Injuries and the 4000+ yard passing issues.

[CraigSca]

Rushing totals for 2002 season:

Rushing Attempts: 14102
Yards: 59459
Yards per carry: 4.21

[cthomer5000]

Originally posted by SkyDog
Right, but I don't see league totals, just team totals.

add them up. do I have to hold your hand?

[Samdari]

Originally posted by SkyDog
If anyone can find, or calculate, league rushing and passing totals for 2002, it would be VERY helpful.

Funny, I was coming here to do that

Pass Attempts: 17292
Completions: 10314
Percentage: 0.5965
YPA: 6.72
Rush Attempts: 14102
Rush Yds: 59459
YPC: 4.21

Two things jump out at me from those numbers. The number of pass attempts and rush attempts over a season are stunningly accurate for SkyDog's 2003 season and for the 2002 NFL season.

Looks like the comp% and ypa are a bit higher than the NFL's, and the YPC are a bit lower. Note that the actual 2002 NFL stats shatter one of my long held beliefs on NFL statistics - that yards per rush was almost exactly 4.0 from 1963 through 2002.

[CraigSca]

It seems to me...the YPC is causing the teams to pass more often on 2-3rd down. However...since the quarterback has to throw longer (as well as the defense expective a pass more often in these 2nd/3rd and long situations), you would think comp. percentages would go down. So...after the yards per carry gets fixed, it seems that QB completion percentages need to be toned down across the board as well.

[Daimyo]

Originally posted by amdaily
Daiymo - I updated my chart with three seasons of your stats.

While they are inline with FOF2001, pass attempts are way to high still. As SkyDog has suggested, as better QB's arrive though the draft they will be able to exploit those increased attempts for more yards and a higher rating. So it doesn't look like that the player files are at fault.

Are you accounting for 2k1 having 31 teams and 2k4 having 32 teams when comparing total pass attempts?

EDIT: Or did 2k1 have expansion after the first season? Can't even remember...

[CraigSca]

One other thing...I noticed in the NFL 2002 data that the league average 10.5 yards per completed pass. Just looking at the FOF2004 statistics from someone's league posted on the first page of this thread, and after an admittedly cursory look, I saw not a single quarterback who posted an average of 10.5 or lower. ALL of them were higher.

It just seems that passes are being completed too often and for too many yards.

[Daimyo]

I don't have access to FOF2k4 at work, but didn't someone post above that 50% of the starting QB's in the real roster file are rated 5 or above? Accordoing to Jim's help file it should be 3.9% of all total QB rated 5+. At most that would probably be something like 10% of starting QB's rated 5+. If 50% of starting QB's are really 5+ it seems that must obviously be the problem, in which case I invite you try my roster file (sorry to sound like a broken record here, but it seriously fixes the problem for me). If HR based his roster file off the real players file like he said than it will also be inflated just like the real rosters.

For reference here are the player ratings, their meaning, and the frequency they SHOULD occur:

9 - Rare Superstar - 0.3%
8 - Superstar - 0.4%
7 - Near Superstar - 0.6%
6 - Elite Starter - 1.1%
5 - Excellent Starter - 1.5%
4 - Very Good-Plus Starter - 2.6%
3 - Very Good Starter - 3.9%
2 - Good Starter - 12.2%
1 - Replacement-Level Starter (decent backup) - 23.9%
0 - Roster Filler - 53.4%

I would interpret an excellent starting QB to be a guy who could pass for 4000+ yards... (We're talking the top ~4 or 5 QB's in the entire NFL here! That's like half the QB's being Manning quality!) if half the starting QB's are at least that good, you'll probably see 10 of them get 4000 yards every season...

[Ben E Lou]

Originally posted by Daimyo
I don't have access to FOF2k4 at work, but didn't someone post above that 50% of the starting QB's in the real roster file are rated 5 or above? Accordoing to Jim's help file it should be 3.9% of all total QB rated 5+. At most that would probably be something like 10% of starting QB's rated 5+. If 50% of starting QB's are really 5+ it seems that must obviously be the problem, in which case I invite you try my roster file (sorry to sound like a broken record here, but it seriously fixes the problem for me). If HR based his roster file off the real players file like he said than it will also be inflated just like the real rosters.

