New Realities, Part III: The Academic Index

It's funny: as soon as I'm just about ready to unveil my blog posting on academic indexes, the New York Times scoops me. With their article "Ivy League Crunch Brings New Cachet To Next Tier" (which highlights Lehigh as the prototypical school of this "next tier", right below the Ivies), Alan Finder reports that it's harder than ever to get into schools like Lehigh.

A subscription is required to get the full text, but a key snippet of the article follows:

At Lehigh, known for its strength in engineering and business, about 12,000 students applied this year. That is a whopping 50 percent increase in applications over seven years ago and more than 10 times the seats available in a freshman class of 1,150. The median SAT score of admitted students has climbed about 10 points a year in recent years, officials said.

...

The result, said Jonathan Miller, a senior at Mamaroneck High School in suburban Westchester County, N.Y., is that many classmates perceive institutions like Tufts University, Bowdoin, the University of Rochester and Lehigh in a new light. “I would say that high school students are looking more and more at these schools,” he said, “the way they used to look at the Ivies.”

The article goes on to explain that it's not only Lehigh which is experiencing this academic crunch. Bucknell, Lafayette, and Colgate are also prominently mentioned - in a nutshell, the academic credentials of all Patriot League schools are rapidly getting better and better. Like you, I've been wondering lately if I would have gotten into Lehigh with my high school grades.

Unlike other institutions, this increase in academic credentials of the entire incoming class has a direct effect on the composition of our athletic teams. As you'll see, this is causing folks all over the Patriot League to perhaps rethink the Academic Index (or AI for short) and how it's used in athletic recruiting.

Academic Indexes
The best way to think of the AI is that it's a tool to determine whether schools are admitting students that are similar academically to the rest of the incoming class. It’s a formula which attempts to measure an athletic recruit’s academic record, which dates from the founding of the league. Although the Ivy League's formula changed slightly in the mid-1990s, the way the Patriot League has computed its AI has remained largely the same since the league's inception.

In the Patriot League, the AI is roughly half made up of a percentage of class rank (or GPA if the school doesn’t share that information) and roughly half made up of standardized test scores. These scores are then compared against the rolling four-year average of the AI for the entire incoming class of the institution. Per sport, as long as the entire recruiting class is one standard deviation below the AI for the entire incoming class, everything should be copasetic.

Yet very recently, the executive director of the Patriot League, Ms. Carolyn Femovich, announced that the leagues’ AI system would be going under review. “We’ve engaged the services of an outside research entity to look at our AI, to look at any issues. Do we have the right measures in place? Is it working in the intended way? Are we being too restrictive? Are we really reflecting the entire student body? Is there a better way to go about doing it?”

It’s especially interesting that this review is coming now – and it shows that maybe, just maybe, folks at the Patriot League are taking a fresh look at the AI calculations because as they stand right now they are not working the way they should be. More specifically, it looks like it is being undertaken for three reasons – all which require some explanation. Although the league is undergoing the study in regards to all sports, the first one is definitely about football.

Reason #1: Presumptive Floors
The Ivy League’s AI calculation for football is different than the Patriot League, but it also differs in one significant way: it has a hard floor. This means that if an athlete scores less than the AI floor for the league, he can’t be admitted to any Ivy League school.

The Patriot League doesn’t have any such hard floor. But it does have a “presumptive floor” below which an individual school could not realistically recruit. Using SAT scores as a simplistic example, let’s say a Patriot League school has an average incoming SAT score of 1350. In order to have the football team’s incoming class have the same average SAT score, you could recruit an athlete with an 1100 SAT – but you’d also have to have a kid with a 1600 SAT to balance him out. In this example, the “presumptive floor” for this student is an 1100 SAT score. (By definition, this number would be one standard deviation below the average AI score.) Although it would be extremely difficult, you could theoretically recruit him.

However, the Ivy League’s system is also different in this regard. Recruits are placed into AI “bands” (unique for each school), where a school can recruit a certain number of athletes in four “bands” down to the floor of the entire league. Most Ivy schools are not restricted, like the Patriot League, to recruit within one standard deviation of the incoming class’ AI. Some can recruit athletes which are more than three standard deviations below the average.

Although Ivy League teams are restricted to only two football players from the AI floor to the top of their first band, they can target players that the great majority of Patriot League teams cannot. Furthermore, this means that for almost every Patriot League school, their presumptive floor is actually above the hard floor set by the Ivy League.

Currently, this puts Patriot League schools at a systemic disadvantage to the Ivies – their pool of athletes is smaller due to the restrictions of the AI that exist today.

Reason #2: Better Academics
As seen in the New York Times article, private colleges and universities across the board have seen their academics profiles get, on average, better and better in recent years. This is making it even more challenging to find Division I athletes that fit this new academic mold in AI leagues. Although the New York Times article seems to show that Patriot League schools are not that far off in terms of admissions choices by incoming freshmen, coaches around the league don't share that view.

