FOF2K4: Low Avg. Yards Per Carry for RB

[castorius]

Out of curiosity, what was the verdict on the low YPC (yard per carry) for RBs? (avg yards per rush)

Was this fixed in the patch? Does this need to tweaked in a patch?

I just noticed that my first year simming, YPC for RB was quite low. Not extremely low, but low enough for me to notice. Only one player broke over 5.0 YPC and he only had 680 yards rushing total. Currently in the real NFL, there is 5 players in the Top 10 leading rushers with over 5.0 YPC. I think that's a bit high than would be expected in a normal year, so looking at last season's statistics, there was 3 players in the Top 11 leading rushers with over 5.0 YPC.

What do you guys think?

[Vince]

If only 3 did it last year, and 1 did it in your test sim, I'd say that's just an off year. Where there any/many players with high 4.X YPC?

[castorius]

Well, most were hitting around the 3.6-3.8 area, a few in the 4s. The thing about my statistical comparison to last season is, I only mentioned 3 players over 5.0 YPC in the Top 11 rushers (in real life). I didn't mention how many players were over 5.0 YPC among all rushers. If I mentioned all rushers, it would probably be something like 6-7 players over 5.0 YPC. Whereas in my FOF2K4 season, only 1 player in the whole league had over 5.0 YPC, and he wasn't even among the Top 15 rushers. Nobody in the Top 10 rushers, had over 5.0 YPC. And that is quite a rarity in real life.

I'm just wondering if other people who have simmed more seasons are getting different results. The main reason I'm concerned after only one season is, people on this board have mentioned this problem before (before the patch). I was wondering if this problem still exists.

[Ben E Lou]

In extensive post-patch testing, this problem has been shown to be eliminated.

[Ben E Lou]

Here are the numbers from a 30-year test. The ranges on the left are the league ranges for the past five years in the NFL. Note that the averages for 30 years of FOF all fall within the real NFL range, and rushing yardage in particular is actually a hair higher than in the past five seasons of the NFL.

And a final comparison for the key stats...

League YPC: 3.90-4.21 (Mean: 4.05, FOF2K4: 4.09)
League YPA: 6.72-6.85 (Mean: 6.77, FOF2K4: 6.74)
League YPComp: 11.26-12.10 (Mean: 11.66, FOF2K4: 11.4)
Comp %: .5660-.5965 (Mean: .581, FOF2K4: .590)

If you want to see more detail on this study, go to this thread

[castorius]

Thanks SD for the info,

I simulated a few more seasons (about 8 seasons), and I observed that, overall I guess the YPC average of the whole league in FOF2K4 is pretty well aligned with the real NFL whole league. But generally, you don't get as many extremes (ie. over 5.0). The league overall average is equal, but individually there are variations. I think the NFL tends to have more guys with that are in the extremes (ie. high and low YPC, 5.3 & 3.4) of the YPC average, whereas FOF seems to come to the YPC average with players who tend to be closer to the mean (ie. 3.9-4.4). Either way, I don't see it as big statistical issue. I'm sure others have looked into this more than I have, so I'll trust you guys on this one.

[Ben E Lou]

Castorius:

I'm pretty sure no one has studied it from the standpoint of variations. Question: What player file are you using? I ask because there is some speculation out there that the initial player file doesn't have an accurate distribution of talent, which would skew stats somewhat in the early going. I'm looking at a career that is at the end of 2020 right now, and therefore has no players from the original file, and there are two qualifying RB's with ypc >5.0, and 12 who are 4.5 or better, and there are 8 with ypc <3.9, three of whom are <3.5. I don't know what those numbers should be though.

[Godzilla Blitz]

Originally posted by castorius
Thanks SD for the info,

I simulated a few more seasons (about 8 seasons), and I observed that, overall I guess the YPC average of the whole league in FOF2K4 is pretty well aligned with the real NFL whole league. But generally, you don't get as many extremes (ie. over 5.0). The league overall average is equal, but individually there are variations. I think the NFL tends to have more guys with that are in the extremes (ie. high and low YPC, 5.3 & 3.4) of the YPC average, whereas FOF seems to come to the YPC average with players who tend to be closer to the mean (ie. 3.9-4.4). Either way, I don't see it as big statistical issue. I'm sure others have looked into this more than I have, so I'll trust you guys on this one.

