It’s late in the fourth quarter, the score is tied and a potential playoff berth may be on the line. Are there quarterbacks in the NFL who excel in this situation?

I decided to take a look at the fourth-quarter comeback in the NFL in 2008. I defined a pressure win as follows:

1) The game-winning score took place with less than eight minutes to go in the game.

2) There was at least one first down or an offensive touchdown scored on the drive.

With those limitations, I was surprised to find that there were 67 pressure wins in 256 NFL regular-season games last year (26%), and four in eleven playoff games. That’s more than one in four games coming down to one key series, even one key play.

NFL quarterbacks have to get used to pressure very quickly if they’re to survive, long-term.

The leaders in pressure wins in 2008:

Quarterback	Pressure Wins	Total Wins
Matt Ryan	4	11
Jake Delhomme	4	12
Peyton Manning	4	12
Ben Roethlisberger	4	12
David Garrard	3	5
Gus Frerotte	3	7
Jay Cutler	3	8
Eli Manning	3	12

Was it any harder to pull off these wins on the road, with an enthusiastic crowd sheering on the defense? Of the 67 pressure wins, 33 came from visiting quarterbacks. When the game’s on the line, top players forget about the crowd.

Only three teams - Cincinnati, Kansas City and Detroit - did not have a pressure win from a quarterback. Of course, these teams combined for only six wins all season. Of the league’s playoff teams, three - Arizona, Baltimore and Philadelphia - had only pressure win all season. Baltimore and Arizona each added one during the playoffs.

No team had more than four pressure wins all year. But when I looked at the defenses, the results were far different. Only four teams did not allow a pressure loss all year - Arizona, Dallas, Denver and Tennessee (Arizona and Tennessee did in the playoffs). The leader board, however, was a little more interesting.

Green Bay led the way with an astonishing eight pressure losses among their ten regular-season defeats. The Packers could have been a high seed in the playoffs with a better fourth-quarter defense.

Kansas City allowed five, and San Diego, New Orleans and St. Louis each allowed four.

One final note: two quarterbacks pulled off an unusual double during the regular season, beating a division rival twice with late-game drives. Ben Roethlisberger swept Baltimore, and Philip Rivers swept Kansas City.  Those were Rivers’ only two pressure wins of the entire season, though he earned an additional pressure win over Indianapolis in the playoffs.

Finally, to round off my preview of the 2009 draft, I wanted to provide a list of which teams have done the best job drafting quarterbacks over the last ten years.

My scoring system is somewhat arbitrary. It attempts to reward teams on a gradual scale for great picks, and penalize harshly for early-round picks who fail.

Team	Total QBs	1st Round	Total Points
New England	5	0	13
New Orleans	3	0	9
San Diego	5	1	8
Baltimore	6	1	8
Atlanta	4	1	7
Philadelphia	4	1	7
Tampa Bay	4	0	6
New York Giants	2	1	5
Washington	6	2	5
Denver	4	1	4
Green Bay	4	1	4
Pittsburgh	4	1	4
Indianapolis	1	0	3
Arizona	4	1	3
New York Jets	4	1	3
Kansas City	2	0	2
Buffalo	2	1	2
Minnesota	3	1	2
St. Louis	4	0	2
Cincinnati	6	2	2
Miami	2	0	1
Detroit	4	1	0
Seattle	5	0	0
Cleveland	5	2	0
San Francisco	6	1	0
Jacksonville	2	1	-1
Oakland	3	1	-2
Chicago	4	2	-2
Dallas	2	0	-3
Carolina	3	0	-3
Tennessee	2	1	-8
Houston	4	1	-8

Included were the nine drafts from 1999 through 2007. It’s probably too early to judge 2007 in any significant manner, but definitely too soon to make heads or tails out of 2008. Joe Flacco might have Baltimore jumping to the head of this list soon enough.

Here’s a quick and very tough trivia question for everyone (or perhaps anything can be googled these days):

Who is only the only quarterback currently in the NFL who has a 1.000 winning percentage as a starter?

Bonus for naming the high school he attended.

I thought I’d continue my preview of the 2009 quarterback draft, as well as review of quarterback drafting in the NFL over the last ten years with a list of great first-round busts and draft reaches.

