Predictions for this week in college football can be found on the College Schedule.

Best bet for the week? I’ll go with Michigan (-14) over Miami of Ohio. The Redhawks did not look like a capable team at home against Vanderbilt last week, and the system was only one point off (Utah by 3 instead of 2, failing to factor in the blocked extra point, I guess) with the Michigan game. It’s certainly difficult to underrate the Wolverines after that weak performance, but this is a team they would beat by 50 in an ordinary year.

Predictions for this week in the NFL can be found on the NFL Schedule.

My best bet for this week is Philadelphia (-7) over St. Louis. The Rams were just dreadful all of last season, and I don’t see them up to the McNabb challenge right off the bat.

The Frontier was 31-10 overall picking games between Division I-A teams, including a decent 22-18-1 against the spread. The Best Bet was successful, Northern Illinois losing, but covering the spread at Minnesota.

I’m not planning on any expansive previews this season. Mostly, I plan on keeping the Frontier updated. I will continue to gather detailed data, and will have three full seasons to analyze next summer.

Because of this, I’ll scale back the best bets I did last season. I’ll just pick one game each week, and see how that works out for me.

My system seems to find the Big Ten overrated going into 2008. So, for week 1, my pick of the week is as follows:

Northern Illinois (+8) at Minnesota.

I’ve added a little more sophistication to my strength of schedule model for the new season.

More people have come to this site in search of strength of schedule information than for anything else over the past year.

So I’ve decided to make strength of schedule a permanent part of the Football Frontier.

My model takes the ratings of each team played, and modifies them by 15% for neutral-site games and 30% for road games. Analysis shows this is the proper modification, based solely on conference play over the past two seasons.

Here is the final strength of schedule chart for 2007:

2007 College Football Strength of Schedule

The SoV columns refer to Strength of Victory, which is what the NFL uses to break ties for playoff positions. The idea behind SoV is that it measures what you’ve accomplished, rather than what you’ve attempted.

I prefer SoS (Strength of Schedule) for college football because schedules are so vastly different in the college game, while in the pro game, the league tries to make all schedules roughly equal.

Here’s this season’s college football schedule strength chart:

2008 College Football Strength of Schedule

Note that the Pac-10, as usual, leads the way. That’s because they play nine conference games, while most other conferences play eight. This eliminates one mediocre non-conference opponent. Keep in mind the SEC, ACC and Big XII all have conference championship games, which improves the schedule strength of the division winners. Georgia, therefore, might well wind up playing the hardest schedule in the entire country.

The 2008 season is just two weeks away, believe it or not. All across the country, students from hundreds of college are strapping on pads and helmets and trying to figure out how to translate Xs and Os from playbooks into action.

I have my season preview at the ready, and I believe the Georgia Bulldogs are set to continue right where they left off last season. With Matthew Stafford and Knowshon Moreno leading the way, Georgia will score a lot of points. And led by an experienced secondary, the Bulldog defense will be hard to beat as well.

Football Frontier 2008 Pre-Season Ratings

Rounding out the top five are Ohio State, Southern California, Florida and Louisiana State.

There’s still time to enter our Sixth Annual NFL Draft Contest, which takes place on Saturday, April 26.

To enter, just fill out the form and submit your guesses for the 31 picks made during the first round of this year’s amateur draft.

If you entered early and did not have Michigan offensive tackle Jake Long as your first pick, please feel free to go ahead and enter again. We’ll just delete your original entry.

To wrap up this series, I’m presenting the list of discretionary calls made, by referee.

This includes all the discretionary calls not attributed to the umpire in part four: defensive and offensive pass interference, illegal contact, roughing the passer and intentional grounding.

Of these calls, the referee is only directly responsible for the latter two. Based on zone, the five officials (field judge, back judge, side judge, line judge and head linesman) who ring the defense study the individual receiver/defender matchups and throw the interference and contact flags.

Since separating out each official’s calls is impossible based on common play-by-play data, I assigned these calls to the referee, who is the crew chief. Generally, crews work together for the entire season.

Last season, this subset of calls accounted for only 1.65 penalties per game (13 percent of all penalties). When you think of the impact of interference and contact calls alone, that’s a surprisingly low number.

