Football Outsiders: Is it worth benching Romo in the finals?

Championship week presents a conundrum to even the best fantasy owners: Do you take advantage of a late-season matchup that could yield a huge game, or do you start the guys that brought you here and hope that they keep it going?

The answer is no different than any other week: It depends. Week 16 of last year yielded some huge busts from fantasy studs. Adrian Peterson had four fantasy points against the Redskins. Fantasy MVP Tom Brady mustered 17 points, one of his worst totals of the season. Vince Young had four points; Jay Cutler had one. Eli Manning, weeks away from leading the Giants to the Super Bowl, busted out a minus-3.

Nothing is guaranteed from these guys, which is why playing the matchups is the ideal option. Take Tony Romo this week, for example. Romo plays a Ravens pass defense that is second in the league in defensive pass DVOA (Defense-adjusted Value Over Average), performing 28.5 percent better on defense than the average team. (Because it's a defensive figure and fewer yards and points are the ideal situation, their defensive DVOA is actually -28.5 percent). The DVOA of the Cowboys' pass offense this year is 10.6 percent; that includes the time without Romo, so we bump up the figure some in our regression analysis thanks to Romo's individual DVOA of 22.4 percent.

One of the advantages of DVOA is that it measures teams against the average, so teams from different years are more easily comparable. With that being the case, our matchup system looks into the past and measures how similar pass offenses have done against pass defenses of the Ravens' caliber.

The average quarterback under these conditions produced 16.8 fantasy points, ranking an average of 13th the week they faced such a defense with this offense behind them. Not bad, but certainly worse than we'd expect Romo to perform. Romo has produced an average of 20.6 fantasy points per start this year; the difference in the two numbers is 19 percent. After accounting for the myriad injuries facing the Ravens' defense, the fact that the Cowboys are at home and Romo's performance in previous seasons, that figure is down to 13 percent, which is the expectation for how much worse Romo will perform than he would against an average defense.

Is it possible that Romo could have a better game than expected? Of course -- football is not played on paper or on a spreadsheet. But he could also do worse. The primary reason you draft players in fantasy football is because you expect them to score more points than any other player available. Those points aren't evenly distributed from week to week, though; if Adrian Peterson has 304 points this year, for example, you don't get 19 fantasy points a week; you get 28 some weeks and 8 some other weeks. Identifying the weeks in which Peterson is more likely to get eight points than 28 and playing a backup who's more likely to score 20 points turns the 304 points a year you'd get out of Peterson's spot into 350 or 360 points. Just because it's Week 16 doesn't mean those facts are any different.

Best and worst matchups for Week 16


Drew Brees (Saints, plus-15 percent): Expecting this figure to be higher against the woeful Lions? Yeah, we were, too. Then we noticed Brees' astounding splits; at home, he's averaging 25.7 fantasy points per game. On the road, he drops all the way to 16.6 PPG. His touchdown-to-interception ratio at home is a pristine 19-4; on the road, it's 9-12. It's still the Lions, so expectations are high, but there's the possibility that Brees just ends up having a good day as opposed to winning you a title by himself.

Kurt Warner (Cardinals, plus-14 percent): Warner may struggle some with the weather Sunday -- temperatures in Foxborough are expected to be in the mid-20s, with snow possible -- but the Patriots' secondary is so abysmal that he should be able to throw the ball downfield successfully wearing 15 gloves if he wants to. On the other hand, it's a small sample, but the only two December games in cold weather our database shows Warner playing in are against Philadelphia in 2002 and Seattle last year; in those games, Warner threw three touchdowns against seven interceptions. We're going to project that the Pats' corners are worse than the weather.

Brett Favre (Jets, plus-14 percent): No obvious hazards for Favre; the Seahawks' pass defense remains among the worst in football, and Favre should have eons of time to throw Sunday considering Seattle's paucity of sacks this year.

Tony Romo (Cowboys, minus-13 percent): We addressed this above; Romo's day could very well be similar to his performance against the Steelers, a similar defense, in Week 14.

Kerry Collins (Titans, minus-12 percent): You're either in a really deep league or one that rewards checkdowns and swing passes if you're drafting Collins. He had a poor game a week ago against a weak Texans defense, struggling when they sold out on stopping Chris Johnson. Whether Collins can adjust to that is yet to be seen, but the Steelers have been able to stop pretty much every offense they've faced this year. In a game that most likely will determine who wins the AFC regular-season title, don't expect Collins to be putting the ball up for grabs very often.

Matt Ryan (Falcons, minus-7 percent): There's a huge gap between Collins and Ryan, who merely has a difficult matchup, not an impossible one. Minnesota's pass defense has finally caught up with its rush defense this year, likely owing to the impact Jared Allen has had on their pass rush. If Allen can't go Sunday, Ryan's matchup becomes an average one; if he can make it, he could make it a long day for the combination of Todd Weiner and Sam Baker. That would make things difficult for Ryan.

Running backs

Pierre Thomas (Saints, plus-23 percent): Sometimes, everything falls into place. Deuce McAllister is on the outs. Reggie Bush went on IR. So Thomas gets to play the Lions, the worst rush defense in the league, essentially all by himself. If you mined the waiver wire for him a few weeks ago, this week is your reward.

