Tuesday, July 16, 2013

Patriot League Schedules (Pt. 1): How Lehigh Stayed At Home In November, 2012

I really thought I'd have been over it by now, but I'm not.

Last year's Lehigh football team, who went 10-1, was denied an at-large spot to the FCS playoffs, becoming the first school with 10 Division I wins to be denied the honor since the playoff field expanded from 16 teams.

How did this happen?

The most common complaint from some Lehigh fans and many other FCS fans was that Lehigh's "schedule strength" in 2012 simply was not cutting it when comparing them to other at-large candidates like New Hampshire, Wofford, or Illinois State.

Is it true that Lehigh scheduled themselves out of a playoff game last season - and will this season's Lehigh schedule make it any easier to qualify as an at-large team should the Mountain Hawks be in that position again?

In recent years, the Patriot League has been a one-bid conference to the FCS playoffs, even with the expansion to 20 teams.

But there was a time when the Patriot League didn't just qualify for the playoffs - they also provided at-large teams to the party on a fairly regular basis.

Let's take a closer look.

From 1998 to 2005, the Patriot League provided four at-large teams to the playoffs in an eight year span, and studying those at-large qualifying schedules provides some insight as to how they were selected.

(Home game in caps)

1998 Colgate
  • 8-3, lost only to undefeated Lehigh in Patriot League play  
  • Out of conference: at Navy (Indep. FBS, L), UCONN (Atlantic 10, L), HARVARD(Ivy, W), at Yale (Ivy, W), DARTMOUTH (Ivy, W)
1999 Lehigh
  • 10-1, lost only to Colgate in Patriot League play
  • Out of conference: MONMOUTH (NEC ,W), at Princeton (Ivy, W), COLUMBIA (Ivy, W), at Dartmouth (Ivy, W), at Delaware (Atlantic 10, W).

2004 Lehigh
  • 9-2, losing only to Lafayette in Patriot League play
  • Out of conference: STONY BROOK (NEC, W), VILLANOVA (Atlantic 10, L), at Liberty (Big South, W), ALBANY (NEC, W), at Yale (Ivy, W).
2005 Lafayette
  • 8-3, losing only to Colgate in Patriot League play
  • Out of conference: at Marist (Non-scholarship MAAC, W), RICHMOND (Atlantic 10, W), PRINCETON (Ivy, L), COLUMBIA (Ivy, W), HARVARD (Ivy, L)

What do these at-large years have in common?

In every instance, the Patriot League at-large bid only had one conference loss.  (The flipside of this is: no Patriot League team with two conference losses has ever qualified for an at-large bid, something to keep in mind for upcoming seasons.)

And in each of these four cases, each school also had at least one Atlantic 10 opponent on the schedule.  (Since then, the Atlantic 10 has reformed as the CAA football conference, home of former FCS national champions Delaware, Villanova, James Madison, and Richmond.)

In two of those cases (1999 Lehigh and 2005 Lafayette), the Patriot League school won that game.

The broad theme is clear: A one-loss team in the league with a "strong schedule" will get a Patriot League team strong consideration for an at-large bid to the playoffs.

But therein lies the rub.

How does a team get a "strong schedule" for an at-large bid?

In 1999 and 2005, the "schedule strength" came from a CAA.

And looking again at Lehigh's 1999 team and Lafayette's 2005 team, it's hard to overlook two giant strokes of luck to help those teams qualify for the playoffs.

In the case of 1999's Lehigh win over Delaware, the Hens that season were very much on the bubble for the playoffs at 8-3.  Was it the Mountain Hawks' head-to-head win over the Blue Hens that allowed Lehigh to leapfrog over the Delaware for one of the final playoff spots?  It sure seems like a good bet that was the case.

In 2005, Richmond was an Atlantic 10 co-champion, qualifying for the playoffs as an at-large team. Did Lafayette's win over them early in the season give them a lot of "schedule strength", allowing them to nab an at-large spot in the eyes of the committee?  Again, it sure seems so.

