The Guy Posted January 23, 2022 Share Posted January 23, 2022 12 minutes ago, lancerman said: DVOA is a fantastic stat at isolating different units on their quality. It's issue is that it cannot account for symbiotic a football game is and how different units effect others. An amazing DVOA defense can be let down by it's offense. That's why I always so that DVOA is fantastic at telling you if a unit is good, it's mediocre at being a predicative stat Nothing is great at being strongly predictive with regard to the NFL because the game has a sizeable degree of randomness involved. DVOA, like other quality statistics, provides a signal, but nothing can eliminate the significant amount of noise (randomness) sufficiently to provide strong predictive ability. 1 Quote Link to comment Share on other sites More sharing options...
childofpudding Posted January 23, 2022 Share Posted January 23, 2022 In a one-and-done playoff format, there is no regular season stat that is going to consistently predict who wins the SB. I've never seen any DVOA supporter argue otherwise. That said, in the last five years, the SB winner was ranked 2nd, 4th, 7th, 5th and 1st in DVOA. That's pretty darn good. Profootballreference's SRS is about as good as DVOA in predicting team success in the playoffs. 2 Quote Link to comment Share on other sites More sharing options...
childofpudding Posted January 23, 2022 Share Posted January 23, 2022 6 hours ago, Awsi Dooger said: Years ago I through it into Excel toward forecasting the playoffs. It was moderate but other foundational angles were superior, namely YPPA Differential. And once again YPPA Differential dominated today. The 49ers were #1 in the league. The two fraud playoff teams were the Titans and Steelers, both at net negative of -0.5. I mentioned that here a couple of weeks ago. The only problem with YPPA Differential is that the league itself is so congested toward good but not great. The 49ers +1.9 is first this season but would have been well down the list in other years. It's like an Grade 2 horse favored to win the Kentucky Derby. You're going to get somewhat arbitrary results. I'm not sure if DVOA is reporting the same thing but I would be very skeptical of any numbers-based approach that wasn't revealing lower caliber teams overall. I'd bet that a straight-up yards/play differential would be decent. If it's a good indicator, then the Rams, 49ers and Bucs look strongest this season. Quote Link to comment Share on other sites More sharing options...
11sanchez11 Posted January 23, 2022 Share Posted January 23, 2022 6 hours ago, Awsi Dooger said: And once again YPPA Differential dominated today. The 49ers were #1 in the league. I don't think the 49ers winning had anything to do with YPPA differential. Might have been the least important aspect of the gm tbh 1 Quote Link to comment Share on other sites More sharing options...
The Guy Posted January 23, 2022 Share Posted January 23, 2022 4 minutes ago, 11sanchez11 said: I don't think the 49ers winning had anything to do with YPPA differential. Might have been the least important aspect of the gm tbh Certainly offensive and defensive pass efficiency -- which is what that statistic is measuring -- provided a parameter within which those teams would be predicted to play, and then of course the outcome of the game could've been determined also by other variables. Had the YPA differential been significantly different, those other variables may have contributed to a different outcome. For example, two teams wildly discrepant in YPA differential may not have the outcome of their game hinge on turnover differential for example. Two teams far less discrepant on YPA differential may very well have their game hinge on turnover differential, however. Analogous to turnover differential, yesterday the issue between GB and SF was special teams. But the same special teams differences may have been powerless to overcome a wildly different YPA differential -- in terms of how those teams typically play. Here is what Awsi Dooger is getting at with YPA differential -- it's essentially the predictive ability of offensive and defensive pass efficiency: https://www.opensourcefootball.com/posts/2020-08-23-exploring-wins-with-nflfastr/ Quote Link to comment Share on other sites More sharing options...
kramxel Posted January 23, 2022 Share Posted January 23, 2022 9 hours ago, DontTazeMeBro said: For years I’ve been seeing this fake statistic touted as the be-all end-all metric for measuring how good teams are. So why does the team that leads the league in it almost never win the Super Bowl? Why are we supposed to care about it? If you're looking for stat instant gratification, look for SB victories. That stat usually does a good job showing SB victors. Quote Link to comment Share on other sites More sharing options...
Matts4313 Posted January 23, 2022 Share Posted January 23, 2022 9 hours ago, SkippyX said: I like everything about Football Outsiders. Even their articles that say the run game doesnt matter? Truck/wagon debate ringing any bells? 😁 Quote Link to comment Share on other sites More sharing options...