1st-round busts:

Akili Smith (1999, 3rd pick, 1st round, Cincinnati, 3-14 record as a starter, currently out of the NFL).

Alex Smith (2005, 1st pick, 1st round, San Francisco, 11-19 record as a starter, currently backup with San Francisco).

David Carr  (2002, 1st pick, 1st round, Houston, 23-56 record as a starter, currently backup with the New York Giants).

Cade McNown (1999, 12th pick, 1st round, Chicago, 3-12 record as a starter, currently out of the NFL).

Vince Young (2006, 3rd pick, 1st round, Tennessee, 18-11 record as a starter, currently backup with Tennessee, by far the highest pre-draft grade of any quarterback since Peyton Manning).

J.P. Losman (2004, 22nd pick, 1st round, Buffalo, 10-23 record as a starter, currently out of the NFL).

Joey Harrington (2002, 3rd pick, 1st round, Detroit, 26-50 record as a starter, currently backup with New Orleans).

Tim Couch (1999, 1st pick, 1st round, Cleveland, 22-37 record as a starter, currently out of the NFL).

Patrick Ramsey (2002, 32nd pick, 1st round, Washington, 10-14 record as a starter, currently backup with Tennessee).

Byron Leftwich (2003, 7th pick, 1st round, Jacksonville, 24-22 record as a starter, currently backup/competing for starter role with Tampa Bay).

Rex Grossman (2003, 22nd pick, 1st round, Chicago, 19-12 record as a starter, currently out of the NFL).

Biggest reaches:

Quincy Carter (2001, 2nd round, Dallas, did not pay off).

Joe Flacco (2008, 18th pick, 1st round, Baltimore, looks like it paid off).

Joey Harrington (2002, 3rd pick, 1st round, Detroit, did not pay off).

J.P. Losman (2004, 22nd pick, 1st round, Buffalo, did not pay off).

Cade McNown (1999, 12th pick, 1st round, Chicago, did not pay off).

Marques Tuiasosopo (2001, 2nd round, Oakland, did not pay off).

Kevin Kolb (2007, 2nd round, Philadelphia, hard to tell yet).

Kevin O’Connell (2008, 3rd round, New England, hard to tell yet).

Tarvaris Jackson (2006, 2nd round, Minnesota, hard to tell yet).

Kellen Clemens (2006, 2nd round, New York Jets, did not pay off).

Sage Rosenfels (2001, 4th round, Washington, paid off).

Generally, picking a player far before his draft grade indicates he should go is a poor proposition in the NFL, at least in the first couple of rounds.

After the second round, since a quarterback even staying in the NFL longer than the initial contract is fairly rare, there is more value in picking based on individual scouting and hunches. Both Marc Bulger and Tom Brady, sixth rounders who have fared well in the NFL, were chosen ahead of their draft grades.

In the first of these articles previewing quarterbacks and the 2009 NFL draft, I compared Matthew Stafford to Carson Palmer. That’s not meant to be a comparison of style, it’s simply based on both players having a 6.60 draft rating from Pro Football Weekly.

Stafford will be treated as a franchise quarterback when entering the league. It’s likely two other quarterbacks will receive the same treatment. Mark Sanchez has a grade of 6.30 and Josh Freeman a grade of 6.10.

This is an average year for top-tier quarterbacks in the draft. The three will bring the number of 6.01 or better (”should become a quality NFL player” according to Pro Football Weekly) to 30 since 1999.

Of the other 27, all have started at least one game at quarterback in their careers. Other than the newer apprentices (Brady Quinn and Matt Leinart among others), few have failed to become regular starters. I would say eight have been busts - Tim Couch (6.80, 59 starts), Byron Leftwich (6.50, 46 starts), Kyle Boller (6.40, 42 starts), Rex Grossman (6.10, 31 starts), Alex Smith (6.75, 30 starts), Patrick Ramsey (6.06, 24 starts), Akili Smith (6.80, 17 starts) and Cade McNown (6.02, 15 starts).

Sanchez’s 6.30 grade is closest to Joey Harrington’s 6.25 grade in 2002, when he was the third pick. Third was a bit of a reach for Harrington. Boller was picked 19th with a slightly higher grade. I can see Sanchez going anywhere from around 8th to 23rd.