In both cases, Lehigh and Lafayette scheduled a CAA team - and got lucky.  Not only because both Patriot League teams won those games, but both CAA teams did extremely well that year as well.

What if, in 1999, Lehigh had scheduled a road game against Richmond instead of Delaware?  Would a win against a 5-6 Atlantic 10 team have been enough?

What if, in 2005, Lafayette was filling out the end of a home-and-home with 5-6 William and Mary instead of Richmond?  Would it have been enough?

I'd argue that it would not have been.

Through the years, the Patriot League has had to rely on scheduling strong out-of-conference teams in order to have the perceived "schedule strength" to qualify for the playoffs as an at-large.

In years past, it was enough.

But last year, Lehigh's out-of-conference sweep of Central Connecticut State, Liberty, Monmouth, Princeton, and Columbia was not deemed strong enough by the committee - despite 10 Division I wins, and the fact that Liberty was a preseason Top 25 team.

*****

I'd argue that the business of FCS "schedule strength" has changed quite a lot over the last 10 years, thanks to the increased presence of "guarantee games" against FBS teams.

It was always possible for FBS schools to schedule FCS teams in the late 1990s and early 2000s, but their number remained limited during this timeframe.

There was a good reason for this: FBS schools could only count wins against FCS teams for bowl eligibility once every two years.  FBS teams needs to schedule the games carefully, and rarely.

In 2005, however, the NCAA rule that set the guidelines (18.7.2.2.1) was amended in such a way that greatly increased the number of FCS vs. FBS games.

It was changed to read something closer to how it reads today:
Each year, a Football Bowl Subdivision institution may count one victory against a Football Championship Subdivision opponent toward meeting the definition of a “deserving team,” provided the opponent has averaged 90 percent of the permissible maximum number of grants-in-aid per year in football during a rolling two-year period.
What this did was allow FBS schools to allow one of these games a season - with the important highlighted portion.  Provided the FCS opponent has averaged 90 percent of the permissible maximum number of grants-in-aid per year in football.

90% of 63 scholarships - the NCAA-mandated maximum number of scholarships at the FCS level - is 56 1/2 scholarships ("grants-in-aid").

At the time, the Patriot League's need-based aid structure made it a challenge for those schools to reach this number.  Students needed to be needs-tested to qualify for aid, so the overall number of players receiving aid varied fairly significantly year to year.

At the same time, schools that were already offering 63 conventional football scholarships, like Delaware, Northern Iowa, and Montana, could schedule one such game a year - effectively changing these games from being an occasional fun game for alumni to becoming a consistent revenue source for the program.

Like many rules, it was devised to eradicate a nonexistent problem.  Very few schools under the arbitrary 56 1/2 scholarship limit were scheduling games against FBS squads, and those that were consisted of competitive, Top 25 teams.  No non-scholarship teams like Marist scheduled FBS games.  No Ivy League schools in the 1990s and late 2000s were scheduling FBS teams.  Only Patriot League teams, like Lehigh, Colgate and Holy Cross, fit the description in any way.

Another unintended consequence of that rule, eventually, was to greatly bolster the perceived "schedule strength" of, well, every 63 scholarship school at the expense of the Patriot League and Ivy League.

It is this perceived loss of "schedule strength", I would argue, that has done more to undermine the Patriot League's qualification for at-large teams than anything else.

In Western FCS conferences, before the bowl counter rule was amended in 2005, it was a challenge for FCS schools from conferences like the Big Sky, Southland and Great West to fill out their football schedules.

Most of the larger institutions in the area are big-time Pac 10 schools, or mid-major Mountain West or WAC schools.

If Big Sky schools wanted any games, especially home games, they had to look often at small, relatively local schools, and more often than not they'd be Division II schools.

But once the rules changed, these schools could now promise their alumni that they will play at least one of these "money games" a year.  Extra home dates disappeared, and trips to Tennessee, Nevada, and Indiana showed up instead.

Best of all for these schools, they could do so with extremely little risk in terms of postseason qualification.

