Wednesday, July 05, 2006
MAS NUMEROS - QUE LASTIMA!
- - - - - - - - - - - - - - -
For all claims and evidence in this post and others, readers are encouraged to remember: skepticism is the antidote for the confirmation bias
SMQ prefers to keep comments in their place, but a recent response to his rather haphazard analysis of big passing stats' correlation to victory drew a subsequent barb requiring a response too illustrative to be confined to the bickering undercard.
Because he's criticized the commenter before, SMQ wishes the "rebuttal," as it were, had come from someone other than Heisman Pundit, because there's none of the common petty friction between SMQ and HP and no desire to create any by piling on.
But then, SMQ's conclusion that big passing stats are often - though certainly not always - synonomous with losses, and at a much higher rate than big rushing performances are, ended with a "for example" throwaway line warning stat-watchers to beware big totals from Arizona State quarterback Sam Keller against "some generous PAC Ten defense this fall," a phrase that rubbed HP the wrong way:
The short answer is, 'It meant he lost.' And he did: assuming HP is refering here to Arizona State's opener against hurricane-rocked LSU last September, the only SEC opponent Keller has faced in his career, the Sun Devils fell to "that crappy SEC defense," 35-31, despite Keller's very impressive 461-yard, four-touchdown, zero-interception passing night.
The more complete answer is, "It meant he threw for 461 because he lost." Or rather, because he was losing.
It should be noted that it's certainly not Sam Keller's fault his team was losing: the Devils were comfortably ahead entering the fourth quarter, 17-7, when their very special teams proceeded to give up consecutive blocks on a field goal and a bizarre, screwed up-looking punt, both of which LSU ran in easily for touchdowns and a sudden 21-17 lead.
Looking at the play-by-play, the Tigers went ahead with 13:27 remaining in the game. From this point, the possessions went as follows:
ASU: 8 plays, 75 yards, Touchdown (ASU 24, LSU 21)
LSU: 7 plays, 80 yards, Touchdown (LSU 28, ASU 24)
ASU: 13 plays, 80 yards, Touchdown (ASU 31, LSU 28)
LSU: 10 plays, 90 yards, Touchdown (LSU 35, ASU 31)
ASU: 6 plays, 53 yards, Turnover on downs
That's three drives for Keller after his team fell behind in the fourth quarter, and his passing total on those final drives: 210 yards. Which leaves 251 yards on ASU's previous nine possessions during the first 46:33 of the game, when his team was ahead or tied.
Keller should be commended for his late heroics, incomplete as they were (SMQ did, in fact). But if not for his teammates' colossal kicking game gaffes, the heroics would have been entirely unnecessary: ASU would have remained on top, turned to a more conservative gameplan, and Keller would have finished with around 280 yards passing and two touchdowns. Still not a bad night, and his team would have won. You'll also recall that quarterbacks who finished in this range, between 250 and 299 yards in any given game, had one of the best aggregate win percentages of any yardage total.
Instead, ASU fell behind, and he had to pull out every stop to catch it up. Voila: 461 yards, a bunch of touchdowns. And an 'L.' Right in line with the general slight losing trend of single-game 400-yard passers in 2005 as a group.
What we also know about Arizona State last year is that the Devils as a team - with both Keller and Rudy Carpenter - threw for an average of 380 yards in seven wins. That's a lot. But they also threw for an average of a little over 364 in five losses - still a lot; a difference of only a little under 4 percent between wins and losses, in fact.
On the other hand, ASU rushed for more than 174 yards per win, to 105.4 in losses. That's a difference of about 40 percent (and 1.27 more yards per carry in wins, too, if you were wondering). In the game in question, Arizona State rushed for 99 yards, LSU for 190.
SMQ thinks this example wholly, completely encapsulates his point in its purest form, and regrets only that he hadn't entered it as evidence sooner. Very good of you to bring it up, HP.
