Wednesday, January 30, 2008

Super Bowl 42 Referee Mike Carey Great for Favorites

The referee for Super Bowl 42 between the Giants and Patriots is Mike Carey. He will have an all-star group of officials working with him that is mixture of officials from other referees' regular season crews.

How does he and the referee crews from which the Super Bowl officials come from look for favorites?

To study this, we had to assume a baseline of what makes a favorite. We looked at the actual season results using the footballoutsiders DVOA difference for the visiting and home teams. We made a model of the overall expected wins for the favorites in each of the games and added up the expected and actual wins.

Mike Carey comes out tops out of the 17 referees for the winning percentage for favorites and for the difference between the expected winning percentage and actual winning percentage for favorites.

Favorites won 93% of Mike Carey's games in the regular season, far and away the best. The next best was down at 88% for Ed Hochuli.

Favorites did 17% better than expected in Mike Carey's games (93% winning rate as opposed to the expected 76%). That was tops for the 17 referees, with Bill Leavy coming in second at 14% (favorites won 82% as opposed to the expected 68%).

Let's take a look at the percent boost that favorites received above the expected winning percentage for the regular season crews from which the other officials (other than Mike Carey) come.

Bill Leavy was 2nd best of 17 for favorites at 14% above expected (82% actual, 68% expected).
Ed Hochuli was 5th at 8% above expected (88% actual, 79% expected)
Scott Green was 11th at -1% (73% actual, 75% expected)
Gerald Austin was 13th at -5% (69% actual, 73% expected)
John Parry was 17th and last at -12% (60% actual, 72% expected)

Do these numbers really mean anything at all? Perhaps some referees call the game in a standard kind of way (which would help the favorites because they are the better team) while other referees call a more haphazard game or use their own pecular interpretation of the rules, which might hurt the favorites? Post a comment and of course, feel free to say that you think we are completely out of it with this proposed statistic!

All we do is crunch the numbers, it is up to you to decide whether the numbers are really worth anything. But we do hope you appreciate our number-crunching. We seem to be offering statistics and analysis that hardly anyone else is offering!

By the way, the regular season DVOA difference between the Patriots and Giants was 52%. That is huge. When one team had a 50% or more advantage over the other team in the regular season, the favorites went 20-1. Maybe you can argue that the Giants' playoff form means the regular season DVOA is not a good reflection, but it does not look so great from the DVOA standpoint. Then again, the Giants have bucked the trend. When they visited Dallas, Dallas had over 20% better DVOA in the regular season. Home teams with 20-30% DVOA advantage went 24-4 in the regular season (winning 86% of their games), yet the visiting Giants won anyway.

Sunday, January 27, 2008

Super Bowl Giants-Patriots: Referee Statistics

In this post, we will take a look at the number of penalties called per game, total points per game, and the two teams' experience with the different parts of the Super Bowl crew. We track penalty statistics by the entire referee crew, not by the particular individual in that crew. So let us take a look at how the entire referee crew that each official was a part of did in the regular season:

Referee: Mike Carey (Carey)
Umpire: Tony Michalek (Parry)
Head linesman: Gary Slaughter (Leavy)
Line judge: Carl Johnson (Austin)
Field judge: Boris Cheek (Green)
Side judge: Larry Rose (Green)
Back judge: Scott Helverson (Hochuli)

By the way, there were 17 referee crews in the 2007 regular season.
Referee: (Carey) -- 9th in total penalties, 6th in total points, NE 2-0, NYG 0-1.
Umpire: (Parry) -- 7th in total penalties, 7th in total points, NE 1-0.
Head linesman: (Leavy) -- 6th in total penalties, 14th in total points, NYG 1-0.
Line judge: (Austin) -- 17th and last in total penalties, 5th in total points, NYG 1-0.
Field judge and side judge: (Green) -- 11th in total penalties, 4th in total points, NE 1-0.
Back judge: (Hochuli) -- 5th in total penalties, 3rd in total points, NE 1-0, NYG 0-1.

If you average all of the crew members equally, you'd come out with 11.2 penalties per game which would be 10th if it were a regular season crew. You'd come out with 44.7 total points per game, which would have been 7th.

In the 2007 regular season, New England had experience with 5 of the 7 officials (one of them twice) and the Giants had experience with 4 of the 7 officials. Advantage: New England Patriots.

