Friday, December 29, 2006

The Real Conference Record

Hypothetical: one wildcard "bubble" club is in a tough position because of its bad record in a tough division, despite its strong record in inter-divisional play. Another bubble club is in a tough position because, while it has a good record in a weak division, it's pathetic in inter-divisonal play. They never played each other, and overall conference record will determine which gets the wildcard.

Is this fair? Should being good in a bad division translate into being good in the conference? I think one way to control for division disparities in the conference is to judge a team's conference performance not by the twelve conference games (six division - 2 each against three team, six inter-division - one each against six teams) played, but by the nine conference teams (three division, six inter-division) that the team played.

I think the inter-divisional record should play a more significant role in the conference record tie-breaker. I think the inter-division games should be weighted heavier, perhaps two-thirds, in evaluating the conference record, and the division games one-third.

This would apply only to the conference tie-breaker, and not to division standings or overall standings.

So Many Comebacks

In tonight's Insight Bowl, Texas Tech came back from 31 down to beat Minnesota 44-41 in overtime. This is the biggest comeback ever in a bowl game, breaking the 30-point deficit record set in 2001, which broke the 25-point deficit record set in 2000. Why big comebacks increasingly common? Five reasons:

1. More and more bowl games, leading to
2. More and more average to bad teams getting into Bowl games,
3. Average to bad teams tend to be so because of their defenses,
4. Offenses are better, and the game is more offense-oriented.
5. Overtime - it's only been in existence since 1996. Games like tonight, in which Tech tied the game at the end of regulation on a field goal, would have ended in a tie.

Friday, December 22, 2006

Favre's Thought Process

I think Brett Favre was torn last spring:

- The Packers were in rebuilding mode. If Favre was to be part of that rebuilding, he should commit to three more years .
- but if the young team totally sucked - not just because they're inexperienced but because they're bad, Brett Favre wouldn't want to commit to more years of losing. He wanted the option of quitting if it was another bad, 4-12 season.
- On the other hand, if the team showed signs of improvement, so as to make play-off/championship runs the following two years, Favre would want to keep playing, for two reasons: a) it would be fun again, and b) it would be unfair to the other players if, just when they started getting good, they had to go through another "rebuilding" year adjusting to a new young quarterback and his inevitable growing pains.

I believe this is was Favre's dilemma last spring. I believe early on this year, and in the mid-season slump, there was reason for Favre to doubt whether he would come back. If he lost confidence in the organization and saw nothing but more years of losing, why should he?

But now with the Packers still in the play-off hunt and at worse a 7-9 team, Favre has an obligation to come back and see this thing through. This was a rebuilding year, and it was a successful rebuilding year. If Favre quits now, he will have set the program back by a year, as the Packers could have installed a young quarterback and grown with him.

So I predict that this week, or immediately following the last game of the year, that Favre will announce that he'll be back.

Wednesday, December 20, 2006

How Brett Favre is Having a Good Year

Brett Favre came off his worst year. He leads the league in passing attempts, but at 56% is well below his career completion percentage. 15 INT's is pretty good considering he puts the ball up 37 times per game, but he's barely at six yards per pass, has just 17 touchdowns, and a rating of 74.8 - which is 25th among the top 30 leaders in passing yards.

Plus, the team sucks overall. It got whipped by all the good teams it faced, and lost to some pretty bad teams as well. Ahman Green isn't the same as he was 2000-2003, and the team is just 19th in rushing. Favre has just one consistently reliable receiver, Donald Driver.

So why do I say Favre is having a good year? Through it all, the Packers are still just a game out from the play-offs. Yes, it's a long-shot, but they're in the strongest position of all the 6-8 teams. And while Favre leads the league in passing attempts, the team is sixth in fewest sack yards allowed, and fifth in fewest sacks allowed - even though the team starts three rookie offensive linemen.

The man refuses to be sacked. He's thrown a lot of balls away to prevent the sack. A lot.

That's the sign of a good quarterback.

Some Rankings

There are twelve play–off teams, six in each conference. Not all positions have been filled, but I am counting as a “play-off team” the teams that would go to the play-offs if they started today, based on the conference standings.

Of the top twelve teams in points per game, ten are play-off teams.

Of the top twelve teams in rushing yards, just five are play-off teams, though all but one are in the play-off hunt.

Of the top twelve teams in rushing attempts, seven are play-off teams, and just one has a losing record.

Of the top twelve teams in passing yards, just six are play-off teams, though all of the top five are.

Of the top twelve teams in passing attempts, just four are play-off teams, and just one of the top five.