- - - - -
Since most college teams strive to achieve offensive balance in their game plans, it's easy to see where large rushing totals will produce more victories than large passing totals. As you've no doubt pointed out, these results will also depend on a team's offensive identity, its key players, mismatches identified throughout the game, and the game situation.
We should have teams' bloggers study this issue (passing/rushing totals in regard to game situation) in depth throughout the season and share the results.
We should also keep in mind that last year, two of the most balanced teams offensively played for a national title (UT 235 pass/275 rush, USC 315 pass/260 rush).
We should explain the concept of good defense=winning (in most cases) to HP sooner or later as well. Sometimes I wonder if he realizes defense is actually played in college football. If defense didn't mean anything, then Washington St. would win more than four games in a season in which they averaged 490 yards a game (275 pass/215 rush). That team had the third best offense in the Pac-10 (not far behind ASU, 365 pass/140 rush).
High passing stats = loss
There are teams that do nothing but pass and exercise no ball control. I can clearly see how that would lead to a loss. Clearly the goal most teams should be to have a balance attack that leans to a running attack.
Loss = High passing stats
In the ASU example losing produced the high passing stat. Going into catch-up mode was pretty much the option at that late point in the game.
First time to view your blogg and am impressed, it's quality shames most professional sports writers.
This is one inherent difference between running the football and passing the football - every running play keeps the clock going (if the player stays in bounds), whereas an incomplete pass stops the clock. This has nothing to do with the play selection, personnel or scheme. It's a rule of the game that dictates what teams do. And it dictates the game in two very important ways: (1) if a team wants to keep the ball from the other team, it has to chew up the clock, and to do that, running the ball is a better method than passing in general. (2) And more importantly, when a team is up as the game progresses into the 4th quarter, it will rather run out the clock than take chances to score more points. The game of football, like a lot of sports, is based on a set amount of time (unlike baseball), and the clock is a team's biggest ally when they're winning late.
So this is how I think the rule affects the game (and it's really pretty simple): teams that are winning, run the ball more to keep the clock going. This increases their running stats and decreases their chances of getting more yards via passing. But why are they winning the game? It's simply because they have more points vs. their opponent at a juncture of the game. But by no means do we know HOW they got there. They could have gotten there because their offense was simply on fire that day. And it could have been either the running game or passing game or both. OR, it could have been their great defense. Alternatively, it could have been the other team's ineptitude on either offense OR defense. Lastly, it could have been turnovers or special teams. So the score could have been 14-0 or 28-14, but the effect is the same if you're entering into the 4th quarter; you want to run out the clock by running the football
The problem with trying to correlate offensive stats with winning percentage is that it leaves out defense and special teams altogether, and those two components, I believe, determines how much a team runs or passes the football because, again, running = keeping the clock running and incomplete passes stops the clock.
Imagine if the rules were different and that an incomplete pass KEEPS the clock going. If that were true, I don't think teams would run the ball as much in the 4th quarter even when they're winning, knowing that a QB can always throw the ball away if he doesn't find a receiver and the clock will keep running regardless. Thus, I believe this arbitrary rule affects play calling which in turn affects stats.
Maybe a good stat is how many rushing yards a team gets in the 4th quarter vs. how many passing yards a team gets in the 4th quarter as a better determative factor on winning because of the rule difference. Like you pointed out in the ASU game, they had to pass to play catch up so the passing stats were inflated. They passed because they were losing/had less points. And an incomplete pass stops the clock. The rule dictates how a team tries to come back and how another team tries to run out the clock
So I'm saying that it may not be so much the ACT of rushing or passing the ball as much as its EFFECT on the clock due to football rules.
The same goes for basketball or soccor. When you're winning, you keep possession and run out the clock and worry about scoring less. When you're losing, you take chances and shoot as many times as you can. In football, running out the clock just so happens to be done by rushing better than passing. That's why I believe there's a stronger correlation between running the ball and winning.