Pro Football Reference Blog

Internet message boards, twitter feeds and sports media are blowing up over Bill Belichick's decision to go for it on 4th and 2 from his own 28-yard line, up by six points, with 2:08 remaining. First, a quick review of the play.

The Pats came out with three WRs to Brady's left, with Wes Welker the nearest receiver to Brady on that side of the field. On Brady's right was Randy Moss isolated out wide, with Kevin Faulk in the backfield. Indy came out looking like they were going to blitz six -- they had the four WRs in tight, man-coverage, and safety Melvin Bullitt about fifteen yards deep to Brady's right. Brady then sent #33, Kevin Faulk, to go line up as the inside WR on the right side, and #33 (Bullitt) for the Colts came in to line up against him. Indy was now going to rush six against NE's five, while NE knew all five of their WRs were in single coverage. Brady recognized that he was going to have to make a quick and accurate pass.

It turns out that Faulk was the primary read all along, as he took twp steps, did a quick fake left, and then curled right just a yard or so past the first-down marker. Since the drive started after a touchback, the Pats needed to get to exactly the 30-yard line for the first down. The ball hit Faulk's hands but he bobbled the ball; Bullitt pushed Faulk backwards, and by the time he landed with possession of the ball, he was on New England's side of the 30-yard line.

I had a feeling about this at the time, and Brian Burke of Advanced NFL Stats only confirms my initial belief: if you crunch the numbers, Bill Belichick made the correct decision when he decided to go for it on 4th-and-2 at his own 28 with 2:08 to go in the game vs. Indy last night...

"With 2:08 left and the Colts with only one timeout, a successful 4th-and-2 conversion wins the game for all practical purposes. A conversion on 4th-and-2 would be successful 60 percent of the time. Historically, in a situation with 2:00 left and needing a TD to either win or tie, teams get the TD 53 percent of the time from that field position. The total win probability for the 4th-down conversion attempt would therefore be:

(0.60 * 1) + (0.40 * (1-0.53)) = 0.79 WP (WP stands for win probability)

A punt from the 28 typically nets 38 yards, starting the Colts at their 34. Teams historically get the TD 30 percent of the time in that situation. So the punt gives the Pats about a 0.70 WP."

And if you consider that Manning is an assassin from anywhere on the field, it will actually make it look like a better decision by Belichick. As Brian says, "you can play with the numbers any way you like, but it's pretty hard to come up with a realistic combination of numbers that make punting the better option. At best, you could make it a wash."

Now, I know Rodney Harrison and Trent Dilfer were ripping Belichick pretty hard on TV, but that's only because hindsight is 20/20. Had the conversion succeeded, they would have praised the coach for his "guts" and "decisive leadership". It's the great thing about being able to analyze decisions like this after the fact -- you can tailor your analysis to fit the result, rather than what the right choice was at the time. Unfortunately for Belichick, he didn't have that privilege. But maybe he can take some solace in knowing that, statistically, he made the right call, talking heads be damned.

EDIT: Here's an Excel sheet where you can enter the probabilities you think were correct last night, and see how often the Patriots would win. "1stDprob" is the probability of getting the 1st down on 4th-and-2; "1stdwp" is the Pats' win probability if they convert on 4th-and-2; "failed1stdcltwp" is the Colts' win probability if the Pats' 4th-and-2 fails; "puntprob" is the probability of the Pats getting off a punt; and "coltsdrvwp" is the Colts' win probability if Belichick punts and asks them to drive down the field. As you can see, even if you believed that Indy would win 70% of the time if you failed on 4th-and-2 and only 30% if you punted, it still comes out a wash... There's no doubt Belichick made the right decision.

With just one week left in the college football regular season, the college football rankings should be more precise than ever this week. And, as members of the national media are becoming more apt to notice, it appears we've got a Big 4, not a Big 3. With a blowout victory over rival Utah last night, TCU now is within half a point of Florida and Alabama, while the Horned Frogs' in-state rival now have a field-goal lead over all schools.