Our next project if we get time will be to assign a rating for each referee crew about how good they were for favorites -- or whether there was a greater chance of upsets in their games. All we need to crunch these numbers is create a model to identify which team is the favorite and by how much, then to compare the game results and add up the season-long totals for each crew(!)

Tuesday, January 22, 2008

Super Bowl Giants-Patriots: Mike Carey's All-Star Crew

Several bloggers are reporting who is on the Super Bowl All-Star crew (if you want to know which bloggers and posters are reporting this, just search on google).

The NFL usually combines officials from various referee crews for the Super Bowl. As best I could, I've put in parentheses the last name of the head referee that the particular official worked with during the NFL regular season.

Referee: Mike Carey (Carey)
Umpire: Tony Michalek (Parry)
Head linesman: Gary Slaughter (Leavy)
Line judge: Carl Johnson (Austin)
Field judge: Boris Cheek (Green)
Side judge: Larry Rose (Green)
Back judge: Scott Helverson (Hochuli)

In a few days, I will try to take a look at the overall referee statistics for the various referee crews that the officials usually worked on. I will also try to take a closer look at Mike Carey's regular season statistics, in case the crew somehow functions similar to how his regular season crew did. (Maybe the head referee to some degree sets the tone for the rest of the crew?)

Monday, January 21, 2008

Why Not Give Back Time After Replay Official's Challenges?

In the Chargers-Patriots game and the Giants-Packers game, there were challenges by the replay booth assistants that seemed to cost the offense time.

In the Chargers-Patriots game, with 1:20 left in the 2nd quarter, Darren Sproles ran up the middle for 26 yards. The Chargers got up to the line and were about to snap the ball. The assistant challenged the play and they upheld the ruling on the field -- he did not fumble the ball. The ref started the clock and the Chargers lost time while doing a second pre-snap read. The play started 46 seconds after the previous play started.

In the Giants-Packers game, with 1:59 left in the 4th quarter, Eli Manning completed a short pass to Steve Smith for 14 yards. The Giants got up to the line and were about to snap the ball. The assistant challenged the play and they overturned the ruling on the field -- ruling instead that he did not make a first down. The ref started the clock and the Giants lost time while doing a second pre-snap read. The next play started 42 seconds after the previous play started.

In each case, it seemed that if the assistant had not called for a review, they would've snapped the ball in 1-2 seconds, not requiring the additional 10-15 seconds of pre-snap preparation. Do you agree?

Why can't the referees either give back the time that ran off during pre-snap preparation before the assistant buzzed down for a review -- or why can't the refs say that because the offense was about to snap the ball, they would not wind the clock for that particular play until the offense gets out of its huddle and is ready to snap the ball.

That way, the assistant still gets to wait to the very last second pre-snap to buzz down for a review without costing the offense an extra 10-15 seconds for having to do its pre-snap preparation twice.

The lost time seemed to put the Chargers in a more difficult spot (1st and 10 from the NE 31 with two timeouts and 34 seconds left rather than 45-50 seconds) and the Giants too (3rd and 1 from the GB 39 with no timeouts and 1:17 left rather than 1:30 or 1:35).

Super Bowl: Giants and Packers with Mike Carey

Super Bowl 2008: New York Giants against the New England Patriots. Several newspapers are reporting that Mike Carey will be the referee. They are taking the angle that this is the first time an African-American will be the referee in a Super Bowl. We're focused on statistics of on-field performance, as opposed to these other angles. Let's take a look at our initial view of some of Mike Carey's statistics from the 2007 NFL regular season.

In the playoffs, the NFL usually mixes together members of different referee crews, so Mike Carey's regular season statistics might not be that important for the Super Bowl. Also, both teams have a bye week to prepare so it's not the usual type of game preparation.

On top of all that, the game is played at a neutral site so it is hard to think that regular season statistics about a potential bias toward or against the home team would play out in the Super Bowl.

Some initial statistics about Mike Carey -- we will take another look at Mike Carey's statistics as we get closer to the Super Bowl and will try to analyze the other members of the Super Bowl referee crew when we find out who they are.

In the regular season, Mike was a good referee for visiting teams. Home teams only won 40% of their games with Mike's crew, which was second-worst of the 17 referee crews.

