tag:blogger.com,1999:blog-38600807.post6121555124985753284..comments2023-11-05T04:16:44.937-05:00Comments on Advanced Football Analytics (formerly Advanced NFL Stats): Passing Predictability Part 1Unknownnoreply@blogger.comBlogger11125tag:blogger.com,1999:blog-38600807.post-59173953019868872562010-08-04T04:55:27.071-04:002010-08-04T04:55:27.071-04:00Teams who stink at running can just as easily face...Teams who stink at running can just as easily face a lot of 2nd and 10 and 3rd and 10. Also, teams who move the ball will have longer possessions and see more 2nd/3rd and 10 although the proportion may be different.J. Rollahttps://www.blogger.com/profile/05366283834813239621noreply@blogger.comtag:blogger.com,1999:blog-38600807.post-89803185881873760732009-04-27T14:39:00.000-04:002009-04-27T14:39:00.000-04:00I disagree with this conclusion: "Which teams are ...I disagree with this conclusion: "Which teams are more likely to face a lot of 2nd and 10s and 3rd and 10s? The ones that stink at passing."<br /><br />What about the teams which very predictably ran into a wall on 1st and 2nd downs, forcing 2nd and 10 or 3rd and 10? The teams who stunk at passing didn't try to throw on 1st down, and probably not on 2nd down either.Anonymousnoreply@blogger.comtag:blogger.com,1999:blog-38600807.post-32401589587158351602009-04-09T20:17:00.000-04:002009-04-09T20:17:00.000-04:00Nice work and interesting to read. Please keep it ...Nice work and interesting to read. Please keep it up. I know it would mean more work but it would be nice to see the breakdown for 1st and 10, 2nd and long cases (8 to 10+ yards), 2nd and Medium (4-7), 2nd and short (3 or less), 3rd and long (8 to 10+), 3rd and medium (4-7), and 3rd and short (3 or less). On the stat side that would be interesting to look at if possible.<BR/><BR/>On the play calling side though I don't think this type of study ever really gets us anywhere. Coaches have to call a specific type of run (iso, counter, inside zone, outside zone, power, etc.) to a specific gap (strong side a, b, c, d, or weak side a, b, or c, etc.) What run they call depends upon the type of front faced (under, over, even, odd, bear, etc.) That is based off of tendency reports and team strength.<BR/><BR/>Same is true for passing game (type of pass vs. type of coverage). A lot of plays work due to good execution and not just surprise or maintaining equilibrium. Texas ran Zone Option Read with Vince Young to a national championship...everyone knew it was coming...just could not stop it.Anonymousnoreply@blogger.comtag:blogger.com,1999:blog-38600807.post-88682503129567482232009-03-31T14:02:00.000-04:002009-03-31T14:02:00.000-04:00Brian, did you do anything to adjust for the fact ...Brian, did you do anything to adjust for the fact that bad teams will get more 2nd and 10s and 3rd and longs thus biasing the averages?<BR/><BR/>A set of 2nd and 10 and 3rd and long plays will be over represented by bad teams. You could adjust for this by looking at the changes in average by team by year from first down to second down or change in average from first to second down by team by game.Anonymousnoreply@blogger.comtag:blogger.com,1999:blog-38600807.post-90995204274522481652009-03-29T20:11:00.000-04:002009-03-29T20:11:00.000-04:00>>Are defenses just plain "better"...>>Are defenses just plain "better" on 2nd and 10? <BR/><BR/>I think that's kinda the point of the article. Defenses aren't really "better", but their proficiency on 2nd down increases because the offense becomes more predictable. As you pointed out, an offense on 1st-and-10 tends to exploit their "whole arsenal", thus making them the hardest to defend because they could conceivably attack any area of the field.Unknownhttps://www.blogger.com/profile/07433838358019660341noreply@blogger.comtag:blogger.com,1999:blog-38600807.post-86928285849211671632009-03-26T22:32:00.000-04:002009-03-26T22:32:00.000-04:00Brian,Interesting. I'm fascinated by the fact that...Brian,<BR/><BR/>Interesting. I'm fascinated by the fact that the YPA for passing on 2nd down drops by almost a full yard while the increased gain from running is only .2. Are defenses just plain "better" on 2nd and 10? (maybe skewed data to the "losers"?) Or is this not as significant as I think.