A quick review of the SRS system. All games are included, even games against FCS opponents. The SRS is simply the sum of a team's margin of victory and its strength of schedule. Strength of schedule is easily calculated; it's the average SRS rating of all of a team's opponents (iterated hundreds of times to ensure that we get "true" strength of schedule ratings). Margin of victory could be easy to calculate, except I've decided to adjust the typical MOV in order to give us more accurate results. In addition to giving each road team three points, there's a cap and a floor for each game. Wins of fewer than 7 points (after including HFA) are counted 7-point wins; wins of greater than 24 points are counted as the average of 24 points and the actual margin of victory (after including HFA). The only exception is FCS games are uncapped.

Courtesy of Matt Knobbe and the Tecmo Super Bowl Repository, here's your Tecmo Super Bowl Game of the Week for Week 10, featuring the New England Patriots and the Indianapolis Colts. The highlights:

(How did we do this? Matt and the other dedicated folks at the Knobbe.org message board have spent a lot of time over the years updating this classic Nintendo football game, including the introduction of a 32-team ROM a few seasons ago. Sounds complicated, but don't worry, it's easy for you to enjoy the fruits of their labor: just get yourself an NES emulator, download the 2009 version of Tecmo here, and play to your heart's content. And be sure to check back at Matt's site for roster updates and more Tecmo-related goodness all season long.)

You may have missed it, but Ahman Green re-signed with the Packers earlier this season, and last week he surpassed Jim Taylor as the Green Bay Packers' all-time rushing yardage leader. Very few team rushing records have stood the test of time with the expanded schedules and increased scoring. In fact, now that Green has surpassed Taylor, only two teams have an all-time rushing yardage leader who pre-dated the AFL-NFL merger. While Green benefited from the increased offensive emphasis of the modern game, he actually played in twenty fewer games than Taylor to reach the milestone. The last few yards may have come quietly as a backup for the Packers. Every yard counts the same in the end, though, and gaining over 8,000 yards, particularly with a storied franchise like Green Bay, is a notable milestone. Congratulations, Ahman Green.

I'm sure that every player wants his milestone records to stay in place forever. Sometimes, though, the previous record holder may be more remembered by the surpassing of his record. Jim Taylor is well known to fans who grew up in the 1950's and 1960's. I wonder how many modern fans under the age of 40, though, know about him. I wrote about Taylor recently when I compared his difficult schedule to his contemporary Jim Brown. Brown is one of the two players who still holds the all-time franchise rushing mark and never played after the merger. His 12,312 rushing yards for the Cleveland Browns is probably safe for a long time to come. The longest standing franchise rushing record, though, may be in some jeopardy soon.

College football polls don't offer much to see behind the curtain: not only are there no official guidelines for voters to use when submitting their ballots, but there are no explanations given for their ballots, just a list of their rankings. While every week I publish the NCAA SRS ratings, they're clearly designed to be a predictive measure of what will happen in the week(s) to come. In no way could you justify ranking Oklahoma ahead of Iowa in a real poll, even if you (and Vegas) think OU would beat the Hawkeyes if they met head to head.

Now that we're ten full weeks into the season, I'm ready to release my first set of real rankings based on what each team has accomplished. Strength of schedule and margin of victory still play a role in these rankings, but for the most part, "a win is a win." If there is one guiding principle behind why each team is slotted where it is, it is this: each team has accomplished slightly more this season than the team behind them, and has accomplished slightly less than the team behind them.

Generally, my formula was this. Sort the teams by losses, from fewest to most. Within each loss-group, sort the teams by SRS rating, but recognize that that SOS matters more than MOV (so a team with an SRS of 20 points above average based on a +15.0 MOV score and a +5.0 SOS score might be behind a team with the same number of losses and has a +9.0 in both the SOS and MOV categories). This is because beating good teams by a little bit is more of an accomplishment for a team than blowing out a bad team, even if barely beating a good team does not make you a better team than devouring a cupcake. Further, in cases where teams with X number of losses have a tough SOS, move them past a team with X-1 losses if they have an easy SOS. Finally, give a bump to the strong defensive teams, as defensive teams are generally underrated in the SRS because a 30-17 win is treated the same as a 13-0 win.

The Results:

1. Texas - looks to be the most complete team in the nation. Elite on both sides of the ball. Had the least frightening "toughest scare" of the big three (Alabama barely beat Tennessee, Florida barely beat Arkansas, while Texas hasn't trailed in the 4th quarter all year and were never seriously threatened in their toughest game against the Sooners).