Actually, I would think the standard deviation in YPC could have a huge impact on gameplay. If the standard deviation is small, it really dictates against spending draft picks or cash for a top running back, since the YPC difference is relatively small between the a top back and a substandard one. Of course, this doesn't factor in receiving ability, so your best bet would be to get any cheap back with good receiving skills and low fumbles. I think Jim really takes a good look at numbers like this when coding FOF, though, and would be surprised if the standard deviation is truly off. I would guess we'll find that everything is ok.

However, it'd be interesting to see someone take a look at this. A comparison between the NFL for the past five years and a likewise sample in a league that's 20 years out would be one way to answer the question.

[Subby]

Originally posted by Godzilla Blitz
Actually, I would think the standard deviation in YPC could have a huge impact on gameplay. If the standard deviation is small, it really dictates against spending draft picks or cash for a top running back, since the YPC difference is relatively small between the a top back and a substandard one. Of course, this doesn't factor in receiving ability, so your best bet would be to get any cheap back with good receiving skills and low fumbles. I think Jim really takes a good look at numbers like this when coding FOF, though, and would be surprised if the standard deviation is truly off. I would guess we'll find that everything is ok.

I, and others (like Daimyo), have always maintained that spending high picks and heavy cash on a running back is a poor strategy. Invest in a good offensive line and then find backs in the late rounds with great speed and handles and you can just rotate backs from year to year with a minimal expenditure of resources.

As multi-player leagues begin to get rolling I think this will bear out over time - particularly with the intense impact the salary cap seems to have.

[HornedFrog Purple]

I do pretty much the same thing using a running back by committee. I try to find guys that are good at what my offensive line is good at. There are backs in the game who specialize in what you are trying to do and even though the rest of their attributes are below average they work for me.

Going the super back route gives me problems if he gets hurt. If my starter gets hurt, I just plug in the next guy and he does about what the starter was doing and it does save a lot of money to upgrade other positions.

[Godzilla Blitz]

Originally posted by Subby
I, and others (like Daimyo), have always maintained that spending high picks and heavy cash on a running back is a poor strategy. Invest in a good offensive line and then find backs in the late rounds with great speed and handles and you can just rotate backs from year to year with a minimal expenditure of resources.

As multi-player leagues begin to get rolling I think this will bear out over time - particularly with the intense impact the salary cap seems to have.

I found myself stumbling onto this same approach when playing an FOF4 game. I added some financial restraints and never had enough money. I just kept getting cheap, old backs with a couple of years left in them. Seemed to work well enough, all things considered.

I think the interesting part of all this is that there are probably many in the NFL that would argue that this is a good strategy to follow in the real world as well.

Still, I think it makes for a better game if the variation in backs is comparable to that of the NFL.

[castorius]

Originally posted by Godzilla Blitz
Actually, I would think the standard deviation in YPC could have a huge impact on gameplay. If the standard deviation is small, it really dictates against spending draft picks or cash for a top running back, since the YPC difference is relatively small between the a top back and a substandard one. Of course, this doesn't factor in receiving ability, so your best bet would be to get any cheap back with good receiving skills and low fumbles. I think Jim really takes a good look at numbers like this when coding FOF, though, and would be surprised if the standard deviation is truly off. I would guess we'll find that everything is ok.

However, it'd be interesting to see someone take a look at this. A comparison between the NFL for the past five years and a likewise sample in a league that's 20 years out would be one way to answer the question.

