Friday, May 15, 2009

Number crunching

Today I’m going to talk about quarterbacks – again.

What I’m not going to get into is who I think should be the Vikings starting quarterback in 2009. What I am going to get into is some basic statistical benchmarks any Viking starting quarterback has to reach to give the team a decent chance of challenging – and winning – a Super Bowl.

But first, some context. Since the NFL changed its rules to outlaw the outright mugging of wide receivers and tight ends on pass routes in 1977 and 1978, quarterback passing numbers have gone way up. And because of the influence of Bill Walsh and the success of the San Francisco 49ers – starting with their first Super Bowl win in 1981 – the West Coast offence Walsh helped create has set a new standard that quarterbacks are expected to meet. Passing for 3,000 yards is a given. Completing at least 60 per cent of your passes a must. And in a perfect West Coast offence world, your ratio of touchdown passes to interceptions should roughly be 2:1.

The Vikings under Brad Childress run a version of Walsh's West Coast offence. But if Walsh were alive today, he'd be appalled that the term was applied to what the Vikings have been running under Childress. That’s how poorly the unit has looked most of the time.

Now some of the blame for that goes to the unimpressive group of wide receivers we've had. Some of the blame goes to an offensive line that struggles to pass block. But a lot of the blame falls at the feet of a collection of quarterbacks - from Brad Johnson to Tarvaris Jackson to Gus Frerotte – who haven't been able to get the job done.

This year, the Vikings could have Jackson back as the number one guy. Or it could be Sage Rosenfels. Or it might even be the retired (for now) Brett Favre. So what kind of numbers will the starter have to produce to get the Vikings within sniffing range of Super Bowl contention?

To get an idea, I compiled the averages in completion percentage, passing yards, touchdowns and interceptions since 1981 of the quarterbacks on Super Bowl-winning teams. Those averages are: 61.5; 3,090; 21.8; 11

Now here are the averages in those categories posted by Viking quarterbacks during Brad Childress' three-year reign: 59.4; 3,185; 15.7; 17

That’s not as large a discrepancy as I thought there would be. And as bad as we think the Vikes quarterback play has been under Childress – and it’s been pretty bad – in completion percentage and passing yards, the Vikings are close to where they need to be.

But Viking quarterbacks have thrown too many interceptions and too few touchdown passes the past three seasons. If the Vikes want to make the playoffs again and be more than a one-and-done team, that will have to change. No NFL team has won a Super Bowl with a quarterback who threw more interceptions than touchdowns since the New York Giants' Phil Simms did it in 1986. And the last team to accomplish that feat before the Giants was Pittsburgh – in 1974. So it doesn't happen very often.

At least the Vikings appear to be going in the right direction. In 2006, Viking quarterbacks threw 13 touchdown passes and 20 interceptions. In 2007 it was 12 TD passes and 14 INTs. In 2008 the tally was 22 TDs and 17 INTS – even with interception machine Gus Frerotte starting 11 games.

If Tarvaris Jackson or Sage Rosenfels can edge that completion percentage up around 62 per cent, throw for about 3,000 yards and add a couple of touchdown passes and subtract four or five interceptions from the 2008 totals, voila, the Vikings could very well be Super Bowl bound. Of course, that’s a big “if.”

And for those who are hoping the Vikings can get Brett Favre to make another comeback, here are his averages in the aforementioned statistical categories the past three seasons: 62.7; 3,837; 22.6; 18.3

Those seem like pretty decent numbers, certainly good enough to take a team with a very good running game and a very good defence deep in the playoffs. But note that in three of his last four seasons, Favre has thrown at least as many interceptions as touchdown passes. And if reducing the number of interceptions our quarterbacks throw is critical to the Vikes playing in January and February, then Favre’s probably not our guy.

More bad news
If the Vikings do convince #4 to come out of retirement and sign with the team, let it be known that no NFL team has ever won a Super Bowl with a 40-year-old starting quarterback (Favre turns 40 in October). John Elway was the oldest QB to help his team win a Super Bowl. He was 38 when Denver won it all in 1998. Remember that season, Viking fans?