Mike was 6th (of 17) in total points scored per game (45.0), 4th in visitors' points (23.5) and 12th in home team points (21.5).

Mike was middle of the pack (9th) in penalties per game at 11.8. Ranked 10th in penalty yards per game (90 yards per game). Ranked near the bottom (14th) in average penalty yards per penalty at 7.6. Mike was tied for the highest in the percent of his games where the visiting team had more penalties called than the home team (70%). He is not the only referee where most games had more penalties called against the visiting team yet the visiting teams did well in his games. The same quality of usually more penalties called against the visiting teams yet the visiting teams nevertheless usually winning the games also applied to Ron Winter this year.

New England is more familiar with Mike Carey, having had him in their week 6 victory 48-27 over Dallas and their week 17 victory 38-35 over the Giants. The Giants only had him once, in their game with the Patriots.

So not only is Super Bowl 2008 going to be a rematch between the Giants and Patriots, it is also going to be a rematch with the same head referee -- Mike Carey. In the Giants-Patriots game in week 17, there were 5 penalties accepted for 42 yards against the Patriots and there were 5 penalties accepted for 53 yards against the Giants.

Check back later for more analysis of Mike Carey and -- we hope -- the others in the Super Bowl referee crew. Post comments if you have questions, ideas, or if you find out the other members of the Super Bowl crew.

Saturday, January 19, 2008

Giants Do Not Typically Win Through High-Yardage Penalties

In the footballoutsiders preview of the Giants-Packers game (, they predict that the Giants offense might get some free yardage because the Packers' secondary led the NFL in pass interference and illegal contact penalties, which gave 34 yards per game to the opposing offense. (I am not tracking those types of statistics, so I can't say whether that's correct.)

This led me to take a look at this season's statistics for the Packers and Giants to take a quick look at whether the teams' results depended to some degree on a large number of penalties or a large number of yards per penalty. A large number of penalties suggests the referee crew that game called lots of penalties, perhaps lots of defensive penalties as part of the mix. A large number of yards per penalty might be a sign of calling defensive pass interference (which can be a large number of yards), but might be a sign of other large-yardage penalties. Not a perfect way to analyze this, but the best I can do.

Looking at the 2007 regular season, the Giants did not typically win in games where there was a large penalty yards per penalty called. If we look at the Giants' games with the largest actual yards per penalty, the Giants went only 3-5. Meanwhile, in the eight games with the smallest actual yards per penalty, the Giants went 7-1.

Looking at the expected yards per penalty (using the referee's season-long statistic as opposed to the actual figure in the game) does not yield any trend -- the Giants went 5-3 in the eight largest and 5-3 in the eight smallest games. So it seems like the trend depends on the actual yards per penalty called in the particular game and is unrelated to the referees' season-long trends. You get equal 5-3 splits when you look at the actual number of penalties called per game and when you look at the expected penalties per game using the referees' regular season statistics.

Turning to the Packers: actual yards per penalty, 7-1 for largest 8 and 6-2 in smallest 8. expected yards per penalty, 6-2 in largest 8 and 7-1 in smallest 8. For actual penalties per game, 6-2 for largest 8 and 7-1 in smallest 8. For expected penalties per game, 7-1 in largest 8 and 6-2 in smallest 8. So for the Packers, no obvious trends.

So if past performance holds up on Sunday, the Giants typically do best when the average yards per penalty is less than 8 -- and is not typically characterized by a bunch of large yardage defense pass interference calls. Still, the Giants didn't lose all their games with large yards per penalty -- they did go 3-5 so it's not impossible, just not their typical type of win.

Post-game update: the initial box score of the Giants OT victory over the Packers says there were 6 penalties for 50 yards against the Giants and 7 penalties for 37 yards against the Packers, for an average yards per penalty of 6.69 yards per penalty. It was below 8 yards per penalty on average, which was typical of the Giants' victories this season. Of course, the Giants only barely won and the average yards per penalty is not exactly a statistic many people focus on so this probably was just a statistical oddity as opposed to anything deep...

Wednesday, January 16, 2008

NFL Conference Championships (i.e. the semifinals): Chargers at Patriots and Giants at Packers

Various bloggers are listing the likely referees for the upcoming NFL Conference Championships in the San Diego Chargers at New England Patriots game and the New York Giants at the Green Bay Packers.

