For those who missed it (everyone, I’m assuming), I had an article published on The New York Times football blog The Fifth Down titled Keys to Defeating Dallas, From a Cowboys Writer. You can (and should) click on the link to read it.
Before I receive the inevitable backlash for writing such blasphemous content, let me point out that this article is not much different than my traditional “DOs and DON’Ts for Dallas” weekly post. Instead of explaining what Dallas needs to do for a win, I simply flipped the script. "Protect Doug Free” would have been sound advice for Dallas this week, and you will find a corresponding “Attack Doug Free” bullet point in my New York Times piece.
Anyway, here I wanted to take a look at the broader picture, assessing both the Cowboys’ and Giants’ odds of making the playoffs after the contest. I will conduct a schedule analysis, estimating the odds of each team winning their remaining games and determining how that relates to the subsequent importance of this particular matchup for each squad.
In the beginning of the season, the importance of the strength of a team’s schedule is very overblown, and for a variety of reasons.
First, when comparing the schedules of two division opponents, the 14 games which are not head-to-head contain just two dissimilarities, i.e. the Giants and Cowboys had just two opponents which differed.
Secondly, the constant talent flux in the NFL makes preseason predictions in regards to a team’s strength of schedule almost useless. Who would have thought the Raiders, Bengals, Bills, Lions, etc. would not be “easy wins” in 2011?
Lastly, the overall strength of a schedule tends to even out over the course of a 16-game season. Simply put, preseason strength of schedule sucks as a method by which to analyze a team’s playoff hopes.
By Week 14, however, a team’s remaining schedule has a huge influence in their odds of making the playoffs. Over the course of four games, it is certainly possible to have three or four strong/weak opponents, meaning the schedule doesn’t necessarily have time to “even out.” Plus, we already know the relative strength of each NFL team, so estimating each team’s win probability for a single game is far easier.
With all of that said, let’s take a look at the Giants’ and Cowboys’ remaining schedules, along with the probability they win each game:
Using their Generic Win Probability (the odds a team will win a game against a league-average opponent at a neutral site), we can estimate the chances of Dallas and New York winning each remaining game by combining the teams’ GWP and factoring in home field advantage.
Thus, while I am “guessing” the odds of wins for the teams in their remaining contests, those probabilities are likely not too far from reality.
If we give each team a win total that directly correlates with their probabilities, i.e. the Cowboys’ 54 percent chance of winning this week’s matchup equates to “.54 wins,” you can see the Cowboys’ theoretical win total for the rest of the 2011 season is 2.37, while the Giants’ is 2.40.
That difference is obviously not statistically significant, meaning the small errors which are undoubtedly contained within my win probability projections are surely enough to “make up for” this difference.
In other words, the Cowboys and Giants can be expected to win the same number of games moving forward. Note that this conclusion does not mean the teams will win the same number of games, but rather that their chances of doing so are the most likely of all possible outcomes.
So what does this all mean for their playoff chances? The incredible similarity between the teams and their win probabilities moving forward means that each team's respective chances of making the playoffs, even after we factor in the schedule, are roughly the same as if they were both of league-average quality.
Think of it this way: If a team of the same quality of the Colts was 7-5 (like Dallas) and one with a talent level comparable to that of Green Bay was 6-6 (like New York), we’d still expect Green Bay’s odds of making the playoffs to be greater due to their far higher win probability in each game.
The role of chance in that situation is less powerful than in the case of the ‘Boys and G-Men.
In the latter comparison, we have two teams whose win probabilities for the remainder of the season are roughly equal. This means that the chances of the Giants making the playoffs, taking their one-game deficit into account, are roughly the same as if we flipped a coin to determine the outcome of each contest.
Perform 10,000 coin-flip simulations to allow the numbers to regress to the mean, and you will be left with each team’s playoff chances.
Luckily, we have computers which can simulate seasons, so we don’t need to waste time flipping a coin. Over at NFL-Forecast.com, they have determined each team’s chances of making the playoffs based on thousands of simulations with the GWP provided by Advanced NFL Stats.
If you head over there, you will see the odds of Dallas winning the NFC East are around 64 percent, with their overall playoff chances adding up to nearly 69 percent (meaning there is little chance they earn a wild-card spot). The Giants’ odds are 35 percent to win the division and 36 percent to make the playoffs.
In terms of particular scenarios, the Cowboys will win the division if they beat the Giants in both of the clubs’ remaining games. If the Giants perform the same task, they will win the division as long as they do not lose both of their other two games or the Cowboys do not win both of their other games.
Simply put, if one squad wins both head-to-head games, they are effectively the NFC East champs. If the more probable splitting of head-to-head games occurs, the Cowboys are in with one more win.
Let me break it down for you.
If the teams split and the Cowboys lose to Philly and beat Tampa Bay, for example, they would finish 9-7 and possess the same record as the Giants if they beat both the Redskins and Jets. Then, intricate tiebreakers would come into play. As shown on NFL.com, those are:
- Head-to-head (best won-lost-tied percentage in games between the clubs).
- Best won-lost-tied percentage in games played within the division.
- Best won-lost-tied percentage in common games.
- Best won-lost-tied percentage in games played within the conference
The first tiebreaker would obviously end in a tie following a split. If the Cowboys lose to the Eagles and Bucs (and the Giants win against the ‘Skins and Jets), the teams would finish 9-7 and both have a division record of 3-3, meaning the third tiebreaker would be a factor. In common games (all but two, remember), the Giants and Cowboys would again be tied at 9-5 (including their head-to-head split).
Thus, the fourth tiebreaker is needed. Here, the Cowboys have the advantage, as they will have a superior conference record following a split, no matter how the other games shake out.
- If the Cowboys sweep the head-to-head matchups, they are in the playoffs.
- If the Giants sweep the head-to-head matchups, they are almost assuredly in the playoffs. They would miss out if they lose both remaining games and the Cowboys win both.
- If the teams split, the Cowboys are in if they win one of their other games. If the Cowboys lose both of those games, the Giants are in if they win both of their remaining contests.
For Dallas, all of the math equates to one simple conclusion: Win this week, and you will almost certainly be NFC East champs.