Anyway, this might be just another stupid statistic. But still, in 44 seasons – a 40-year-old starting quarterback playing for a Super Bowl winner – it’s never happened.

Meanwhile, ESPN’s Kevin Seifert points out that Favre has been brutal down the stretch since 2005. This suggests that as Favre has reached his mid-to-late 30s, his body wears down as the year goes on. So despite his Hall of Fame credentials and reputation for producing in the clutch, Vikings fans might not be able to depend on him to play well in a game outdoors (or indoors) in late December and January.

Non-Viking quarterback/Favre stuff
There was no court decision on the disputed suspensions of Kevin and Pat Williams on Thursday. That could happen next week.

But if the suspensions are upheld, I don’t think it's the end of 2009 for the Vikings, despite what others are going to say and write. If there's a four-game stretch the Vikes can afford to be without a couple of keys players, it's the first four of the 2009 season, when they're playing Cleveland, Detroit, San Francisco and the Packers (in the Metrodome).

And if the Vikings lose the Williamses and come out of that stretch at least 3-2, the suspensions might actually be good for both players, particularly the ageing Big Pat. Four games off could keep both players fresh later in the season. While their opponents are sore, tired and breaking down from the stress of a physical 16 game season, the Williams boys could be healthier, stronger and better than the opponents they are facing during the playoff stretch.

Finally, you learn all kinds of interesting stuff when you do a little statistical research for a post like this. For instance, you learn how bad quarterback play was back in the day.

In 1967 Green Bay managed to go 9-4-1 and win two playoff games and a Super Bowl even though Bart Starr completed just 54.8 per cent of his passes, threw for 1,823 yards and had nine touchdown passes and 17 interceptions.

In 1968 Joe Namath led the New York Jets to an 11-3 record and a Super Bowl victory despite completing only 49.2 per cent of his passes and throwing 15 TDs and 17 interceptions (he did throw for 3,147 yards though, a very high total in those days).

Those kind of numbers today would not only keep you far away from any Pro Bowl, you’d lose your job. Unless, of course, you were quarterbacking the Vikings from 2006-2008.

4 comments:

Peter said...

the 60% completion rate makes sense to me, the 3000 yards seemed low at first (but makes more sense if you accept that a team's running game and defense are paramount to winning championships), but the 2:1 TD-to-INT ratio is the killer. Minnesota hasn't had that in a long while.

I think Rosenfels can throw 60% and 3000 yards. To throw a 2:1 ratio, though, what kind of volume is a must? I'll bet he could keep his INTs down to 7 if he's only asked to throw 14 TDs. Can 14 regular season passing TDs get the job done? With enough defense and rushing yards, (and quality ST play of course), it just might be. If Rosey is asked to throw 22 TDs, however, I don't see him keeping his INTs down to 11.

Cool analysis, DC. I like the way you dig into the numbers. Do you put your findings into Excel before you crunch? Sometimes I feel that I'd be interested in seeing the results and doing further research. For example, take your data set and grab the top half for defense and/or running performances in the regular season and see what the QB averages change to. Perhaps it's a little easier on the QB with a stout D and big runners, but how much easier? And can Minnesota be considered above average D and runners than the SuperBowl winners for the last few decades?

DC said...

Peter:

It seems with Rosenfels' statistical history (as small as it is), that its reasonable to expect he'll always throw more INTs than we'd prefer. I don't think it's necessary for him to have a 2:1 ratio for us to challenge for a Super Bowl, but that's what you're shooting for. However, if Rosenfels can complete 60 % of his passes, throw for 3,400 yards or so and throw 24 or 25 TDS and only have 15 or 16 INTs, with this running game and defence, that could take the Vikings pretty far. I don't think Jackson can produce those kind of numbers.

As for the stats, nothing in an Excel sheet. I'm usually not a numbers guy. I just mark it all down on a sheet of paper or papers and add it all up.

Peter said...

ok, mail me your work.

just kidding.

DC said...

Peter:

I'll try to dig it up at the local recycling depot.

Also kidding.