tag:blogger.com,1999:blog-38600807.post4715266120302194238..comments2023-11-05T04:16:44.937-05:00Comments on Advanced Football Analytics (formerly Advanced NFL Stats): Team Rankings: Success RateUnknownnoreply@blogger.comBlogger14125tag:blogger.com,1999:blog-38600807.post-92109807621983255302010-10-18T10:07:52.103-04:002010-10-18T10:07:52.103-04:00Jim - High SR & high EPA/P would probably be t...Jim - High SR & high EPA/P would probably be the the '07 Pats, '99 Rams, '98 Vikings...Jameshttps://www.blogger.com/profile/01838293735141324662noreply@blogger.comtag:blogger.com,1999:blog-38600807.post-50835086524497992862010-10-17T23:04:03.678-04:002010-10-17T23:04:03.678-04:00William-SR is based on EPA, not WPA. So no, it'...William-SR is based on EPA, not WPA. So no, it's not sensitive to blow-outs.Brian Burkehttps://www.blogger.com/profile/12371470711365236987noreply@blogger.comtag:blogger.com,1999:blog-38600807.post-60000988111291952532010-10-17T22:56:48.003-04:002010-10-17T22:56:48.003-04:00Brian,
Would this success rate get skewed at all ...Brian, <br />Would this success rate get skewed at all for bad teams? It seems that if a team is getting blown out and has an expected winning% 1% there is nothing they are going to be able to do to decrease their chances of winning so they cant have any negative plays.Unknownhttps://www.blogger.com/profile/05862971655648617649noreply@blogger.comtag:blogger.com,1999:blog-38600807.post-34168347371819082502010-10-16T19:19:15.538-04:002010-10-16T19:19:15.538-04:00I notice that teams like BAL, CIN, and NO are much...<em> I notice that teams like BAL, CIN, and NO are much higher in terms of SR than they are in the efficiency rankings. PIT, HOU, and CHI are much lower. I'm betting these teams are the ones that tend to rely on the big play, either offensively or defensively, for their success. </em><br /><br />I'd think that comparing EPA/P to SR would also reveal that. E.g., take Rex Ryan's huge blitzing D of the Jets, last year's #1 D and pass D by most counts, certainly near the league top.<br /><br />So far this year it is only #23 at pass D EPA/P, but #10 at pass D success rate.<br /><br />Think of this week's game vs the Vikings: For most of the game it completely shut down the Vikings' O, 3-and-out, 3-and-out, etc. ... but it gave up two very long TD passes when heavy blitzing on 3rd and 17 and 3rd and 19. Very high success rate (Favre completed only 14 of 34, 41%); but a much less good EPA/P (Favre had 264 yds and 3 TDs) I would expect because of the few very big EPA plays it gave up.<br /><br />On O, high SR & lower EPA/P = Lombardi's Packers. High EPA/P & lower SR = maybe the run-and-shoot Os of the '90s? High SR & high EPA/P = ???<br /><br />Last year at PFR.com there was an interesting discussion on the relative contribution of average yards/attempt versus completion pct to the passing game.<br /><br />It might be interesting to look at the same issue -- the value of big plays v higher consistency at a more modest level -- in terms of EPA/P v SR. (And it would seem easier to do since EPA/P and SR are both defined in terms of EPA, so there's no need to work out what a couple points of completion pct is equivalent to in yardage.)Jim Glassnoreply@blogger.comtag:blogger.com,1999:blog-38600807.post-76533298128400774782010-10-16T19:15:42.642-04:002010-10-16T19:15:42.642-04:00The league's median offensive success rate is ...The league's median offensive success rate is a tad under 44%.<br /><br />The league is offense deficient!Jim Glassnoreply@blogger.comtag:blogger.com,1999:blog-38600807.post-53357202746876315702010-10-16T12:43:48.692-04:002010-10-16T12:43:48.692-04:00These success rates are more reflective of the foo...These success rates are more reflective of the football of <i>seen</i> this year.