Using statistics as a means of describing the past is easy. We do it all the time when arguing who was the greatest player at a particular position or which team from a certain era would beat another team from another era. We site touchdowns or yards or wins or even the quality of the opponents.
Using statistics as a means of predicting future results is a bit trickier, as I’m discovering.
Last week my prognostication skills produced a 5-11 record overall with a 1-5 tally for the picks featured in the column. For the season, I’m 26-35 for a whopping 42.6% accuracy rate.
Needless to say, I’ve spent a lot of time this week pondering whether advanced football measurements of efficiency like Defense-adjusted Value Over Replacement (DVOA) can be used to predict outcomes on a week by week basis. So, I turned to my statistical guru, often referred to by fans of the Tony Kornheiser show as “the person to whom I’m related by marriage,” for input on descriptive versus predictive statistics.
Her response? Have you checked other sites to see how well they are doing?
What a novel idea. So, I did. And guess what I found? I’m really not doing as poorly as I thought I was.
It’s pretty easy to find NFL pick em sites, from ESPN to CBS Sports to even NFL.com, but most of the “celebrity” picks are straight up – who will win – who will lose. And while some of those who have just picked straight up are in the 60% range, there are still quite a few under .500 so far this season.
Of the sites with picks against the spread, most of them are devoted exclusively to gambling and would prefer you pay them to find out their picks and their results. But, I did find picks against the spread on a site I sometimes use for fantasy news and mock drafts – WalterFootball.com. The site has been making picks against the spread since 2008.
Do you want to know what their record is so far this year? Well, its 25-34-1. One win and one push better than I’ve been doing.
So, I’m going to soldier on in my grand experiment to see if advanced football statistics can have a predictive quality and aid in picking against the spread each week.
This week, as usual, I’ll be focusing on the NFC West and the Sunday night and Monday night games. For further explanation of the statistics I’m using, see my Week Two column or visit Football Outsiders on the web.
St. Louis at Philadelphia -7
St. Louis (1-2) is coming off a bye while Philadelphia (3-1) is reeling from a game against San Francisco in which the offense did not score. Philly’s 21 points came entirely from special teams and defense.
The Eagles performance against the 49ers dropped their offensive efficiency from 12th to 22nd. However, the 49ers D ranks 6th in DVOA, whereas the Rams defense ranks 30th. Expect a bounce back by the Eagles offense.
With an overall DVOA that ranks 9th at 10.5% compared to the Rams 29th place ranking of -31.6%, it’s reasonable to assume the Eagles can cover a touchdown. My only reservation is that the Rams are 2-0 under Jeff Fisher coming off the bye. Still, take the Eagles.
Pick – Philadelphia Eagles
Arizona at Denver -7
Both Arizona (3-0) and Denver (2-1) are coming off their bye week.
The Broncos rank 3rd in overall DVOA at 30.3%, while the Cardinals rank 8th with 16.1%, making this one of the better matchups of the week. Denver’s 4th ranked offense going against Arizona’s 3rd ranked defense should be a treat. What might not be so much fun is watching the Cardinals’ 21st ranked offense work against the Broncos’ 4th ranked defense. I’m just not certain the Cardinals can score enough points to cover the 7 points.
Also, its worth considering that the Broncos were 5-3 at home last year against the spread (one of the losses was when Vegas made the Jaguars 23.5 point dogs – a little silly even for a Jags game).
John Fox led Broncos teams are 3-0 coming off the bye. This will be Bruce Arian’s first game after a bye as head coach for the Cardinals.
Once again, I’m taking the home team to win and cover.
Pick – Denver Broncos
Kansas City at San Francisco -6
Kansas City (2-2) thrashed the Patriots 41-14 on Monday night while San Francisco (2-2) reinserted themselves into the NFC West conversation with a win over the Philadelphia Eagles. The 26-21 victory by the 49ers was conveniently a half point under the spread.
The two teams match up nicely on paper with Kansas City ranked 18th in overall DVOA at 2.3% and San Francisco ranked 19th in overall DVOA at 0.9%. The 49ers are 17th on offense while the Chiefs are 19th in defense. But a greater advantage goes to the 49ers defense which is ranked 6th compared to the Chiefs offense that is ranked 12th.
Statistically, it would appear the game could easily be closer than 6 points, but both of these teams make me queasy in that I don’t think either is really that good. But, I’m here to go with the numbers, not my gut. I’ve got San Francisco winning but not covering.
Pick – Kansas City Chiefs
Cincinnati -1 at New England
The New England Patriots have been underdogs at home only one time in the previous 20 games, dating back to December 4, 2011 (20 games is as far back as several sites will allow me to search, so it could be longer). That was +1 against Denver on November 24, 2013. The Patriots won 34-31. But there does seem to be something much different about this year’s Pats, especially on offense.
Cincinnati (3-0) is coming off a bye, while New England (2-2) was humiliated in Kansas City on Monday night 41-14.
Based on three games and weighted 55% with pre-season projections (DAVE), the Bengals rank 1st in DVOA with 41.7%. On the other side of the field, New England’s 23rd ranked overall DVOA of -5.8% is based on four games and weighted 45% DAVE.
The Patriots are decent on defense (ranked 10th) and special teams (ranked 9th), but are surprisingly abysmal on offense, ranking 28th. Tom Brady ranks 32nd at QB in DVOA at -30.2% with a 30th ranked QBR of 46.5. So, while much has been made of the lack of offensive weapons in New England, Brady is simply not Brady, and this week he’s going up against the number 1 rated defense – 1st against the pass but 32nd against the rush. New England has to run the ball to have any chance of winning.
While I have no real confidence in Marvin Lewis, as he is 5-6 with the Bengals when having two weeks to prepare, and it’s really difficult to pick against Brady and Belichick at home, I’m going with the Bengals.
Pick – Cincinnati Bengals
Seattle -7.5 at Washington
Seattle (2-1) is coming off a bye, while Washington (1-3) probably wishes they were on a bye after being destroyed by the New York Giants 45-14 with Kirk Cousins throwing 4 interceptions and coughing up a fumble for good measure on his way to a 13.9 QBR.
I’m actually surprised the line is only -7.5. Look for the line to increase over the weekend as people bet the Seahawks.
Seattle ranks 2nd in overall DVOA with 32.7%. Washington is actually not as bad as you might think following the Giants debacle, ranked 22nd at -5.5%. But, the Seahawks are better than Washington in every facet of the game, ranking 2nd in offense, 8th in defense, 21st in special teams, compared to 16th, 16th, and 31st respectively for the Washington DC professional football team.
This seems like the easiest pick of the week, so watch Washington’s resurgence this week.
Pick – Seattle Seahawks
And just remember, if you’d taken the exact opposite of me last week, you’d be rolling in style this week. Have fun and enjoy the games.
Add The Sports Daily to your Google News Feed!
