![[DEV] ISFL Forums](http://dev.sim-football.com/forums/images/emerald/logo.png "[DEV] ISFL Forums"){kind=logo}
+- [DEV] ISFL Forums (http://dev.sim-football.com/forums)
+-- Forum: Community (http://dev.sim-football.com/forums/forumdisplay.php?fid=5)
+--- Forum: Media (http://dev.sim-football.com/forums/forumdisplay.php?fid=37)
+---- Forum: Graded Statistical Analysis (http://dev.sim-football.com/forums/forumdisplay.php?fid=153)
+---- Thread: *ANY/A (/showthread.php?tid=4172)
*ANY/A - JuOSu - 09-05-2017
ANY/A is a very interesting statistic. It means adjusted net yards per pass attempt. It is a very good indicator of how good a quarterback is and it is being used more and more in the NFL. The formula is:
ANY/A = (PassYards + (20*PassTD) - (45*INT) - SackYards) / (PassAttempts + Sacks)
The average NFL quarterback is usually around 5.35 for this statistic. As we can see, the NSFL is still much worse. Only one quarterback is actually above that average NFL quarterback and most are far below it:
C. Orosz: 5.45
M. Boss: 5.08
E. Hunt: 4.79
K. Bronko: 4.65
S. Kyubee: 4.54
J. Christ: 4.10
L. Noble: 3.62
J. Bercovici: 2.48
Again we can say that some of this is due to bad offensive lines, but there is more to it than that. Orosz has been fantastic and we can see the difference in quality between quarterbacks really well here.
Also if anyone has any other statistics they would love to see, please recommend some! I have been researching and I was going to try DVOA but that would require an intense amount of work of going through every single play. I would love more mathematical challenges. In this case I had to find sack statistics myself in order to make it work but I loved doing it. Please throw other advanced statistics my way! I couldn't find many more interesting ones.
Edit: Thanks to @kckolbe for the suggestion to make TANY/A as well which includes running. You can find the formula in the comments.
This changes the results to:
C. Orosz: 5.25
M. Boss: 4.88
K. Bronko: 4.54
S. Kyubee: 4.46
E. Hunt: 4.22
J. Christ: 3.92
L. Noble: 3.58
J. Bercovici: 2.48
I will also just include this here, but this is mostly due to E. Hunt having the most fumbles by far at 6. Christ had 3 fumbles, Boss, Bronko and Orosz had 2. Noble and Bercovici had one fumble and Kyubee had zero.
In terms of difference that means that the less the difference, the better the QB was in terms of running and not fumbling, so here is how much everyone lost:
L. Noble: 0.04
S. Kyubee: 0.08
K. Bronko: 0.11
J. Bercovici: 0.16
J. Christ: 0.18
M. Boss: 0.20
C. Orosz: 0.20
E. Hunt: 0.57
So despite having the most rushing yards, it actually hurts Hunt the most due to six fumbles.
*ANY/A - kckolbe - 09-05-2017
I would like to see TANY/A as Bercovici had to scramble a lot, and Hunt had a lot of designed runs.
*ANY/A - JuOSu - 09-05-2017
(09-05-2017, 07:47 AM)kckolbe Wrote:I would like to see TANY/A as Bercovici had to scramble a lot, and Hunt had a lot of designed runs.
Can you send me the formula? I cannot find it via google! I don't know what it is! Haha
*ANY/A - kckolbe - 09-05-2017
(09-05-2017, 08:50 AM)JuOSu Wrote:Can you send me the formula? I cannot find it via google! I don't know what it is! Haha
Same as ANY/A, only it includes runs as well. Also, count all fumbles (lost or recovered) as one half of an interception.
*ANY/A - JuOSu - 09-05-2017
(09-05-2017, 07:51 AM)kckolbe Wrote:Same as ANY/A, only it includes runs as well. Also, count all fumbles (lost or recovered) as one half of an interception.
I would need an exact formula, then its no issue. Or at least a link
I cant find anything
*ANY/A - kckolbe - 09-05-2017
(09-05-2017, 08:55 AM)JuOSu Wrote:I would need an exact formula, then its no issue. Or at least a link
TANY/A =
{passing yards + rushing yards - sack yards + 20 * (TD passes + TD rushes) - 45 * (INT + 0.5 * Fumbles) }
________________________________________________________________________________
(passing attempts + sacks + rush attempts)
*ANY/A - JuOSu - 09-05-2017
(09-05-2017, 07:57 AM)kckolbe Wrote:TANY/A =
{passing yards + rushing yards - sack yards + 20 * (TD passes + TD rushes) - 45 * (INT + 0.5 * Fumbles) }
________________________________________________________________________________
(passing attempts + sacks + rush attempts)
Added it in the OP!
*ANY/A - kckolbe - 09-05-2017
(09-05-2017, 09:26 AM)JuOSu Wrote:Added it in the OP!
Thanks! I'm really trying to get TANY/A used more as QB rushes become more common. Rushing and sacks accounted for about 25% of Russell Wilson's plays in 2015, for example.
As passing gets more effective, mobile QBs should see an even bigger drop off between ANY/A and TANY/A. In the NFL, this is due to being hit full speed more often (causing more balls jarred loose) and due to the average ANY/A for rushing (a little over 4) and the average ANY/A for passing (just under 6). Mobile QBs often look far more efficient than they really are when only ANY/A is used.
I expected more drop off on Bercovici, but with his ANY/A so low, I guess it couldn't get much worse. However, by percentage, Bercovici suffered second-most from TANY/A (after Hunt) with 6.45% (Hunt's was 11.90). Boss and Orosz, despite falling by more, lost much lower percentages (both under 4%).
*ANY/A - JuOSu - 09-05-2017
(09-05-2017, 08:35 AM)kckolbe Wrote:Thanks! I'm really trying to get TANY/A used more as QB rushes become more common. Rushing and sacks accounted for about 25% of Russell Wilson's plays in 2015, for example.
As passing gets more effective, mobile QBs should see an even bigger drop off between ANY/A and TANY/A. In the NFL, this is due to being hit full speed more often (causing more balls jarred loose) and due to the average ANY/A for rushing (a little over 4) and the average ANY/A for passing (just under 6). Mobile QBs often look far more efficient than they really are when only ANY/A is used.
I expected more drop off on Bercovici, but with his ANY/A so low, I guess it couldn't get much worse. However, by percentage, Bercovici suffered second-most from TANY/A (after Hunt) with 6.45% (Hunt's was 11.90). Boss and Orosz, despite falling by more, lost much lower percentages (both under 4%).
Thanks for the comments and love! Feel free to suggest more statistics. I love doing them.