Date of Release of Patch 5.0C??? [Archive] - Front Office Football Central

Icy

02-03-2004, 09:34 AM

Do you mean in multiplayer, where the lowest numbered team gets an FA, regardless of bid? That happens only outside of the normal FA period.

If that's not what you were talking about, what is the problem with the way the computer handles the FA period?

It happends too in single player, you will notice that during the season, if you offer any contract to an FA, he signs with you inmediately without any negotiation or allowing the AI teams ofer him anything, the same when an AI team signs FA, you can't offer them anything as they sign inmediately.
All the FA negotiations should be like in the offseason FA periods, where you and the computer make offers and the FAs answer after a week, so everybody has a chance to sign them. This is the same bug that happends on multiplayer but it's mroe easy to notice there for the nature of the MP game.

albionmoonlight

02-03-2004, 10:17 AM

It happends too in single player, you will notice that during the season, if you offer any contract to an FA, he signs with you inmediately without any negotiation or allowing the AI teams ofer him anything, the same when an AI team signs FA, you can't offer them anything as they sign inmediately.
All the FA negotiations should be like in the offseason FA periods, where you and the computer make offers and the FAs answer after a week, so everybody has a chance to sign them. This is the same bug that happends on multiplayer but it's mroe easy to notice there for the nature of the MP game.
Actually, in terms of SP, I prefer the way it works, even if it is a bit more unrealistic. If I am signing a guy in mid season, it is because I am filling out my roster or because I need a guy to replace an injured guy. I don't want the hassel of a FA period to get a guy rated 13/17 just so I can have a legal roster. I don't think that mid season pickups are worth the trouble for a full FA period each week.

Ben E Lou

02-03-2004, 10:27 AM

Actually, in terms of SP, I prefer the way it works, even if it is a bit more unrealistic. If I am signing a guy in mid season, it is because I am filling out my roster or because I need a guy to replace an injured guy. I don't want the hassel of a FA period to get a guy rated 13/17 just so I can have a legal roster. I don't think that mid season pickups are worth the trouble for a full FA period each week.Amen. The AI does a good enough job picking up those with any real value whatsoever in the eight-step process that these guys are pure roster filler almost exclusively for me. I think I've had two mid-season pickups start more than two games in a season in 50+ seasons of FOF2K4.

mmarra82

02-03-2004, 02:52 PM

I was talking about how the computer lets so many of its players to end up in the FA Pool, and are left with 15 guys on their roster from the previous season. A bit unrealistic.

Ben E Lou

02-03-2004, 03:07 PM

I was talking about how the computer lets so many of its players to end up in the FA Pool, and are left with 15 guys on their roster from the previous season. A bit unrealistic.I may be wrong here, but I am under the impression that it has to be this way for the AI teams to build better rosters. I'd *much* rather see large player turnover and a more competitive AI, if the alternative is less player turnover and a weaker AI. From my experience, the marquee players on AI teams typically don't switch teams more than once or twice in a career, and that is good enough, immersion-wise, for me.

Vinatieri for Prez

02-03-2004, 10:31 PM

hukarez

02-03-2004, 11:06 PM

It's also oh so sweet when you sign a division rival's marquee linebacker to your squad, after having spent years watching him excel against your team :D

Flasch186

02-04-2004, 12:11 AM

in MP, however, a change is absolutely necessary.

amdaily

02-04-2004, 07:09 AM

I agree SD. Interesting stat: There were 30 (of the 53) players on the Pats SB 38 winning team that were NOT on the SB 36 winner. That's pretty high turnover for a team that won the SB two years ago. Except for Ty law, the entire secondary was new (including nickel and dime backs). FOF doesn't seem that far off.
So were taking about losing about 19 players per year in the NFL. FOF teams lose 35-40 players a year. Not even close to realistic numbers.

Subby

02-04-2004, 07:17 AM

So were taking about losing about 19 players per year in the NFL. FOF teams lose 35-40 players a year. Not even close to realistic numbers. Yes, but I see it as a compromise between having a competitive single player AI and having realistic roster turnover.

Of course, single player anything sucks Rosie O'Donnell's left nut...

Ben E Lou

02-04-2004, 07:17 AM

So were taking about losing about 19 players per year in the NFL. FOF teams lose 35-40 players a year. Not even close to realistic numbers.Maybe not, but keep in mind this is a Super Bowl team...we don't have numbers for the dregs. Further, if my presumption above is correct, sacrificing a little realism for challenge makes all the sense in the world. It is a minor thing if a bunch of medicore players change teams every year. The stars don't. For example, looking at the Top 10 rushers of all time in my universe, and top 10 passers of all time as well, only one out of those 20 guys played for more than 3 teams, and that was because he bounced to three different teams in the final three years of his fading career. Of all these 10-18 year careers, no others played for more than 3 teams, and several stayed with their original teams for the duration. If you recall my story about the greatest back in my league's history, Bob Stephenson, he played for six years in Washington, but the AI 'Skins let him become a free agent for very obvious reasons, and the moves that the team made were perfectly sensible, in fact very smart for computer AI.

Ben E Lou

02-04-2004, 07:23 AM

Dola--

Here (http://dynamic2.gamespy.com/%7Efof/forums/showthread.php?t=18372&highlight=stephenson+skydog) is the thread where Stephenson was discussed.

amdaily

02-04-2004, 07:37 AM

I don't think foregoing the swapping of 15 scrubs is going to result in a excessively worse AI, but it would help out tremendously in terms of league countinunity. And it's not just scrubs swapping teams every year... as I've had a drastically different result than you.

In my 35th year, 6 of the top 10 passers have players for 3 or more teams including one guy who played with 6 different teams; 4 of them over 3 years each. This says to me he signs a contract with team A, it expires; he signs a contract eith team B, it expires, and so on. The AI most likely never even makes an attempt to keep him. This particular player is one of the top 5 QB's ever in my league. It would make no sense for multiple teams to keep letting him go.

-edit- I have the same issue on defense. 6 of my all time top 10 tackle leaders have players for 3 or more teams, and this is excluding single year stints with yet another team.

