There are four clear favorites this week (I thought, see odds below). This makes most peoples picks hard, but mine easy.
"All four favorites won't advance. There will be one upset, but I don't really see more than one happening"-everyone (Bill Simmons and Mike Golic specifically)
Yeah, you're probably right, but even if that is true you can either:
Try to guess which game is the upset and have a 1 in 4 chance at getting all 4 games right, but a 3 in 4 chance of getting two games wrong (the wrong upset pick + the actual upset)
Take all four favorites and lock in a 3 of 4 (75%, better than anyone gets long run)
And that's assuming a favorite loses. They are the favorites for a reason. They are probably going to win. Using Brian Burke's Win Probability from here The odds of all four favorites winning 26.4% (.74 x .80 x .55 x .81)
The odds of the most likely one upset combo 21.6% (.74 x .80 x .45 x .81)
The odds that the popular upset pick (Giants) is right 6.2% (.74 x .80 x .55 x .19)
If someone wants to bet you that all the favorites will win, take it, because then you get all four underdogs working for you. If you're just picking winners though the odds are with the favorites.
So give me the favorites. Pats, Colts, Packers and Cowboys
Bonus odds: Pats v Colts AFC title game next week 59.2% (.74 x .80)
Showing posts with label Playoff Picks. Show all posts
Showing posts with label Playoff Picks. Show all posts
Saturday, January 12, 2008
Friday, January 4, 2008
Wild Card Saturday
Redskins @ Seahawks
The Seahawks are the better tea, but Washington has put together a great streak of games to sneak into the playoffs. The difference in DVOA stats between the 5-7 Redskins team that played the first 12 games of the year and the 4-0 Redskins that won
all of it's last 4 is almost entirely an improvement in third down defense. FO
readers know that a team performing differently on 3rd down than on 1st and 2nd is
likely to regress to the mean (1st and 2nd down performance). So a team playing very well on 3rd down, compared 1st and 2nd is probably going to stop playing over it's ability level soon*. That's the Redskins. The regression might not be this week, but it's (likely) coming. Seahawks.
*=The Peyton Manning led Colts have defied this general trend for most of Peyton's career. This is why FO's preseason predictions underestimate the Colts. It's factoring in a likely regression that isn't actually likely.
Jags @ Steelers
The Jags went into Heinz Field and dominated the Steelers for 3 quarters (22-7 through Q3). Then the Steelers got on track and put together two defensive stops each followed with successful drives , tying the game at 22 (one short drive following a Garrard pick and the other in true 07 Steelers fashion, overcoming a stuffed Parker run on first down and a sack on second, to convert 3rd and 18 to keep the TD drive alive).
The Jags outplayed the Steelers on their own field a month ago. They are likely the better team. They aren't the overwhelming favorite they are being made out to be, but they did win the previous game with, 3 quarters of dominance and held off the desperate comeback. The Steelers have relied on Roethlisberger to perform miracles being a awful O-line, usually in long yardage situation because Willie Parker is incapable of running for any gains between 2 and 15 yards (42% success rate, -11.3% DVOA). Parker is out, so the more consistent (52% success rate) Najeh "The Dump Truck" Davenport is the starting runningback. He could give them an offense that actually sustains drives. It's not the "obvious game" some make it out to be, but the Jags showed that they are the better team a month ago and did nothing to suggest they aren't the same team that beat the Steelers, while the Steelers have gotten worse if anything due to injuries. Jaguars.
The Seahawks are the better tea, but Washington has put together a great streak of games to sneak into the playoffs. The difference in DVOA stats between the 5-7 Redskins team that played the first 12 games of the year and the 4-0 Redskins that won
all of it's last 4 is almost entirely an improvement in third down defense. FO
readers know that a team performing differently on 3rd down than on 1st and 2nd is
likely to regress to the mean (1st and 2nd down performance). So a team playing very well on 3rd down, compared 1st and 2nd is probably going to stop playing over it's ability level soon*. That's the Redskins. The regression might not be this week, but it's (likely) coming. Seahawks.
*=The Peyton Manning led Colts have defied this general trend for most of Peyton's career. This is why FO's preseason predictions underestimate the Colts. It's factoring in a likely regression that isn't actually likely.
Jags @ Steelers
The Jags went into Heinz Field and dominated the Steelers for 3 quarters (22-7 through Q3). Then the Steelers got on track and put together two defensive stops each followed with successful drives , tying the game at 22 (one short drive following a Garrard pick and the other in true 07 Steelers fashion, overcoming a stuffed Parker run on first down and a sack on second, to convert 3rd and 18 to keep the TD drive alive).
The Jags outplayed the Steelers on their own field a month ago. They are likely the better team. They aren't the overwhelming favorite they are being made out to be, but they did win the previous game with, 3 quarters of dominance and held off the desperate comeback. The Steelers have relied on Roethlisberger to perform miracles being a awful O-line, usually in long yardage situation because Willie Parker is incapable of running for any gains between 2 and 15 yards (42% success rate, -11.3% DVOA). Parker is out, so the more consistent (52% success rate) Najeh "The Dump Truck" Davenport is the starting runningback. He could give them an offense that actually sustains drives. It's not the "obvious game" some make it out to be, but the Jags showed that they are the better team a month ago and did nothing to suggest they aren't the same team that beat the Steelers, while the Steelers have gotten worse if anything due to injuries. Jaguars.
Subscribe to:
Posts (Atom)