For reference here are the player ratings, their meaning, and the frequency they SHOULD occur:

9 - Rare Superstar - 0.3%
8 - Superstar - 0.4%
7 - Near Superstar - 0.6%
6 - Elite Starter - 1.1%
5 - Excellent Starter - 1.5%
4 - Very Good-Plus Starter - 2.6%
3 - Very Good Starter - 3.9%
2 - Good Starter - 12.2%
1 - Replacement-Level Starter (decent backup) - 23.9%
0 - Roster Filler - 53.4%

I would interpret an excellent starting QB to be a guy who could pass for 3500-4000+ yards... if half the starting QB's are at least that good, you'll probably see 10 of them get 4000 yards every season...

Daimyo:

I started with HR's file, which worked just fine in the early seasons, but when the newly-generated draftees started playing, the passing totals started going up. I'll give it a try with yours tonight/tomorrow morning, but I think it is a bigger issue than just the initial player file composition.

[McSweeny]

well if we could get some draft files that fit the suggested rating frequencys wouldn't that fix things? Of course presuming that this problem is talent based

[Ksyrup]

This would explain why QBs are more successful on 2nd and 3rd down passing. The remaining question in my mind is, why are yards per carry so low? I think that is what is causing the passing numbers to be inflated, in combination with inflated QB ratings and possibly fewer injuries, leading to more successful passing attempts by players rated above the norm.

[korme]

Does anyone agree that rushing averages are a bit low?

[Ben E Lou]

Well, ypc has gone up very recently in football apparently, but up until the late 90's, I agree that 4.0 was about the norm. I'm still getting the same rushing yardage averages (around 4.0) that I got in all previous versions of FOF. It was 4.2 in '02, but I wonder if that is an anomaly. Anybody got '01?

[Ben E Lou]

Just found 2K1

4.059ypc

[TroyF]

I'm finding the same things wrong as other people are.

The longer I go into a career, the more bigtime passers there are. In 2020 of my career, there are now 15 4,000 yard passers. To me, the problem isn't the yardage stats. Those can be skewed by the injury setting. I'm having a problem with the QB "ratings"

In 2002, the NFL had 5 guys at a 90.0 or higher rating. In 2003, that number is at 6.

In my FOF2004 league, I'm seeing an average of 12-15 QB's with a 90+ rating. I'm usually seeing 5 at over 100. I'm routinely seeing guys finish their careers with ratings in the mid 90's. That just doesn't happen in real life.

I'm also seeing the lower rushing totals. Rarely do I see a guy go over 1600 yards. Even top notch backs don't have those 1800+ yard seasons.

I think it's a combination of a lot of things, but I do think some of it is in the core game engine. I hope Jim can look into this.

TroyF

[CraigSca]

I'm seeing QBs in my quick-simmed FOF2004 league (I'm in 2008) who aren't really considered great (ratings wise). However, they still complete 50-60% of passes for close to 4000 yards (or more). It seems ALL quarterbacks in FOF2004 can pass for 4000 yards easily, throw a lot of touchdowns, complete 55%-60% of their passes and average higher than the 2002 NFL average of 10.5 yards per completion. The only differentiating statistic I see when it comes to quarterbacks is the number of interceptions thrown.

[Ben E Lou]

2K1 Passing:

58.97 Completion Percentage
6.776 yards per attempt
11.490 yards per completion

[Ksyrup]

It just occurred to me that at some point in the past, I recallJim suggesting that football in the future would become more of a passing-oriented game, with rushing taking a clear back seat to passing in terms of game planning. Does anyone else recall this? I don't remember when this was, or what form it took - it might have been a post, part of an interview, a chat, I just don't remember.