There is no reason the AI needs to go into the offseason with $50 million under the cap. 75% of that should be spent on extendig contracts and resigning their own players before the FA period even begins.

QuikSand

02-04-2004, 08:08 AM

While I wouldn't agree that the sky is falling, I do think that the pendulum has swung a bit too far in this direction (teams letting go too many players each year).

A major criticism of the FOF games up to this point was that the AI teams would sit idly by, with their starting LT rated as a 20/20 guy, while there was a standout LT in the free agency pool, asking for money that wasn't much more than they were paying their current guy. There seemed to be a bias toward standing pat -- as long as there was a player on the roster, the team considered that position "filled" and didn't do much to try to improve upon their current players. Tat caused all sorts of havoc in the league - including (most notably) far too many quality players going unpursued in the FA process.

I think in an effort to fix this, we have seen things swing in the other direction. AI teams sign a lot of players to short term deals, and don't make any particular effort to re-sign their own players. If their guy is the best available, he might be pursued for a new deal, but there is no inherent bias toward keeping the familiar guy around. It is, in a way, the opposite of where we were before.

I recognize that striking a balance here is a difficult thing to do -- it's a very complex system we're talking about. But in my judgment, a modest tweak back in the direction of some loyalty (especially to players of some real consequence - successful starters) probably would make some sense.

Ben E Lou

02-04-2004, 10:02 AM

I recognize that striking a balance here is a difficult thing to do -- it's a very complex system we're talking about. But in my judgment, a modest tweak back in the direction of some loyalty (especially to players of some real consequence - successful starters) probably would make some sense.The highlighted word would be the key here.

cthomer5000

02-04-2004, 10:28 AM

can we maybe get primelord to get in communication with Jim to see what his timeline on the patch is?

JonInMiddleGA

02-04-2004, 10:50 AM

The highlighted word would be the key here.
Amen to "modest" being the key word.

Somewhere around here, I believe there's one or more threads with comments about all the teams that were losing draft picks because they were over the cap.

So, either teams aren't keeping enough players around _or_ they're spending too much money re-signing players, but I don't think it can be both at the same time.

amdaily

02-04-2004, 12:55 PM

With a consensus on that reached, I'd like to reopen the injury proneness debate.

As it stands now, we either have zero-prone or super-prone players. Super proneness seems to be attained once a player is listed as "out" for more than 10?, 12? weeks with a serious injury. After this initial injury the player will never again play a complete season (even if he suffers a zero reduction in ratings) usually participating in 1, 6, maybe 11 games per year due to a variety of injuries odd ball injuries.

In the NFL a player with an ACL tear has 3 career trajectories:

1 – Will continue on as if it never happened
2 – Will end his career
3 – Will reinjure the ACL at a later point and face options 1 or 2

But in FOF a player has 1 career trajectories after an ACL tear (even if be as short a recovery time as 18 weeks, as I recently experienced).

1 – Player will return after recovery, become super-prone, and never again complete a season due to ankle ligaments problems, strained back, hamstring issues, etc.

Just because a player injuries his ACL in a freak basketball accident, or walking back to the sideline after a penalty, does not mean he is injury prone. Why does FOF, to an absurd degree at that, treat it as such?

yabanci

02-04-2004, 02:46 PM

I'm in the camp that thinks there needs to be a modest swing back in the direction of some loyalty.

The other thing I think is really important for the patch is to provide a new "multiplayer log," meaning a log that gives you the text that would have shown up on the scoreboard. As it stands now, after you send in your multiplayer stage file all you get back is the box score and game log. The game log is useful for seeing what happened in the game, but it leaves out key information like offensive and defensive formations, blitzers, yards after catch, etc. This doesn't seem like it would be a difficult option to add (and if you don't want it, you should be able to turn the option off just like you can with game logs and box scores). It would add to the size of the file you get back from the commissioner each stage, but I think it would be worth it to most people. Maybe I'm wrong though, so any comments are appeciated.

Subby

02-04-2004, 03:04 PM

Great posts by yabanci and amdaily...

PunkyQB

02-04-2004, 03:24 PM

I may be the board's resident "roster continuity for AI teams" extremist ... but I just want to chirp in some support for the "modest" or whatnot tweak in the retention rate of key players on teams.

I know I am not the norm ... but my experience has been such that I have lost all interest in playing the game because of this issue. Over decades of history, even hall-of-famers average more than 4 or 5 teams each! I can find plenty of challenging games ... but the challenge is unimportant to me if I don't care about the universe I am competing in. I hate having all of my budding rivalries abort and disappear when, routinely, the other teams in my division/conference/league just let go of nearly all the players who have a history against me and shuffle in a new team.

I wonder if my obsession with this issue is related to the fact that I follow my team's play by play (a different discussion, I know).

Anyhow, FOF plays like an expedited version of fantasy football to me right now ... my "immersion" satisfaction with the game is nearly zero because the teams have no identity whatsoever. I say this as a loyal supporter of the game since its birth.

So ... whatever consensus camp comes anywhere close to my feelings on this ... no matter how modest a tweak people want to tolerate ... I'll take whatever I can get!!!

Ben E Lou

02-04-2004, 03:27 PM

This is a really interesting direction to me. I couldn't possibly care less about what players are on the AI teams. It is hard enough (and quite enough immersion) to keep details about 53 different players in my head, let alone all the players on other teams as well. There's no way I could memorize the rosters of the entire league--or even the teams in my division I don't think.

mmarra82

02-04-2004, 04:07 PM

I think the reply about tweaking the loyalty is probably more of what I have in mind. To me looking at the NFL Teams roster compared to the Front Office Rosters the turnover is a tad too high. Computer teams are not making an effort to keep their superstars. For example, Peyton Manning becomes a FA after the 2003 season. He is and All-Pro the second highest rated QB, according to my scout, and the Colts make no effort or even an offer to sign their franchise player. That is what I find unrealistic.

Taur

02-04-2004, 11:19 PM

Any idea when patch 5.0C will be out?
Did I miss the answer to the original question?