SkippyX Posted January 23, 2022 Share Posted January 23, 2022 Just now, Matts4313 said: Even their articles that say the run game doesnt matter? Truck/wagon debate ringing any bells? 😁 Yeah, Dak just had a playoff game with a run game that went for 77 yards. Pollard and Zeke combined for 16 carries and 45 yards. The other team ran for 168. How'd that go? You can watch him destroy the Covid WFT and the Eagles backups on a loop I guess. Quote Link to comment Share on other sites More sharing options...
Matts4313 Posted January 23, 2022 Share Posted January 23, 2022 1 minute ago, SkippyX said: Yeah, Dak just had a playoff game with a run game that went for 77 yards. Pollard and Zeke combined for 16 carries and 45 yards. The other team ran for 168. How'd that go? We self destructed with 14 penalties. I know you know this as we live rent free in your head 24/7 1 minute ago, SkippyX said: You can watch him destroy the Covid WFT and the Eagles backups on a loop I guess. Thats not even a clever or funny quip. Quote Link to comment Share on other sites More sharing options...
SkippyX Posted January 23, 2022 Share Posted January 23, 2022 I'm pretty sure you are the one who showed up to troll here. You got dunked on. Move along now. Quote Link to comment Share on other sites More sharing options...
The Guy Posted January 23, 2022 Share Posted January 23, 2022 3 minutes ago, SkippyX said: Yeah, Dak just had a playoff game with a run game that went for 77 yards. Pollard and Zeke combined for 16 carries and 45 yards. The other team ran for 168. How'd that go? You can watch him destroy the Covid WFT and the Eagles backups on a loop I guess. Single games can go any number of ways. Almost nothing is strongly predictive of the outcome of a single game. You can find games in history in which punting was responsible for the outcome of a single game -- doesn't mean punting is supremely important with regard to winning and teams should be drafting punters in the first round. If you want to know what's important with regard to winning over the long haul in the NFL, see here: https://www.opensourcefootball.com/posts/2020-08-23-exploring-wins-with-nflfastr/ This is why QBs, WRs, LTs, DEs, and CBs -- and not RBs, interior OL, interior DL, and LBs -- are the most sought-after and highest-paid players in the league. The NFL knows what wins over the long haul, and what provides the strongest probability of winning in any single game. The fact that single games can go differently than predicted doesn't change the strongest predictors of winning in the league. Dallas may have been outrushed in the playoffs, but if the two teams played again the strongest likelihood of the outcome would revolve around offensive and defensive pass efficiency nonetheless. Quote Link to comment Share on other sites More sharing options...
SkippyX Posted January 23, 2022 Share Posted January 23, 2022 That's theory. Dak Prescott is practice. He needs 120+ rushing yards. He thrives at 140+ Short of that he is hot garbage stuffed into clown shoes. Oh boy does that reality hurt some feelings. Quote Link to comment Share on other sites More sharing options...
SkippyX Posted January 23, 2022 Share Posted January 23, 2022 The argument was never really about DVOA and rushing vs passing for expected wins. It was about the value of Dak vs Zeke on the Cowboys. Its now Zeke/Pollard but the answer is obviously the same. (unless Dak gets to play a really bad team) Quote Link to comment Share on other sites More sharing options...
The Guy Posted January 23, 2022 Share Posted January 23, 2022 (edited) 5 minutes ago, SkippyX said: That's theory. Dak Prescott is practice. He needs 120+ rushing yards. He thrives at 140+ Short of that he is hot garbage stuffed into clown shoes. Oh boy does that reality hurt some feelings. If you really want to support that theory, then just do a simple correlation between the Cowboys' rushing yardage and Prescott's performance, game-by-game. If Prescott needs more rushing yardage to perform adequately (or better, or whatever), that correlation should be at least moderately strongly positive. You'll have to control for late-game, clock-killing rushing yardage, however, as that may be attributable to Prescott's effectiveness earlier in the game. In other words, Prescott's effectiveness may cause rushing yardage, rather than vice-versa. Edited January 23, 2022 by The Guy Quote Link to comment Share on other sites More sharing options...
SkippyX Posted January 23, 2022 Share Posted January 23, 2022 1 minute ago, The Guy said: If you really want to support that theory, then just do a simple correlation between the Cowboys' rushing yardage and Prescott's performance, game-by-game. If Prescott needs more rushing yardage to perform adequately (or better, or whatever), that correlation should be at least moderately strongly positive. No, you do that. I already know it to be true through observation and review. I don't care if others pretend its not reality. I just bring it up every time a fly gets off a big pile in Jerrah World and buzzes by my face on random threads. Quote Link to comment Share on other sites More sharing options...
Recommended Posts
Join the conversation
You can post now and register later. If you have an account, sign in now to post with your account.