However, if I'm not just dreaming this, perhaps Jim has purposely set the game up so that the further out we go, the more passing-oriented teams get. Now, I'm not sure that in and of itself would explain why the passing attacks are more successful. But, maybe game plans are being set up to maximize each QB's efficiencies, so more and more of them are successful.

Just a thought.

[korme]

I am in 1979, the 15th season of my OPU career. I will examine my own passing/rushing totals after the season and post them here for you all.

[Daimyo]

As anyone else had considerably more luck developing QB's than other positions? I've developed two QB's into studs (one picked 17 and the other at 31) despite neither looking all that special at draft time. Conversely I've had very little luck with other positions. I had a maxed out potential OT who ended up being something like 99/40/40/40.

The OT didn't play much his first two years because of poor currents and he quickly lost potential whereas my QB's didn't lose it so fast. In the help Jim lists QB's as having 5 seasons to peak whereas most other positions happen in years 2 or 3. I wonder if that is partially why drafted QB's tear it up? Maybe studs at other positions get buried in the depth chart early and it stunts their development while QB's almost always eventually develop?

Or maybe the ai puts a premium on OC's with good skill developing QB's since its such a primo position?

[Ben E Lou]

Hmmm...interesting thought.

Just FYI, I'm up to 2033, with the last 8 seasons having been done with injuries=200. Passing has come down a bit, but not to present-day totals. Most recent season totals (2033):

Completion Percentage: 61.2% (higher than normal)
Yard/Completion: 11.5 (right in line with real-life totals)
Yards/Attempt: 7.0 (a little higher than real life)
Rushing Yards/Attempt: 3.9 (in line with pretty much every season but 2K2)

Looks like the problem is becoming isolated to being too high of a completion percentage. Passing and rushing attempts 31 years into the game are still so ridiculously close to real-life, it ain't even worth talking about attempts any more.

[Samdari]

Originally posted by CraigSca
One other thing...I noticed in the NFL 2002 data that the league average 10.5 yards per completed pass.

Be careful here when using team stats. Team net yards passing have sacked yards lost subtracted, but the individual passer getting sacked does not have that subtracted. At least, that is what happens in the NFL does, I have not verified this for FOF2004.

So, we have to be careful when discussing these things, that we are comparing the same numbers. The numbers I used for 2002 only counted the individuals passing totals, thus the 10314 completions for 116201 yards equals 11.27 yds per completion. I believe that most of these discussions are the numbers without sack yards subracted from the team totals.

To preemptively settle the inevitable "the NFL does not do that with team passing stats" posts, look here. Note that the Colts are shown as having 2433 passing yards, but that Colts QBs have passed for 2488. The Colts have been sacked 8 times, presumably for 55 yards. This discrepancy occurs on every team's stat page.

[Ben E Lou]

Yards per carry so far in the 2003 real NFL season is 4.10.

[amdaily]

Originally posted by SkyDog
Passing and rushing attempts 31 years into the game are still so ridiculously close to real-life, it ain't even worth talking about attempts any more.

I beg to differ. I purchased a game meant to simulate the NFL. Not a game which I have to sit back and sim 31 years, or three generations of players, in order to start playing. There is no valid reason why passing attempts should decrease over the span of 30 years. They should be approximently the same in 2003 as they are in 2034.

[Ben E Lou]

Originally posted by Samdari
Be careful here when using team stats. Team net yards passing have sacked yards lost subtracted, but the individual passer getting sacked does not have that subtracted. At least, that is what happens in the NFL does, I have not verified this for FOF2004.

So, we have to be careful when discussing these things, that we are comparing the same numbers. The numbers I used for 2002 only counted the individuals passing totals, thus the 10314 completions for 116201 yards equals 11.27 yds per completion. I believe that most of these discussions are the numbers without sack yards subracted from the team totals.

To preemptively settle the inevitable "the NFL does not do that with team passing stats" posts, look here. Note that the Colts are shown as having 2433 passing yards, but that Colts QBs have passed for 2488. The Colts have been sacked 8 times, presumably for 55 yards. This discrepancy occurs on every team's stat page.