Suppose you were Southern Utah in 2004.  You want a successful season, and a chance to compete in the postseason.  But you can't travel all across the country hoping to find games - you have to find schools relatively close by to play, which invariably meant playing a school like D-II Dixie State, for example.

For the sake of the playoffs, a game against Dixie State was not a very good deal for these schools in regards to the postseason.  Wins against D-II schools didn't count towards the requirement for the playoffs that a qualifying school had at least seven Division I wins.  Furthermore, a loss could be devastating to their chances of postseason.  It was an all-risk, little reward proposition.

Fast forward to 2013.  Now, the Thunderbirds have an option for multiple FBS trips on the game every year.

It doesn't only mean one, two, or even three six-figure guarantee check from the FBS school in question - effectively bankrolling the trip, or perhaps even the entire travel budget for the season.

And it's not only an exciting ticket for players, friends and alumni, either - the game that many fans of the school circle on the calendar in the summer.

It's also a possible free ticket for "schedule strength" if the team ends up with a good overall record when the committee is looking for at-large teams - whether they win or lose.

This year, Southern Utah scheduled two FBS teams: South Alabama, and Washington State, sandwiched around a game against a local D-II, Fort Lewis, and a non-counting Big Sky opponent, Sacramento State.  (This is an unusual year when it is possible for FCS schools to have 12-game schedules.)

If Southern Utah goes 7-1 in Big Sky play, beats Fort Lewis and Sacramento State, and plays South Alabama and Washington State close, the Thunderbirds will be 8-4 with 7 Division I wins.

A casual look by a committee member at Southern Utah for at-large consideration will reveal a strong performance in Big Sky play, and those two FBS games.  "This schedule looks strong to me," you can picture the unnamed committee member saying, after looking at a 9-2 Lehigh squad that features win over a host of .500 teams out of conference.

If Southern Utah wins one of those two games - not out of the realm of possibility in the slightest, as South Alabama and Washington State were two of the weaker FBS prograns last year - they may as well be able to write their ticket for the playoffs as long as they goet to 7 D-I wins.

There is no better indicator for at-large selection to the playoffs than a school with at least 7 D-I wins and a victory over an FBS school, no matter how pathetic the FBS team is.

The committee member might not look at Southern Utah's wins over 1-11 Idaho State or 3-9 Weber State.  But they will see the FBS games, especially if one is a Thunderbird victory.  And it's something that Lehigh would not be able to offer the committee members should they be fortunate enough to make the playoffs.

Think that a computer formula could tilt the playing field towards the Patriot League, and let the committee members see the light?  Think again.

There are a lot of varied rating systems for college football.  As the number of FCS vs. FBS games have exploded, the ratings systems have had to come up with ways to rate all the Division I teams together.

They can do this in one of two ways, neither which favor the teams that don't have such games on their schedules.

One way is to weight all FBS teams the same, one generic number that is essentially plugged in for any FBS team that plays an FCS team.  By necessity, it will have to weight the FBS wins ridiculously high.  Whether they play Ohio State or Eastern Michigan, games against them will be treated identically.

When these systems look at "schedule strength", they do what the hypothetical committee member does, sees that they're an FBS school, and plugs in a ridiculously high number in the schools' "strength of schedule" value.

And when a school like, say Illinois State "upsets" Eastern Michigan - even when it's not exactly an upset - any "power rating" of the school would go up dramatically, even when the FBS team in question would have a rough time keeping up in the FCS Top 25, let alone FBS.

The other way to attempt to gauge the quality of these FBS wins is to rank all FCS and FBS teams together in one big list, and try to figure out, through common opponents, one common ranking of all of Division I.

he Simple Rating System, which the committee itself has said it would use in at-large selection, uses this method to determine strength of schedule.   (This isn't as extensive as some of the more granular systems try to figure this out across all subdivisions, from NAIA to Division I, but it at least makes an attempt.)

But even this method, too, gets easily skewed.

While the number of FCS vs. FBS games has increased over time, it's still too small a sample size to accurately judge the relative strengths with FCS and FBS teams during any given season.