Of the top twelve teams in turnover margin, eight are play-off teams are all are technically in the hunt.

Of the top twelve teams in 3rd down percentage, eight are play-off teams and the rest are at .500 or better and in the hunt.

Of the twelve least-penalized teams, just five are in the play-offs.

Of the five least penalized in yards, just five are in the play-offs.

Of the top twelve in time of possession, seven are in the play-offs.

Of the top twelve teams in first downs, nine are in the play-offs.

Of the top twelve teams in yards/play, seven are in the play-offs and four more in the hunt.

Tuesday, December 19, 2006

Is Yards Per Pass Play THE Stat?

Here are some more Week 15 numbers:

The team with the greatest average gain per offensive play went 12-4.
The team with more yards rushing went 11-4-1.
The team with more Yards Per Pass Play (including sacks) went 14-2. On this stat, the Raiders were better than the Rams, but were -5 on turnovers and lost. And the Jags were better than the Titans, but were -4 on turnovers, and allowed three returns for touchdowns.

I'm beginning to suspect that Yards Per Pass Play is THE stat, the best offensive predictor of victory. This would suggest that the most important positions on the offense are the quarterback and the two tackles, who block the pass rushers.

Monday, December 18, 2006

Third Down Efficiency

(Updated Tuesday, including Monday night's game)

In the Week 15 games the teams with a superior 3rd Down conversion rate went 11-5.

But when considering the total number of unconverted third downs, the team with the fewest went 12-3. Example: the Bears went 5 of 16 on third downs, and the Bucs were a slightly more efficient 6 of 18. But that translated to eleven unconverted third downs for the victorious Bears, and 12 for the losing Bucs.

The most dramatic example of this was the Titans-Jacksonville game. The Titans were a pathetic 0-8 on third downs; the Jags a remarkable 10-19. But the losing Jags had one more unconverted third down.

I'm not suggesting that this was the deciding factor of the game - three turnover returns for touchdowns is what gave the Titans the victory. But one thing leads to another.

The Rams shut out the Raiders despite being worse on third downs in both categories, but had a +5 turnover advantage. Indeed, the six teams that were -2 or worse in turnover margins were all losers, and they averaged just 10.5 points per game.

The team that won the turnover battle went 11-3. The three that lost the turnover battle but won the game were each only -1 in turnovers. Significantly, their third-down conversions were 7-15, 7-15, and 7-14; their opponents were 0-11, 2-12, and 2-13. Blowing too many third downs probably evens the turnovers out. Not to mention the fact that the three winners: the Ravens, Packers, and Jets, were clearly superior to their opponents: Browns, Lions, and Vikings.

The overall lesson? Don't get into third down situations, but if you do, convert them. And don't turn the ball over.

Saturday, December 16, 2006

On Field Position

How important is field position? I looked at week 14. The numbers are rough (they don't add up exactly), but close:

In the 16 NFL games last week, there were 358 drives, leading to

-74 TD's
-42 FG's
-49 turnovers
-15 turover on downs downs
-11 missed FG's
-132 punts

-12 drives ending at halftime
-14 drives ending at the end of the game (I may have miscounted half-ending and game-ending drives).


Of these, 38 drives began across the 50 - that is,in the opponent's territory. Three were game-enders by the winning team. Of the remaining 35:

- 26 ended in scores
- 4 turnovers
- 3 punts
- 1 turnover on downs
- 1 missed FG

Signifcantly, teams that began more drives in their opponent's territory went 12-1. No wonder: teams that begin drives inside their opponent's territory score at a rate of 75%, as opposed to 33% overall.

One interesting statistic is the number of plays that originated at or inside a team's own 20. There were about 117, of which 37 ended in scores. This is close to the overall scoring rate. The 15 turnovers from drives beginning in this territory is also close to the overall turnover rate.

This does not mean, however, that it makes no difference where a drive starts. Teams that found themselves at or inside their own 20 more often than their opponent, won just three games and lost nine. Since most drives end in a punt or turnover, you don't want your drive to stall in the shadow of your own goalposts. That just makes it easier for the other team to score. If you have to punt, better to punt at your 40 instead of your 20.

Although it is worse to begin near one's own end zone, there was one interesting, though not surprising, statistic. The team that scored more points in drives beginning at or inside their own 20 than their opponent did, went 10-4.

This is not surprising. The team that is able to put together a long drive ending in a touchdown, when the opponent is unable to put together a long drive, is obviously executing better that day.