As in other playoffs, the NFL is mixing up the crews. Let's take a look at the crews for each game. In parentheses is the crew that the official was part of during the regular season:

Chargers at Patriots:
R- Jeff Triplette (Triplette)
U- Butch Hannah (Anderson)
HL- Steve Stelljes (Triplette)
LJ- Gary Arthur (Carollo)
FJ- Tom Sifferman (Hochuli)
SJ- Greg Meyer (Triplette)
BJ- Greg Steed (Triplette)

So the Chargers game is pretty much a Triplette crew with just a few substitutes. Can we analyze it as if it will be similar to the usual Triplette crew or will the substitutes mess up our ability to predict how this crew will perform?

Giants at Packers:
R- Terry McAulay (McAulay)
U- Roy Ellison (Winter)
HL- Jim Mello (McAulay)
LJ- Jeff Bergman (Boger)
FJ- Scott Steenson (Boger)
SJ- Rick Patterson (Anderson)
BJ- Perry Paganelli (Parry)

Wow, only the head linesman is from McAulay's crew. This is really a mixture of lots of different crews -- the seven officials come from five different crews. We can analyze the regular season numbers of McAulay's regular season crew, but this bunch might really be different from how McAulay's crew did in the regular season. As if trying to predict the tendencies of referee crews weren't hard enough, now we have to factor in the mixture of the crews.

Let's look at the regular season statistics for the crews led by the same referee:

Chargers at Patriots: Jeff Triplette. Tied for 5th out of 17 for best ref for the home team's winning rate (67%). Near the bottom in penalties per game (14th) and yards per penalty (15th) so of course near the bottom in penalty yards per game (15th). 4th-most for percent of penalties called against the visiting team (55%). Middle of the pack (9th) in total points scored per game and near the top (4th) in home team points scored per game. Near the bottom (16th) in amount of penalty yards called against the home team at just 36.8 per game. Tied at the top in the percent of his games in which the visiting team had more penalties than the home team (70%). Overall, home teams did pretty well in Jeff's games during the regular season with not so many penalties called per game -- especially few called against the home team. Note: see a separate blog post for 2008 statistics. These are the 2007 statistics.

Giants at Packers: Terry McAulay. (Well, Terry has a mixed crew but let's look at this as if it were Terry's regular season crew so we can draw on our statistics. Why look at the real crew when the only stats we have are for the regular season crew, right?) Tops for the home team winning rate at 75%. This guy was great for home teams in the regular season! Strangely, the lowest total points scored per game of any of the refs at 35.5 per game. Second-lowest for visiting team points per game (15.9) and worst for home team points per game (19.6). For penalties, middle of the pack on number per game and a bit on the high side for average yards per penalty. Number one in calling the largest percentage of penalties and penalty yards against visiting teams (59% by penalties and 60% by penalty yards)! On average 18.9 more penalty yards against visiting teams per game than against home teams.

Let's look at the Giants-Packers crew for home team win rate, knowing it's a mixed crew. The referee and heads linesman were on the top crew for home teams in the regular season. The line judge, field judge, and back judge were on crews that tied for second-best for home teams. The side judge was on the 9th best out of 17 for home teams. Only the umpire worked on a pro-visiting team crew, which was 15th out of 17 for home team win rate in the regular season.

Overall, if you assume the referees control everything and that the players on the field had nothing to do with the results of any of the games (have I made it clear that this whole analysis requires a huge disclaimer?), then I'd say that the crew on the Giants-Packers game would be a very good one for the home team (Packers) and that the crew in the Chargers-Patriots game is pretty good for the home team (Patriots). If the Chargers fall behind early, I might suggest that they start committing penalties and hope that Jeff Triplette (somewhat near the bottom in penalties per game) lets both teams get away with a little holding here and there.

It's like knowing the strike zone for the particular umpire in baseball -- if the referee has certain tendencies, you should use that knowledge to your advantage during the game! Just make sure you aren't videotaping the other team's sideline. Or go ahead if it'll help you win, just pay a little money afterwards.

Sunday, January 13, 2008

NFL Quarterfinals: New York Giants-Dallas Cowboys and Pete Morelli

According to the blog hw's World of Sports (, the playoff game between the New York Giants and Dallas Cowboys today (January 13, 2008) in the NFL playoffs will be covered by referee Pete Morelli using the John Parry's crew.

