(Ed. Note: Cold, Hard Football Facts reader Mike Stickles, one of our most fact-filled e-mailers, recently ran a study of data of the past year or so of games to discover which mattered more, establishing the run or establishing the pass. Once again, the findings of a fact-filled study skewer conventional wisdom and will come as a huge disappointment to Merril Hoge and Co. from the "establish the run" crowd. Stickles finds that "Establishing the run" is about five shades of meaningless. Check out his spreadsheet of all kinds of cool data here.)
By Mike Stickles
Cold, Hard Football Facts reader with nothing better to do
I know you guys have already skewered the myth that teams need to "establish the run" in order to win.
But just for fun I thought I'd give that one a serious detailed analysis. Turns out, you're right on the money (no surprise there).
I figured a team's performance in the first half was a good measurement of what they "established."
I checked first-half statistics for both teams in the last 218 regular season games, from Week 9 2006 through Week 5 2007 (nothing magical about that selection, I just got tired of reading play-by-plays and figured I had enough data for good results).
The summary stats are in the attached spreadsheet, but here's the upshot.
Establishing the run might be the most meaningless thing a team can do ... UNLESS they can establish a truly dominant running game (to the tune of 120+ rushing yards in the first half).
Teams which ran for 120 to 139 yards were 9-5 (.643); teams which ran for 140 or more yards were 9-2 (.818). That's only 25 out of 436 team performances (5.7 percent). Obviously, not many teams can dominate the first half on the ground (using 120 or more yards in the first half as the definition of dominate).
Teams in the vast middle land of 20-119 yards went 187-183 (.505), with almost no variation in chance of winning between low and high yardage within that window. Teams stuffed for less than 20 yards in the first half went 13-28 (.317).
Yards per attempt rushing has even less variation. Teams that were totally incompetent, meaning less than 1.0 YPA, went 2-7 (.222). At 1.0 YPA and above, records were close to .500 in every grouping. 
And just running better than your opponent didn't mean much, either. Teams rushing for more yards were 115-100 (.535); teams with a more yards per attempt went 107-108 (.498). Run for more yards or run for more yards per attempt, and it really just doesn't seem to matter.
So, when it comes to running the ball, all you have to establish is that you're not totally incompetent. Beyond that, if you can't be totally dominant, it's not worth worrying about.
Passing, on the other hand, has a real correlation to winning football games. Teams with more net passing yards (gross passing yards minus yards lost on sacks) at halftime went 137-78 (.637).
Passing yards per attempt is even more meaningful. Teams with more passing yards per attempt (using the CHFF formula) went 148-70 (.679). 
In terms of absolute performance, each stat broke down with surprising neatness into three divisions.
Passing yards:
  • Teams passing for less than 70 yards in the first half went 41-84 (.328)
  • Teams with 70 to 139 yards went 114-105 (.521)
  • Teams with 140 yards or more went 62-29 (.681) 
Passing yards per attempt:
  • Teams with fewer than 2.0 YPA went 5-21 (.192).
  • Teams with 2.0 to 5.99 YPA went 81-119 (.405).
  • Teams with 6.0 YPA a or higher went 132-78 (.629)
Even better: in the 62 times teams managed to exceed both 140 yards passing and 6.0 YPA, while not throwing any interceptions, they went 46-16 (.742). Exclude the four games where strong passing teams played each other, and the record was 42-12 (.778). 
Strong passing might not be quite as big a winning edge as a totally dominant running, but far more teams can pass the ball well than can generate a truly dominant running game. As noted above, over the last 218 games, just 25 teams (5.7 percent) have been truly dominant on the ground (using my definition of 120-plus yards in the first half).
The only thing more important than establishing a strong passing game is (of course) establishing a lead, preferably a big one.
Teams with a halftime lead went 156-45 (.776). Teams which led by 8 points or more at the half went 77-9 (.895).
There are some other stats noted in the summary spreadsheet (big plays, SPR, etc.).  I hope Cold, Hard Football Facts readers get something out of it.