In fact, including the line calls attributed to the umpires, discretionary calls account for 3.73 penalties per game. But as I’ve written more than once in this series, these are the calls that dictate to coaches and players what’s allowed within the rules. These are the calls that fans complain about on talk radio Monday afternoon. These are the calls that have coaches fuming in post-game conferences, trying to walk the line between expressing frustration and getting fined.

While this list is not a true indication of how a referee controls the game, it is relevant information.

Averages were weighted by the average number of those calls made during each season.

Referee	Calls per Game	Compared to Average
Gene Steratore	2.32	+31.2%
Terry McAulay	2.55	+23.3%
Tony Corrente	2.41	+17.6%
John Parry***	1.93	+17.0%
Ed Hochuli	2.29	+11.3%
Ron Winter	2.21	+7.2%
Tom White*	2.38	+6.9%
Mike Carey	2.13	+3.8%
Larry Nemmers**	2.10	+0.9%
Jeff Triplette	2.10	+0.3%
Scott Green	2.06	-0.8%
Gerry Austin**	2.01	-1.1%
Walt Anderson	2.01	-1.3%
Bernie Kukar*	2.13	-4.9%
Walt Coleman	1.90	-8.3%
Bill Vinovich*	1.98	-9.3%
Jerome Boger	1.60	-10.2%
Johnny Grier*	2.06	-11.1%
William F. Carollo	1.74	-16.4%
Peter Morelli	1.69	-17.3%
Ron Blum*	1.70	-17.3%
Bill Leavy	1.71	-17.9%
Bob McElwee*	1.67	-18.6%
Dick Hantak*	1.38	-32.8%

* - retired before last season
** - retiring before 2008 season
*** - 2007 was first season as umpire

In the future, I’d like to focus more deeply on pass interference calls, since these are often the most controversial penalties. To do so, I need to find a way to isolate the referees actually making these calls.

With six seasons of data, I don’t have enough material for a focused study. There are only 0.80 defensive pass interference calls made per game in the NFL. With five different officials throwing these flags, all reporting to the crew chief, this would be a fairly serious undertaking.

I hope Football Frontier readers have enjoyed these reports, and are looking forward to the 2008 season, when the Frontier will have more frequent new material.

There’s a significant variation in how each officiating crew approaches discretionary penalty calling.

Since these penalties are based more on rule interpretations than concrete “yes or no” violations, each call has implications far beyond the one play it affects. Coaches and players often have to make immediate or long-term adjustments, possibly even determining strategy based on which crew is assigned to the game.

When penalty calls dropped from 16.2 to 13.1 from 2005 to 2006, a decrease of 19.0%, the small subset of key discretionary penalties, which made up 36.3% of all calls in 2005, dropped from 5.9 per game to 4.3 per game, a decrease of 27.7%.

The drop in discretionary calls made up 52.9% of the entire decrease in penalties.
Even more convincing: when penalty calls dropped another 3.3% from 2006 to 2007, discretionary calls dropped 12.6%. There was actually an increase in penalty calls outside the subset - the discretionary call drop made up 122.3% of the entire decrease.

Umpires are responsible for most penalty calls involving linemen. So I separated out the discretionary calls they usually make - offensive and defensive holding, and illegal use of hands - and studied the number of calls each individual made over the last six years.

Averages were weighted by the average number of those calls made during each season.

Umpire	Calls per Game	Compared to Average
Carl Paganelli	4.03	+39.0%
Bob Wagner*	3.55	+28.9%
Roy Ellison	3.30	+21.9%
Bruce Stritesky**	2.43	+9.5%
Ed Coukart*	3.43	+9.1%
Undrey Wash	3.12	+8.5%
Garth DeFelice	2.97	+4.1%
Scott Dawson	2.93	+3.5%
Dan Ferrell	2.88	+3.5%
Butch Hannah	2.68	+3.5%
Chad Brown	2.78	+2.4%
Darrell Jenkins	2.81	-0.6%
Jim Quirk	2.71	-1.4%
Steve Wilson	2.67	-2.3%
Brian Balliet*	2.84	-3.1%
Carl Madsen	2.70	-4.4%
Tony Michalek	2.48	-9.8%
Ruben Fowler**	1.96	-11.6%
Bill Schuster	2.45	-15.3%
Richard Hall	2.25	-20.8%
Jeff Rice	2.11	-28.4%
Jim Duke*	1.87	-32.1%

* - retired before last season
** - 2006 was first season as umpire

Since referees are crew chiefs, they do have significant control over the calls made during their games. So I also looked at the breakdown of these line calls for each referee.

