As we head into the divisional playoff weekend, I thought I’d see how the eight remaining teams stack up in terms of our little model. Here’s an accounting for the four games taking place this weekend:
The columns on the left show the regular-season differences between the home and away team on each of our three factors — yards gained, sacks, and turnovers — while the second set of columns uses the predicted point value of these differences based on our model. For instance, the Baltimore Ravens outgained their opponents by 28.0 yards on average. That compares to a figure of 21.4 yards for the New England Patriots, for a net 6.6 yard advantage for Baltimore. According to our model, that figure translates into an estimated 0.4 point advantage for the Ravens; the two teams are pretty much even in terms of net yardarge.
However the teams diverge on the other two measures, sacks and turnovers. The famously tough Ravens’ defense has averaged one (net) sack more per game than the Patriots, which is worth about two points on the scoreboard. In contrast, the Patriots hold the advantage in the turnover department, averaging 0.6 (net) turnovers per game more than the Ravens. The Ravens collected eleven interceptions and eleven fumbles in the regular season, but they gave up twelve interceptions and eight fumbles as well. Tom Brady was intercepted just nine times while the Patriots defense collected sixteen. The Patriots also collected nine opponents’ fumbles while losing just four themselves. While the Patriots’ net advantage in turnovers is smaller than the Ravens advantage in sacks, it has a greater effect on the score, since turnovers were worth about five points on average in 2014 while sacks generated a bit over two points.
When we add up all three factors we find the Patriots with a minuscule predicted lead over the Ravens. Of the four games this weekend, this is by far the closest match-up using the teams’ regular-season performance on our three factors.
Looking at the other three games, Indianapolis has the toughest road; our model makes them a nine-point underdog to Denver. The Broncos lead on all three measures, but the biggest gap between the teams comes in the turnover category. The Broncos collected eighteen interceptions and seven fumbles in the regular season, while giving up only five each. The Colts figures were twelve versus sixteen for interceptions, and fourteen versus fifteen on fumbles. The model rates the Broncos’ turnover advantage as worth about five points.
In the NFC, Green Bay’s advantages over the Dallas Cowboys look rather similar to Denver’s advantages over the Colts. The teams were about equal in terms of yards gained, but the Packers have solid leads in sacks and turnovers. Out in Seattle, the Seahawks outgained their opponents by over a hundred yards on average. Comparing that difference to Carolina’s meagre seven yard advantage over its opponents predicts Seattle should score nearly a full touchdown more than Carolina. The Seahawks also gain from a lower turnover rate, but it has a much smaller effect on the point spread than does Seattle’s advantage in yards gained.