Freeman’s 6.10 grade matches that of Grossman (22nd in 2003) and Chad Pennington (18th in 2000). I see him fitting safely into the bottom half of the first round.

After the top three, there’s a huge drop to the next rated quarterback, John Parker Wilson, whose 5.37 grade puts him in the fourth round. Thirteen other quarterbacks had grades from 5.35 to 5.40 in the last ten drafts. They were picked anywhere from midway through the second round to the sixth round.

The two second rounders were Kevin Kolb in 2007 and Tarvaris Jackson in 2006. Of the others, only Kyle Orton (4th round in 2005) has significant starting experience. Six of the 13 have never appeared in a game, though in fairness, four were drafted last year.

Stephen McGee (5.33) and Rhett Bomar (5.30) are also on the potential draft list. Nine quarterbacks have had ratings from 5.30 to 5.33 in the last ten years. There was one spectacular second-round bust pick (Marques Tuiasosopo in 2001) and one nice find - David Garrard in the fourth round in 2002. It’s likely McGee and Bomar will also go around the fourth round.

While quarterbacks have been drafted with grades as low as 4.80, generally 5.10 is considered the bottom of the seventh round. There are 12 quarterbacks with grades from 5.04 to 5.17. It’s likely some will get picked near the end of the draft. Generally, these players are lucky to even hold a clipboard for a couple of seasons, but every once in a while someone with a grade in that range makes good.

Tom Brady (5.09, 6th round, 2000) heads that list. Others include Tony Romo (5.10, undrafted, 2003), Marc Bulger (5.04, 6th round, 2000), Derek Anderson (5.16, 6th round, 2005),  Sage Rosenfels (5.11, 4th round, 2001) and Aaron Brooks (5.05, 4th round, 1999).

There are even a pair of current starting quarterbacks who had 4.95 grades coming in - Matt Cassell (7th round, 2005) and Shaun Hill (undrafted, 2002).

Drafting quarterbacks is a risky proposition. This is considered a weak class, after the first-round talent. But you never know where the next Tom Brady may come from.

Now where was I before I was so rudely interrupted? My host (probably soon-to-be-former host), Hostway, shut down my server for almost four days fighting off some random or imaginary internet problem. Customer service was less than forthcoming, less than helpful, denying that there was an issue after the second day. Seems to be resolved now.

I was discussing the drafting of quarterbacks. You’ll notice, even on a day when a Pro-Bowl left tackle changes teams, there’s very little trading done in the NFL. Think back to the last starting quarterback who got himself traded. Oh, Jay Cutler. Well, before that. Oh, Matt Cassell. Sage Rosenfels? Kyle Orton?

What’s going on here? We have four starting quarterbacks moving in trade in one off-season. Now, look at the other 24 starters in the NFL who were not undrafted as rookies.

Well, 16 of them are still with the teams who drafted them (for all intents and purposes, Eli Manning and Philip Rivers should count in that group because that trade was announced very shortly after Rivers was drafted).  Of the eight who weren’t, we have four free-agent journeymen (Brian Griese, Kerry Collins, Daunte Culpepper and Chad Pennington), a pair of sixth-rounders who made good (Matt Hasselbeck and Marc Bulger), an unusual injury/release case (Drew Brees) and Matt Schaub, who was traded before the draft a couple of years ago because the Falcons thought they were set, long-term, with Michael Vick.

Is this a new trend? It’s not unusual (Cassel, Rosenfels) to trade a quarterback who saw some time because of an injury to the starter, showed a little promise, but won’t play because the starter returned to health. And the Cutler/Orton trade is one of the more bizarre temper-tantrum cases in NFL history. You expect that sort of thing from a wide receiver, not a quarterback, since quarterbacks are expected to be leaders.

Still, given the importance of the quarterback to a team, I think it’s significant that the percentage of teams starting a man they drafted has suddenly dropped to 50%.

Overall, there are 79 quarterbacks currently on NFL rosters who were drafted into the league. Of those, 43 (54%) are still with the team who drafted them.

Of the 79, 24 quarterbacks were drafted in the first round. And of those 24, 16 (67%) are still with the team who drafted them. And that group includes 12 of 16 starters still with their original draft team.