Jeff Sagarin's rating system, popularized through the USA Today, is not identical to the SRS, but they do have a similar method of determining "schedule strength", so it serves as a good proxy for the system that will be used.

If you looking at Sag's ratings today, it includes the results of the FCS playoffs, which skews the final numbers.  But through the Massey historic ratings we can get a snapshot of where Sag, and other, rating systems had the teams when it was time to look at at-large consideration.

As any reader of this blog will know, Lehigh went 10-1 last season.  They went 6-1 in league play, losing only to the league champion, Colgate, and going 5-0 against their out-of-conference schedule.  They beat the co-champions of the Big South, Liberty, who finished at 6-5, and also beat Princeton, who went 5-5.

Yet Lehigh, who didn't play any FBS teams, finished as the 50th best team in the Sagarin ratings - only one spot above Weber State, who went 2-9.

Weber State did have FBS teams on the schedule.  In fact they had two: BYU and Fresno State, games which they lost by a combined 82-23 score.  They also had two conference wins, both on the road: beating 1-10 Idaho State and 5-6 Southern Utah.  Every other game was a defeat.

How does something like this happen?

First of all, Weber State's "schedule strength" is "strong" because they, of course, played two FBS teams that were good enough to play in bowl games.  Despite the fact they got creamed by both schools, they get "schedule strength" credit for it.

Less obvious, however, is the "schedule strength" they got last season from Sacramento State.

In Week 1, Sacramento State "upset" Colorado 30-28.  The Buffs were nowhere near a bowl game last season, finishing at 1-11 and were clearly one of the very worst programs at the FBS level.

Yet Sacramento State's (No. 37) win over them helped the Big Sky's overall conference rating immensely.

Four conference teams beat the Hornets last season - Montana State, Eastern Washington, North Dakota, and UC Davis.  Since they beat a team that beat an FBS school, their Sag ratings went up, and when those teams got beaten, other teams got extra credit.

I'd argue that this one win is an outsized schedule strength boost  for the entire conference.  Another example is Western Carolina in 2007.  The Catamounts went 1-10, with their only victory coming against a non-Division I foe.  But the mere fact that they were present in the same conference as Appalachian State lifted their final rankings in Sagarin to ridiculous heights at the end of the season.  Their ranking (186) was higher than ten FCS teams with .500 or better records!

Even if you're not a Patriot League homer like myself, you have to admit something is wrong when a 2-9 team is rated above a 10-1 team.  But the problem is very easy to see: the "schedule strength" in the form of FBS games on the schedule, and the inordinate spillover from "wins", is the culprit.

Depressingly, it's a weakness in the formulas that doesn't seem likely to be able to solved with such a small sample size of games.  FBS schools have eleven shots to play one another to boost their Sagarin raiting, but each FBS school generally only plays one such game.  Colorado went 1-11, but with 10 shots at an FBS victory, they finally came through against 3-8 Washington State.  That's the equivalent of an FCS Top 25 win in Sag's ratings.

Does that mean Patriot League schools without FBS games have no hope at all?

Let's say, most years, FBS games are not an option for Patriot League teams.  Might the old standby of scheduling a "strong" local team, like Delaware or New Hampshire, play the part of an FBS game in terms of "schedule strength"?

The short answer is yes, it can - but it's a much riskier proposition than an FBS game is in terms of "schedule strength".

In 2013, Lehigh plays one out-of-conference team that made the FCS playoffs last season - New Hampshire.

If Lehigh beats the Wildcats, you would think it would help the Mountain Hawks' playoff chances next season.

But not necessarily.

What if this is the year that New Hampshire struggles in the CAA, and goes 6-5 or 5-6 and is out of the playoffs?

Think it can't happen?  Think again - last season, Liberty started out ranked in the Top 25, but crashed out to an 0-3 start before losing a heartbreaker to Lehigh and then proceeding to win six of their last seven games to finish at 6-5, co-champions of the Big South.

Last season, the difference in Sagarin's rating for a championship-winning squad in the CAA and a middle-of-the-pack squad was huge.  11-2 Old Dominion was at 117.  5-6 Delaware was at 185.