My 2001 stats were calculated by adding up the passing totals of every individual QB in the league.

Fyi, the reason I was having copy and paste problems into Excel was that I was using Opera instead of IE. When I switched to IE, everything copied very easily.

Here is a great site for NFL stats for a whole bunch of seasons, easily copied into excel.

[CraigSca]

Originally posted by SkyDog
Yard/Completion: 11.5 (right in line with real-life totals)

FYI...Yards/Comp. in NFL 2002 was 10.5. 11.5 may look "in-line" but 1 yard is actually a huge difference.

[cuervo72]

Originally posted by Ksyrup

However, if I'm not just dreaming this, perhaps Jim has purposely set the game up so that the further out we go, the more passing-oriented teams get. Now, I'm not sure that in and of itself would explain why the passing attacks are more successful. But, maybe game plans are being set up to maximize each QB's efficiencies, so more and more of them are successful.

Just a thought.

That was my initial thought.

[Ben E Lou]

Originally posted by amdaily
I beg to differ. I purchased a game meant to simulate the NFL. Not a game which I have to sit back and sim 31 years, or three generations of players, in order to start playing. There is no valid reason why passing attempts should decrease over the span of 30 years. They should be approximently the same in 2003 as they are in 2034.

Ummmm....you have GOT to be kidding me.

NFL 2K2: 14102 rushing attempts
FOF 2K3: 14074 rushing attempts

NFL 2K2: 17292 passing attempts
FOF 2K3: 17480 passing attempts


(The FOF 2K3 numbers came from my saved career with real player names file.)

Are you seriously trying to suggest that you have a problem with that minute of a variance????

[CraigSca]

I'd have to give more though whether I accept that Jim purposely made the game to evolve to a more passing-oriented style. However, the still doesn't explain two things.

1) This passing-oriented style hits on day 1 of the simulation...not 20-30 years into it.

2) EVERY QB excels in this situation and rarely gets hurt. Ok...not EVERY QB...but, is anyone finding it hard to draft a pretty darned good quarterback in this game? Is that representative of the NFL in ANY time period?! Since when is a good quarterback a commodity?

[Samdari]

Originally posted by SkyDog
My 2001 stats were calculated by adding up the passing totals of every individual QB in the league.

All of mine were. In fact, I did the 2001 season and got the same numbers as yours.

[Ben E Lou]

Originally posted by CraigSca
FYI...Yards/Comp. in NFL 2002 was 10.5. 11.5 may look "in-line" but 1 yard is actually a huge difference.

No. Actually it was 11.266, and 11.490 in 2001. You used team stats, which take into account sacks. FOF team stats to not take into account stats. To compare apples to apples, you've got to add up the individual players' passing yardages, not the team totals.

[korme]

Originally posted by SkyDog
Hmmm...interesting thought.

Just FYI, I'm up to 2033, with the last 8 seasons having been done with injuries=200. Passing has come down a bit, but not to present-day totals. Most recent season totals (2033):

Completion Percentage: 61.2% (higher than normal)
Yard/Completion: 11.5 (right in line with real-life totals)
Yards/Attempt: 7.0 (a little higher than real life)
Rushing Yards/Attempt: 3.9 (in line with pretty much every season but 2K2)

Looks like the problem is becoming isolated to being too high of a completion percentage. Passing and rushing attempts 31 years into the game are still so ridiculously close to real-life, it ain't even worth talking about attempts any more.

Disagreed.3.9 is close, but on the low end of realism. Get that average consistently to 4.0-4.2 and I will be pleased with rushing. But right now it is dipped too low.

[Fritz]

Originally posted by CraigSca
EVERY QB excels in this situation and rarely gets hurt. Ok...not EVERY QB...but, is anyone finding it hard to draft a pretty darned good quarterback in this game? Is that representative of the NFL in ANY time period?! Since when is a good quarterback a commodity?

how do you define a "good qb", or a good player for matter in FOF terms?