<BR/><BR/>It does seem to lend credence to the conventional wisdom that a 2nd down and 10 run is a good play call to get half of it, but not for the reason I would think.<BR/><BR/>The other behaviorialist issue I can think of (but can't think of a good way to test it) is this "get half of it" conventional wisdom: on first and 10 teams tend to use their whole arsenal, i.e. go deep, short medium; on 2nd and 10 maybe they are just going for shorter passes? Maybe that could be answered by looking at the variance for those YPA numbers.<BR/><BR/>Sorry for the speculations: good post.Chrishttps://www.blogger.com/profile/07204245083374821812noreply@blogger.comtag:blogger.com,1999:blog-38600807.post-6912033491001767372009-03-25T15:44:00.000-04:002009-03-25T15:44:00.000-04:00Interesting article. What about the other anchor ...Interesting article. What about the other anchor point (i.e. near 100% run vs. run defense)? 3rd or 4th and 1 seem like a good place to look, but then adjusted YPA becomes even more misleading because of the utility of small gains.<BR/><BR/>Also, I'm curious as to why you drew the blitz line the way you did.Anonymousnoreply@blogger.comtag:blogger.com,1999:blog-38600807.post-55272119615339339432009-03-25T13:04:00.000-04:002009-03-25T13:04:00.000-04:00Sorry. I wasn't clear at all above. The YPA I used...Sorry. I wasn't clear at all above. The YPA I used is actually "net YPA," which includes all yardage gained or lost on a passing play including sacks.Brian Burkehttps://www.blogger.com/profile/12371470711365236987noreply@blogger.comtag:blogger.com,1999:blog-38600807.post-4548862112441861782009-03-25T12:50:00.000-04:002009-03-25T12:50:00.000-04:00are going to factor in sacks in the next analysis?...are going to factor in sacks in the next analysis?Anonymousnoreply@blogger.comtag:blogger.com,1999:blog-38600807.post-32421238554722539432009-03-25T11:22:00.000-04:002009-03-25T11:22:00.000-04:00Will-complete agree. Everything I'm doing is predi...Will-complete agree. Everything I'm doing is predicated on an over-simplified run/pass dichotomy.Brian Burkehttps://www.blogger.com/profile/12371470711365236987noreply@blogger.comtag:blogger.com,1999:blog-38600807.post-69047815707444034932009-03-25T10:25:00.000-04:002009-03-25T10:25:00.000-04:00Brian, Interesting, and I'll definitely be back to...Brian, <BR/><BR/>Interesting, and I'll definitely be back to read the next installment, but I submit that the offense's choice between running and passing is a false dichotomy propagated largely by under-informed television sports analysts. The effect on the defense of a quick-pitch sweep to the corner and of a 3-step swing pass to the same area are, for all practical purposes, identical. There is no actual utility in "running to set up the pass" or vice-versa: this line of thinking is accidentally correct some of the time, because the true choice facing an offense is which area of the field to attack on a given play. By attacking one area of the field the offense forces the defense to adjust, opening up another area of the field. The reason the run/pass view of the world is accidentally correct part of the time is that the deep areas of the field are much harder to attack with running plays, so when an offense is attacking the short frontal area to open up the deep outside areas, doing the first part with run plays and the second part with pass plays, then tv analysts are not wrong when they say they are running to set up the pass. However, the offense could do exactly the same thing with shovel passes or shallow crosses instead of runs. I suspect that if you do not somehow account for the similarity of some runs to some passes (eg. shovel passes and draws), you will bias your results. But I will definitely be interested in the results, either way.Willhttps://www.blogger.com/profile/02178230449052059046noreply@blogger.com