In Part 1 I showed that UMass is no I-AA pushover and likely won’t suddenly become one in I-A. In Part 2 I examined the composition of Northeast team rosters and how they differ from national norms. In this final segment I’ll discuss the details of my modeling and see if we can draw any conclusions on the Northeast recruiting landscape on a quantitative basis.
The gist of the maps I’ll show here is pretty simple: each player is considered a point-source of population 1 at their hometown. Using the "decay function" described in Part 2 which considers the probability of any given recruit to be willing to travel X miles to a school, an "effective recruit population" from each player is added to every possible location. For example, using the numbers mentioned above, the "effective recruit population" for a player to a point 50 miles away from their home is about 0.80. After summing the contribution from each player at all pixels on the map (yes, the map even tells you how different spots in the North Atlantic Ocean are for recruiting) and normalizing, the map is color-coded according to the scale bar in the bottom left corner.
I've also considered an extension to this method in which the effect of the competition from other schools is considered. To do this, I considered each school as a negative population source (equal not to one, but to the average roster size). You'll note the effect on Ohio is quite striking when this is factored in.
A nice feature of these methods is that they generate a continuous map and are not restricted to calculating a score for individual schools. The maps tell you where to build a school if you wanted to optimize the football recruiting potential, which is partially apt for looking at the transition of a program to the I-A level. In both methods you can rank the schools based on their access to recruits, and I'll discuss what it may mean for UMass below the maps.
A caveat: given the relative simplicity of this model, there is a natural bias against sparsely populated areas. E.g. in Utah a recruit may be hundred of miles from both Utah and BYU, but there's still a high probability he's going to one of those two schools (I deal with this issue in the more predictive method I used here).
Figure 1: (click for larger image) Map neglecting the impact of competition between schools for players. The white dots indicate the locations of I-A schools (and UMass, in central Massachusetts). The numbers indicate the top ten concentrations on the map (separated from one another by a reasonable distance).
Figure 2: (click for larger image) The map is strikingly different when competition is included. In particular observe how schools in the Midwest are now in a rather barren territory. Although there are many recruits in the area, there are not enough for the many I-A programs in the region. As per the scale bar, red next to orange is bad, red next to pink is good.
The impact of competition between schools is easy to see in comparing the two maps. In particular, note how the Midwest looks to be in pretty good shape when the density of I-A teams in the region is not considered. Simply put, although there may be many recruits in this area, there aren’t enough to fill the schools therein. On the other end, confirming conventional wisdom, is Florida. Florida’s dominance is even more dramatic when only the top recruits -- i.e. Rivals 250 players -- are considered. It's also rather fascinating to look at the maps, with and without competition, when the general population (from the 2000 Census) is used as opposed to football players.
When we rank the schools using the 2010 rosters it’s not surprising that UMass falls well behind schools in the southeast. In the two tables below I’ve shown a few schools of note and their ranking when competition is and is not factored in. One obvious trend you’ll note is that coastal schools make a big jump when competition is included (they don’t get recruits from the ocean but they don’t lose them to the ocean either). It’s also interesting to observe the big drop in the MAC conference average (not including UMass) when competition is included; this effect on the Ohio-heavy league can easily be understood by looking at the map.
|No Competition||Including Competition|
|1||Georgia Tech||1||Florida Atlantic|
|89||Texas Tech||19||Georgia Tech|
While UMass falls well into the bottom 50% of the rankings for recruiting potential, this forecast is certainly not destiny, as indicated by the teams that fall near and behind them. Particularly of note is that UMass is well above the MAC average when competing schools are considered. This indicates that UMass has an opportunity to quickly rise up the conference pecking order. The only relatively close conference teams they’ll be recruiting against are Buffalo and Temple.
Currently very few players from New England wind up in MAC schools—only 11 over all 12 schools in 2010—although a decent number come from New York (59) (but about two-thirds of these players are at Buffalo). As mentioned in Part 2, more than half of UMass’s team is from the Northeast, a much higher percentage than the I-A teams from the region and while UMass must have strong ties to Northeast high schools in order to pull in so many local players, the essential question is how many of these players are I-A talents. It will be interesting to see if UMass’s recruiting reach soon extends into MAC territory (further depleting the region) and vice-versa—will the other MAC schools see a bit of an influx of players from the Northeast? Is the area currently under-scouted for mid-major I-A football players?
The presence of UMass will also affect the other Northeast schools’ ability to bring in recruits, and their ranking (when competition is considered) drops when a school is added in Amherst, which is shown in the table below. The changes are noticeable, but it should be stated that the schools are pretty crowded in this part of the rankings, so a small change in recruiting prospects can produce a large fluctuation in ranking.
The UMass Effect
|School||Pre-UMass Rank||Post-UMass Rank|
Finally, just how ambitious is UMass with its football program? Will they soon be competing for the top players in the area? From 2006-2011 there were a mere 21 Rivals 250 recruits from the Northeast, and only 6 of them stayed there for college. The table below shows the breakdown by state with the destination colleges. Even Connecticut, a program on the rise, has not been able to lure in any local Rivals 250 recruits. It is way too early to foresee UMass getting commitments from high level recruits, but it will be worth watching if UConn, a program with little tradition but recent success, and Syracuse, a program with great tradition and seemingly back on the upswing, can pick up steam on the recruiting trail as they are likely harbingers of UMass’s future potential. Of course, they are also competition.
Destination Schools of Rivals 250 Recruits from the Northeast
New York (10) -- Boston College (2), Notre Dame (2), Florida, Maryland, Penn State, Pittsburgh, Stanford
Connecticut (6) -- Penn State (2), Boston College, Florida, North Carolina, Virginia
Massachusetts (5) -- Boston College (3), Georgia, North Carolina
In conclusion, I've taken a look at an often-overlooked area of the college football nation. While UMass might not make the headlines, their transition will be interesting to watch -- basically, can they duplicate UConn’s successes, and then perhaps surpass them? That may be a tall order given UMass will be playing in a non-BCS conference, but don’t forget the example of Boise State. As my modeling demonstrates, the rest of their future league has to compete desperately in a deep, but finite recruiting pool. UMass on the other hand, while not in a talent-rich area, has much less competition. If UMass can tap their region consistently (and steal some players from Ohio, for example) their location could prove to be a great source of stability for the program -- a rare asset in the up-and-down MAC. UMass has two other attributes that will serve it well: status as the flagship university in a populous state and a history of consistent success (only two losing seasons in the past 13 years).
The broader question is whether or not UMass can help catalyze greater interest in college football in New England. UMass will have natural rivalries with their neighboring state school (UConn) and in-state private school (BC). I can definitely imagine a game between UMass and UConn/BC held at Gillette Stadium selling lots of tickets -- 30,000 went to see UMass play New Hampshire -- and generating some hype (also, let’s be honest, it may help if the Patriots start losing). UMass may not have Floridian talent at its doorstep, but their recruiting pool should be more than adequate for them to compete. The primary determinant therefore of UMass’s success in I-A will be the ability of the head coach Kevin Morris to adapt to the higher level of competition.