<br /><br />The efficiency ratings, at least for the Redskins, don't jive with on field performances I've watched.Erichttps://www.blogger.com/profile/02735177249996910485noreply@blogger.comtag:blogger.com,1999:blog-38600807.post-12564006566157886312010-10-15T17:51:54.962-04:002010-10-15T17:51:54.962-04:00Have you looked at whether the success rate varies...Have you looked at whether the success rate varies by the situation - i.e. down, quarter...That may provide an even better correlation with wins if you could identify the SR in "key" situations.Briannoreply@blogger.comtag:blogger.com,1999:blog-38600807.post-9464941842787798322010-10-15T14:27:22.595-04:002010-10-15T14:27:22.595-04:00Almost 100%. It's really about 98% because the...Almost 100%. It's really about 98% because there is a tiny fraction of plays that actually end up at exactly 0.00 EPA.Brian Burkehttps://www.blogger.com/profile/12371470711365236987noreply@blogger.comtag:blogger.com,1999:blog-38600807.post-53849735447848152772010-10-15T14:22:41.783-04:002010-10-15T14:22:41.783-04:00so im guessing that 100% is the break even average...so im guessing that 100% is the break even average, which would have oaklands win against san diego not as much a shock as your efficiency rankings had it.Shaunnoreply@blogger.comtag:blogger.com,1999:blog-38600807.post-69471959066057348252010-10-15T11:47:13.066-04:002010-10-15T11:47:13.066-04:00For NO, I can say that THE one part missing from t...For NO, I can say that THE one part missing from the stats is missed FG's. The Saints have 4--1 of which resulted in a loss (vs. ATL) and another which would have resulted in OT (@ ARI). A 5-0 or 4-1 with OT loss being 6th in your rankings wouldn't look out of place at all.Josephnoreply@blogger.comtag:blogger.com,1999:blog-38600807.post-66700111812432453052010-10-15T11:16:03.979-04:002010-10-15T11:16:03.979-04:00Do the Cowboys really have a tendency to give up a...Do the Cowboys really have a tendency to give up a big play, or is that just random chance? I don't think there's anything predictive about a fumble return for a touchdown, a blown coverage on a blitz leading to a TD, or a 73 yard kickoff return which all resulted in a game deciding TD.<br /><br />The numbers make sense to me, and in your example the Cowboys have simply been unlucky.Jameshttps://www.blogger.com/profile/01838293735141324662noreply@blogger.comtag:blogger.com,1999:blog-38600807.post-42971724384456652682010-10-14T18:58:26.164-04:002010-10-14T18:58:26.164-04:00I don't know if the correlation is there or no...I don't know if the correlation is there or not. But I would define a 'big play' as anything that is outside +/- 1 std. deviation of EPA averaged over the entire league.<br /><br />The first step would be to see if there are teams that actually have more then their share of plays. Then see if you can find any correlation between results and the rates, especially when combined with the above SRs.<br /><br />There is alot of tuning that can be done with this type of type of model which may make it suspect. I'm sure that you can pick a std dev cutoff and a coefficient on SR and 'big play rate' to make it line up, but thems the breaks with modeling.epvhttps://www.blogger.com/profile/01563873887885098509noreply@blogger.comtag:blogger.com,1999:blog-38600807.post-75354673260755039912010-10-14T18:17:27.842-04:002010-10-14T18:17:27.842-04:00On the 'big play' factor you mention, is t...On the 'big play' factor you mention, is there a way of seeing how predictive 'the big play' is?<br /><br />You'll probably get a multiple end point argument depending on what you choose to be a big play, but if we say that a big play is one in which your expected points increases by 1 pt (or any other arbitrary point increase), is the rate at which teams get these 'big plays' consistent throughout the season, as SR seems to be.<br /><br />This could be the football equivalent of small ball v big inning.Ian Simcoxhttps://www.blogger.com/profile/01518825067469269377noreply@blogger.comtag:blogger.com,1999:blog-38600807.post-84050644510717134132010-10-14T17:47:37.065-04:002010-10-14T17:47:37.065-04:00It seems like to be more encompassing you need to ...It seems like to be more encompassing you need to take in a second aspect of a 'failure' rate for plays that go truly bad. Dallas has a high SR for both offense and defense, but has a tendency to either give up big plays on D or give up turn overs and drive ending penalties on O.epvhttps://www.blogger.com/profile/01563873887885098509noreply@blogger.com