PD% and PR%

[danimal]

This may be a stupid question, but how does the game calculate pass defense % and pass rush %? I understand what these stats mean, but I'm curious about how they are calculated.

[Ben E Lou]

Morgado ran a regression and came up with a very close approximation to the PD% calculation, and I know I figured out PRPct at one point. Gimme a bit to dig those up. Or cuervo might bve able to help you here, too.

[QuikSand]

I am pretty sure that once it got patched, PR% is simply the % of pass plays in which the QB was sacked, the QB was hurried, or the pass was blocked.

PD is tougher, but it's been cracked, as SkyDog mentioned. It clearly is centered on 82.5, moves up with passes defensed and interceptions, and down with catches allowed (all normalized to number of pass plays, I believe) but I think in an effort to make this formula more mysterious they are no longer linear calculations and are instead logarithmic or somesuch.

[danimal]

I am pretty sure that once it got patched, PR% is simply the % of pass plays in which the QB was sacked, the QB was hurried, or the pass was blocked.

Awesome, thanks. I suspected as much, but it's good to get the confirmation.

PD is tougher, but it's been cracked, as SkyDog mentioned. It clearly is centered on 82.5, moves up with passes defensed and interceptions, and down with catches allowed (all normalized to number of pass plays, I believe) but I think in an effort to make this formula more mysterious they are no longer linear calculations and are instead logarithmic or somesuch.

I see. That's why I had so much difficulty figuring this one out. So a defensive back with PD% above 82.5 would be considered above average, and a defensive back with PD% below 82.5 would be considered below average. Is that the best way to read this?

[cuervo72]

82.5 + (505 * sum(ints) + 240 * sum(passdefended) - 135 * sum(passcaught))/sum(passplays)


At least, that's what Morg came up with.

[Ben E Lou]

I am pretty sure that once it got patched, PR% is simply the % of pass plays in which the QB was sacked, the QB was hurried, or the pass was blocked.Heh. Now that you mention it, that is correct. Hadn't thought about it in a while.

but I think in an effort to make this formula more mysterious they are no longer linear calculations and are instead logarithmic or somesuch....or it could be FOF's rather quirky rounding of these sorts of ratings. That was the case with the old 20-based PDPct formula in FOF2K4, if I recall correctly. Morg's formula gets it close enough for the purpose of MP leagues displaying it, though.

[QuikSand]

I see. That's why I had so much difficulty figuring this one out. So a defensive back with PD% above 82.5 would be considered above average, and a defensive back with PD% below 82.5 would be considered below average. Is that the best way to read this?

Conceptually, yes, but I think the weightings used (see above) tend to penalize more than they reward. If you have a full-time defensive back with a PD% that's above the 82.5 mark here, he very likely either (a) had a lot of interceptions, or (b) had a really, really good season. I would guess that a truly average CB in full-time play probably posts around a 74-78 or so here.

[danimal]

82.5 + (505 * sum(ints) + 240 * sum(passdefended) - 135 * sum(passcaught))/sum(passplays)


At least, that's what Morg came up with.

What an arcane formula...

Thanks again, guys.

[Ben E Lou]

I would guess that a truly average CB in full-time play probably posts around a 74-78 or so here.Gimme about 3 minutes, and I can answer this one...

[Ben E Lou]

Average for all CBs = 78.68

Edit to Add...

FS: 79.46
SS: 80.34

[Ben E Lou]

(That's over a 26-year SP career. Extender is da BIZOMB!!!)

[QuikSand]

Thanks.

[Ben E Lou]

Average for all CBs = 78.68

Edit to Add...

FS: 79.46
SS: 80.34
Bump to reflect additional calculations, for anyone interested in what an average player is.

[Narcizo]

Average for all CBs = 78.68

Edit to Add...

FS: 79.46
SS: 80.34

Interesting. I've always been a bit harsh on my cornerbacks then as I've always thought around 80 was the mark they should be aiming for. Looks like I'm alright with my safeties though.

Cheers SkyDog.

[Ben E Lou]

Interesting. I've always been a bit harsh on my cornerbacks then as I've always thought around 80 was the mark they should be aiming for. Looks like I'm alright with my safeties though.
Same here.

Incidentally, when I looked at just the "full-time" corners, the number was pretty much identical. I presume that's because the better corners are assigned coverage responsibilities on the better receivers.

[danimal]

Average for all CBs = 78.68

Edit to Add...

FS: 79.46
SS: 80.34

Thanks again, SkyDog. This is great color.

[Narcizo]

Out of interest what are the numbers for linebackers. I'm guessing the average is around 73 or so.

[Firefly]

Very interesting thread. I was just wondering the same thing.

I'd love to know the number for LBs too, if someone has it.