The four who have moved include Culpepper, Pennington and Collins - all who suffered release and disinterest until resurrecting their careers (OK, Culpepper may not quite be resurrected). And Cutler.

The media, for now, has stopped talking about the Cutler trade. To me, it’s one of the biggest player personnel stories in NFL history.

I’ve been taking a long look at the NFL draft this week, as it applies to quarterbacks. So I have a few notes that might be useful as you prepare for the 2009 draft, which takes place on April 25-26.

My study focused on the last ten drafts, going back to 1999. This includes most of the quarterbacks currently playing in the league. I’m using only the last ten drafts because I don’t have any draft preview data going further back, though I think we all remember Peyton Manning and Ryan Leaf were 1-1A in almost everyone’s draft book in 1998. Yes, it’s been eleven years since we were all wondering Manning or Leaf.

I used the Pro Football Weekly draft preview ratings from 1999-2008. These were done by Joel Buchsbaum until he died in 2002, and Nolan Nawrocki since. I’ve found these ratings an insightful and accurate reflection of the views of NFL coaches and scouts.

This year, Georgia’s early entry, Matthew Stafford, has the highest grade for a quarterback, 6.60. According to Pro Football Weekly, that means he’s a “sure-fire first rounder who could become a Pro Bowl-caliber player.”

Only twice in the past ten drafts has the top quarterback been rated lower than 6.60, though that pair (Michael Vick, 6.40, 2001 and Chad Pennington, 6.10, 2000) have not been terrible by any means. In 2003, the top quarterback was also a 6.60 and was chosen first. That was Carson Palmer.

Quarterbacks have gone first in seven of the ten drafts.  Of those seven, three were fairly certain busts (David Carr, 7.10, 2002, Tim Couch,  6.80, 1999 and Alex Smith, 6.75, 2005) and one looks like he’s heading bustward (JaMarcus Russell, 6.65, 2007). Vick, Palmer and Eli Manning (6.75, 2004) have had a more positive influence on their teams.

The Detroit Lions, picking first in the draft, could definitely use a franchise quarterback. But even though Stafford is a near-consensus pick, it’s far from certain he will meet expectations.

As important as this pick is for teams, you rarely see teams bucking the consensus. Since 1999, 17 quarterbacks have been rated 6.60 or higher. Eleven were picked in the top four,  four in spots ten and eleven (none five through nine) and two dropped considerably (Brady Quinn, 7.00, 22th in 2007 and Aaron Rodgers, 6.60, 24th in 2005).

Below Vick’s 6.40, only two quarterbacks were picked higher than 18th during those ten drafts. Both busted (Joey Harrington, 6.25, 3rd in 2002 and Cade McNown, 6.02, 12th in 1999). There’s no reward for reaching for your quarterback at the top of the draft.

If the Lions choose a quarterback, history dictates it will be Stafford, especially since the Lions were the team burned by the Harrington reach.

Do teams need a quarterback chosen in the top 11 picks? Only ten of the 32 NFL teams have one expected to start this season, and six others have one slated as a backup (including the Lions, who have Daunte Culpepper on their roster). Four teams are expected to start a quarterback who was never drafted, and another four (including New England) are expected to start a quarterback taken in round six or seven.

I doubt the Lions, having gone through about as miserable an experience the last few years as any team in modern NFL history, have the guts not to pick Stafford, but if they did decide on Jason Smith or Aaron Curry instead, history indicates that might not be as bad a move as fans would claim.

A Frontier reader, QuikSand, recently commented that it might be interesting to see how the system fares when there is a significant difference between the predicted result and the point spread.

Since the system did not perform very well simply picking every game, perhaps it did better only when counting those games. After all, why bet money on the underdog when the system says a team will lose by one point and the spread is at +2?

I ran the numbers. One difficulty in using these numbers is that my system does not take injuries into account. If a starting quarterback goes down, that will affect the point spread, but not the system. For that reason alone, even if you believe my system has some validity (I’m not sure I do, it just weights correlations based on historic results), you should take it with a grain of salt.

In a sense, my weekly best bets were based on this concept. In most cases, my best bet was the game where there was the most difference between predicted result and spread, taking into account significant injuries and trends. This season, I was 10-7 on best bets. Last season, with four games chosen each week, I was 35-33 against the spread. No, still not about to give up the imaginary day job.

