TABRankings 2012: Postseason

With the 2012 Regular Season TABRankings posted, the basis for rankings each NFL team is now established. However, before one can move ahead to 2013, there must be some postseason adjustments. Obviously, every team plays to get into the postseason in hopes of surviving the tournament to win a Super Bowl championship. While the spirit of TABMathletics would not allow the Super Bowl XLVII champion Baltimore Ravens to be named the unquestioned number one overall team, they should get a significant boost for making the most of the “second season.” That’s what these rankings will do.

To see what adjustments need to be made, each individual statistic should be considered. Will there need to be simple sum adjustments (re: adding on the postseason total to the regular season total), or will there need to be an overall rate that combines the regular season and postseason numbers (with some extra weight applied to the postseason)? Remember, each team made the postseason, so it’s an added bonus. Teams should NOT be penalized significantly for losing in an extra season:

• Overall Win Percentage: People refer to the NFL postseason as the “second season” because the playoffs contain in entirely different beast. In the regular season, a team advances to the next level by either winning the division or finish with one of the two best records in the conference among non-division winners.In the postseason, a team advances by winning a head-to-head matchup. Respect the playoff run! This stat should use a simple sum adjustment. (Example: The 2012 Ravens went 10-6 in the regular season and 4-0 in the postseason. Their non-weighted totals will be 62.5 points and 100 points, respectively.)
  **NOTE: For the sake of rewarding teams for earning a postseason bye week, that week will count as a victory in these rankings.**
• Pythagorean Win Percentage: There will be a simple sum adjustment made to this statistic because a loss would hurt this stat’s points total if it’s applied as an overall rate adjustment. Also, with more games in this sample, the mean converges closer to 50 points. Theoretically speaking, playing in the postseason hurts teams in the stat if it’s applied as an overall rate adjustment.
• Divisional Win Percentage: Consider three things. First, playoff games rarely feature intra-divisional games. There can only be a maximum two such games in each conference. Second, a playoff game shifts the dynamic of the conference structure, but doesn’t shift the division structure. The 2007 Giants defeated the Cowboys in the NFC Divisional round, but the latter team still win the 2007 NFC East. Third, there needs to be a concern for overvaluing a playoff win. Is it fair to reward teams for defeating a divisional opponent in the playoffs significantly more than a team that defeats non-divisonal opponents? No, therefore it would be best to use an overall rate adjustment with no extra weight to playoff games. (Example: The 2012 Packers went 5-1 against the NFC North in the regular season and defeated the Vikings in the Wild Card round. Their non-weighted total will be 85.71 points to reflect the team’s 6-1 overall record against the NFC North in the 2012 season.)
• Quality Win Percentage: Save for the 2010 Seahawks, every postseason team ever counts as a “quality opponent” in these rankings. Therefore, this stat will use an overall rate adjustment with no extra weight to playoff games.
• Third/Fourth Down Differential: Teams very frequently arrived at the third or fourth down situation during the regular season. In that sense, the outcomes were normalized. Because this is a basic stat, it doesn’t hold as much importance as a normalized record stat like Pythagorean win percentage. Furthermore, of the basic stats, this does not correlate as strongly to victory as the non-normalized turnover margin. Therefore, this stat will use an overall rate adjustment with no extra weight to playoff games.
• Turnover Margin: Because turnover margin is a quantitative stat that simply requires counting, there is no choice but to make a simple sum adjustment. There will definitely be a weighted emphasis on the postseason, though.
• Fumbles Kept Differential: Here’s another quantitative stat, so once again a simple sum adjustment will be made.
  
• Sack Differential: Once again, this quantitative stat will be a simple sum adjustment.
• Time Ahead Average: As the rankings move to the advanced stats, there will be no more quantitative stats. However, this stat is only a rate because the total was intended to converge to a certain number as the regular season weeks increase. Thus, there can be a simple sum adjustment without using an average for the time ahead.
• Time Ahead Big Average: This stat is a strict subset of the previous one, so this too will use a rate adjustment. Getting a big lead in the postseason is a bit more impressive than doing so in the regular season, no matter who the opponent may be. Therefore, this stat will use a heavier-weighted simple sum adjustment without using an average for the time ahead.
• Real Quarterback Rating Differential: QBRD perhaps does the best job of any stat to show correlation to victory, especially in the postseason. A whopping 85.2 percent of teams with the QBRD in 2012 won their games. This includes 9-2 postseason record. It would seem THAT important to make it a simple sum adjustment, but singular-game QBRD can get fluky. Therefore, this stat will use an overall rate adjustment with twice the weight added to playoff games.
• Relative Points Average per Game: This stat creates an issue different than all of the above, simply because it focuses on relativity. Remember, the Pythagorean win percentage adjustment will be used as a simple sum adjustment, so that makes working on a relative points average adjustment via an overall rate adjustment an enviable task. However, it should be noted using relative points average per game singularly for the postseason creates wayward statistical results. With all these adjustment already made, there will be no adjustment made to this stat.

The TABRankings formula:
(Regular Win Pct x 2)+(Post Win Pct)+(Regular Pyth Pct)+(Post Pyth x 0.2)+(Overall Div Pct x 0.5)+(Overall Qual Pct x 0.5)+(Overall 3rd/4th Down Diff x 2.25)+(Regular TO Margin x 3)+(Post TO Margin x 5)-(Regular FK Diff x 0.75)-(Post FK Diff)+(Regular Sack Diff x 0.75)+(Post Sack Diff)+(Reg Time Ahead Avg x 1.2)+(Post Time Ahead x 0.05)+(Reg Time Ahead Big Avg x 1.5)+(Post Time Ahead Big x 0.1)+(Weighted Real Quarterback Rating Diff x 1.2)+(Regular Relative Pts Avg x 1.8)

Of course, we need to apply strength of schedule adjustments. For the purpose of these rankings, these adjustments will feature both the regular season and postseason games. This should be the best way to reward the playoff teams.

The TABRankings opponent adjustment for the 2012 postseason:
(Gross Opponent TABR / Games Played) – League Average TABR
NOTE: The multiple will be determined to be a number X near to the variable that fits the equation:
(1/6) * TABR IQR = Y * Adjustment IQR. For these results, the multiple was determined to be 1.85.

With now all that put together, the results for the 2012 postseason are available below. Note that “RS” refers to team’s ranking from the 2012 regular season. What’s the concern is how little change there was in the rankings. Much of this had to do the upset-less Wild Card Weekend. In fact, outside of the Ravens, no team really did much beyond expectation, according to TABR. We’ll wait until next year to see if this postseason formula truly works.

Meanwhile, stay tuned for the 2013 TABRankings, which will be published weekly during the upcoming regular season!

The TABRankings top eight:
1, 2, 3, 4
5, 6, 7, 8

True TABRankings — Postseason