By and large, officiating crews stick together for an entire season. But crews are often switched around during the off-season.

I found that the standard deviation from average was 16.8 for umpires and 14.8 for referees. So umpires have more effect on the number of line calls made, but referees clearly have impact.

As an example, umpire Roy Ellison calls as many holding penalties as anyone. He has been paired with Ron Winter, among the penalty-calling leaders, the last two seasons. In 2005, he was paired with Gerry Austin, who calls less penalties than any referee. Ellison was close to the league average in discretionary line calls in 2005.

You know the drill by now. The crowd roars as the visiting quarterback sets up under center. Players look uncomfortable as they struggle to hear the signals. A guard jumps too early, resulting in a five-yard penalty against the offense.

The crowd gets even louder as the announcers celebrate their contribution.

There’s no doubt visiting teams are called for more false start penalties than home teams. Last year, that edge was 1.56 per game to 1.30. Over the last six seasons, the visitors were called for 1.63 false starts per game to 1.37 for their hosts.

Credit the fans, on average, for an extra five yards every two home games. Maybe not as much as the announcers would have you believe, but significant.

Since the overall penalty numbers are about even, however, that brings up an interesting question. Do officials compensate, overtly or subconsciously, for this crowd advantage?

In most cases, a penalty is a penalty. But when looking at the small set of discretionary calls, which encompass about 33% of all penalties, there’s opportunity to counter a crowd that may be perceived to alter the delicate balance between the teams.

The following chart shows the number of calls that went against home teams and visiting teams over the last six seasons, splitting out false starts and key discretionary calls.

For reference, false starts make up about 21% of all penalties, defensive pass interference about 6%, defensive holding about 5% and offensive holding about 18%.

Year	Visitors	Home	Pct.	FSt.	DPInt.	DHold.	OHold.
2002	7.16	7.61	48.5	51.5	43.6	44.2	49.0
2003	7.29	7.51	49.2	57.3	43.5	37.2	45.0
2004	8.30	7.67	52.0	57.0	45.0	50.5	48.4
2005	8.04	8.17	49.6	53.3	41.2	46.6	48.2
2006	6.71	6.42	51.1	54.6	50.8	47.3	50.0
2007	6.38	6.31	50.3	54.5	44.8	53.9	44.0
Overall	7.31	7.28	50.1	54.7	44.6	46.1	47.5

Visitors = Average number of penalty calls against visiting teams, per game.
Home = Average number of penalty calls against home teams, per game.
Pct = Percentage of total penalty calls against visiting teams.
FSt = Percentage of False Start calls against visiting teams.
DPint = Percentage of Defensive Pass Interference calls against visiting teams.
DHold = Percentage of Defensive Holding calls against visiting teams.
OHold = Percentage of Offensive Holding calls against visiting teams.

Looking at those penalty calls, the answer was found in offensive holding in 2007. The home team was called for 1.23 holds per game while the visiting team was called for only 0.97 holds. That made up for the false start chasm. It was a league-wide change. Remarkably, 15 of the 17 umpires in the NFL called the home team for holding more often.

In 2006, however, the false start advantage went uncompensated, and visiting teams were penalized slightly more frequently than home teams.

In 2005, visiting teams and home teams were penalized at about the same rate, with the false start compensation spread out over several types of penalty calls. Most notably, almost 59% of all defensive pass interference calls in 2005 went against the home team - only three of the 17 crews called more pass interference against the visitors.

The numbers were startling taken as a whole. Over the six-year period, despite 438 more false starts against visiting teams, they were only penalized 48 more times than home teams. That’s 390 subconscious or overt make-up calls.

While there were variations, year to year, in the three discretionary calls studied above, the home team was flagged 425 times more than the visiting team. Now that’s only about one flag every four games, but it’s a real factor.