Breaking down results against the spread based on the difference between predicted result and point spread, here’s how I fared:

Difference	Record
10	1-1-0
9	2-1-0
8	9-4-0
7	3-4-1
6	6-6-1
5	10-12-3
4	12-21-0
3	17-19-0
2	23-23-2
1	32-30-3
0	0-0-10

The numbers where the system was 8 or more points off the spread look promising, 12-6-0. Otherwise, just looking random.

So, I looked at the 2007 results. The system performed better in 2007, 24 games above .500. Thought I may have had something there.

On games where the system was 8 or more points off the spread, 8-14-1 (though, in fairness, there were several of those tricky week 17 games in that group in 2007, while there was only one this year). And in that 3-to-5 point range, where the system fared so poorly this season, 44-27-3 in 2007.

Interesting idea, measuring the variation, but I don’t think I answered any questions.

The 43rd Super Bowl is now part of football history. The biggest surprise was how Arizona stormed back in the fourth quarter and came within inches of pulling off the upset.

The Cardinals took their time gaining momentum in each half. In the first and third quarters combined, they netted just under ten minutes of possession time and only 54 total yards. In the fourth quarter alone, they racked up 220 total yards (128 in the second quarter).

Until the fourth quarter, Arizona’s longest gain on first down was only seven yards. The Cardinals had six first-down plays of ten or more yards in the fourth quarter. The announcers made a big deal of Pittsburgh’s safeties playing deeper in the fourth quarter to prevent the big play. That, in retrospect, was an almost disastrous mistake.

Arizona’s offense was extraordinarily effective on second down and ten or less yards. The Cardinals passed on 13 of those 16 plays, and gained the first down 12 times (they gained 219 of their 402 total yards on those 12 plays). They were only 3-of-8 on third down, 1-for-5 on third and ten or more yards.

The Steeler offense, looking back, must feel some relief in winning despite almost complete failure on first-and-goal. The first four times it got inside the 10-yard-line, Pittsburgh ran the ball, netting -2 yards. The fifth time, with just seconds remaining the game, Santonio Holmes almost came up with a terrific catch in the corner of the end zone (he did a little better on second down that time).

Pittsburgh gained 145 yards on its first two possessions and 88 yards on the game-winning drive, which started with just 2:37 remaining in the game. The Steelers had six possessions in between, with leads ranging from 10-7 to 20-16. They gained just 72 yards on those six possessions, though the Cardinals helped out by adding 44 yards in penalties.

So, where does Pittsburgh rank among Super Bowl winners? On offense, they scored 3.3% more than the average team in the NFL this year. That’s last among the 43 winners, and 81st among the 86 teams to play in a Super Bowl.

Arizona was 23.8% better on offense than the average team. That places 20th among the 43 teams to lose a Super Bowl, and 43rd among the 86 teams to play in the big game.

While it doesn’t always hold up in the NFL, the cliche that defense wins Super Bowls did hold true tonight. Arizona’s defense was 17.0% worse than the average in the NFL this year. That’s last among the 86 teams ever to reach the Super Bowl (the next lowest was 4.1% worse than average).

Pittsburgh, on the other hand, had a defense 32.1% better than the NFL average. That ranks 13th among Bowl winners, and 16th among the 86 Super Bowl participants.

The team with the better defense has won 32 of 43 Super Bowls, and the team with the better offense has won 26. In the 20 Super Bowls where one team has ranked better on defense and the other better on offense, the team with the better defense has won 13 times.

Last year, with all the attention given to the undefeated Patriots, I developed a ranking algorithm for every team in NFL history. Of the 1357 teams since the modern football era began in 1960, I would place this Steeler team 73rd, the relatively low rank (34th among Super Bowl winners) almost completely due to the mediocre offense - the same offense that was just good enough for victory tonight in the closing seconds. The best of all-time? Still the ‘85 Bears.

Not a great post-season for the prediction system. Including the Super Bowl, it was 4-7 straight up in the playoffs and 168-98-1 for the year.

Against the spread, the system was a beyond-dismal 1-10 during the playoffs, and 116-131-20 for the year. If I had a day job, I wouldn’t be quitting it right now. And if you don’t have a day job and you’re using my system to make bets, you probably need one now.