  • As we learned from week 13, teams that punt more, lose more games. And, the team with more turnovers loses almost all the time. Each time a team has a long drive ending in points, the likelihood of losing the turnover and punt battles are reduced.
  • A long drive indicates that the running game is clicking. Teams that rush for more yards usually win.
  • A long drive also indicates that the passing game is efficient. The team with more efficient passing (Yards Per Pass Play) usually wins.
  • A long drive exhausts and demoralizes the opponent while it gives the team's own defense a longer rest.
Conclusion: Turnovers and field position are the most important factors in an NFLgame, but the team that can overcome bad field position with long, sustained drives ending in points is probably playing better that day and will be the likely winner.

At some point, I will look at more games from more weeks to see if the stats I looked at in weeks 13 and 14 were normal or aberrant

Thursday, December 14, 2006

Great Teams Are Good Teams With Great Players

Michael Lewis's Moneyball answers the question of how the Oakland A's win a lot of games every year despite a low payroll.

But they haven't gone to the World Series.

I think it's because the A's acquire players that give them statistical advantages, particularly against average and below-average teams. Although I don't follow baseball closely, I could theorize why the A's fail in the play-offs.

1. GM Billy Beane frowns on steals and sacrifices, believing that a team built on "small ball" won't go far.The focus is on on-base and slugging. But in the play-offs, where the pitching is superior, the "little things" are the difference betweeen victory and defeat.
2. Because of the A's low payroll, the A's can't afford great players, and great teams are good teams with great players.

Because of revenue sharing and the salary cap, NFL teams don't have the obstacles that teams like the A's face. Many teams over the years have, like the A's, sought to play the percentages. A disciplined, conservative team that doesn't turn the ball over or commit penalties will win a lot of games - and make quick exits in the play-offs. Ask Chuck Knox. But teams can still be conservative on offense and win the Super Bowls, but only if they have an aggressive defense. And a defense can be aggressive only if it has great players.

If you make predictions going into the play-offs, favor teams with great defenses. That doesn't mean favor teams that lead in defensive statistics, but rather teams whose defenses scare the heck out of you. They will likely be defenses that are
a) solid all-around,
b) with great individual players.

If when picking a play-off game, both defenses scare the heck out of you (or if neither defense scares the heck out of you), only then should you look to the offenses. And, again, don't favor the team whose offense leads statistically, but rather the offense that scares the heck out of you.

A "scary" offense doesn't need a lot of pizzazz. It can be quite conservative. Truly, the scariest offense is one that
a) runs the ball well, and
b) is most capable of converting any third down between one and fifteen yards.

In other words, it has a great qb-rb-wr combination.

This advice applies only to the play-offs. Picking games between successful teams on Any Given Sunday in the regular season is more of a roll of the dice.

Wednesday, December 13, 2006

The Most Valuable Player, the Outstanding Player

How does one decide on a Most Valuable Player award? I would consider the following factors, in order of importance:

1. The Indispensible Man: based on performace, where would the team be without this player?
2. The Best Man: if you were asked to select a first-place team from scratch, then based on this season's performances only (disregarding age, potential, previous performance, or reputation), who would be the first player you'd pick?
3. The Winner: Was the team in fact successful? Was it successful without the presence of other great players?
4.. Consistency: is he often guilty of bone-headed or sloppy play? Has he cost games?
5. The Playmaker: does he come through in the clutch?
6. Leadership: does his presence improve team morale?
7. The Stat Man: is he a leader in meaningful statistics?
8. Hype: Effect on ticket sales, ratings, etc.

Obviously, factors 3-8 play into the consideration of both 1 and 2. I consider them things to consider in the case of close calls.

Now, for an Outstanding Player award, like the Heisman Trophy, I would reverse factors 1 and 2 in order of importance, but otherwise consider the same criteria in the same order.

This year LaDanian Tomlinson's season was tragically cut short by an off-field accident tomorrow, I'd still probably give him the MVP this year, and an Outstanding Player award if there is one. It wasn't just setting the touchdown record in just 13 games. He comes out on top in nearly all categories, and in consideration of indispensiblity and play-making:

Week 8: Chargers 38, St. Louis 24; Tomlinson with 3 TD's.
Week 9: Chargers 32, Cleveland 25: Tomlinson with 3 TD's.
Week 10: Charger 49, Bengals 41: Tomlinson with 4 TD's.
Week 11: Chargers 35, Broncos 27: Tomlinson with 4 TD's.
In the last two games, the Chargers became the first team to come back from 17 or more points down and win two weeks in a row. And over the course of the four games, the Chargers became the first team to win four straight while giving up 24 or more points in each game.
Week 12: Chargers 21, Raiders 14: Tomlinson with 2 TD plus one TD passing.
Week 13: Chargers 24, Bills 21: Tomlinson with 2 TD's.