That's my biggest concern with this issue, whether it effects game balance and gameplay. I agree that using a running back by committee is a valid strategy and one that real NFL teams use. However on the same token, the large majority of the teams in the NFL, use a feature running back as their strategy. I don't want that aspect being taken away in this game. I believe using a feature back should be a valid and successful strategy as well (depending on other factors), or else teams in the NFL wouldn't be doing this for years. It appears that most Human GMs seem to favor the running back by committee strategy (and it works because of perhaps the deviation among running backs), and that may pose as a balance problem because the CPU GMs don't all follow. And if the CPU GMs do follow, it may also pose a gameplay problem because its not realistic to have 'all' teams use a RB by committee strategy because alot of teams do have a 'feature back' in real life.

Like I said, I don't have enough experience playing this game to make the judgement yet. I was hoping that some of you more experience FOFers might know a thing or two about this. Right now, all I can go on is that the consensus I've heard from most experienced players on this board tend to follow with the same "RB by committee strategy". It could be a problem if 90%+ people do use this similiar strategy, and they use this strategy particularly if it has to do with the overly high success rate of this strategy because it deems other strategies obsolete (ie. feature back). If 'RB by committee' was truly this successful in real life, you'd probably see 90% of the teams in the NFL use it, which I don't believe is the case.

[castorius]

BTW, I don't know if its a coincidence, but I also use/favor the RB committee approach. I always pride myself on building that big, skilled O-Line, and just having a fair skilled RB run behind it (ala Emmitt Smith.... )

[Bee]

Originally posted by castorius
That's my biggest concern with this issue, whether it effects game balance and gameplay. I agree that using a running back by committee is a valid strategy and one that real NFL teams use. However on the same token, the large majority of the teams in the NFL, use a feature running back as their strategy. I don't want that aspect being taken away in this game. I believe using a feature back should be a valid and successful strategy as well (depending on other factors), or else teams in the NFL wouldn't be doing this for years. It appears that most Human GMs seem to favor the running back by committee strategy (and it works because of perhaps the deviation among running backs), and that may pose as a balance problem because the CPU GMs don't all follow. And if the CPU GMs do follow, it may also pose a gameplay problem because its not realistic to have 'all' teams use a RB by committee strategy because alot of teams do have a 'feature back' in real life.

Like I said, I don't have enough experience playing this game to make the judgement yet. I was hoping that some of you more experience FOFers might know a thing or two about this. Right now, all I can go on is that the consensus I've heard from most experienced players on this board tend to follow with the same "RB by committee strategy". It could be a problem if 90%+ people do use this similiar strategy, and they use this strategy particularly if it has to do with the overly high success rate of this strategy because it deems other strategies obsolete (ie. feature back). If 'RB by committee' was truly this successful in real life, you'd probably see 90% of the teams in the NFL use it, which I don't believe is the case.

I think this is a huge leap you're making. If there's a problem with the deviation from the mean for rushing yardage, it certainly doesn't appear to be so great as to affect gameplay from what I've see.

I also seriously doubt 90%+ use the "RB by committee" strategy all the time (I certainly don't use it all the time). Also, I doubt you'll find that the majority use it because of the "high success rate". It would be my guess you'll find most use it because it's cheaper, not that it makes the use of a feature back obsolete.

[Godzilla Blitz]

Originally posted by castorius
That's my biggest concern with this issue, whether it effects game balance and gameplay. I agree that using a running back by committee is a valid strategy and one that real NFL teams use. However on the same token, the large majority of the teams in the NFL, use a feature running back as their strategy. I don't want that aspect being taken away in this game. I believe using a feature back should be a valid and successful strategy as well (depending on other factors), or else teams in the NFL wouldn't be doing this for years. It appears that most Human GMs seem to favor the running back by committee strategy (and it works because of perhaps the deviation among running backs), and that may pose as a balance problem because the CPU GMs don't all follow. And if the CPU GMs do follow, it may also pose a gameplay problem because its not realistic to have 'all' teams use a RB by committee strategy because alot of teams do have a 'feature back' in real life.