[Fritz]

Ben,

my pet theory is that there may be too many plays in a season.

[Samdari]

Repetitive post deleted.

[CraigSca]

Originally posted by SkyDog
No. Actually it was 11.266, and 11.490 in 2001. You used team stats, which take into account sacks. FOF team stats to not take into account stats. To compare apples to apples, you've got to add up the individual players' passing yardages, not the team totals.

Whoops...Ok, thanks! That's one less thing to work on then

[Ben E Lou]

Originally posted by Shorty3281
Disagreed.3.9 is close, but on the low end of realism. Get that average consistently to 4.0-4.2 and I will be pleased with rushing. But right now it is dipped too low.

In the four season of my "real" career, it has been:

2003: 4.0
2004: 4.1
2005: 3.9
2006: 4.1

Haven't seen a 4.2 yet, but I'm not sure there's been a 4.2 besides 2002 in the NFL.

[Samdari]

Originally posted by Fritz
Ben,

my pet theory is that there may be too many plays in a season.

Well, initial results suggests that is the number that best matches real NFL data.

[amdaily]

Originally posted by SkyDog
Ummmm....you have GOT to be kidding me.

NFL 2K2: 14102 rushing attempts
FOF 2K3: 14074 rushing attempts

NFL 2K2: 17292 passing attempts
FOF 2K3: 17480 passing attempts


(The FOF 2K3 numbers came from my saved career with real player names file.)

Are you seriously trying to suggest that you have a problem with that minute of a variance????

Perhaps that came out a bit to strong. It turns out Daimyo pointed out a huge flaw in my chart on page one - I did the FOF2001 control without the 32nd team . So disregard that post. The attempts do check out when adding in the first expansion team.

That also tosses out my theory on run % in the gameplans as being the culprit.

[Ben E Lou]

Originally posted by Fritz
Ben,

my pet theory is that there may be too many plays in a season.

NFL 2K2: 14102 rushing attempts
FOF 2K3: 14074 rushing attempts

NFL 2K2: 17292 passing attempts
FOF 2K3: 17480 passing attempts

That's 31,394 plays in real life, versus 31,554 in an FOF season. Pretty close. I don't think an extra 160 plays, spread out over 256 games in a season is going to make the big difference we're seeing.

[CraigSca]

Originally posted by Fritz
how do you define a "good qb", or a good player for matter in FOF terms?

Loaded question Well...not by ratings, by their statistical output in the game.

Again...just looking at my 2008 statistics, there are 32 qualifying quarterbacks. 16 of them have a passing rating of 90 or more. There were 4 in the NFL last year. Passing is just too efficient in FOF2004.

[Ben E Lou]

Got it amdaily. At this point I think it is a fairly simple combination:

1. completion percentages are too high
2. QB injuries are too low

[Fritz]

then it would seem to be yards per attempt ought to be what is high.

[CraigSca]

Ok....so it's solved then.

Jim, if you could add the following line to your code, everything will be fine:

completionPercentage -= 0.05

[Ben E Lou]

Originally posted by Fritz
then it would seem to be yards per attempt ought to be what is high.

It is. In FOF2004, I am seeing a four-season average of 7.0 yards per attempt, while in 2001 and 2002 in the NFL, it was 6.7.

[amdaily]

Originally posted by SkyDog
Got it amdaily. At this point I think it is a fairly simple combination:

1. completion percentages are too high
2. QB injuries are too low

Not to throw a wrench into that theory (or potentially bring more embarassement upon myself because I leave out a variable ), but have you added up the total TD's thrown by all QB's?

In FOF2004, I totaled 755 TD's, vs a FOF2001 league (yes, with all 32 teams ), that threw for a total of 586 TD's.

Could it be the yards/completion that makes for the 4,000 yards seasons, and this TD discrepiecy that accounts for the high QB ratings?