Most notable was the spread in defensive pass interference calls. Only in 2006 did visiting teams receive more than 45% of those flags. And, of the 24 referees who led crews over the last six years, only three have called more defensive pass interference against visiting teams during that span.

Those three crew chiefs: Gene Steratore (58.3%) who has only been a referee for two seasons, Tony Corrente (55.3%) and Ron Winter (50.7%). Of the other 21 crew chiefs, eight flagged visiting teams for less than 40% of all defensive pass interference calls. Of current referees with at least six years of crew-leading experience, that includes Terry McAulay (39.2%), the retiring Gerry Austin (38.0%) and Walt Coleman (37.1%).

Is there anything strange going on in those NFL officiating rooms? Hard to tell. The numbers are very odd, and suggest officials may be using discretionary calls to partially mitigate the home-field advantage.

Before coming to any serious conclusions, however, this topic warrants more study.

The easiest way to look at an officiating crew is by simply counting the number of penalties it calls.

The following chart shows every officiating crew that has worked over the last six years, ranked by how many penalties it calls related to the league average. Since some crews worked only in different years, the average is adjusted to the league average for each year the crew worked.

I’ve also included a split of calls against visiting teams and against home teams.

Referee	Calls per Game	Average	Visitors	Home	Difference
Johnny Grier*	17.17	+10.9%	8.61	8.56	+0.3%
Ron Winter	15.90	+9.5%	7.91	7.99	-0.5%
Jerome Boger	14.10	+9.2%	7.27	6.83	+3.1%
Larry Nemmers**	15.79	+7.7%	7.82	7.97	-0.9%
Terry McAulay	15.54	+6.8%	7.87	7.67	+1.3%
Ed Hochuli	15.42	+5.9%	7.71	7.70	+0.1%
John Parry***	13.20	+4.0%	6.33	6.87	-4.1%
Gene Steratore	13.39	+3.9%	7.13	6.26	+6.5%
Walt Anderson	15.04	+3.7%	7.75	7.29	+3.1%
Tony Corrente	14.96	+3.2%	7.69	7.27	+2.8%
Ron Blum*	15.03	+1.7%	7.73	7.30	+2.9%
Jeff Triplette	14.81	+1.6%	7.37	7.44	-0.5%
Tom White*	15.52	+0.3%	7.62	7.90	-1.8%
Mike Carey	14.30	-1.9%	7.06	7.24	-1.3%
Bernie Kukar*	15.12	-2.5%	7.53	7.58	-0.3%
William F. Carollo	14.29	-2.5%	7.04	7.24	-1.4%
Scott Green	13.60	-4.4%	6.98	6.62	+2.6%
Walt Coleman	13.83	-5.9%	7.08	6.76	+2.3%
Bill Leavy	13.55	-7.5%	6.97	6.58	+2.9%
Peter Morelli	13.49	-8.1%	6.87	6.63	+1.8%
Bill Vinovich*	13.59	-9.5%	6.46	7.13	-4.9%
Bob McElwee*	13.33	-9.7%	6.40	6.93	-4.0%
Gerry Austin**	13.03	-10.5%	6.18	6.85	-5.1%
Dick Hantak*	12.38	-16.2%	5.94	6.44	-4.0%
League Average	14.60	0.0%	7.31	7.28	+0.2%

* - retired before last season
** - retiring before 2008 season
*** - 2007 was first season as referee

Looking at this chart, you can see that an officiating crew can deviate by about 10% from the average number of penalties called. This means that in 2007 terms (12.69 average calls per game), you’d expect about 11 calls from lenient crews and about 14 calls from strict crews.That difference only amounts to about three calls per game per crew. Small, but significant, as about half of those calls come from the discretionary call base I described in the introduction and will discuss more in future entries.

I included the home/road splits just to give people a little confidence in the impartiality of NFL officials. There is almost no difference in the total number of penalties called against home and road teams.

Officials who call less overall penalties seem to call fewer penalties against the visiting team. There is a 21% correlation between those two figures. That can’t be explained simply by a higher tolerance for false starts with noisy crowds, as the referees topping that list called far fewer discretionary penalties against visiting defenses.

That’s a mystery that may be worth pursuing down the road.