Like I said, I don't have enough experience playing this game to make the judgement yet. I was hoping that some of you more experience FOFers might know a thing or two about this. Right now, all I can go on is that the consensus I've heard from most experienced players on this board tend to follow with the same "RB by committee strategy". It could be a problem if 90%+ people do use this similiar strategy, and they use this strategy particularly if it has to do with the overly high success rate of this strategy because it deems other strategies obsolete (ie. feature back). If 'RB by committee' was truly this successful in real life, you'd probably see 90% of the teams in the NFL use it, which I don't believe is the case.

Well said. You've said it much better, but your point is what I was hinting at by saying that I think it would be nice if FOF mirrored the real NFL with the variety of YPC by running backs.

One other thing to consider is that the down side matters just as much as the up side. If the NFL has a number of main backs that are only averaging 3.4 - 3.6 yards per carry, those teams ground games will significantly hurt their offense. One of the reasons that NFL teams might always look for top quality at running back is that you can't be guaranteed that your running back will even be mediocre. In FOF, however, if the vast majority of backs can crank out 3.9-4.4 yards per carry, you've got good odds of throwing just about anyone in there and not hurting yourself.

I think, in part, though, that this issue is influenced by the nature of the running back position. In general, running backs mature quickly and have short careers. Part of the reason many teams use average running backs for relatively short lengths of time is that they're just aren't that many good ones with long enough careers to build an offense around. It's one thing to get one or two 4.5+ YPC seasons, it's a rare creature that can do it for 5+ seasons.

[Daimyo]

My RB's almost always come from the 5th round or lower and I've never had any problems generating a rushing attack. In fact if I run a '75' offense I'm pretty confident my team will finish in the top 5 in rushing nearly every season with RB's like that. It has been the same since FOF2k1 at least (who knows with FOF2, it was too incredibly easy to have all maxed out guys to ever really think about strategy).

My gut tells me this is pretty realistic on a macro level. RB seems to be one position where players can come out nowhere and become standouts in the NFL.

[Daimyo]

DOLA, here is a thread from two and a half years ago talking about the importance of various positions in fof2k1. Alas, the graphs and raw data are all gone and the discussion isn't as interesting as the '75' offense thread linked above, but it might be somewhat interesting. The basic conclusion with regards to RB was that while a good RB made a noticable difference, you could pretty easily get by with even a zeroed out guy as long as the rest of the team was solid.

here

[castorius]

Originally posted by Daimyo
My RB's almost always come from the 5th round or lower and I've never had any problems generating a rushing attack. In fact if I run a '75' offense I'm pretty confident my team will finish in the top 5 in rushing nearly every season with RB's like that. It has been the same since FOF2k1 at least (who knows with FOF2, it was too incredibly easy to have all maxed out guys to ever really think about strategy).

My gut tells me this is pretty realistic on a macro level. RB seems to be one position where players can come out nowhere and become standouts in the NFL.

Well I have my personal beliefs on the value of a 'good RB', but that's a whole other issue entirely. I actually believe that almost any able-bodied football player can be a successful RB if they have a good Offensive Line (ie. Broncos producing 4 different 1000 yard rushers the past 5-6 years?). Sure a RB who is actually good & truly talented could help (ie. Barry Sanders), but its not necessary in order to have a successful running attack.

In that regard, you may be right that it is 'realistic' in that sense. However, realistic it may be, but does it fit with the balance of the game itself? I may have this personal belief of truth (that a good RB is not highly valuable), but does it mean its neccesarily true? The answer is, it doesn't really matter, because I don't believe FOF is a game that tries to answer the theoretical truth, but rather tries to emulate what is known and what is there. And right now, in the current NFL, feature backs are considered valuable and useful to a successful running attack (otherwise you wouldn't be seeing NFL teams adopt this strategy or shell out so much for a top RB). It doesn't mean that its an absolute neccesity, in some cases, 'RB by committee' works, but at the same time, RB by committee can fail, and actually often does if you don't have RBs with decent YPC. It also fails because if a RB is successful in a limited role like 'RB by committee', its very hard to keep them happy at that role, most eventually want to become a featured RB.