Lafayette head coach Frank Tavani was interviewed in February by The Morning Call about this subject, and his frustration is evident. “It's more difficult for us because we aren't cloaked in that magical Ivy League aura and reputation,'' Tavani said. ''We're fighting for the same athletes and it's highly competitive. Quite frankly, there aren't that many players of this caliber that carry those kinds of credentials and academic standards. In the end, because the pool has gotten smaller because of the increased standards, we are diluting the talent level of the league.”

Confirming this impression, “What used to be a 1200 board score is now a 1280 board score,” one Patriot League official told me. “As a result, the qualified pool of athletes shrinks because your academic profile continues to increase – and the competition for these kids is higher than ever. Every institution in the country wants that scholar-athlete. Merit scholar, top athlete, and can play at the Division I level – everybody wants those kids.”

In order to get these kids, Patriot League schools need to travel further away – requiring the spending of more money to get the type of athlete that can be admitted. Some critics point to increased athletics spending as a reason why athletic culture is anathema to elite institutions as Ivy or Patriot League schools. Ironically, the AI restrictions are requiring more spending to get the same type of kid that used to be available (in many cases) somewhat locally. Far from spending more money to get better athletes, folks are spending more money to get academically qualified athletes – and then are criticized for spending so much money.

No school has a smaller window for recruits in this system than Georgetown. In 2005, the freshman class averaged more than a 1450 SAT score – and had, by far, the toughest AI in the entire league. No wonder then that Georgetown has struggled to field football teams – and to reach the same number of football equivalencies as other schools -- that can compete for Patriot League titles. If all Patriot League schools are getting squeezed, no school is more squeezed than Georgetown due to the rising standards.

Reason #3: Ensuring Academic Standards Are Working
To some academic folks, of course, football coaches’ concerns about competitiveness fall on deaf ears. They are busy insisting that tools such as the AI are actually flawed by nature. William Bowen, author of The Game of Life and inventor of the AI system used today by the Patriot League, argues passionately in his book (and comes armed with reams of data) the fact that athletes are allowed to be a “standard deviation” from the rest of the class means that the AI system is, in effect, “affirmative action for athletes,” admitting students with lower academic standards who do worse in school. According to Mr. Bowen’s book, presidents ought to be more strictly insisting that students are all held to the same academic standards.

Patriot League coaches snap back: First of all, athletics at our level enhances the educational experience for students, players and alumni. It’s not like we’re sliding players into the program to get a shot at the NFL – only the most exceptional kids might get that chance. Yet these same kids have a lot more to juggle than other students – practices, road games, and even media coverage. It’s a chance for them to be stars and to handle pressure. You have to cut them a little slack. Not much, but a little.

In this emotional debate about academic standards, who is right? Is the academic achievement of Patriot League athletes not up to snuff with their incoming classes? Are the standards working? Oddly, more than ever it’s harder to say since the ways that high schools are reporting academic performance has changed dramatically over the past twenty years.

First, take the SAT’s. Are they a fair way to rate incoming recruits? Jay Rosner, executive director of The Princeton Review foundation, published in 2003 in The Nation that it is his “hypothesis is that every question chosen to appear on every SAT in the past ten years has favored whites over blacks.” Furthermore, other groups such as FairTest (a nonprofit advocacy group) have also claimed that “standardized tests such as the SAT are biased toward students from families with higher incomes who can hire coaches to help prepare.”

At the very least, this puts into question the fairness of relying on the SAT as a method of rating incoming freshmen. Because of studies like these, an increasing number of schools have stopped requiring SAT scores for admission, like Holy Cross.

Using class rank, then, can be a more effective judge of academic achievement. Unfortunately, fewer and fewer schools are doing it. “The problem with ranking class is the high schools know that this disadvantages some of the kids,” one official told me. In other words, if you’re in an elite academy with academic champions, yet “only” are in the top 50% of your class, it hurts the chance of you getting accepted. As a result, many high schools have made the choice to not rank their students.

For the AI, the fallback method is ranking by GPA (or equivalent). But then how do you determine that a student is taking truly challenging classes? Some school districts, eager to boast they have many of their students getting accepted to Ivy and Patriot League schools, sometimes have “half the class with a 3.5 or better”, one Patriot League coach told me.

This grade inflation doesn’t only make it more difficult to find the really good recruits. It also brings up the academic numbers of the entire incoming class – making it a continuous cycle of tightening every year as the “standards” go up. Sure, the measurements are going up every year. But are the candidates better? Are they smarter? Are the measurements honestly working? The study that the Patriot League is undergoing seems to imply that they may not be working as intended - and therefore, need to be updated.

You'll have to wait until next week for Parts IV and V: Scholarships, and Some Conclusions.