1. Welcome to CowboysZone!  Join us!  Come on!  You know you want to!

Off-Season Priority #1: Find A Running Game

Discussion in 'Fan Zone' started by TheFinisher, Jan 1, 2013.

  1. AdamJT13

    AdamJT13 Salary Cap Analyst

    15,016 Messages
    493 Likes Received
    Not necessarily. All that matters is that you score. If you get a first-and-goal at the 1 and run four straight times before punching it in, that counts just as much as if you score on first down. If you run twice, then throw a TD pass, that counts, too. If you throw a TD pass on first down, that counts, too.

    Keep fishing, though.
  2. Zman5

    Zman5 Well-Known Member

    5,653 Messages
    1,344 Likes Received
    You can say the same thing about ANYPA formula Adam is using. His fomula doesn't account for how the Ints were thrown.

    The team that lost could have had higher ANYPA all through out the game but in the end come out with lower ANYPA because of the last minute hail mary int.

    You can claim the winning team won because it passed better according to ANYPA but that's not completely true.
  3. phildominator

    phildominator Member

    679 Messages
    1 Likes Received
    Regardless of how efficiently we pass, it still has to be better than our opponents. in other words, even if we field a 'D-' in passer efficiency, we'll beat our opponents with the passer efficiency of an 'F'.

    To make the cowboys win more games, what facet of Passing Efficiency do we need to improve - our offenses own passing efficiency, our defenses ability to decrease opponents' passing efficiency, or just timing?

    1. How does our season overall passer efficiency stack up versus the 8 division winners? If we're comparable, then our offense seems to good enough to win games.

    2. If our opponents' passing efficiency averaged over the season scores high, then we need to improve our defense.

    3. Timing - we may very well have a case where both our offenses and defenses passer efficiencies are fantastic, we just have bad timing for when it occurs.

    Say in Game #1, Even if we have an 'A-' in passing efficiency, if our opponents' passing efficiency was an 'A+', we're going to lose 80% of the time. Then, in Game #2, our defense improved and opponent's passer efficiency was a 'B', but then our offense had bad timing to put up a 'C+'. Over the course of a season, our efficiencies look sufficient, but we still lose too many games.
  4. phildominator

    phildominator Member

    679 Messages
    1 Likes Received
    The point is that a team with a "higher" ANYPA throughout the game wouldn't need to throw a last minute Hail Mary.


    This scenario only seems plausible if the two teams' ANYPA were close and being the "higher" ANYPA doesn't make a difference. Or if the higher ANYPA team was losing the turnover battle or Devin Hester goes for 2 punt return TDs on us...but that's why it was stated 80% of the victors have a better rating, not 100%.
  5. AdamJT13

    AdamJT13 Salary Cap Analyst

    15,016 Messages
    493 Likes Received
    First of all, the flaw in xwalker's ridiculous formula isn't because of a correlation, it's because it doesn't measure what he wants to claim that it measures. So it's not even close to the same thing as the flaws in most statistics, such as the "cheap" Hail Mary interception, or the interception that wasn't the quarterback's fault affecting his passer rating, etc.

    Secondly, in most games, the ANYPAs for both teams aren't close enough that one fewer interception by the losing team would make any difference in which team ended up with the higher ANYPA. In only five out of the 256 games this season would it have changed who finished higher. Even if all five of those were because of a "cheap" INT at the end of the game, that would hardly affect the correlation -- it would mean that "only" 77.5 percent of the teams that had a higher ANYPA won the game.

    It's also more likely that the losing team could have had a lower ANYPA all game but bumped it up in the final minutes with meaningless yards against a prevent defense. But it probably happened only a few times, and it likely is balanced out by the cheap stats going the other way. In the end, the correlation probably doesn't change much at all -- and probably is affected even less by those types of things than by all of the other oddities that can happen in games and affect who wins and loses.
  6. AdamJT13

    AdamJT13 Salary Cap Analyst

    15,016 Messages
    493 Likes Received
    Here are our game-by-game ANYPA numbers for offense and defense --