[RPI-Fan]

I'm going to run a study on this - will report back some point.

[castorius]

Basically, I just don't want this to become an ultimate strategy. You know the one in Real Time Strategy games, where everybody does it because it works all the time. Why do anything else, if this one strategy is highly successful?

Don't get me wrong, I'm not reaching a conclusion on this issue, but I'm fearful that I might (unless convinced otherwise). If it is proven to be an ultimate strategy, I think its not only consequential to the realism (in trying to emulate the NFL), but also consequential to the variability of the game itself. This severely limits the valid strategies, if proven to be true.

[RPI-Fan]

TerribleRB, GreatOL:

Definition of Terrible RB = High Breakaway Speed + Endurance, Low Everything Else								Using "75 Offense"		No Injuries

Trial	RB Carries	RB Yards	RB YPC	Team YPC	League YPC	League Qualifying RB's <3.5	3.5<=x<4.0	4.0<=x<4.5	x<=4.5	Team Wins
1	502	2179	4.34	4.11	4.07	9	18	13	10	8
2	510	1912	3.75	3.7	4.09	5	17	16	7	7
3	482	2045	4.24	4.69	4.09	7	16	18	10	8
4	510	1781	3.49			5	15	19	7	7
5	525	1923	3.66	3.66	4.05	8	24	13	8	4
Avg	505.8	1968	3.90	4.04	4.075	6.8	18	15.8	8.4	6.8

[RPI-Fan]

Keep in mind that I fucked up, doing these trials with the default player files. Nevertheless, it should provide some insight as to the effects of various strategies.

[RPI-Fan]

I'll try and finish this by the end of the week - if someone else wants to, feel free to let me know and I'll gladly send you the Excel file I used to record data (things went along pretty quick with it).

I have finals next week and will need to get studying.

[Godzilla Blitz]

I think there are two questions to examine.

The first is whether the standard deviation in YPC in FOF accurately reflects that of the the NFL. I feel like we're assuming that it's off target, but I there's no evidence that shows this.

If FOF's YPC standard deviation is too low, then it would be pretty easy to guess that a study similar to Daimyo's examination showng that running backs aren't an important factor in total wins would produce the same results. They are all too bunched together to make much difference.

If the YPC deviation is accurate, then comparisons to the NFL are more valid.

I'm starting to wonder if the deviation is off, though, and that might explain the tendency for a lot of the people that have played the game a lot to have concluded, consciously or not, that spending high picks or cash on RBs is not worth it.

[RPI-Fan]

I'm pretty much re-doing Daimyo's study. I think there is a big enough gap in versions that it might be interesting to see any possible new results.

Incidentally, I may start a new standard deviation study. It wouldn't be all that hard (start with an HR file, sim out 8 years), then just study all the qualifying RB's year-by-year.

[castorius]

Originally posted by SkyDog
Castorius:

I'm pretty sure no one has studied it from the standpoint of variations. Question: What player file are you using? I ask because there is some speculation out there that the initial player file doesn't have an accurate distribution of talent, which would skew stats somewhat in the early going. I'm looking at a career that is at the end of 2020 right now, and therefore has no players from the original file, and there are two qualifying RB's with ypc >5.0, and 12 who are 4.5 or better, and there are 8 with ypc <3.9, three of whom are <3.5. I don't know what those numbers should be though.

Answer to an old question, but oh well.

I'm using the CSV Version Of Real Names Player File in your "ALL-INCLUSIVE REFERENCE THREAD". I simply used the PlayerFile.exe to make it a .fdt file and then simply renamed the file to player.fdt and put it into my Front Office Football 2004 folder to replace the old original players.fdt

From there, I answer 'No' to "Begin Game Using Alternate Player File" and I still get the real players.

[castorius]

Originally posted by RPI-Fan
I'm pretty much re-doing Daimyo's study. I think there is a big enough gap in versions that it might be interesting to see any possible new results.

