tag:blogger.com,1999:blog-38600807.post1923209737944323593..comments2023-11-05T04:16:44.937-05:00Comments on Advanced Football Analytics (formerly Advanced NFL Stats): Team Efficiency Rankings - Week 10Unknownnoreply@blogger.comBlogger12125tag:blogger.com,1999:blog-38600807.post-70586987146954854302012-11-12T11:21:00.530-05:002012-11-12T11:21:00.530-05:00I keep harping every week on this model overrating...I keep harping every week on this model overrating the Panthers, but it's also way underrating the Ravens.SlackerInchttps://www.blogger.com/profile/08275358994906136088noreply@blogger.comtag:blogger.com,1999:blog-38600807.post-4508164846349890502012-11-08T11:47:07.456-05:002012-11-08T11:47:07.456-05:00If the tables are sorted by team name, the efficie...If the tables are sorted by team name, the efficient rank put SEA above of SF. But sorted by team name, the efficient components puts SF above of SEA.Anonymousnoreply@blogger.comtag:blogger.com,1999:blog-38600807.post-12764920709991478202012-11-08T00:59:57.056-05:002012-11-08T00:59:57.056-05:00I'm not sure I understand the argument set for...I'm not sure I understand the argument set forth in the comments above; Carolina had <i>one</i> game that wasn't close (against the Giants) and racked up garbage yards at high efficiencies? And this is the reason why their rank is inflated?<br /><br />First of all, I think it may be worth noting that, yes, defenses may sacrifice "yards allowed efficiency" for time off the clock in games they are ahead more than 1 score, and certainly more than 2 scores. I don't know if that's exactly what happens (the numbers won't lie), but it's at least conceivable. <br /><br />But it doesn't apply to Carolina. They've been in a LOT of tight games. Using your logic, they should actually benefit from some sort of "garbage time" correction. They've probably had the least qualifiable garbage time out of any team.Anonymousnoreply@blogger.comtag:blogger.com,1999:blog-38600807.post-23474643743322591232012-11-07T21:24:42.181-05:002012-11-07T21:24:42.181-05:00or maybe they are getting 'easy' yards whe...or maybe they are getting 'easy' yards when they are trailing...Anonymousnoreply@blogger.comtag:blogger.com,1999:blog-38600807.post-65485218951454077592012-11-07T20:29:50.218-05:002012-11-07T20:29:50.218-05:00No way do I think Carolina is the 4th best team in...No way do I think Carolina is the 4th best team in the league but they aren't as bad as their 2-6 record suggests. <br /><br />The simple rating system of 0.5 and FO's ranking of 17 feels more accurate. <br /><br />Newton has been terrible under pressure (around 31st) according to pro football focus and terrible on 3rd downs (according to football perspective blog post). How predictive these things are going forward I don't know. <br /><br />That and their success rate ranking for both run and pass is much less than their EPA/P. Maybe they are trying for the big play a little too much this year. <br /><br />Their special teams and turnovers aren't helping them either as they have one of the worst net starting field positions. <br /><br /><br />Anonymousnoreply@blogger.comtag:blogger.com,1999:blog-38600807.post-76525403223276107492012-11-07T18:49:58.056-05:002012-11-07T18:49:58.056-05:00As I have stated weeks back the best way to improv...As I have stated weeks back the best way to improve this metric is to adjust for state of play...or game state..Some combination of this model and Brian's expected points...?<br />My theory is teams play different defenses when they are up by a certain amount at a certain time...teams trade time off the clock for yardage and not just in garbage time Teams leading also trade 3 points for time off the clock..I wish we could find posted datat on how long a team has played from x points behind this would help greatly.(we see this in ice hockey ) This increases Opass and Dpass and helps to explain outliers like Carolina (the NYG game is a clear example)..It wouldnt be that hard to calculate with the data. What we need is the average Opass <br />at different states when we find a diff. that is statistically significant we adjust. It is IMO the only remaining 'major' flaw in Brian's model.<br />And explains those teams that we know aren't as good as Brian's projections...Anonymousnoreply@blogger.comtag:blogger.com,1999:blog-38600807.post-55890557365913012372012-11-07T13:05:54.282-05:002012-11-07T13:05:54.282-05:00Given that the efficiency model is based on per pl...Given that the efficiency model is based on per play statistics, I don't see how following a more optimal 4th down strategy would necessarily improve things from the model's standpoint.<br /><br />You go for it on 4th down when it's better than punting or kicking. But if the odds of success (when running) or the yards per play (when passing) isn't higher or lower than the average type of play, then it shouldn't affect the efficiency stats. Even if it does, 4th down plays are relatively rare so I don't see how they would move the needle much anyway.<br /><br />That being said, I doubt there are enough data points to make coaching decisions a reliably predictive metric. It would be a lot of effort for very little (if any) gain.Michael Beuoyhttps://www.blogger.com/profile/03960600491528993233noreply@blogger.comtag:blogger.com,1999:blog-38600807.post-3815603224068832722012-11-07T12:16:52.262-05:002012-11-07T12:16:52.262-05:00I think it would be fair to say that the current o...I think it would be fair to say that the current offensive metric has been mathematically optimized to encompass as much of the parameters which determine offensive success as we have the statistical know-how to capture. If something were missing, and we had the ability to catch it, a more predictive coefficient would show itself, no? Thats the magic of well done regression. <br /><br />But it could be interesting from an evaluation standpoint to try to break out the coaching piece in a similar manner as we try to break out individual players. So we can say, based on these metrics, Coughlin seems to have added X EPA/play or Y WP. I leave it to better minds to determine exactly how that would be isolated.jditorohttps://www.blogger.com/profile/07836618836509204410noreply@blogger.comtag:blogger.com,1999:blog-38600807.post-38515564875756256112012-11-07T11:24:25.906-05:002012-11-07T11:24:25.906-05:00Keith
Yeah, at some level, but a team in the butt...Keith<br /><br />Yeah, at some level, but a team in the buttom of offensive eff, but in top i WP-forefitted on 4th down, will perform better than pure off eff. would suggest. that being said, it would be very hard to quantifyMartinhttps://www.blogger.com/profile/01751939331151507964noreply@blogger.comtag:blogger.com,1999:blog-38600807.post-49893847242147722452012-11-06T19:28:56.157-05:002012-11-06T19:28:56.157-05:00James -
That coaching component will be factored...James - <br /><br />That coaching component will be factored into the efficiency rankings. If the coaches are making better decisions (like going for it on 4th down when they should) their offensive efficiency will most likely be higher because they are exposing an inefficiency in the modern NFL.Keith Goldnerhttps://www.blogger.com/profile/16510947295485321744noreply@blogger.comtag:blogger.com,1999:blog-38600807.post-45043419050211981872012-11-06T18:49:06.542-05:002012-11-06T18:49:06.542-05:00Hey Brian, have you ever considered adding more an...Hey Brian, have you ever considered adding more and more meaningless variables until your rankings match my intuition?It would be hard to evaluate but I think it's worth considering factoring in because right now your rankings differs from what the ESPN pundits are saying.Anonymousnoreply@blogger.comtag:blogger.com,1999:blog-38600807.post-8950270848602538272012-11-06T18:17:42.208-05:002012-11-06T18:17:42.208-05:00Hey Brian, have you ever considered trying to add ...Hey Brian, have you ever considered trying to add a coaching component to the efficiency rankings? It would be hard to evaluate but I think it's worth considering factoring in. Something like propensity for going on it on 4th down, optimal run/pass balance, etc. all controlled for game state.Jameshttps://www.blogger.com/profile/01838293735141324662noreply@blogger.com