    Code:
    [B]Opp.	Offense		Defense		Net[/B]	
    NYG	7.194		4.600		2.594	W
    SEA	4.707		5.227		-0.520	L
    TB	4.302		0.733		3.569	W
    CHI	2.444		9.962		-7.518	L
    BAL	5.324		8.370		-3.046	L
    CAR	6.676		3.821		2.855	W
    NYG	2.924		4.600		-1.676	L
    ATL	8.417		8.595		-0.178	L
    PHI	6.103		5.047		1.056	W
    CLE	3.526		4.703		-1.177	W
    WAS	4.766		7.400		-2.634	L
    PHI	9.828		6.543		3.285	W
    CIN	3.478		3.263		0.215	W
    PIT	7.326		5.659		1.667	W
    NO	8.800		8.415		0.385	L
    WAS	1.179		4.579		-3.400	L
    
    Avg.	5.437		5.720	

    As you can see, the team that had the higher ANYPA won 14 of the 16 games (and both of the other games were decided overtime). We had some losses when we passed OK but were not good enough on defense and some when we didn't pass well. One thing that sticks out, though, is that we had only one game when our opponent's ANYPA was really low. Teams finish with an ANYPA of 3.2 or worse almost 20 percent of the time, but we held only one opponent to a number that low.
  7. xwalker

    xwalker Well-Known Member

    12,941 Messages
    5,330 Likes Received
    What is your equation?
    Are these your definitions of those terms?

    unadjusted yards per pass play = (total passing yards) / attempts

    pass success rate = completions / attempts
    If you can correlate without a complicated formula like the simple "equations" above, then this is a non-issue.
    My equation was intended to be silly to make a point. The point being that you can manipulate equations to get a desired result.

    ANYPA is defined as passing efficiency by a group of people. Stating that it is an absolute measure of passing efficiency is misleading.
    The point of my equation is that randomly defined equations can be wrong even if they correlate to a set of data.
  8. xwalker

    xwalker Well-Known Member

    12,941 Messages
    5,330 Likes Received
    A great set of data to analyze would be from all the games where neither team throws an interception.

    Does passing yards or passing average or some other simple stat correlate to winning in those games?
    That is not my impression of what you have been saying up to this point.

    I think most people would agree that the passing game in football has more importance than the running game in the modern era.

    I think people could believe:

    importance of passing / importance of running = 60/40

    I don't think anyone is going to believe:

    importance of passing / importance of running = 99/1
  9. Doomsday101

    Doomsday101 Well-Known Member

    77,946 Messages
    3,448 Likes Received
    When you can do both very well you take the predictablity out of the offense. For some teams that running game means a lot more than to other teams it is part of who they are as an offense and means a lot in terms of their winning and losing. Stats are great but watch the games and the importance some coaches place on their ability to run.

    Funny Aikman talked of the importance of it during the Skins vs Seahawks but then what does he know right? I doubt he is a stats guy. :laugh2:
  10. percyhoward

    percyhoward Research Tool

    8,503 Messages
    2,494 Likes Received
    We get it -- it doesn't matter how you score. None of those situations has anything to do with the one I described.

    If you go 10-of-18 instead of 6-of-18, those were necessarily four more (different) drives that you scored on.

    Four touchdowns that you otherwise would not have had, unless for some reason you want to say that all 4 would have been passing TD.

    Address the points that are being made, please.
  11. phildominator

    phildominator Member

    679 Messages
    1 Likes Received
    Do you have the stats for the 8 division winners?
  12. AdamJT13

    AdamJT13 Salary Cap Analyst

    15,016 Messages
    493 Likes Received
    I already posted it in this thread --

    (Net passing yards - (50*INT)) / (Attempts + Sacks)


    Either yards per pass or net yards per pass play (subtract sack yardage from total yards and add sacks to attempt) also correlate highly. They don't take interceptions into account, though.


    No, that's completion percentage. Success rate is the percentage of plays that are deemed "successful" based on the down and distance. There are various sites that track success rates by their own definitions.

    And what's the point of that?


    Nobody ever said it's the only measure of passing efficiency.


    So are you saying that ANYPA does not measure passing efficiency?
  13. wick

    wick Active Member

    745 Messages
    44 Likes Received
    What are your thoughts on including quarterback fumbles and/or receiver fumbles as part of that equation?
  14. AdamJT13

    AdamJT13 Salary Cap Analyst

    15,016 Messages
    493 Likes Received
    Net yards per pass play is the same as ANYPA when neither team throws an interception. There were 42 games this season when neither team threw an interception. The team that averaged more net yards per pass play went 29-12-1 (a .702 winning percentage).