2. Alabama - the hardest SOS of the undefeated teams, so you could easily put them at #1. Slightly more impressed with Texas' much better MOV and lack of close games.

3. Florida - top notch D, great special teams, and the offense is rounding into form. If not for the Arkansas game, the biggest blemish on the faces of any of the big three, they'd likely be #1 on my ballot.

4. TCU - what's not to like? Elite D, blowing out the bad teams, and that win @Clemson is looking better and better. Showing that they're not one trick horned frogs, TCU won by 31 in Provo.

5. Cincinnati - great on both sides of the ball. Only behind the front four because of their SOS, but their MOV tells me they're elite. How good is Brian Kelly's defense? After losing 10 starters on defense, the Bearcats sport the #1 scoring defense in the Big East. How good is Brian Kelley's offense? Heisman hopeful Tony Pike ranked in the top 10 in the country with a 155 QB rating before going down with injury; his backup Zach Collaros has posted a 210 QB rating on an even 100 attempts. For those not familiar with the college system, Colt Brennan (Hawaii, 2006) holds the single-season record with a 186.0 rating.

6. Boise State - a very ugly (and getting uglier) SOS, but good enough MOV to make me think they're legit. They're blowing out most of their opponents. Haven't accomplished as much as the big five, but that's not exactly the Broncos fault.

7. Georgia Tech - streaking at the right time, at least before having to go to overtime with Wake Forest. Could make the argument for them over Boise State, in my opinion, if they had just handled their business more convincingly against the weaker opponents. Beat the two best three-loss teams in the country.

8. Oregon - the offense certainly didn't let them down against Stanford, and Chip Kelly's unstoppable attack makes the Ducks an elite team. Oregon has played one of the hardest SOSs in the country, and I doubt many teams would have had only one loss against this schedule. Losing on the smurf turf is not a bad loss, and Oregon is both the best and the most accomplished two-loss team in the nation.

9. LSU - hard to knock a team for losing to Alabama and Florida. They're a rich man's Penn State, with two losses to the only good teams they played and with no signature wins against good competition (best win is against Auburn or Georgia). Still, I tend to give teams a pass when they lose to undefeated teams.

10. Pittsburgh - two Big East teams in the top 10? Pitt hasn't loss a conference game, and 9-1 is 9-1. Beating nobodies, but winning is winning. They've similarly dominated four of their five common opponents with Cincinnati, and both teams struggled with Connecticut. Cincy comes to Heinz Field the last Saturday of the regular season in what should be one of the marquee games of the year; the same does not go for the Steelers matchup against the Raiders the next day.

11. Iowa - no, they're not a top 11 team in the country. Heck, I'm not even sure if they're a top 20 team in the country in terms of ability. And without Ricky Stanzi, they might not be top 40. But they've managed to win nearly every game against a decent enough schedule, and they've accomplished more than nearly every team in the country. As friend of P-F-R Dr. Saturday points out, Iowa is the only team to beat three of the BCS top-25 teams this year. But a loss to Northwestern is going to knock you out of the top ten.

12. USC - Behind Oregon and LSU, they're the next best two-loss team in the country. And frankly, I don't even think they're that good. But because of impressive road wins in Columbus, South Bend and Berkeley, they've got a top-twelve resume.

13. Ohio State - underrated this year; six wins of 17 points or more this year, while both of their losses came down to the final minute. More talented than people give them credit for, and capable of beating anyone (remember they took the Longhorns to the final minute in the Fiesta Bowl last season).

14. Miami - I think they've been a bit overrated this year; that exciting opener against Florida State looks a lot less impressive now, as does the win over Oklahoma. In addition to winning at the very end against FSU, the Hurricanes beat Wake Forest and Oklahoma by one point each and looked horrendous against the Hokies. The win over Georgia Tech looks terrific, though, and that keeps them ahead of the best of the rest.

15. Penn State - the Big 10's version of Nebraska; fantastic defense but little offense. Have overpowered most of their weak schedule and won every game they should, which is more than the teams below them can say. They're ranked third in the nation in scoring defense and have won seven games this year by 18 points or more.

16. Arizona - no shame in a road loss to Iowa or losing to Washington by three. That victory over Stanford is looking good. The Wildcats haven't been tested often-- they've yet to play USC, Oregon or even Cal (now without Jahvid Best). Despite that, I've slotted them ahead of the Cowboys because of their tougher SOS.

17. Oklahoma State - best win is against Georgia, and Georgia isn't very good. That said, OSU fared better against UT than the final score indicated, and the loss to Houston isn't a bad loss. Like Arizona, they've got a backloaded schedule, as they play Texas Tech this week and end with a watered down version of the Bedlam rivalry on November 28th.

18. Houston - yes, behind Oklahoma State. It's hard to argue that you're not a product of a weak schedule when you lose by 17 to UTEP. Houston's an oddity, as they beat the two best opponents -- by far -- on their schedule, but lost to one of the worst. With virtually nothing else of note on their schedule, I can't put them above the Cowboys who at least play a decent B12 schedule. Not convinced? Take a second and think about what Houston's record would be if, like Oklahoma State did, Houston had to play Texas. The Cougars don't have a defense, and Tulsa and Southern Miss lit them up. It's absurd that they're ahead of Iowa in the AP poll.

19. Utah - is there a shred of difference between Houston and Utah? Equally bad schedules, equally solid but not great MOVs, one loss each. A loss to Oregon is hardly embarrassing, but I'll take Houston's two solid wins and one awful loss over Utah's 1 very good loss and uh, wins against Air Force and Louisville?

20. Virginia Tech - losses to Alabama and Georgia Tech would put Virginia Tech in nearly the same category as LSU, if not for the loss to UNC. Still, because of their good wins (Nebraska, Miami, blowout win over BC), I'd consider putting them at #15. The problem? I have trouble throwing them over the 1-loss teams but want to put the PSU-Arizona-OSU triumvirate ahead of those teams.

21. BYU - a poor man's Oklahoma State. Their losses to Florida State and OU were more embarrassing than OSU's losses to UT and Houston, otherwise they'd be ahead of them. Win over OU is big, but the next best team on BYU's schedule is.... San Diego State? Utah State? Colorado State? Wyoming? UNLV? Ugh.

22. Clemson - like Virginia Tech, much better than their ranking, but with three losses you can't put them higher. Four point loss to TCU and a 3-point loss to GT are not knocks; losing to Maryland is. Wins over Miami and BC are nice, but I'm more impressed with them going toe to toe with two top seven teams and nearly winning both games. With their schedule, more than a couple of the teams above them would have three losses.

23. Wisconsin - like Penn State, their only losses are to OSU and Iowa. Ranked this far down because of razor thin margins against Fresno State, Minnesota, Indiana, and 8-point home wins over UM and Northern Illinois don't make me sympathetic to the Badgers.

24. Nebraska - Only the Texas Tech game was a clunker; losses to Iowa State and Virginia Tech could have easily gone the other way. #2 scoring defense in the country and the win over OU still counts for something. Perhaps the only hope left for those rooting for chaos and a non Texas-Floribama BCS championship game.

25. Stanford - win over Oregon shows that this program has arrived. The Wake Forest loss sounds bad, but really all three of the Cardinal's losses were close, road losses against decent teams. They beat up on rest of the schedule, and their game @USC this week will be a chance for Jim Harbaugh to show the world his team belongs.

Every Tuesday, I'll be writing a short column for the New York Time's Fifth Down blog. I'll be writing a short piece of original content along with a quick recap of what's been happening on the P-F-R blog over the prior seven days. This week, I wrote about DeSean Jackson's record-setting performances for a young wide receiver.

I'm a big fan of Mike Leach and Texas Tech's offensive scheme, especially the way he's been able to plug in seemingly random guys at QB and consistently produce video-game numbers no matter who's under center. This year, obviously, they've had some trouble settling on a QB to run the Air Raid offense, but if you add up the composite stats for Taylor Potts, Steven Sheffield, and Seth Doege, you get a quarterbacking group on pace for these stats when it's all said and done (bowl game included): 446-for-657, 5206 yds, 41 TD, 16 Int. Compare that to some other numbers in the Leach era:

So this year doesn't really represent as drastic a drop-off as you might think for Leach's passing attack, it's just that he's using 3 QBs instead of having one rack up all the eye-popping numbers. Anyway, all of this is an excuse to direct you towards Chris Brown's terrific archive of Air Raid-related reading material at Smart Football. Kinda makes me want to go out and throw 60 passes right now, you know?

Last week, I posted some quick numbers on the new stadiums from 1997-2003, and the winning percentage of the road team based on the number of visits they previously made to that new stadium. Commenter "Guy" had a concern about team quality:

Don't you need to account for the strength of these 12 teams, and the visiting teams? These twelve teams, in their 2nd, 3rd, and 4th seasons in the new stadiums (when many of the first-time visits must have occured), were a combined .564 overall. If you assume a HFA of 70 points, we'd expect them to be about .634 at home -- not that much lower than you found. If the first-time visitors happened to be slightly below average (which they likely were, since these 12 home teams are above average), that could account for the remaining gap.

As a quick correction, I need to point out that there were actually thirteen teams in my study, but I forgot to list the Houston Texans and Reliant Stadium, though they were included in the data set. But I'm sure Guy would have the same question even knowing the Texans were part of the data set, and rightly so. When I first began looking at this several months ago and then stored it away for a rainy day, I did not have "against the spread" records readily available. However, we do have "against the spread" data for all the games in question, so I figured I would dig back in and try to account for team strength.

Here are some tips and tricks for using our Player Game Finder tool:

Yesterday, I argued that defenses shouldn't be thought of as good against the pass or good against the run; defensive statistics should be considered fluid measures, because most defenses can choose how they'll let the offense beat them.

So is this true? Is this even possible to measure? Assuming that defensive disparities (i.e., being very good at defending the run or pass and very bad at defending the pass or run) are random, then we shouldn't see teams consistently being good (or bad) in one just area of defense. One way to measure this is to break out each team into two half-seasons. To avoid injury issues, I spit teams into "odds" and "evens"; games 1, 3, 5, 7, 9, 11, 13 and 15 go into one group, and games 2, 4, 6, 8, 10, 12, 14 and 16 go into another. We would expect a metric like quarterback yards per pass to be relatively consistent in "odd" and "even" games, because averaging more (or fewer) Y/A than average is a repeatable skill; but we wouldn't expect teams that did well in even weeks on opponent missed field goals to again do well in odd weeks, because that's (generally) not a repeatable skill. So, are defensive disparities narrative statistics that merely described what happened or are they repeatable and predictable events?

I looked at every team-season since 1988, giving us 21 full seasons worth of data. I measured run defense by yards per carry allowed (relative to league average) and pass defense by adjusted net yards per attempt allowed (relative to league average). I also noted the raw number of points allowed by each defense (relative to league average). This gave me run, pass, and scoring defense grades for each half-team in the study. Assuming that defensive disparities were real and consistent, we should see them for teams in both the even and odd splits. If defensive disparities were merely a mirage, and defenses really force offenses to pick their poison, then there should be no correlation between the even and odd splits.

Kurt Warner has made an entire career out of bouncing back, and he did it again Sunday, following one of the worst games of his career with one of the best, a 5-TD effort that carried the Cards past the Bears in Chicago. With that performance, Warner joined Tom Brady, Drew Brees, and Carson Palmer as the only QBs to toss 5 or more TDs in a game so far this season:

I've got a theory that there aren't necessarily teams that can only play the run well or can only defend the pass; rather there are just good defenses and bad defenses (and everything in between). Certainly there are teams that have played well against just the run and not the pass, and vice versa; there's no denying that looking backwards, certain teams in certain years were great in one area of defense and terrible in the other. But looking backwards, there are teams that recover lots of fumbles and teams that faced opposing kickers who missed a high number of field goal attempts, but that doesn't mean that those things will happen again in the future.

My claim is that theoretically -- i.e., if we played 1,000,000 games so that randomness would be eliminated -- defenses should just be defenses. No good run-D, bad pass-D teams; no average run-D, excellent pass-D teams; just defenses. That (I think) is a bold claim, so I better have some good reasons behind it. What are they?

1) Defenses are like chains: they're only as strong as their weakest links (for the flip side to this argument, see Brian Burke's article that offenses are like chains). Picture an unbelievable run-D teamed with an awful pass defense. That defense isn't going to be very good, as almost every team in the league could pass on them all day long. Flip the script, and nearly every team could control the game with power football against a defense that can't stop the run. On defense, if you have a weakness, almost every opponent can exploit it.