[John Galt]

Originally posted by CraigSca
Ok....so it's solved then.

Jim, if you could add the following line to your code, everything will be fine:

completionPercentage -= 0.05

Until someone tests Daimyo's league file, can this conclusion be reached? It could be that the player ratings is the culprit and not the statistical output. Lowering the completion percentage would be a bandaid fix for the modified default roster, but I'm just wondering how it would affect fictional leagues with rosters using the Help file distribution.

[CraigSca]

Originally posted by John Galt
Until someone tests Daimyo's league file, can this conclusion be reached? It could be that the player ratings is the culprit and not the statistical output. Lowering the completion percentage would be a bandaid fix for the modified default roster, but I'm just wondering how it would affect fictional leagues with rosters using the Help file distribution.

I thought Skydog was testing a 20 year old league. If that's the case, Daimyo's fictional league would be a nice start for a future dynasty, but would have very little impact on what's wrong with the statistical output of the game (although it would comfirm what Skydog is seeing 20 years down the road). Of course...the talent level of draft picks could be skewed as well as the initial player file. If that's the case, all bets are off and the game is, for all intents and purposes as an NFL simulation, "broken" for lack of a better word. I hope not, and I have all the confidence in the world that if that *IS* the case, it will be corrected.

[JonInMiddleGA]

All this talk made me curious (it's been a busy day around here, I hadn't been able to look sooner)

My only completed season, with the current NFL players as my starting crop.

One 4k+ passer (Favre, 4,338, league MVP)
Nine other 3500+ (Brunnel, Gannon, Manning, Carr, Harrington, Culpepper, Green, B.Johnson, Collins)

Looking at my QB's by game
19 QB's played 16 games
4 played 15 (Carter, Weinke, Collins, Maddox, Bulger)
1 played 14 (B.Johnson)
1 played 13 (Garcia)
1 played 12 (Chandler: 9 with Chi, 3 w/Buf)
1 played 11 (Warner)
1 played 10 (Vick)
3 played 9 (Bledsoe, Boller, Dilfer)

I also had 12 RB's at 4.0 YPC or higher (no idea how that matches reality) and 14 w/ 1000+ yards

[Ben E Lou]

Originally posted by amdaily
Not to throw a wrench into that theory (or potentially bring more embarassement upon myself because I leave out a variable ), but have you added up the total TD's thrown by all QB's?

In FOF2004, I totaled 755 TD's, vs a FOF2001 league (yes, with all 32 teams ), that threw for a total of 586 TD's.

Could it be the yards/completion that makes for the 4,000 yards seasons, and this TD discrepiecy that accounts for the high QB ratings?

Let's see....

Just looked at one FOF2004 season, got 781 TD passes. There were 634 in 2001, and 694 in 2002. Of course, this may be the result of more factors being added in. Most significantly, I wonder if the AI is throwing inside the red zone more often than in real life, and the jacked up completion percentages result in more TD's. Lower rushing ypc totals than in 2001 and 2002 in the NFL could come into play here as well. The AI may be facing more 3rd and goal from the 4 yard line situations, and then hitting more TD's. That one is probably too hard for us to isolate.

[Ben E Lou]

Originally posted by CraigSca
I thought Skydog was testing a 20 year old league. If that's the case, Daimyo's fictional league would be a nice start for a future dynasty, but would have very little impact on what's wrong with the statistical output of the game (although it would comfirm what Skydog is seeing 20 years down the road). Of course...the talent level of draft picks could be skewed as well as the initial player file. If that's the case, all bets are off and the game is, for all intents and purposes as an NFL simulation, "broken" for lack of a better word. I hope not, and I have all the confidence in the world that if that *IS* the case, it will be corrected.

I'm looking at two different leagues now (desktop and laptop running). One is in 2034, the other is in 2007. Most of the specifics I've given have been from the 2007 league (Primelord rosters), BUT the 2034 league numbers are very, very close to numbers at the beginning of the simulation, so there's a problem either way.