    Interestingly, games were at least 2.9 times more likely to go into overtime this season when neither team threw an interception. Out of those 42 games above, eight went into overtime. Out of the other 214 games this season, only 14 went into overtime. (It's possible that in several of those, there were no interceptions thrown until overtime, which would make it even more than 2.9 times as likely. Our win over Pittsburgh was one example of that -- the first interception of the game was Carr's in overtime.) In non-overtime games with no interceptions, the team with the higher NYPA went 26-8 (.765). In overtime games, the team with the higher NYPA for the entire game went 3-4-1, but the team with the higher NYPA *in overtime* went 5-2-1.


    Then you haven't paid close attention to what I've said for years.


    First of all, your use of "passing" and "running" is quite general. I've been very specific -- that how well you run the ball or stop the run, as measured by yards per carry, has almost no effect on winning and losing in the NFL. If you're averaging 3.0 yards per carry, you're getting virtually the same "benefit" from the running game that you would be getting if you were averaging 5.0 YPC. It's hard for some people to believe, but it is true.

    And, if you don't pass the ball better than your opponent, then running it better than your opponent has almost no effect on your chances of winning the game. If you pass it worse than your opponent, running it better still has almost no effect on your chances of winning.

    So, whatever percentages you want to come up with for the importance of each is fine with me, but it doesn't change the facts about what wins games in the NFL.
  15. AdamJT13

    AdamJT13 Salary Cap Analyst

    15,016 Messages
    493 Likes Received
    "Placing importance" on running doesn't change whether running it well or poorly affects the outcome of the game. Lots of teams run the ball a lot, and lots of teams pass it a lot. For both types of teams, running it well or poorly hardly affects their chances of winning. If they pass better than their opponents, they almost always win. When they don't, they almost always lose.

    For example, the team that passed more effectively went 14-2 in the Redskins' games, while the team that ran more effectively went 9-7. The team that passed more effectively in the Seahawks' games went 14-2, while the team that passed more effectively went 9-7. "Placing importance" on the running game doesn't change the correlation.
  16. AdamJT13

    AdamJT13 Salary Cap Analyst

    15,016 Messages
    493 Likes Received
    Those aren't all on different drives, you know. If you run it four times from the 1 and finally score on fourth down, you scored just as many TDs on that possession by going 1-for-4 as you would have if you had scored on first down (1-for-1).

    We ran the ball 18 times on 15 different possessions inside the opponent's 10-yard line. We scored touchdowns on 10 of those possessions. Another was Romo's kneel-down after Carr's interception against Pittsburgh, putting the ball in the middle of the field for Bailey's winning field goal.

    That leaves four possessions all season when we ran the ball inside the opponent's 10 and failed to score a touchdown. One of those was Murray's fumble against the Steelers. On each of the other three, we ran the ball one time inside the 10 from no closer than the 6-yard line -- a third down from the 6 against Tampa Bay, a second down from the 9 against Carolina and a second down from the 6 against Atlanta. Do you know how often runs from the opponent's 6-9 yard line go for touchdowns? About 12 percent of the time. Adrian Peterson had seven carries from there and did not score. Marshawn Lynch had seven carries from there and did not score. Frank Gore had five carries from there and did not score. We had a total of six and did not score (although two of them were followed by TD passes), and *that* is what you say our big problem was this season?
  17. AdamJT13

    AdamJT13 Salary Cap Analyst

    15,016 Messages
    493 Likes Received
    I do; it will just take me a little while to put them into this type of chart format.
  18. Zordon

    Zordon Well-Known Member

    10,117 Messages
    5,596 Likes Received
    My goodness Adam, who are you?
  19. AdamJT13

    AdamJT13 Salary Cap Analyst

    15,016 Messages
    493 Likes Received
    I'm all for making stats more accurately reflect what happens -- factor in lost fumbles on runs or passes, take out spikes and kneel-downs, include certain penalties, etc. It's just difficult to get the information for every game.
  20. JustDezIt

    JustDezIt Formerly sm0kie13 ROY

    3,647 Messages
    16 Likes Received
    ether

Share This Page