If that's correct, then it's hard to know whether to apply the regular season statistics for Pete Morelli (today's referee) or for John Parry (the officials other than the referee)! Let's take a quick look at both, but focus on Pete Morelli because he is just so much more interesting!

Pete Morelli: second-least penalties per game (9.7), second-least in penalty yards per game (77), 16th of 17 in percent of penalty yards against the visiting team (50%), 16th in difference in visitor penalty yards greater than home penalty yards (0.1), 2nd in total points (47.7), 2nd in visitor points (25), 15th in visitor penalty yards per game (38.6), last in home penalties per game (4.4), last in home win rate (33%).

John Parry: 7th in penalties per game, 3rd in penalty yards per game (100), 4th in yards per penalty, 2nd in percent of penalty yards against the visiting team (58%), 2nd in difference in visitor penalty yards greater than home penalty yards (15.7), 7th in total points (43.8), tied for second with two others in home win rate (73%).

What to me is the most interesting about Pete Morelli is that he had by far the worst win rate for the home team in his games this season. Home teams only won 33% of the games he handled. The two next-lowest refs were up at 40%. Home teams lost seven games in a row in Pete's games (weeks 3-11) and then three in a row (weeks 14-16). The losing home teams were Houston, Carolina, Buffalo, Miami, Chicago, Oakland, Baltimore, San Francisco, Giants, and New Orleans. Very interesting. But it could be interesting because the NFL might pair up Pete Morelli with John Parry's crew and home teams did very well in John's games this year. Will this mix and mash effort somehow attempt to even off the skewed regular season statistics?

There are several areas where Pete Morelli and John Parry were on different extremes during the regular season. Pete was near the top of all of the following categories while John was near the bottom: fewest penalty yards per game, percent of penalty yards called against the home team, and penalty yardage differential called against the home team.

Only in the playoffs would a referee be torn away from his usual crew and transplanted with a new crew. So there are few examples to help us draw a conclusion about what might happen in the Giants-Cowboys game later today.

So, for the Giants fans, keep rooting for the Morelli Curse to strike the Cowboys as today's home team! And Cowboys fan should root like crazy that the Morelli Curse was actually embodied by Pete's crew and not Pete himself so that the Cowboys can escape the curse.

Maybe the location of the Morelli Curse will have more of an impact on today's game than the location of Jessica Simpson?

Saturday, January 12, 2008

Jaguars-Patriots NFL Playoffs: Referee Jerome Boger

THe Jacksonville Jaguars visit the New England Patriots tonight in the NFL playoffs and it looks like Jerome Boger is the referee. Let's take a quick look at Jerome's statistics during the season.

Remember, playoff referee crews are all mixed up from various groups that did not work together during the year. Jerome's statistics were racked up with a specific set of referees he worked with all season. So the regular season statistics might not be so important in the playoffs. Enough of these explanations already --

Jerome Boger had the fourth-most accepted penalties per game out of the 17 crews this year at 12.5 (average was around 11.5). Middle of the pack on total penalty yards. Somewhat near the bottom in average yards per penalty at 7.6 (13th out of 17).

Third-most percent penalties called against the visiting teams at 56%. Also third-most in percent of the penalty yards being called against the visiting teams at 56% too. Not a surprise second-most in the raw difference in penalties (1.4) called on the visiting teams and 4th-most in raw different in penalty yards (10.8) called on the visiting team.

His games were pretty low-scoring, ranking second-lowest at 37.7 total points per game with the lowest visiting team scoring per game at 15.7 points per game.

Home teams did well in Jerome's games this year -- winning 73% of the games, tying him for second-best for home teams with two other referees. That certainly would be a great sign for the Patriots if only you ignored reality that the quality of the teams probably has a great deal more to do with the odds a team will win than whether a referee might have a tendency to have the home team win more often!

Post your comments on the refereeing in the game and any questions on Jerome's referee statistics as a comment to this posting.

Saturday, January 05, 2008

Jacksonville Jaguars and Pittsburgh Steelers Wild Card Game: Scott Green

Let's take a quick look at the wild card playoff game between the Jacksonville Jaguars and the Pittsburgh Steelers. The referee is Scott Green. Let's take a look at Scott Green in the 2007 NFL season.