Conversely, even scheduling an awful FBS team would help anyone's strength of schedule, regardless of outcome.  For example, 3-9 UTEP finished at 139 last season.  That's a very winnable game for any FCS squad - and arguably, an easier out than Delaware would be at the Tub.

With the FBS game, you get near-certainty of outcome.  For "schedule strength", it's very low-risk, potentially high reward.

Handicapping football team strength isn't like computing the RPI of men's hoops, where there's a multitude of out-of-conference games across many leagues to sample and position the teams.

In football, your "RPI" is largely based on one or two games, three or four if you're lucky.

Worse, the Patriot League schools' historic relationships with the Ivy League greatly hurt their playoff chances as well.

Why?  Because the Ivy League, like the Patriot League, cannot lard up their schedules with FBS games to up THEIR "schedule strength".  You can't get "schedule strength" by proximity to FBS wins when the teams you are playing don't play FBS teams themselves.

Any Patriot League watcher knows that Harvard, year in and year out, has a near-FBS-caliber team suiting up every season.  Every season there are one, two, or three players that make it to NFL training camps, more than many, many scholarship FBS schools.

But Sagarin doesn't know this.  Why?

Last year, Harvard finished at 8-2.  Like Patriot League schools, Harvard scheduled zero FBS teams.

Harvard's out-of-conference schedule consisted of San Diego, Holy Cross, and Bucknell.  None of these teams played FBS opponents, and as a result, San Diego (Sag FCS rating: 72), Holy Cross (Sag: 87) and Bucknell (Sag: 106) don't help Harvard's computer standing one whit, despite San Diego's 8-3 record.

At playoff selection time, 8-2 Harvard sat at No. 39.  To put this in perspective, this put Harvard behind teams like 5-6 Southern Utah (27), 5-6 Northern Iowa (8), and 5-6 Murray State (27).

When there was a handful of FBS games on any FCS schedule, perhaps these sort-of lapses would be forgivable.  But these days, the great majority of FCS schools have FBS games on them.  And in a larger conference, it means more shots at "schedule strength" through proximity through FBS wins.  The eleven-member Big Sky has 16 shots at an FBS upset this season (17 if you count Appalachian State, on their way to the Sun Belt, as an FBS team).  The six-member Big South only has 10.

This season, only one championship-contending Patriot League team can get an FBS win.  On opening weekend, Colgate travels to Colorado to take on Air Force.  If Colgate shocks the world and beats the Falcons, the Patriot League will, perhaps, be in the game for an at-large bid.

Even so, the Patriot League might still be behind other conferences.

Overall. you can expect at least a handful of wins over FBS schools every year, so it's not outrageous for a conference with a double-digit number of these games to have at least one FBS upset themselves.  (The Missouri Valley conference last season had three, which entirely explains the fact that Northern Iowa, at 5-6, was deemed the 8th best team in all of FCS last season.)

If a team is on the bubble, it's been proven time and again no matter how good the FBS opponent, a win against any FBS school will trump almost any other consideration in terms of playoff qualification.  Last season, Illinois State was ranked below Lehigh, Towson, and Richmond in human polls, but thanks to that one FBS win (and little else, frankly), the Redbirds were in the playoffs, and the Mountain Hawks, Tigers, and Spiders all were sitting at home.

And for a conference that has one or more FBS wins, it's a bonanza of unearned "schedule strength" for all members, not just the schools that make it happen.

The message is clear from the playoff subcommittee - with "schedule strength" comes extra playoff bids, for those conference and it's at the expense of the conferences that don't lard up their schedules with these games, like the Patriot League and NEC.

The fact that the committee values "schedule strength" over pretty much everything else is a gigantic flaw in their whole process.  Is it wrong?  I think it is.  But sadly, it's the way the teams are going to be picked for the playoffs.

It's a game that is seriously stacked against the Patriot League the way things are now, because it's not at all clear how leagues can get "schedule strength" by any other means than to schedule a whole lot of FBS games and hope they win some of them.

Tomorrow: How should the Patriot League schedule?

No comments: