Question about off-season changes

[Marco_Frankfurt]

This goes mainly to Paul, but I think its something we all should discuss.

So Paul, did you have a lot of games simulated many times to get 'normal' stats or was it only some
at the beginning of the season ?

The running game was (in the official stats) a bit worse then last seasons and we only have one 1000 yard
rusher, however 3 more were on the way before they stoped playing for different reasons.
(Huerta was injured, Aska changed to FB, Simonton rested)

So if it was only a couple of bad games early in the season, then I would say we also make only a slight
change (a -1 to defense or so). However if it still was a lot of extra work for you Paul, then we should maybe discuss a more complex change.

I.e.: reducing DL SP,AC,AG stronger then the overall defense reduction, creating new maximum numbers for
the potentials of defense players, maybe also changing some offense values and so on.
This is a lot more work but may result in more satisfying results.
It has the risk to change the stats stronger so we may see a lot new records.

Whatever we do, I think we should discuss it with all (active) GMs and plan it in advance of the new season,
so everybody knows what is coming. If we decide to make a bigger change then of course also every single bit of  input from the more experienced GMs is necessary (Jason H, David ...)

[Paul-London]

Well its a difficult one as last offseason i wanted some decent 1st round picks as i thought some of the skills were too low.  I think in Round 1 at least 8 should be instant starters.  However of course last year i upped many players and David was right, these players were probably too good and i think that doesnt help the overall stats of the league.

Throughout this season i did have to re-run a few games, there was one game i had to rerun about 10 times which was frustrating as anything.  EVERYTIME the stats for BOTH HB's were something like 20 carries for -4 yards.  Totally unrealistic.

I would personally like to make small changes, i dont want 2500 yard rushers but one 1000 yard rusher is a bit poor.  I think David/Jason are the best ones to ask about this, me putting rookies above the limits last year didnt help so i should have listened to David/Jason more.

I would like to see at least 5 or so 1000 yard rushers and some healthy averages, not 3.1 yards a season for some HB's etc.  On the other hand because of gameplans we have HB's catching a ton of passes, i do like the introduction of gameplans however.

The rookies this year i would say will be a notch below last year for sure, i still want a couple of instant starters but just like in the NFL you have some 'off years' with the draft and then a great year. Difficult for teams with tons of draft picks as obviously if they knew the draft would be poor they might not have bothered with draft picks so its difficult to reach a happy medium

But i would say the files have to be changed slightly as the rushing games are non existant at the moment as are interceptions.  But this time instead of changing things i'll listen completely to David/Jason as i should have the year before

[Mario ADK]

PLEASE NO CHANGES WHATSOEVER. IT IS WORKING FINE THE WAY IT IS RIGHT NOW!!!

[Paul-London]

Its not Mario, the rushing attacks are pretty poor, one 1000 yard rusher is one of the worst stats we've ever had.  2 seasons ago the output was fine, enough good rushing attacks, the stats were pretty spot on

1 1000 yard rusher in the league isnt realistic, i love the FWL when its realistic, if we have stats like we had much earlier in the FWL (eg season 4) then i get bored with it.  Its not that far off but i think defenses are so good its stopped nearly every rushing attack

[Aztecs]

Its not Mario, the rushing attacks are pretty poor, one 1000 yard rusher is one of the worst stats we've ever had.  2 seasons ago the output was fine, enough good rushing attacks, the stats were pretty spot on

1 1000 yard rusher in the league isnt realistic, i love the FWL when its realistic, if we have stats like we had much earlier in the FWL (eg season 4) then i get bored with it.  Its not that far off but i think defenses are so good its stopped nearly every rushing attack

You wouldn't know it by my team.  Year in and year out Mexico's defense is as stout as a wet noodle.  They typically head for the locker room about mid-way through the 4th quarter.

[Paul-London]

Must admit you have some very good defensive players on your d-line and your LB's are much improved, shocked how bad your secondary was this year.  Perhaps a fraction more speed is needed.  One day that units gonna click well !

[Stuart-Barcelona]

I would say that the rushing stats are not THAT bad this year.
If players hadn't been rested/moved positions during the year we would probably have had 3 or 4 over the 1000, which seems about right.

The one thing i would say though is I don't think there are enough interceptions/Defensive touchdowns, but I guess to do that we would need to make the D better, which would in turn make the offense seem worse, its a vicious cycle.

[Paul-London]

Most seasons theres hardly any interceptions and specially interceptions for touchdowns etc.  I think sadly its just a quirk of the software, i mean Anes a few years back with 17 td's and 0 ints ! Very unrealistic sadly

[Marco_Frankfurt]

I would personally like to make small changes, i dont want 2500 yard rushers but one 1000 yard rusher is a bit poor.  I think David/Jason are the best ones to ask about this, me putting rookies above the limits last year didnt help so i should have listened to David/Jason more.

From my small knowledge it would be better to do a '-3' to DLs SP,AC,AG then an overall -1 to all Defense.
However we have allways the problem, that with better running stats we get also bad sack stats.
The same with more INT we get also again less big plays and so a lower run and pass average.

The rookies this year i would say will be a notch below last year for sure, i still want a couple of instant starters but just like in the NFL you have some 'off years' with the draft and then a great year. Difficult for teams with tons of draft picks as obviously if they knew the draft would be poor they might not have bothered with draft picks so its difficult to reach a happy medium

Maybe we should make it draft good for offense and mediocre for defense ?

The last two drafts were heavy on DBs and also some great LBs and DLs. I think it would be good to have a great year for
OL and some nice WR and RBs.

However I think the league has allways a place for at least one superstar from each position, but we should not
flood a position with 10 superb player of one position. Or only if we want to become the league as a whole better in that are, like for OLmen.

All this is just my opinion and open to discussions

[JH-SanAntonio]

We could certainly debate a variety of changes for defense, but I don't think "only one 1,000-yard rusher" is a good measurement.

Most seasons see quite a few players hovering around 250 carries.  This year we had what, two?  Are we suggesting that guys with 180 carries should get 1000 yards, and defenses should be reduced to make that happen?  Meh.

This season's low turnout for 1000 yard rushers was likely due more to injuries and trades than anything else.

An informed discussion on this matter would also require that actual numbers be given to us by Paul- how many games/week are rerun?  Are you rerunning based on scores or rushing totals?
     Using an already published example- a few weeks ago the program puked and Paul was forced to rerun the weeks games (entirely an error in the game, not Paul's fault).  In the initial game my HB picked up something like 90 yards, in the reran game he ended up with 30.
     Just 3 instances of rerun games could, in theory, turn a 1000-yard rusher into an 800-yard rusher.  To average 3 reruns per team, Paul would have only need to rerun a total of 27 games over the course of the entire year.  That week alone, he had to rerun a minimum of 9 (to correct the bug), and who knows how many more to get the initial winners to win the second time around (which can be huge- imagine a game where the underdog has a 10% chance to win, pulls it off, but then Paul is forced to rerun the game... on average he would need to run that matchup 10 times to get the same winning result).  After week 3 he was complaining about having to rerun a couple of games multiple times- how many were rerun that week?  Over the course of the year?

Without those numbers and the attendant stats from them, we can't really make educated decisions.
    As a further example, I have a defensive lineman who had a fairly weak season- 33 tackles, 6 sacks.  He played poorly a few weeks due to injuries, etc.. In the week that had to be rerun, the final run had him pick up 1 tackle, then get injured for 2 weeks.  In the first run, not only was he not hurt, he picked up 7 tackles and 5 sacks.
    Imagine if he had those numbers for the 2 weeks he was hurt- he would have finished the season with 47 tackles, 16 sacks.  Suddenly a mediocre season turns into an awesome one.  If you look at his 33/6 number, you couldn't possibly make the case that defensive linemen need reducing.  At 47/16, the argument suddenly warrants consideration.
    We CAN NOT make decisions based on the stats we have without considering the stats that were pulled.

So... Does Paul have all those numbers?
    Does he actually have each set of games saved and the stats from them parsed?  I'd say no.

If he does, would he want to share them?
    As a GM for a simulation league, there are things you have to do frequently to keep things running smoothly.  Does Paul really want to say "Look, for good halfbacks I figure they should get at least 40 yards if they run 12+ times.  I know your HB had 9 41-43  yard games that weren't rerun, and San Antonio had 6 39 yard games rerun that averaged 140 yards per game afterwards, and I'm sorry, but that's my cutoff."?  I doubt it.

If he did, would we actually want to know?
    You might, but honestly, I don't, and as both Paul and Dave can attest, 90% of my enjoyment during the season is screwing around with my setups to get different stats out of my players.   Paul and I clash often enough (as I'm sure most of you are aware), but I figure most of what he does is to make the league more enjoyable, and don't really want to know about it.


I've said it before, and I'm sure I'll end up saying it again.  Regulating what you put into the draft and proper capping works great.  Random flailing about and knee-jerk penalties to stats don't.  If we are to make changes, we need to stick to what works.  Using the FIFL  (for those of you who are unaware, a FBPro league very similar to the FWL in which many FWLers play as well) as an example, we put in caps and a draft structure Dave has stuck to at the end of season 5.  We haven't had to hit defense with a penalty since then- and we're entering season 13.

That having been said- it's more work to set up the draft each year, there is no "The computer sets it up randomly, I'm sorry!" excuse for the GM to use when a player is unhappy with the draft list, and ultimately unpopular trends in performance attributable to initial player abilities get dumped at the feet of the GM.  It's not all wine and roses, as I'm sure Dave will grumpily attest.

Some of the caps we use in the FIFL were put in to the FWL 4 seasons back, though I'm not sure whether they have been held to- and I do know that penalties have been assessed since then as well.

Without analyzing all of the thrown out stats etc. as well, we can't make decisions based on the stats we have.  What we CAN do is look to models that have been proven to work.  I'm sure there are many out there other leagues use that work great.  The only one I can vouch for, however, is the one that has help up over many seasons and we have personal experience with, and that's the FIFL model.

1. You MUST regulate the draft.
2. Capping must be implemented, held to, and given time to work.
3. Penalties must be accompanied by corresponding cap reductions.


But for God's sake don't look at stats in a vacuum, and don't make changes that aren't necessary.


This year, rushing numbers are probably down due to trades and injury, and their consequential toll on attempt totals.  However, there were also fewer runaway games this season, which reduces carries and rushing overall.
     Looking at other stats, some teams are relying increasingly on short, conservative passing attacks to move the ball, further reducing carries and the need for dominating running games.  This brings scoring down a bit, but also changes the character of the running game.
     Many players are using offenses that push the ball almost exclusively between the tackles, which provides for a more consistent running attack- but also in lower yards/carry overall.
     Other teams appear to be using more aggressive outside-run attacks, but lack the offensive line to make it work, which can crush a HB's stats early in the game and leave the team throwing the ball for the balance of the game- another bad day for our rushing totals.

As to lower INTs... the software itself tends to generate fewer INTs than you might normally see.  We get a lot less than normal, though.  But again, the vacuum rule applies.
    None of us wants to turn the ball over.  We train up abilities in our offensive players to reduce turnovers.  We try to prevent our runners from fumbling, but spend a great deal of effort to stop our QB from tossing INTs.  If you're upset about the lack of INTs in the league, dump your starting QB and play a guy with 70 IN and 45 DI.
    The last several seasons, DBs have hit the draft that were, in my opinion, silly- over the squalling of yours truly and others who felt the position was already overdone in general to begin with.  In a vacuum, this seems counterintuitive; if they are so good, why aren't they getting more INTs?
     My favorite DB example comes from the good old Raiders.  For several years, the Raiders had a pair of DBs, Lester Hayes (you might remember him as the Stickum guy) and ex-Patriot Mike Haynes.  Hayes was a stats maniac, with all kinds of tackles and passes defended each year; he was also a fan favorite, not only as a long-time Raider but also as a bit of a showboater on the field.  Everyone in the Bay Area knew who he was, and considered him to be just godly at DB.  Haynes was basically ignored by the fans, and even by sportscasters.  Hayes just did so much more!
    Then come the interviews with QBs and WRs of the day.  Who would they rather face?  Hayes!  Haynes didn't get the stats Hayes did- because he was all over his man and QBs refused to throw the ball at him.  They'd much rather throw to Hayes, and would in fact put up wounded ducks for his guy before they'd toss a laser at Hayne's guy.
    Our DBs don't suck, folks.  Our QBs are too smart and disciplined to throw the ball at receivers who are being blanketed by their cover guy, so throw it underneath to RBs and TEs.  Combine that with conservative passing games, and voila!  Low INTs. 
    If the DBs sucked we would see 400-yard games and 90% completions as a rule.

The moral of the story is:  Don't confuse player performance with managerial guidance, and make sure you look at all possible reasons why a given stat seems high/low.


To be honest, I don't think penalties are needed at this point, and certainly shouldn't be done outside of a larger overhaul to integrate planned changes.



JH

[Marco_Frankfurt]

Jason, first of all thank you for your detailed inside information.

I agree with you that there are things Paul has to do for the better of the league, that i as a player/GM
dont want to know.

But I think we GMs should know if and what kind of regulations are done to make the league stats better
(however 'better' is defined) If all GMs can read here about the reasons, they probably will not moan as much
about a 'horrible draft' if there are no all 90s DLmen.

So could you or David give us some numbers what the caps in the FIFL are and how you limit the rookies.

I only know that many years ago there was a set of league files called 'Kang-files' (I think the it was the guys name)
that did get good rushing stats out of the normal plabooks (the ones we use and not any special created ones)
By doing two things:
Increasing OLmen to very high values
Decreasing SP, AC, AG of DLmen

It also did reduce the sacks of course and the other player values were not as high as they are in the FWL
right now. Just take a look at the rosters in season 1-3 then you see what kind of values were 'normal'
for the game.

However i never did test these files so I think the FIFL way is tested and therefore more reliable.

There was also another set of league files called VPNFL-files. They had a complete NFL league
with changed players and playbooks to match the real NFL stats, they also capped the players and
did gave out a rookies file with players suited for this league. But to turn the FWL into a league that
would meet those standards would be a gigantic work !

Miguel had a 8 team league some years ago, that was created from these VPNFL files. Maybe he still
has the files and and informations about it, so he could send these to David/Jason/Paul for information.
Would you do this please Miguel ?

In the end Paul has to decide as he runs the league and its mainly his freetime sacrifized.

Thank you Jason for trying to help our FWL. Please let us all continue to make it better

[Paul-London]

Thanks for the post Jason, very interesting reading.  Good point about the rushing game.

This year i havent had to rerun many games, in the old days there were many many games which had 20 carries for -15 yards.  Games like that i will always rerun

There was a game earlier in the season which was a nightmare, it probably didnt help that the computer always simmed the game in 'rain'.  If we had all domes then this wouldnt happen.  There was one game i think featuring Berlin and i had to sim that about 15 times and it took ages.  EVERY time the stats were 20 carries for -10 yards type of thing and thats soooo unrealistic.

So on the vast vast majority i dont have to re-run many games.  As for re-running  a week like i had to a few weeks back that totally was a one off. Never had that happen before. 

I think the rushing totals are down as more people are using offensive gameplans, this features many more screen passes to HB's and in some cases TE's.  In the old days TE's never caught more than 30 passes on the whole, now some can get 900 yards.  HB's always had about 60/70 catches and now we have many HB's with 80 catches.

Good point about the Raider CB's !! Same with Charles Woodson early in his Raider career, All-Pro constantly but hardly any ints as no teams threw his way.

And yes i admit i did mess up last year with way too high rookies, so this year i'll talk to you about what players to do for this years draft.  I would say offense for sure, as always i would like a good chunk of the 1st round to be decent, i dont want a 1st round pick 5th overall to not be able to come in and start and will take two years of training camps to be able to but i understand that you shouldnt put them too high either.

As for my freetime i dont have as much as i did previously, being married now and trying to start my own business i cant go overboard on the FWL so i obviously want to make things easy for myself.

The stats werent super crazy this year, they werent as bad as i think season 4 or 5 where defenses ruled supreme and on the other hand we dont want 58-55 shootouts.

After last year and not listening to David/Jason i will this year of course do the complete opposite and get their advice for the league.  I still want some instant starters but this year wont be as brilliant as last year.  Its like the NFL, some years you get a great crop, others you dont.

I totally would consider the caps on offense and defense that the FIFL uses.  I know we have a cap on offense and defense (i think some rookies ahem may have gone over that last season) but do you have caps on particular skills ?

The stats in the FIFL look very good both for offense and defense so i would be happy to have the same ones in the FWL.  Might ruffle a few feathers but its better for the longterm.

[JH-SanAntonio]

The most important thing in embarking on a revamp is knowing where you want to go with it.  Even if the changes seem minor to the owners, you'll be revamping a large part of your underlying structure for the league.  You won't want to have to do it again anytime soon.

I'd suggest taking the time to figure out exactly what you want, whether you want the rest of the owners to have a say in how it's set up, and to what extent their input will matter.

Even seemingly simple things can have a huge impact on the draft setups, caps, and trade economy.  For example:

"I dont want a 1st round pick 5th overall to not be able to come in and start"

It's a choice as to what you want this to mean, and can have far-reaching effects when you're looking to develop player structure.

You'll have to decide how stable you want your draft economy to be.  You'll have to decide what sort of basic timelines you want to have.  How much you want to have to micromanage trends.  How much you want different aspects of team management to matter.  How much parity you want.

At any rate, in your case, the bad news is all this stuff is time-consuming to set up properly.  The good news is, you'll hopefully only have to do it once.

And, like I said earlier- there are a LOT of options, such as the ones Marco brought up. 

If you want this to be a league-wide endeavour, you might want to start a new thread (or threads) about various aspects of what we are looking to do, and send out an email so that people who don't normally frequent the forums are informed about it too.

The time aspect might seem a bit daunting, but this is a hobby we all share; if it comes out right I'm sure all of us will be happy to have put enough time into it to make it work.


JH



PS Marco- it's better to wait and see where things go before putting up the specific setups for the FIFL, since the caps and entry ratings are integrated and meaningless without the existence of the other.

[jcompton-Orlando]

It sounds to me like you two are talking past each other.

Jason seems to be saying "You set up draft standards and you run the same sort of draft every year."

Paul seems to be saying "I still want to have different kinds of players coming through every year."

[Paul-London]

Well its a tough one, i always think the first round specially the first 8 or so picks should be instant starters.  I did feel that the year before the draft crop wasnt great and that why i upped the stats.  Granted i did it too much so that didnt help matters.

My problem is i love the FWL to be realistic, i hate it when suddenly Mario starts a 3rd string FA QB as that just wouldnt happen in the NFL.  If the Patriots were 11-1 they wouldnt sign QB Pete Smith and then start him a week later.

But i have to realise not all these things are possible as its just some old software, we cant always have some good first round picks as it will disrupt how things work in the future.  I like some average then good then weak then average drafts as its real, not every 1st round in every draft has superstars, some have some pretty poor players so teams tend to trade out etc.  A few years back there were all those hyped up QB's, eg Culpepper/McCown/Smith/Couch etc, at the time it looked a great time to have a high pick.

Like i say its very difficult.  One factor is the skill levels in the FWL are so high to make a great rookie his skills have to be so high as well.  In Davids league 1st round picks are generally on the whole good.  However last year in the FWL draft before i made the changes i hardly saw any decent players.  Dont get me wrong i was totally grateful for David and he did the sensible thing rather than me acting like the hare out of the tortoise and the hare and increasing stats too much

But like i say i understand that you cant go over the top like i did otherwise it ruins things, tough one it is the whole ratings thing/rookies

[JH-SanAntonio]

Variety in drafts is fine.  Large gaps in quality isn't.

"Realism" includes the ability of NFL teams to see players develop as they advance through college.  When a team trades a player for 2 picks in next year's draft, they have pretty good idea of what players will be in that draft- especially 1st-round quality players.

There are 2 possibly viable ways to simulate the ability to predict things here in the FWL:

1) Post draft lists 2 seasons in advance

2) Close up the gap between draft A and B

Eventually you'll need to choose one or the other- but both work.



The inherent flaw in the "i always think the first round specially the first 8 or so picks should be instant starters" idea is simple enough.
    Lets assume that you make a guy "average" as a rookie- let's say a QB.  So, the guy is good enough to finish the year as,say, #10 on the list.  Great!   Problem is, he bumps QB #18 off the starter list, and increases the average quality of players at the postion.  That means that next year, when you put in an "average" rookie QB, that player will have to be better than the player you put in the year before.   
     As a result, your players will always tend to escalate.  Not only is that obviously bad for the league, but highly unrealistic- you'd have to have rookie stats like this:

Year  Name                   SP  AC  AG  ST HA  EN  IN  DI
1983 John Elway            75  72   75  88  76  81  90  78
1987 Vinny Testaverde   78  79   81  91  84  84  92  80
1990 Jeff George            81  83   85  93  86  87  94  83
2006 Matt Leinart           88   87  92  98  94   97 99  95




Anyway, I'll probably start putting up individual threads in the next couple of days regarding different aspects of the draft and structure.  That way we can keep each discussion on topic a bit more easily and, hopefully, make things a bit more understandable without a degree from Sports Dork University.


JH



PS  Jason:  We aren't talking past each other.  Paul and I have been having similar conversations over the past 5 seasons or so- it's just now we're doing it on the forum

[Paul-London]

Thanks for the input Jason, much appreciated.  Like you say the ideal thing is to put a draft list up well before the actual draft, i do feel bad for teams this year who may be hoping for good players but then we dont have a huge crop of great players.  Obviously in the FWL this is one of those things, in real life you would know the top players for sure so sadly we cant make everything so realistic.

Good point about the QB's, i suppose in the last 20 years theres hardly been any QB's who have come straight in and done well, Marino and Manning probably the exceptions.  Other big name QB's have taken a few seasons.  In the FWL we've had a few too many, Aaron Rodgers had a rocky start but came on well in the end.  Its interesting to see so many veteran QB's still starting in the FWL though, gonna be interesting to see the next stage of starters in the FWL.

Are there any volunteers to create a draft list ??!?

[JH-SanAntonio]

You should probably hold off on asking for volunteers until everything gets sorted out- it might end up being an involved process this season.