DeShaun Foster (49ers, plus-16 percent): If you didn't mine the wire for Thomas and need a running back this week, Foster might very well be your huckleberry. Frank Gore missed practice Wednesday and isn't likely to go Sunday. Foster, meanwhile, remains available in an astounding 88.6 percent of ESPN leagues. This is absolutely an all-or-nothing play; if Gore plays, Foster is likely useless. If Gore is out, Foster has a fantastic matchup against a Rams defense that has allowed an average of just under 17 fantasy points per game to the opposing team's top back during their eight-game losing streak. Gore himself had 20 points against them in Week 11. If you need to set your lineup Thursday, Foster is too dangerous a play; if you can wait and see whether Gore's declared out on the Friday injury report, stash Foster on your bench and wait as late as you can Sunday before deciding whether you can insert him in the lineup.

Justin Fargas (Raiders, plus-13 percent): Another option would be the Raiders' starting running back, who has had 15 or more carries in five of the past six weeks. His biggest problem is that he has only entered the end zone once, but the Texans' rush defense has a way of solving those issues. As a flex play or even as a No. 2 running back in deeper leagues, Fargas could really surprise people this week.

Steve Slaton (Texans, plus-12 percent): Slaton should also be able to make some noise when Fargas is on the sidelines; the Raiders' run defense has gotten worse as the year has gone along, and they allowed the Patriots 277 rushing yards a week ago. That's Sammy Morris and LaMont Jordan -- not exactly two legends. Slaton should have a very nice day, and is a must-start in all leagues.

Marion Barber (Cowboys, minus-19 percent) and Tashard Choice (Cowboys, minus-19 percent): Don't do it to yourself. Barber is very clearly not 100 percent, although he can't be worse than he was a week ago. Choice had 166 yards against the Steelers, but 50 of them came on a blown coverage. The Ravens' record against running backs has been laid out in this column in the past; it would be a small miracle if either of these backs had a game worth putting in your lineup.

LenDale White (Titans, minus-15 percent) and Chris Johnson (Titans, minus-12 percent): As usual with Johnson, he's less likely to suffer from the effects of a good run defense because he can get outside and catch the ball. The Steelers' front seven, on the other hand, shouldn't let White get too far past the line of scrimmage.

Clinton Portis (Redskins, minus-10 percent): Portis has 131 yards in his past three games. Combined. Whether that's due to facing two great run defenses (and the Bengals), his knee injury, his tussle with Jim Zorn or a combination of the above, he doesn't have much going for him into this matchup against the Eagles. Portis had 145 yards against them in Week 5, but without Chris Samuels, expect Portis to struggle to 75 or 80 yards.

LaDainian Tomlinson (Chargers, minus-9 percent): If you own Tomlinson, you probably don't need us to convince you to bench him at this point. His name value is worth approximately zero fantasy points. His actual performance has been worth 13 points a game this year; now he faces a Tampa Bay run defense that, the past two weeks aside, has been very good. Start him at your own peril.

Wide receivers

Marques Colston (Saints, plus-17 percent): After last week's "Double everyone and leave Dallas Clark wide open" strategy failed miserably, expect the Lions to revert back to the same boring Tampa-2 shell that teams have exploited all year. The Tampa-2, in general, gives up bigger plays to the opposition's No. 1 receiver than standard defenses. Colston should have a huge day Sunday. Huge.

Calvin Johnson (Lions, plus-12 percent): Calvin Johnson, of course, always has huge games. The Saints' secondary is no great shakes, and that doesn't even consider that Johnson will be up against either 5-foot-11 Randall Gay or 5-8 Jason David in man coverage all game. Both should have all the impact of gnats.

Braylon Edwards (Browns, plus-12 percent): We know, we know. Eight times bitten, twice shy. Edwards actually had 100 yards last week, and the Bengals don't have anyone who can cover him one-on-one. To quote another famous Edwards, you play to win the game, people. Braylon is going to pop off this week.

Terrell Owens (Cowboys, minus-10 percent) and Roy Williams (Cowboys, minus-10 percent): Not that either of them have been particularly great recently, but this isn't a combination that should fare particularly well against the Ravens' defense. Granted, their secondary is a little weak at the moment, but the pass rush should prevent either of them from getting too far downfield.

Chris Chambers (Chargers, minus-9 percent): Each week, we remind you that the second wideout of the team facing the Buccaneers has an awful matchup. Each week, they fail to perform. Chambers, whose stock has fallen dramatically since returning from injury and forgetting to catch a touchdown every other pass, is no different.

Michael Jenkins (Falcons, minus-8 percent): Jenkins was, in fact, that exact guy a week ago -- he had 55 yards. Nothing you needed to have in your lineup. This week, he faces another excellent defense against No. 2 wideouts, the Vikings. A similar level of performance seems likely.

Tight ends

John Carlson (Seahawks, plus-13 percent): Carlson continues to emerge as a supremely underrated tight end and the Seahawks' second target in the passing game. If your league hasn't picked up on it yet, Carlson has a very good matchup against the Jets this week.

Antonio Gates (Chargers, minus-10 percent): We understand that it'll be close to impossible to find a better tight end than Gates on such short notice. Fair enough. Just keep in mind that the Buccaneers do a great job in coverage on tight ends, and that Gates shouldn't be expected to do a ton.

Bill Barnwell is an analyst for FootballOutsiders.com.