But first, keep in mind that he is not working with his usual crew -- so statistics about his crew in the regular season don't seem to be as valid for the playoffs unless you imagine that the ref sets the tone for the rest of the crew.

And of course keep in mind that we are looking at statistics that only make sense if they to some degree depend on variations among the different ref crews.

Back to looking at Scott Green:
Home teams won 53% of the time (middle of the pack)
Total penalties were 10.7 per game, a little below average
Average yards per penalty was 8.6, which was second-highest and quite above average.
Total points per game was 45.5, which was fourth-highest so maybe it could be higher scoring than you'd ordinarily expect.
By the way, visiting teams averaged 24.1 points and home teams averaged 21.5 points.

Post comments if you have questions or comments on the game!

NFL Teams 2007 compared with referees in their games

What if we pretend that the main "opponent" that a team faces in a game is the referee and whether that referee tends to favor the home or visiting team?

In that imaginary situation, we can check whether a particular team had a ref that favored visiting teams when it was the visiting team and whether it had a ref that favored the home teams in its home games.

We can then add it up across the season and see which teams had the easiest or hardest referee schedule (assuming that referees consistently tend to favor either the home or visiting teams) and we can see how much better each team did against the projected wins!

Let's start with the teams with the easiest and hardest "referee schedule":

Easiest schedules:
Seattle: 9.21 projected wins, 10 actual wins
Indianapolis: 9.17 projected wins, 13 actual wins
Tennessee: 8.89 projected, 10 actual
Minnesota: 8.72 projected, 8 actual
Detroit: 8.72 projected, 7 actual
Dallas: 8.60 projected, 13 actual

Hardest schedules:
St Louis: 6.64 projected, 3 actual
Chicago: 7.04 projected, 7 actual
Atlanta: 7.15 projected, 4 actual
New Orleans: 7.30 projected, 7 actual
San Diego: 7.33 projected, 11 actual (wow, San Diego had a hard schedule yet still made the playoffs)

Here's the list of the teams that did the best over the projected wins, relying only on the referee schedule (which of course is an artifical scenario):
New England: 7.82 (16-8.18)
Green Bay: 5.02 (13-7.98)
Dallas: 4.40 (13-8.60)
Indianapolis: 3.83 (13-9.17)
San Diego: 3.67 (11-7.33)
Jacksonville: 3.20 (11-7.80)
Cleveland: 1.89 (10-8.11)
New York Giants: 1.87 (10-8.13)

at the bottom:
Miami: -6.65 (1-7.65)
Kansas City: -4.01 (4-8.01)
Oakland: -3.85 (4-7.85)
St Louis: -3.64 (3-6.64)
New York Jets: -3.42 (4-7.42)
Atlanta: -3.15 (4-7.15)

Let's use these limited statistics to check out Washington at Seattle in the wild card game:
Washington: 0.46 adjusted wins (9-8.54) versus Seattle: 0.79 adjusted wins (10-9.21)
If perhaps Walt Coleman is the referee (47% home win rate), then you tack in the adjusted wins, that would mean Seattle would have a 49% chance of winning, ignoring such critical factors as how good their players are and how tough their opponents are.

Clearly, there are tons of limitations to this analysis!

NFL Referee Tendencies Across the 2006-2007 Seasons

Let's look at correlations between the distribution of statistics among the 2006 NFL referees and the 2007 NFL referees. This is a rough attempt at figuring out which particular characteristics seem to hold the following season. For example, if referees A, B, C, and D called lots of penalties in 2006 and referees E and F didn't, let's see whether the same kind of distribution exists in 2007 -- whether A, B, C, and D called lots of penalties and E and F didn't in 2007 too.

Running the numbers and checking for a correlation, some seem to hold up across the 2006 and 2007 seasons. In order of the correlation coefficient:
Average yards per penalty (0.595)
Penalties against home teams (0.489)
Penalty yards against home teams (0.417)
Total penalty yards (0.394)
Total penalties (0.392)
Penalty yards against visiting teams (0.332)
Percent penalty yards against visiting team (0.240)
Difference in yards against visiting teams (0.236)

Let's look at those with the worst correlation coefficients across two seasons:
Home team win rate (-0.172)
Points scored by home teams (-0.163)
Difference in number of penalties against visiting teams (-0.146)
Percent of penalties called against visiting teams (-0.120)

NFL 2007 Referee Statistics (including all regular season games)

Here are the final NFL referee statistics for the 2007 regular season, including all regular season games:

Possible relationship between the number of penalties called and the rate the home team wins for a particular referee. The strong correlation between the number of penalties per game and the rate the home team wins as usual came out with a positive correlation -- it decreased a bit toward the end of the year but ended up with 0.216.