Last week, the SRS (along with just about every human being) was unimpressed with Iowa's resume. After an embarassing loss to Northwestern, the Hawkeyes are now at #29 in the SRS.

Before moving on to the results, a quick thanks to Peter R. Wolfe, who makes the results of every college football game available in an easy to use format. Early in the year, I was hesitant to use the results from the FCS (formerly I-AA) teams, as the small sample size was likely to skew the results. With ten weeks in the books, I am no longer lumping all FCS teams into one big group; now, each FBS team (formerly I-A team) that plays an FCS opponent will be given that specific team's strength of schedule grade, as opposed to the rating of the average FCS team. Along with an interesting week in college football, this tweak means the week 10 rankings look quite a bit different than last week's set.

The ratings are going to be a lot higher for each team as a result of eliminating the "FCS Team" and simply iterating every team in all of college football. But we're only concerned with the differentials between teams, and that hasn't been amplified. A quick reminder: these ratings are meant to be predictive, not a list of how college football teams should be ranked. Teams like Virginia Tech, Clemson and Oklahoma shouldn't be ranked that high in the polls, because they've had a number of losses this year. But based on their large victories, close losses, and/or tough strength of schedule, they're much better teams than their record indicates.

The full list of the 120 FBS teams, after the jump. As you will see, Texas finally broke the tie at #1.

Courtesy of Matt Knobbe and the Tecmo Super Bowl Repository, here's your Tecmo Super Bowl Game of the Week for Week 9, featuring the Pittsburgh Steelers and the Denver Broncos. The highlights:

(How did we do this? Matt and the other dedicated folks at the Knobbe.org message board have spent a lot of time over the years updating this classic Nintendo football game, including the introduction of a 32-team ROM a few seasons ago. Sounds complicated, but don't worry, it's easy for you to enjoy the fruits of their labor: just get yourself an NES emulator, download the 2009 version of Tecmo here, and play to your heart's content. And be sure to check back at Matt's site for roster updates and more Tecmo-related goodness all season long.)

I've spent alot of time talking about familiarity and home field advantage. Most recently, I wrote about the New York teams sharing a stadium. Today, I'm going to do mostly a data dump of other research I had done on the familiarity issue.

From 1997-2003, a flurry of new stadiums opened in the NFL, with the following franchises playing in a newly built stadium: Washington, Baltimore, Tampa Bay, Cleveland, Tennessee, Cincinnati, Denver, Pittsburgh, New England, Detroit, Seattle, Houston and Philadelphia. I pulled the results for those thirteen stadiums, in terms of how many times the visitor had previously visited. I didn't use exhibition data, but did use all regular season and playoff visits. These numbers are through the 2008 season and playoffs, and do not include results this year. The win percentage is the wins and losses from the perspective of the visitor.

Visit		W	L		win pct
1		134.5	242.5		0.357
2		60	113		0.347
3		50	54		0.481
4		29	34		0.460
5		20	30		0.400
6		19	25		0.432
7		21	18		0.538
8		6	10		0.375
9		7	7		0.500
10		4	5		0.444
11		3	2		0.600
12		2	1		0.667

These are the overall results. I divided the games out into divisional and non-divisional games. We know that divisional opponents get to play at a venue every year, whereas non-divisional opponents typically do not. There is a bit of a grey area when it comes to teams that were division opponents until the realignment in 2002. I treated games as divisional games if the road team was in the same division that season. So, Tennessee is a divisional road opponent of Baltimore for games before 2002, and a non-divisional opponent after 2002.

Here are the division opponent only results, sorted by number of prior visits to the same stadium:

Visit		W	L		win pct
1		22.5	31.5		0.417
2		17	37		0.315
3		28	24		0.538
4		24	23		0.511
5		18	22		0.450
6		15	24		0.385
7		19	18		0.514
8		9	9		0.500
9		6	9		0.400
10		7	5		0.583
11		3	2		0.600
12		2	1		0.667

There are a small number of games once we get beyond five visits, but the combined win percentage of division opponents when visiting these stadiums for the fifth time (or more) is 79-90 (0.467). Also, with a couple of exceptions, every one of those first visits by a division opponent was in the first year that the stadium opened. We see that the division opponents actually perform worse in the second season.

Non-divisional opponents, on the other hand, will tend to have their first visit to a stadium after the home team has already been playing there for at least a season. Here are the non-divisional results, based on number of visits.