And taking away these six games, Tomlinson still has 11 touchdowns in the other games, where the Chargers are 5-2. Tomlinson also has over 1800 yards rushing and receiving. The Chargers are 11-2 with a first-year starter at quarterback. Tomlinson's presence obviously makes Philip Rivers a more effective quarterback, and the team much better, than they otherwise would have been with an average starter.

Saturday, December 09, 2006

More On Week 13

On Games Where the Margin of Victory Was Two Or More Possessions (9+points)


All 8 teams that lost by 2+ possessions had losing records entering the game.

In seven of these games, the losing team had more turnovers (the 8th was even), and in six the loser had two or more.


In just four of those games did the winner have more total yards.

HOWEVER, in the three games decided by three or more possessions (17+ points), the victor had more yards


In just four of these games did the winner have a longer Time of Possession, but they did in all three three-possession victories. However, only just one (a three-possession winner) have more total plays.


In six (except Houston and Chicago), the winner had a superior Yards Per Pass Play. (5-3 in closer games) Also in six games, the winner had more rushing yards (just 4-4 in closer games)


More on Turnovers


Of the six losing teams (in all games) with more rushing yards, five had more turnovers (the sixth, Min-Chi, was even)

Of the five losing teams with superior Yard Per Pass Play, two had more turnovers and three were even

Of the eight times they loser had more total yards, six times they had more turnovers (other two were even).

Of the ten time the loser had more first downs, eight times they had more turnovers

All seven times when the loser had fewer penalties, they had more turnovers.

Likewise all seven times when they had fewer penalty yards.

And the three losing teams that punted less.

And the four teams that missed fewer field goals.


Turnovers certain negate almost all other advantages.

Wednesday, December 06, 2006

Some Week 13 Information

I've been curious about some football stats, and not knowing where to look them up (and not willing to pay for them), I decided to find them out by myself. Curiously, my sources didn't seem to provide the home team, or the time of possession statistic. I'm sure that's available, but I don't have time right now to look it up.

In any case, here are some things I wanted to find out, and the results of week 13 in the National Football League. First, some statistics that are thought to help a team:

- The team that scored first won ten of the sixteen games.
- Twelve times a two-score lead was taken in a game. Nine times, the lead held, whereas three teams came back to win.
- The team with the most first downs won just five games and lost ten.
- The team with the most total yards went 8-8.
- The team with the most rushing yards, however, went 10-6.
- I assumed that most passing yards could easily favor the losing team, who often throw more when trying to come from behind so I didn't count it. Instead, I look for passing efficiency, and the team with more yards per pass play went 11-5.
- The team with the most field goal attempts, which I think suggests superior control of field position, went 8-4.

As to some "bad" stats:

- The team that was sacked more went just 3-8.
- The team that punted more went 3-6.
- The team with the most penalties went 7-8.
- The team with the most penalty yards went 7-9.
- The team that missed more field goals went 4-5.
- Now here's the most telling statistic: the team with the most turnovers went 0-13.

I knew that turnovers were bad, but I didn't think having more turnovers would be worse than getting behind by two scores early. As to the three teams that came back from 10 or 14 points, one was even on turnovers (Titans over Colts), the Falcons had a +2 turnover margin over the Redskins, and the Seahawks were +4 over the Broncos.

0-13 in turnovers may be flukey, but it is clear that turnover margin is the most important statistic in football.

Saturday, December 02, 2006

Wake Forest wins the ACC?

When Miami, Virginia Tech, and Boston College bolted the Big East for the Atlantic Coast Conference, it looked like the ACC would get more competitive. That is, Florida State would have to lose the title sometimes to either Miami or Va Tech. A Big Three, Little Nine situation.

Now look what happens. Two years in, and its Wake Forest and Georgia Tech in the championship game, with Wake Forest winning! Apparently, the additional team created parity throughout the conference.

I'm reminded of when Penn St. joined the Big Ten in 1993. It was assumed that they would battle it out with Michigan and Ohio State every year. But within the next ten seasons, Wisconsin and Northwestern each won or shared the Big 10 Title three times, whereas the "Big Three" each went to the Rose Bowl just once (though they may have shared the title in other years), as did Purdue, Illinois and Iowa. Only Minnesota, Michigan St. and Indiana have missed out on a championship over the past 14 years.

Perhaps conference expansion can be good for the "weaker" teams in the conference.