Referees by penalties accepted per game (followed by home win rate):
Ron had a huge lead and kept the top spot to be the 2007 leader in penalties per game. Notice many of the referees who call lots of penalties also have a high win rate for home teams.
Ron Winter (14.3, 40%)
Tony Corrente (13.3, 73%)
Walt Anderson (12.7, 60%)
Jerome Boger (12.5, 73%)

and at the bottom, Gerald locked up the title and kept it to the end:
Jeff Triplette (10.6, 67%)
Bill Carollo (9.8, 63%)
Pete Morelli (9.7, 33%)
Gerald Austin (8.9, 50%)

Referees by home win-rate: if you believe the referee has some influence (perhaps an unintentional tendency) on whether the home team wins, then use this list. Here's the list, again with the penalties per game followed by the home team's win-rate. Terry McAulay kept his lead and home teams did the best in games where Terry was the referee:
Terry McAulay (12.1, 75%)
Tony Corrente (13.3, 73%)
Jerome Boger (12.5, 73%)
John Parry (12.1, 73%)

And at the bottom, Green Bay broke through the Peter Morelli curse by winning as a home team one of his games. The home teams went 5-10 in Peter's games this year, winning the first two, losing the next seven, winning two, losing three, and winning the last one.
Bill Leavy (12.1, 45%)
Ron Winter (14.3, 40%)
Mike Carey (11.8, 40%)
Pete Morelli (9.7, 33%)

So, if you like the home team, hope for Terry McAulay, Tony Corrente, Jerome Boger, and John Parry. If you like the visiting team, hope for Pete Morelli but also good are Bill Leavy, Ron Winter, and Mike Carey. Of course, this is only if you think particular referee crews are better for home teams than others.

And here is some end-of-year extra data for you -- in order are the following statistics for each referee -- penalties accepted per game, penalty yards per game, average yards per penalty, percent penalties called against visiting team, total points scored, visiting team points, home team points, home team win rate. Post a comment if you request other types of data or want any particular theory investigated or calculations run!
Anderson, Walt (12.7, 97, 7.6, 50%, 41.7, 20.3, 21.3, 60%)
Austin, Gerald (8.9, 80, 9.0, 50%, 45.4, 19.8, 25.6, 50%)
Boger, Jerome (12.5, 95, 7.6, 56%, 37.7, 15.7, 22.1, 73%)
Carey, Mike (11.8, 90, 7.6, 54%, 45.0, 23.5, 21.5, 40%)
Carollo, Bill (9.8, 72, 7.4, 54%, 40.2, 17.8, 22.4, 63%)
Coleman, Walt (10.7, 81, 7.6, 51%, 43.3, 22.3, 21.1, 47%)
Corrente, Tony (13.3, 105, 7.9, 53%, 43.1, 18.5, 24.5, 73%)
Green, Scott (10.7, 92, 8.6, 53%, 45.5, 24.1, 21.5, 53%)
Hochuli, Ed (12.4, 95, 7.7, 53%, 46.1, 19.9, 26.3, 63%)
Leavy, Bill (12.1, 94, 7.7, 56%, 41.5, 21.0, 20.5, 45%)
McAulay, Terry (12.1, 97, 8.1, 59%, 35.5, 15.9, 19.6, 75%)
Morelli, Pete (9.7, 77, 7.9, 55%, 47.7, 25.0, 22.7, 33%)
Nemmers, Larry (10.8, 90, 8.4, 51%, 43.7, 17.3, 26.4, 67%)
Parry, John (12.1, 100, 8.3, 51%, 43.8, 18.7, 25.1, 73%)
Steratore, Gene (10.7, 82, 7.6, 48%, 50.6, 25.3, 25.3, 50%)
Triplette, Jeff (10.6, 79, 7.4, 55%, 43.5, 18.2, 25.3, 67%)
Winter, Ron (14.3, 106, 7.4, 53%, 42.5, 21.3, 21.1, 40%)