Incidentally, I may start a new standard deviation study. It wouldn't be all that hard (start with an HR file, sim out 8 years), then just study all the qualifying RB's year-by-year.

I think there could be more to it than simply a standard deviation problem. What I've noticed through several season sims is that the best rushers (among the top 15 leading rushers) tend to have lower YPC avg. than would be expected in real life, and the backup RBs tend to have good YPC avg., maybe better than expected in real life. Perhaps the backups getting better YPC avg. may be lifting the overall league YPC average up to real life standards. Sure the weight of a backup RBs yards won't account to as much because they don't get nearly as many yards or rushing attempts, but it still might be significant enough to tilt the figures. However, more analysis needs to be done to verify this.

I think you would have to compare the figures of the top 20 rushing leaders to the top 20 rushing leaders in real life. Simply doing an analysis on the overall league average isn't gonna give you an accurate account in this case. Backup RBs who get high averages may skew the stats a bit (they may compensate for the low YPC avg of the top rushers). I'm guessing that backup RBs get good YPC numbers because they are going against weaker competition when they do play (tired players and backup players).

[castorius]

I've checked up the stats of the Top 15 leading rushers in the past decade or so. I didn't do an objective experiment, I simply took a quick count and wrote it down on a piece of paper. From my observations, I found that in real life there are about 3-5 players with over 5.0 YPC avg in a given season. On average, usually 3 players with 5.0 YPC. I think one season had 2 players only. In FOF2K4, on average, I usually see only 1 player with over 5.0 YPC amongst the top 15 rushers. Alot of times there are none (a more accurate average would be more like 0.8 players with over 5.0 YPC in a given season). I've never seen 3 or more players with over 5.0 YPC in a season. I think I might've seen 2 at most. But in general, the YPC seem to be low amongst the leading rushers in the league (in FOF2K4). And this is not strictly judging from players with over 5.0 YPC. A quick glance at the figures, it seems as if they are just generally low in range when comparing to real life. Again, I didn't do an accurate objective study on this, so I would have to verify this for it to be considered a fact. Call this just a quick observation.

[RPI-Fan]

castorius: I agree that the numbers appear to be slightly low on the top-end, and slightly high on the bottom-end.

Regardless, this isn't a MAJOR problem, it doesn't seem, and if I determine whether or not using a great RB is more effective than a mediocre RB and great OL, we'll know for sure.

[Godzilla Blitz]

Originally posted by RPI-Fan
Regardless, this isn't a MAJOR problem, it doesn't seem, and if I determine whether or not using a great RB is more effective than a mediocre RB and great OL, we'll know for sure.

Another way to think about the impact of a top running back is to consider the statistical differences among levels of running backs.

Take three running backs: A, B, and C. Assume each of them carries the ball 300 times in a season, and that receiving, fumbles, etc., are equal. Running back A gains 5.0 YPC, B gets 4.5 YPC, and C gets 4.0 YPC.

Doing the math, Back A will gain 1500 yards on the season, B will gain 1350, and C will gain 1200.

The statistics of football dictate that it takes about 12 yards of offense to put one point on the scoreboard. Breaking this down, running back A will be worth 125 points of offense, B will be worth 112.5 points, and C will be worth 100 points.

Stats also tells us that a 20-point differential in points scored versus points allowed over the course of a season should average out to one more win in a 16-game season. Again, there are lots of exceptions, and the scale breaks down at the ends of the spectrum, but on average, a 9-7 team will score 20 more points than they give up, an 8-8 team will be even, and a 7-9 team will give up 20 more points than they score.

Putting the two together, our 5.0 YPC back will generate 12.5 more points (roughly one-half of a win) more than the 4.5 YPC back, and 25 more points (roughly one win) more than the 4.0 YPC back.

Bottom line: A great back (5.0 with lots of carries) versus a good back (4.0 with lots of carries) should win you one more game in a season, on average.

If the running YPC stats in FOF are more closely bunched together than in the NFL, then the impact of the running back's performance in FOF will be even less than that of the real NFL.

[Ben E Lou]

Bump.