Visit		W	L		win pct
1		120	211		0.363
2		47	78		0.376
3		22	34		0.393
4		9	12		0.429
5		4	10		0.286
6		4	4		0.500
7		2	3		0.400
8		0	1		0.000

If we exclude the first time visits that occur during the first year of a new stadium (when the non-divisional visitors went 24-29), then we get a 96-182 record (0.345 win percentage) for all non-divisional teams in their first visit to a new stadium more than a year after it opened.

A few weeks ago, I threw out this crazy idea about how to isolate a team's peak performance by SRS:

"Obviously, we can't run the SRS on single-game samples, because it requires multiple opponents to 'work'. But what if we broke each team-season down into 15 to 19 'mini-teams' based on 2-game stretches of the season? Like, the Patriots' win over the Jets in Week 1 of the 2007 season would be part of the 'New England Games 1-2' team, and so would their win against the Chargers the next week. And that Chargers team would be part of the 'San Diego Games 2-3' team, who played 'Green Bay Games 3-4' the next week, who played 'Minnesota Games 4-5' the week after that... and so on and so forth. Now, every 'team' connects to every other team, just like in the regular SRS, but we've also isolated team performance down to the most specific time period possible using the SRS method."

As a follow-up, I calculated the best and worst 2-game stretches by teams this decade (2009 isn't included because not all teams have played the same # of games yet). Remember, the SRS is focused on measuring a team's point differential vs. the point differential you'd expect an average team to have based on the game's location and the strength of the opponent; this method takes it even further and is only concerned with the strength of the opponent at the time of the game, meaning wins against teams with mediocre records can still be positive for a team's SOS if they play them either before or after a strong performance. I think there's definitely some logic to this, because (as Chase pointed out in the post that inspired this series) every game features a different version of the same team; sometimes the differences are so small as to be imperceptible, but sometimes they're huge (think the '07 Giants early in the season vs. late), so it really does matter when you catch a particular opponent.

As Chase pointed out earlier "the U"s alumni have an active 113 consecutive week scoring streak in the NFL, which is pretty astounding when you think about it. That is over six and a half seasons without missing a weekend. I had to check and see if this is a record of some sort, so I wrote a little perl script to find the longest streak for every school with a scoring alum.

Yikes, what can you say about the Browns? Things are getting pretty ugly in Cleveland, following a 30-6 loss to the Bears that left the team having been outscored 61-9 over the past 2 weeks. QB Derek Anderson was brutal again, tossing 2 picks and turning the ball over another time on a botched snap. Amidst Cleveland's myriad woes, it might be easy to write off Anderson's performance so far in 2009 as merely one of many breakdowns, but that's understating just how awful Anderson has been... Through 8 games, here are the worst passer-rating performances by a QB in a season since 1960:

That's right, Eric Mangini benched Brady Quinn in Week 3 so he could receive the 14th-worst 1st-half quarterbacking performance in almost 50 years (!) from Anderson. And keep in mind that passer ratings in the modern game are almost always significantly higher than they were in the sixties and early seventies, since today's defenses are greatly restricted in what they can do when trying to stop receivers from getting open and/or catching the ball. Not since another Cleveland quarterback, Mike Phipps, posted a 27.7 rating in 1975 has a QB opened the season as badly as Anderson this year, and among the bottom 25 half-seasons listed above, Anderson's '09 and Heath Shuler's 1994 are the only ones to come after 1980. You can be sure that this is not the kind of history the Browns wanted to make in 2009.

Something historic happened Sunday at Lambeau Field—something that had never happened, but which appeared like it could be a possibility as of last November, and which we had to wait all offseason and eight weeks into this season to finally witness. That’s right, it was the first ever 38-26 game in the history of the National Football League. Bet you didn’t see that one coming, did you? Well, reader David Herson did.

Remember last year, when the Steelers and Chargers played a game that finished 11-10, and everyone was a-buzz with talk about the first ever 11-10 game in NFL history? Well, Sean Forman whipped up a game score finder so that we could find every game for each score combination. (click here to play around with it). David took that information and calculated that 38-26 was the score combination most due to occur out of those that had never happened. You can see his full list of most likely score combinations in the comments to that earlier post.

Next on his list: 42-16. You’ve been warned.