert admissions officers to factor in expert analysis from specific experts in music, art, theatre, and the Faculty when making difficult admissions decisions. It seems to be an untenable claim that admissions decisions would not be able to identify high-quality athletic talent in the same way.

In what is essentially a multivariate optimization problem, final admissions decisions should rest in the hands of the admissions committee. Among actors in the system, they are the only ones that have an unbiased perspective and the largest amount of information. We trust in the judgement of the admissions committee to map the university's values onto the selection of the next class. As such, we should also trust that they would treat athletic talent in the same way as any other talent - evidence of passion, of leadership, of dedication and of future potential.