Galaril

02-05-2004, 03:58 AM

Are we even sure a 5.0c version is in the hopper currently.Any word I think we are just looking at a rough estimate nothing too exact but some idea.Primelord or Jim? :)

Tryfan

02-05-2004, 04:43 AM

Perhaps the computer could be convinced to handle things the way I do out of expediency - it could renegotiate with any player entering their final year under contract. Those players tend to be rather easy to resign because they almost always take a paycut in the next year (though not always) allowing the team to get control of its cap.

Couple that with the computer having a low line for whether or not a player is worth keeping at all.

This would keep the numbers of players actually going into free agency fairly lower, IMO, and would maintain continuity much better.

FargoFreez aka fof playa

02-05-2004, 12:19 PM

Are we even sure a 5.0c version is in the hopper currently.Any word I think we are just looking at a rough estimate nothing too exact but some idea.Primelord or Jim? :)This was posted by Jim on January 20th. I believe it's all we've got.

I am intending to get to work on 5.0c very soon now. Should I even bother with html given Fido's utility? He's got some features in there that are pretty nice - much more time consuming than I was anticipating.

amdaily

02-06-2004, 06:31 PM

One more note on roster turnover. I've never seen a league in such distress going into the FA period:

http://s92878788.onlinehome.us/problem.gif

Dutch

02-06-2004, 07:52 PM

Wow, a picture is worth a thousand words...

Karim

02-06-2004, 09:11 PM

That says a lot right there.

yabanci

02-07-2004, 01:25 AM

what year is that screenshot from?

amdaily

02-07-2004, 09:22 AM

That screens are from 2041.

The more I look at some of these offseason roster moves, the more I'm convinced the AI has no idea what it's doing :). Consider the following 3 examples from the Arizona Cardinals in the 2043 offseason.

Richie Tamburrino is the backup FL under contract for 2 more years at $840K. The current starter is a unrestricted FA. Upon entering the FA period, the Card’s cut Tamburrino meaning now FL’s are left on the roster.

Leo Baker is a backup RT scheduled to make $1.4 million next season with 33/57 scout ratings. The current starter scheduled to be an unrestricted FA. Upon entering FA, the Card's cut Baker, leaving no RT’s on their roster.

The Card's 3rd year SE Donovan Whitmire is a backup entering the last year of his contract making $680K. Upon hitting "begin free agency," the front office cuts him. The problem is, the previous season's starting SE is a free agent as well. As of this moment they have no split ends on the roster.

The point is, not only are these specific players talented guys, two are making the minimum salary. The AI’s willingness to cut players making minimum salary and go into the offseason with no players at a number of positions needs to be addressed. Not only is that a concern itself, but does the AI even care about cohesion?

Further, upon looking though the transaction logs, the AI rarely extends it’s own players contracts. What happens is, league wide approximately 72 players will sign renegotiated contracts (only a very low average of 2.4 per team). Upon reloading the game and finding these players, you’ll find that the larger percentage of the renegotiated contracts are with players who have 2 + years left on their contract. I’ve seen numerous renegotiations with players who have a full 5 years left on their contract! Renegotiating those that have 5 years left on them is doomed to put the AI in cap hell. So why do they do it? And why can the AI seem to only sign contracts that look like this (which I suspect is the reason for the constant restructuring even when 4 or 5 years are left on a contract):

Year 1: $380K
Year 2: $3.1 million
Year 3: $3.6 million
Year 4: $4.1 million

In year two, the AI will renegotiate this contract again. Then the next season do it again, and again and again offering basically the same contract each season. The problem is, the bonus money catches up with them after 2-3 years thus compounding the problem. Solution: Offering balanced contracts with few exceptions (ie. cap out offer).

In relation to this, why are rookies still on their original contracts signing renegotiated contracts mid way though? Again, this has no effect but to leave a team in cap trouble.

Another problem – some teams will go into training camp with as low as 27 players on under contract. When checking their roster after training camp, you’ll find a large percentage of those spots filled with undrafted FA’s. In my most recent league, Houston went into the season with 20 1st year players, 13 of them on 1 year contracts.

And as for roster turnover directly, not only are expiring contracts not resigned (thus meaning a team loses 15-20 players each season), but upon hitting “begin free agency” every teams releases 15-20 players additional players, thus leaving them with a roster rating of 0 and 13-16 players on the roster going into FA.

So, the renegotiation timetable needs fixed, the dollar distribution in contracts offered needs fixed, the threshold of releasing players needs fixed, and deciphering a teams strengths/weaknesses (ie – releasing all players at a certain position) needs fixed.

[And Skydog, why haven’t you participated in this thread in 3 days? Your trying to find a way to tie all of these issues into injuries aren’t you :)]

Ben E Lou

02-07-2004, 01:09 PM

[And Skydog, why haven’t you participated in this thread in 3 days? Your trying to find a way to tie all of these issues into injuries aren’t you :)]This is a really interesting direction to me. I couldn't possibly care less about what players are on the AI teams. It is hard enough (and quite enough immersion) to keep details about 53 different players in my head, let alone all the players on other teams as well. There's no way I could memorize the rosters of the entire league--or even the teams in my division I don't think.What part of what I said was confusing??? I only care how AI teams roster build insomuch as that I can't get any FA I want, and they field competitive teams. This version does a better job than any I've ever seen in those two areas, so I don't care about the details. The only reason I checked in the case of Bob Stephenson was that someone asked, and it looked good--very good in fact--to me.

So, in short, I haven't participated lately because this has become meaningless conversation.

yabanci

02-07-2004, 03:10 PM

good post, amdaily. While it might be a meaningless convesation to Skydog, it's definitely not to many others who do care about the details. What you are descrbing goes to the heart of the problem and is the reason I no longer play FOF solo. The roster management of the AI teams (i.e., the stripping the team down to nothing, signing the players other teams have stripped away to one-year contracts, then repeating this process over and over) just becomes absurd. The AI teams are too aggressive in stripping away their teams for no apparent reason. It just needs to be tweaked, as many have pointed out repeatedly for the last couple of months. We'll see what happens.

Anthony

02-07-2004, 03:35 PM

i rather the game be harder with the AI teams building superior rosters, if that means high turnover so be it but I *don't* want to be able to win easily every year. I've only won 2 Championships in about 15 years of play, but pretty much have always had a winning record. i like the game being harder.

then again, perhaps if the AI teams kept some of their nucleus intact the cohesion rating might go up and they'll get as much success through that as opposed to cutting aguy and signing a better FA.

like SkyDog - i care only about my team. i'd much rather Jim code some stuff pertaining to more interactive emails (should I suspend PLayer X for a team infraction, should I show support or offer no comment if i get an email from a newspaper about an underperforming guy on my team, etc.). these mean more to me than if the AI teams decide to cut the Billy Scrubs of their team.

amdaily

02-07-2004, 03:58 PM

Ben E Lou

02-07-2004, 04:39 PM

I just loaded up FOF 2001 and was running similar tests to those I did earlier.

It seems the engine in FOF5 was drastically tweaked, as the least amount of players that a team went into FA with was 19. In FOF5 19 is considered "alot" of players to have under contract at that point in the season.

Also, there is a huge difference in the number of players signing renegotiated contracts during the offseason. Whereas FOF5 has a league wide total of 70 odd some, FOF2001 has 180 (2.4 per team vs. 5.8 per team).

There is also a huge difference in the number of players the AI cuts going to the FA period. The average FOF5 team releases an average of approximently 9.9 players while an FOF2001 team releases 5.5.

And of course we're all aware that the FOF2001 AI had the same backloading tendancies that plauge FOF5, so nothing changed in that respect.

But those first two changes alone have severely disrupted league continunity. One aspect that hurt FOF2001, however, was a lack of a second stage of FA. Due to that I remember star players never getting picked up and retiring in their 5/6 year in the league. With that issue resolved, if FOF2005 could revert to the level of player renegotiations and releases that FOF2001 had, we'd be much closer to perfection then we are today.Now this is meaningful conversation again--and hits to the crux of my concern about moving it back in the other direction. FOF2K1, even with very stringent house rules preventing the exploiting of FA's who were left after the 20-stage period and draft, was *FAR* too easy. Even playing under the stringent house rules of the WLC, I won something ridiculous like seven straight Super Bowls. With modest house rules, 10-15 in a row was common. By contrast, I've yet to threepeat in 2K4, and I strongly believe a primary reason is that AI teams build better rosters.

amdaily

02-07-2004, 05:16 PM

Now this is meaningful conversation again--and hits to the crux of my concern about moving it back in the other direction. FOF2K1, even with very stringent house rules preventing the exploiting of FA's who were left after the 20-stage period and draft, was *FAR* too easy. Even playing under the stringent house rules of the WLC, I won something ridiculous like seven straight Super Bowls. With modest house rules, 10-15 in a row was common. By contrast, I've yet to threepeat in 2K4, and I strongly believe a primary reason is that AI teams build better rosters.
Now that does not make any sense. FOF2001 had less free agents available, so therefore, if the 8 step period and other upgrades in the FA process were not implemented, it would be easier to exploit FOF5. But it is not directly because of those upgrades - not because of roster management. There are far more talented and more numerous FA's available in FOF5.

The AI teams would be even tougher to beat then they already are if they resigned their own players and plugged holes through FA, rather than completely rebuild every single season.

Ben E Lou

02-07-2004, 07:06 PM

Now that does not make any sense. FOF2001 had less free agents available, so therefore, if the 8 step period and other upgrades in the FA process were not implemented, it would be easier to exploit FOF5. But it is not directly because of those upgrades - not because of roster management. There are far more talented and more numerous FA's available in FOF5.

The AI teams would be even tougher to beat then they already are if they resigned their own players and plugged holes through FA, rather than completely rebuild every single season.Call it what you want, but I'm results-oriented, and all I know is that I routinely would win the Solecismic Bowl for 5+ years consecutively with the old version, but I have yet to win three straight a single time in this version.

amdaily

02-07-2004, 07:09 PM

Call it what you want, but I'm results-oriented, and all I know is that I routinely would win the Solecismic Bowl for 5+ years consecutively with the old version, but I have yet to win three straight a single time in this version.
And if Jim could find a way to marry the roster AI from FOF2001 with the FA system of FOF5, alot of us would have yet to win two consecutive Solecismic Bowl's.

Ben E Lou

02-07-2004, 07:25 PM

And if Jim could find a way to marry the roster AI from FOF2001 with the FA system of FOF5, alot of us would have yet to win two consecutive Solecismic Bowl's.I seriously doubt that is the case (grammar issues notwithstanding). The WLC success that I mentioned (that was shared by others by the way, I didn't even win) was achieved with house rules that completely prohibited the exploitation of trade and FA issues that existed at that point. Looking back at a few of the updates, several other guys were winning 5+ bowls in a row.

In case you don't remember, here are a few of the house rules for the WLC:

1. franchise tag allowed only once per player in his career
2. no initiating trades
3. no more than one FA signing during the first 10 rounds, then 3 during the last 10 rounds
4. no renegotiating contracts
5. no backloading contracts
6. no signing players after training camp

Point being, the game has gotten a lot harder, and from what I can tell, a large piece of that is because of the new method of AI roster building.

cthomer5000

02-07-2004, 07:35 PM

Call it what you want, but I'm results-oriented, and all I know is that I routinely would win the Solecismic Bowl for 5+ years consecutively with the old version, but I have yet to win three straight a single time in this version.
I've been in the conferecne title game the last 13 years, and have won about 9 bowls in that time wwith 3 different teams. Both those numbers might be off 1 or 2, but you get the point.

I'm not finding this version of the game any more difficult than prior versions.

Ben E Lou

02-07-2004, 07:39 PM

I've been in the conferecne title game the last 13 years, and have won about 9 bowls in that time wwith 3 different teams. Both those numbers might be off 1 or 2, but you get the point.

I'm not finding this version of the game any more difficult than prior versions.Are ou playing fully "all-out", or somewhat limited? There are some AI exploits that have been around since FOF1 that I simply won't do that are natural handicaps.

cthomer5000

02-07-2004, 07:46 PM

Are ou playing fully "all-out", or somewhat limited? There are some AI exploits that have been around since FOF1 that I simply won't do that are natural handicaps.I guess all-out, I havent' really be doing anything out of the ordinary though. I find it very easy to build a deep roster, and my AI set gameplan is apparently enough to steamroll the competiton.

amdaily

02-07-2004, 07:49 PM

The success people experienced in FOF2001 was not directly related to AI roster management. This game has advanced light years in a multitude of areas that add to the difficulty, although, as you mention, most of the “cheats” are still there if players wish to exploit them.

The draft has made it incredibly difficult to build a team through it, numerous new ratings and statistical categories have been added, player personality traits have been incorporated, more coaches each responsible for different aspects of development, etc, etc. I see the more intricate design of the game engine the primary reason for the increased difficulty of FOF5, not AI roster management. The computer is able to decipher all these new varibles much better than you or I (who is some case still don't know exactly what many of them do) thus making it more competitive.

If anything, the new roster management system handicaps the AI teams by making it impossible for them to maintain cohesion and not being able to keep together players with affinities together.

Ben E Lou

02-07-2004, 07:51 PM

my AI set gameplan is apparently enough to steamroll the competiton.Oh wait...aren't you the guy that had that stud running back???

Ben E Lou

02-07-2004, 07:55 PM

The success people experienced in FOF2001 was not directly related to AI roster management.I couldn't disagree more. I almost *never* had a roster rating <100 in that game, while I now have to work very hard to be among the elite rosters in the league.

As far as drafting, several people have whined about it being too random/difficult in this version, but nothing could be further from the truth. With just a tad bit of research combined with logical intuition, and sound drafting strategies become rather obvious.

cthomer5000

02-07-2004, 07:58 PM

Oh wait...aren't you the guy that had that stud running back???
about 2/3 of that success streak i'm talking about came after him. I also had plenty of success before him.

I'm absolutely dominating the league, no matter how you look at it.

Ben E Lou

02-07-2004, 08:01 PM

about 2/3 of that success streak i'm talking about came after him. I also had plenty of success before him.Oh, I wasn't implying that all your success came because of him at all. I was saw the comment about AI-set game plans, and that triggered the context, because I remember reading in the thread that his success came with AI game plans.

amdaily

02-07-2004, 08:23 PM

And I almost never have a roster rating < 100 in this game. The only difference is, all the other teams have ratings of 0. At least in FOF2001 the AI attempted to field a team (and don't get me wrong, I'm not saying the FOF2001 roster AI is the end all, but it did seem more logical than that in FOF%). It, of course, couldn’t win because too many other variables worked in favor of the human.

FOF5 seems to reverse the problem by giving the computer more competitiveness, but completely takes away it’s sense of roster management. The result is we have a game that is nearly as easy to dominate.

This game in its current state more closely resembles fantasy league software where all teams redraft their teams each year. You may be biased towards good results, but what good are they (and even this point is arguable as I still don't like the large number of QB's finishing a season with 4,000 yards and a 100+ rating) if realism is sacrificed to get there?

And I had one more good point to make, but its slipped my mind at the moment :).

Cringer

02-07-2004, 11:24 PM

this is all good....but uh, whens the release?

Dutch

02-08-2004, 09:01 AM

Past Jim saying he's getting ready to start working on it, it's unknown.

Vegas Vic

02-08-2004, 09:18 AM

I'm not saying the FOF2001 roster AI is the end all, but it did seem more logical than that in FOF%).

The FOF2001 roster AI was poor in many areas. It would routinely sign a backup QB to a 5 yr/60 million contract, or draft a RB in the 1st round when it already had the league MVP at RB. As SkyDog pointed out, even with the most stringent of house rules, the human player could dominate the game.

amdaily

02-08-2004, 09:45 AM

The FOF2001 roster AI was poor in many areas. It would routinely sign a backup QB to a 5 yr/60 million contract, or draft a RB in the 1st round when it already had the league MVP at RB. As SkyDog pointed out, even with the most stringent of house rules, the human player could dominate the game.
Yes, those would be two of its many flaws. But as I mentioned earlier, and in terms of the quantity of players on the rosters, FOF2001 renegotiated the contracts of 5.8 players per teams while FOF5 renegotiates only 2.4 per team each year. Also, the average FOF5 team releases an average of approximently 9.9 players while an FOF2001 team releases 5.5 each offseason.

Whatever change was made between the two versions in terms of extendng contracts, and the threshold for releasing players was a change for the worse.

Ben E Lou

02-08-2004, 09:48 AM

Whatever change was made between the two versions in terms of extendng contracts, and the threshold for releasing players was a change for the worse.In your opinion.

Ben E Lou

02-08-2004, 09:48 AM

(and even this point is arguable as I still don't like the large number of QB's finishing a season with 4,000 yards and a 100+ rating)How many do you think there should be?

T-Storm

02-08-2004, 10:32 AM

How many do you think there should be?
Eventhough I wasn't asked, I think on average there should 5 QBs with a 90+ Rating and 3 QBs with 4000yards+.
However, the best way to "cure" this would be a draft-file-creator.
I remember somebody thinking about developing such a thing? A draft file creator where you could set the average ratings for any position up.
Like havin most Qbs excell at very long passing, but with no ability to avoid interceptions.

amdaily

02-08-2004, 11:21 AM

It’s not just me, if you stick up a poll we'll find roster reform is an opinion shared by a large majority of members who seek realism from a product which, up to this point, has always provided it.

As for QB stat distribution, we need look no further than the NFL for a model: http://sports.yahoo.com/nfl/stats/bycategory?cat=Passing&sort=49&qualified=1&conference=NFL&year=season_2003

2003:

2 4000 yards
6 90+
1 100+
5 20+ INT
1 570+ attempts

2002:

4 4000
4 90+
1 100+
2 20+ INT
4 570+ attempts

2001:

2 4000+
5 90+
1 100+
6 20+ IN
1 570+ attempts

2000:

3 4000+
7 90+
2 100+
2 20+ IN
3 570+ attempts

NFL AVERGAE:

2.75 4000 yard passers
5.5 90+ rated passers
1 100+ rated passer
3.75 20+ INT
2.25 570+ attempts

Yet if we look at FOF5, the average is closer to this( stats from a variety of seasons including 2003, 2004, 2045, 2046)

6 4000+
6 90+
1 100+
11 20+ INT
X 570+ attempts

2 4000+
5 90+
0 100+
9 20+ IN
7 570+ attempts

5 4000+
9 90+
4 100+
6 20+ IN
4 570+ attempts

6 4000+
8 90+
4 100+
4 20+ IN
9 570+ attempts

FOF5 AVERAGE

4.75 4000 yard passers
6 90+ rated passers
2.25 100+ rated passer
7.5 20+ INT
6.67 570+ attempts

I will concede the 90+ QB's ratings are averaging out since the 5.0b patch. However, there should be 1 less QB reaching the 100 rating mark each season.

To get to the root of the problem for the other stats (yards, INT's (and if I tested it, TD's), and attempts) we need to look at injuries. There are about 5 more, give or take depending on the season, FOF QB's than real life QB's who start every game (with injuries at 200). This isn’t to say that we need a rating of 300 (there’d be so many players tearing their ACL in one season, then never being able to complete a full season again, that league would have to shut down due to lack of players). But rather this gets right to the heart of the super-prone or zero-prone argument.

In the NFL approximately 14.5 QB's play between 10-15 games each season. FOF has an average of closer to 10 (with injuries at 200). Since a plain old "injury prone" status does not exist in FOF5, it's boom or bust for QB's - they play all games and amass superhuman stats (see yards, attempts, and INT’s), or they play no games because they come down with a season ending injury every single season like clockwork thanks to their super-prone rating. The injury system (in designating a player super-prone after what should be an isolated season ending ACL tear/rotator cuff tar/etc or zero-prone) is the single cause of the skewed QB’s stats and with the numbers posted above I don’t know how you could possibly disagree.

Ben E Lou

02-08-2004, 11:33 AM

Eventhough I wasn't asked, I think on average there should 5 QBs with a 90+ Rating and 3 QBs with 4000yards+.
However, the best way to "cure" this would be a draft-file-creator.
I remember somebody thinking about developing such a thing? A draft file creator where you could set the average ratings for any position up.
Like havin most Qbs excell at very long passing, but with no ability to avoid interceptions.Nope. The best way to "cure" this is to play with realistic injuries. :p

I ran some long-term tests, but have since deleted that universe. However, here are the results from a new career I recently have started:

2003
0 QB's >4000 yards
0 QB's >100.0 rating
4 QB's >90.0 Rating

2004
4 QB's > 4000 yards
0 QB's > 100.0 rating
3 QB's > 90.0 rating

2005
2 QB's > 4000 yards
1 QB > 100.0 rating
5 QB's > 90.0 rating

AVERAGES SO FAR
2 QB's > 4000 yards
.3 QB's > 100.0 rating
4 QB's > 90.0 rating

Those numbers are quite similar to what you seem to be after. The problem is that most people don't want to play with a realistic number of injuries. If you are unwillling for the number of injuries to be realistic, then why should you expect your stats to be realistic?

amdaily

02-08-2004, 11:41 AM

Dutch

02-08-2004, 11:43 AM

I agree with SD on the idea of higher injury rates would produce more realistic numbers (based on his old data).

But before I come to the conclusion that I want more realistic injury rates, I want roster management to be shored up.

Practice Squads could help.

But better than that, having secondary positions for players and still being able to switch players to any position you want (granted with performance penalties) but keeping the master and secondary position skills static would be a must.

I could have a RB that is rated at 30/50 as a RB (his primary skill), 15/35 as a FB (his secondary skill). Perhaps he's a 13/32 at TE, and a 2/5 at QB.

Guys like RB Keith Byars of yesteryear found out they were good at FB and TE because injuries to those positions forced him to play there during a course of a given year and actually ended his career as a TE.

We need to have the flexability to move a player out of position in FOF and then move him back without penalty. If I have 4 RB's on my active roster and one of my QB's get hurt, I should, for emergency situations be allowed to move my 4th RB into the emergency QB slot for a game or two, if I want to.

Ben E Lou

02-08-2004, 11:44 AM

LOL. Cross-post there. I'll add in the numbers amdaily looked at...

2003
>20 INT: 5
>570 ATT: 3

[u]2004[/b]
>20 INT: 5
>570 ATT: 5

[u]2005[/b]
>20 INT: 3
>570 ATT: 4

A little high on the high INT, high attempt guys, but not enough that I would have even noticed before the exact stats were mentioned.

Ben E Lou

02-08-2004, 11:49 AM

I have a gut feeling that if one were to add up the amount of "out" weeks at injury level 200 experienced league wide, they would match up to NFL standards or thereabout.

However, the distribution of that "out" time is the problem here.

FOF5 has the same players, every year, get injured in preseason and be out for 16 weeks as opposed than have 4 players miss 4 weeks. With only 1 player hogging up the injury time, so to speak, it leaves the other three to pursue the abnormal stats.I'd like to see the distribution better, but the total number of injuries is pretty solid. I did five runs with injuries at 200, using the default training camp setting, with a coach who is "Good" at avoiding injuries. After 17 weeks, here were the averages from my team:

Out: 5.4
Doubtful: .6
Questionable: 1.2
Probable: 5.0

That looks about right to me.

Incidentally, I ran some tests at other training camp times as well. I'm not going to give out the details (too many IHOFers reading), but suffice it to say that training camp time settings make a difference, but I'm not going to say how much and at what point there are diminishing returns. :p

Ben E Lou

02-08-2004, 11:57 AM

But better than that, having secondary positions for players and still being able to switch players to any position you want (granted with performance penalties) but keeping the master and secondary position skills static would be a must.Agreed. It is annoying to have to do a position switch or sign a new player because a player is Out for 2 weeks.

yabanci

02-08-2004, 01:31 PM

amdaily, you are doing excellent work. Keep it up.

FargoFreez aka fof playa

02-08-2004, 05:48 PM

SkyDog, you are doing extraordinary work. Keep it up.

Solecismic

02-10-2004, 02:42 AM

Both of you do fantastic work. I should take that into account when choosing beta testers next time around, though that's not to say my current group doesn't do fantastic work as well (all bugs that remain are the fault of the developer).

I'm interested in this discussion, and have looked at both issues quite seriously for 5.0c.

On the statistics issue, I'm happy to say that knocking down the percentage of pass plays called by the AI has fine-tuned that quite nicely. I've also made a tiny tweak up on expected yards per carry to compensate, and notched up the spread of standard deviation on the implementation of ball carrier ratings. Basically, the engine is stable and doing what I want it to do, so this type of fine-tuning carries no risk and would never be noticed by most who play the game.

The other issue is much more difficult, because small changes often have huge effects. I wasn't happy with the FOF2001 AI, or the FOF4 AI, which had some small rewrites here. So I started from scratch in a few places like this for FOF 2004. The end result was a more competitive AI, but it's acting a little less like a pro team's AI.

In general, I'm happier with a competitive AI than a completely realistic AI. But for the reasons amdaily points out, I'm not ecstatic. So, I've spent the last couple of days on this issue, and here's what I've done (in brief).

1) Made teams a little less aggressive about releasing players. They try to do a little better job assessing free agent talent available, and the release threshold is a little lower. It amounts to about 25% less releases during this stage.

2) I've largely rewritten the routine for renegotiating contracts at this stage of the season. The idea being that if a player is valuable, and he's in the last year of his contract, teams will throw out a lot of bonus money. I've notched down the renegotiating for contracts with more than a year remaining, and I worry that this carries some significant risk of caprupting teams. I also worry that teams will stick too much to players they like and free agency will become much less interesting.

Because of these risks, I'm going to be watching discussion of 5.0c in this particular area very closely, with the idea that tweaking may well be necessary. There's one little routine that has a monumental effect on player retention, and I've made broad strokes here. Fine-tuning is going be necessary unless I'm very lucky.

Peregrine

02-10-2004, 02:50 AM

Basically, the engine is stable and doing what I want it to do, so this type of fine-tuning carries no risk and would never be noticed by most who play the game.

Probably only by people who spend their time huddled over their computers in darkened rooms, obsessing over statistical minutiae. We don't have anyone like that here at FOFC I'm sure... :)

Anyway, great feedback Jim, a lot of us have been following this thread with interest. It's a great sign in my opinion that FOF is so solid at this point that these are the biggest problems the game has :)

NoMyths

02-10-2004, 02:51 AM

Jim, your comments assure me you're on top of things. Looking forward to C.

Ben E Lou

02-10-2004, 04:34 AM

Jim:

Your post absolutely thrills me in this area. A few comments:

In general, I'm happier with a competitive AI than a completely realistic AI. But for the reasons amdaily points out, I'm not ecstatic.Geez, I feel like Chief Rum right now. :P You managed to summarize my entire position on this issue in 23 words.

1) Made teams a little less aggressive about releasing players. They try to do a little better job assessing free agent talent available, and the release threshold is a little lower. It amounts to about 25% less releases during this stage.

2) I've largely rewritten the routine for renegotiating contracts at this stage of the season. The idea being that if a player is valuable, and he's in the last year of his contract, teams will throw out a lot of bonus money. I've notched down the renegotiating for contracts with more than a year remaining, and I worry that this carries some significant risk of caprupting teams. I also worry that teams will stick too much to players they like and free agency will become much less interesting.These sound like smart ways to address this issue. Thanks for letting us know specifically how you're addressing it. It will make it much easier for those of us inclined to "spend their time huddled over their computers in darkened rooms, obsessing over statistical minutiae" (very nice one Peregrine!) to look at the specific outcome of the changes.

Because of these risks, I'm going to be watching discussion of 5.0c in this particular area very closely, with the idea that tweaking may well be necessary. There's one little routine that has a monumental effect on player retention, and I've made broad strokes here. Fine-tuning is going be necessary unless I'm very lucky.This entire section is absolutely beautiful. I'm very excited that you're going to give this a tweak, but watch it carefully. That's really all any of us could ask. The primary reason that I pushed back so hard against amdaily was that at some point the next project has to become a priority, and I was afraid that you might tweak this in 5.0c too much, leaving us with a final version of the game that ended up being easier. Thanks for being willing to put this out there for constructive criticism and then finish it up with another tweak if necessary.

--Ben

Eaglesfan27

02-10-2004, 11:55 AM

Nice post Skydog. This is excellent news, the only downside - the patch sounds like it might be so compelling that I will want to start a new career and see how that universe unfolds ;)

amdaily

02-10-2004, 12:05 PM

I'll simply echo what everyone else has said. Despite my criticisms, previous version of FOF aren't even in the same league as FOF2004 in terms of functionality and competitiveness. The fine tuning, specifically in regards to renegotiations and releases, sounds like they will be a real improvement.

With the statistics problem seemingly solved as well, that leaves only the injury system as a concern for me (ex. the same players being chronically injured every year, without fail, while others play an entire career without missing a snap. This holding true for whether the injury setting be 10 or 200). Has, or will, this be addressed in 5.0c?

T-Storm

02-10-2004, 12:27 PM

1) Made teams a little less aggressive about releasing players. They try to do a little better job assessing free agent talent available, and the release threshold is a little lower. It amounts to about 25% less releases during this stage.

2) I've largely rewritten the routine for renegotiating contracts at this stage of the season. The idea being that if a player is valuable, and he's in the last year of his contract, teams will throw out a lot of bonus money. I've notched down the renegotiating for contracts with more than a year remaining, and I worry that this carries some significant risk of caprupting teams. I also worry that teams will stick too much to players they like and free agency will become much less interesting.
I guess it's impossible to make this user-controllable?
Besides that an option to play without Salary Cap andthus enabling the AI Teams to only worry about having talented players wouldn't necesarily a bad idea I guess. But that's porbably something for FoF6 as well.
Anyway, I for one don't care all that much about free agency, because it simply doesn't seem to realistic as it is, with several 80+ (superstars) rated players available. And that in itself somewhat kills it for me.

yabanci

02-10-2004, 12:27 PM

1) Made teams a little less aggressive about releasing players. They try to do a little better job assessing free agent talent available, and the release threshold is a little lower. It amounts to about 25% less releases during this stage.

2) I've largely rewritten the routine for renegotiating contracts at this stage of the season. The idea being that if a player is valuable, and he's in the last year of his contract, teams will throw out a lot of bonus money. I've notched down the renegotiating for contracts with more than a year remaining, and I worry that this carries some significant risk of caprupting teams. I also worry that teams will stick too much to players they like and free agency will become much less interesting.

These two solutions go to the heart of what some of us saw as the problem (especially #1), so I'm very happy to hear this news. And to be fair, when I say "problem" we're talking about taking a 4.5 star simulation and making it 5.0 stars.

One other thing that needs to be looked at on these lines is the franchise player choices. It seems that too many mediocre players (often even backups) are being named franchise players, then the AI team still signs a free agent at the position (for half the price) who inevitably starts over the franchise player because the franchise player has such low ratings.

PunkyQB

02-10-2004, 12:40 PM

1) Made teams a little less aggressive about releasing players. They try to do a little better job assessing free agent talent available, and the release threshold is a little lower. It amounts to about 25% less releases during this stage.

2) I've largely rewritten the routine for renegotiating contracts at this stage of the season. The idea being that if a player is valuable, and he's in the last year of his contract, teams will throw out a lot of bonus money. I've notched down the renegotiating for contracts with more than a year remaining, and I worry that this carries some significant risk of caprupting teams. I also worry that teams will stick too much to players they like and free agency will become much less interesting.

Because of these risks, I'm going to be watching discussion of 5.0c in this particular area very closely, with the idea that tweaking may well be necessary. There's one little routine that has a monumental effect on player retention, and I've made broad strokes here. Fine-tuning is going be necessary unless I'm very lucky.

It is at times like these that I remember why you rule!

Seriously, I remain ever appreciative that you are so attentive to these things and work so hard. Rock on!

Rhone Ranger

02-10-2004, 01:47 PM

1) Made teams a little less aggressive about releasing players. They try to do a little better job assessing free agent talent available, and the release threshold is a little lower. It amounts to about 25% less releases during this stage..

Thanks Jim, this will be a great help. I'm coming in a little late to this discussion but I just want to add my support for everything that amdaily suggested. I think the change you've made will increase the AI competitiveness and give the game a more realistic feel. I would actually prefer a free agency period where there is not a huge amount of talent regularly on the market ripe for the picking. This will make the game even more challenging and interesting, and will indirectly increase the importance of the draft, which I feel is currently underemphasized. (As it stands now, I always know I can pick up a 45- or 50-rated FA player at just about any position for a cheap price virtually all the time - it's a totally safe thing but it shouldn't be).

I also worry that teams will stick too much to players they like and free agency will become much less interesting.

Actually this won't make it less interesting for me at all! Just the opposite, in fact. :D If the game gets closer to the teams really having to fight and scrap over a small pool of talent in the FA period I think that would be great! Currently, the FA period is a little too cushy with too many easy pickings due to the AI being overly happy to release medium to high quality players. So your change sounds wonderful and will only increase my interest in FA.

Because of these risks, I'm going to be watching discussion of 5.0c in this particular area very closely, with the idea that tweaking may well be necessary.

May I suggest something now? :)

It could be effective to have the AI not be the same for each team. Vary it a little, so some teams are conservative and hold onto their players or don't like to pay out big contracts, while other teams throw caution to the wind and go for broke with big-time players. I suggest this because I find that whie the AI can do a decent job for each individual team, in the aggregate it can be too predictable and a human player can stay a step ahead of the pack. This phenomenon is most easily recognized in the free agency period, where one can pretty well predict when AI teams will start to make offers for a given player who's ranked at a given spot on the list of free agents at his position and plan accordingly. It would be fun and make the game more challening if the computer would behave more unpredictably just occasionally. Sometimes a little bit of randomness, though it seems illogical, can make an AI a more effective opponent.

korme

02-10-2004, 01:57 PM

Umm.. *tosses in championship list idea and scampers from discussion*
http://www.jbl.ootp-leagues.com/host/rings.jpg

Peregrine

02-10-2004, 02:33 PM

Heh I've wanted that for a long time Shorty.

yabanci

02-10-2004, 03:26 PM

One other suggestion regarding play-by-play: with the advent of FOF2004, we now have just about every recorded statistic showing up in the pbp, including hurries, pass blocks at the line, passes defensed, and sometimes even identification of which defensive player ruined a running play (e.g., x player ran over y player and ruined the play). One significant recorded statistic that is not mentioned in pbp, however, is the "caught" statistic for defensive players. So when there is a pass reception the offensive player who made the reception is identified and in the statistics that reception is chalked up as being "caught" against one of the defensive players, but the pbp doesn't identify which defensive player. Since the game is already tracking the statistic, would it be possible to have it identified in the pbp? -- for example, after the reception say "the ball was caught against x cornerback" or "x cornerback had the coverage on the play." That way when you have a 60 yard td reception you know who got burned or blew the coverage, which is not only important for understanding what happened on the play but also very useful information in setting depth charts and game planning. I know not everybody is interested in pbp and this might not be feasable if it is too difficult or time consuming to program, but if it is relatively easy to add this would be a great improvement.

Vinatieri for Prez

02-11-2004, 12:23 AM

I like Ranger's idea about randomness in AI. I think that would be awesome and add to the play. Something within a reasonable range, and maybe have it change over the years for each team to reflect new GMs, etc.

JasonC23

02-11-2004, 10:36 AM