Fictional Player/Draft File Generator

[Ben E Lou]

There are a number of things in FOF that are realistic to the modern NFL that just aren't my cup of tea. I'd prefer more offensive variety than the product that the NFL is putting on the field these days. I'm sure that some of what I'm looking for would be provided in the historical files, but I have little/no interest in playing that way. I've been playing text sims long enough to know what I like: fictional player pools. Period. Thus the creation of this utility. It currently just runs on my local machine's php server, but if enough people would want to use it AND I can find time to create an interface for some key user inputs, it's something that I could potentially put on a web server for others to use. At any rate, this utility creates an initial player pool of 3300ish players, and can spit out 100 years worth of draft classes a well that you don't have to import in one at a time. They're just in the csv and the game reads them. Here are the changes that I've made to player creation, and why:



OVERALL CHANGES

• Less overall talent, which reduces offensive output a bit.
• Dice roll upon generation of QBs to potentially reduce accuracy and good decisions a bit (decreases completion percentage, increases interceptions)
• Players are a little more likely to have attended a college closer to home than FOF9's baseline (which, to be fair, is the best FOF has ever done at this by far.)
• Draftees are a bit younger than FOF standards, making careers last a bit longer. (Nothing huge here. You're not going to see 18-year-olds in the draft pool or anything, just a bit more bias toward ages 20-22 than you see in FOF.)
• Endurance and speed more closely correlated with weight/height ratio. (Players whose ratio is higher for their position are more likely to have lower endurance, lower speed, more strength. Vice versa for those with lower ratios.)
• Overall, endurance is higher for non-DL guys, particularly for players who are otherwise crappy. (If a guy is good enough to make an NFL roster as a backup WR, he shouldn't be doomed to 25 endurance just because overall he's not very good. Replacement-level player != out of shape.)

POSITION-SPECIFIC CHANGES


So here is where I'm taking a bit more license. First off, I grew up on Larry Csonka and Eugene "Mercury" Morris, Franco Harris and Rocky Bleier, William Andrews and Lynn Cain. Teams that had a great running back AND a great fullback. Of course, the fullback is a dying/dead breed in modern football, and this is reflected in FOF9: you'll almost never see a FB rated above 60ish overall. Well, I've removed that cap in these files. You'll see total stud FBs in these files from time to time.
Next, while most players are built the standard FOF way with just an overall rating and letting FOF fill in the details, I've implemented--especially for modestly-talented players--some "player types." Basically, these are designed to reflect skill sets that "go together." Maybe a RB is a "1" overall rating (serviceable backup,) but he's a "3rd down specialist" player type with poor running skills, but pretty solid receiving skills. (An overall "1" isn't going to be a stud or even above-average receiver, but he might have 45-55s across the board in route running, hands, adjust to ball, etc., making him worth a possible roster spot.) Here are all of the "player types" that I have:


QB

• Short Passer--As the name suggests, these guys get a boost to timing/accuracy/touch, but they'll likely be low in arm strength/confidence and therefore will keep it short.
• Gunslinger--Exact opposite of the above. Strong-arm guys who like to air it out.
• Runner--By default, often stud QBs in FOF9 have mad running skills and can go for 1000 yards scrambling. But that's not what I'm talking about here. These guys are "what if Jamelle Holieway had been allowed to do his thang in the NFL???" types. (And FOF9 will game plan a LOT of read option runs for these guys. Good times!)
• Field General--"Game manager" types. They don't make a lot of mistakes, but they don't have especially good physical skills.

RB

• Third-Down Back--As mentioned earlier, some guys will have upgraded receiving skills and downgraded running skills.
• Workhorse--Modern NFL RBs don't tote the rock like they used to, and FOF reflects that in their endurance ratings. There's a small chance that any starting-quality RB will be a workhorse back with very high endurance, capable of 300-350 carries per season.

FB

• As mentioned earlier, the guardrails are taken off of these guys, so they can be studs overall. But also...
• Receiver--Like third down backs above, some FBs will be good at pass-catching, but not as good at other aspects of their craft.
• Blocker--Minimal running/receiving skills, but solid blocking backs.

TE

• Receiver--Pretty obvious here. Good at receiving, bad at blocking.
• Blocker--Vice versa

A 100-year sample league is finishing up now, and I'm heading to bed. Screen shots in the morning, and I'll also provide some files then most likely. Let me know if there's interest in testing/playing with this setup.

[NawlinsFan]

I am of that same era and appreciation of what was real football when compared to the flag/touch football played today. I'm interested at giving it a run.

[Jbmoore68]

I'm with you guys on the real football, count me in.

[Ben E Lou]

Testing last night didn't give me the final results I'm looking for before outputting the files. I just made a change to create more variability in the accuracy/decisions nerf. (Short version: I want those overall numbers to be down, but more outliers on either end of the spectrum.) I *think* I'll have at least a test file for folks to play with by this evening.

[Jstraub]

Count me in as far as interested! Love the idea and sounds like an amazing opportunity to enhance the value of career sims. Thank you

[Jstraub]

Not sure you are taking suggestions but if so... here are some other 'archtypes' that could be intriguing.

DL -> run stuffer vs. pass rusher (maybe the default engine produces enough of these already?)

OL -> road grader vs. protectionist

CB -> press specialist vs. man specialist

RB -> in addition to what you suggested... Add a Power back? No speed but great strength (would weight more) and convert short yardage?

LB - > pass rusher vs. coverage specialist vs. run stopper
WR -> Deep threat vs. possession Vs. YAC specialist

[NawlinsFan]

Power back - off on a tangent but imagine having the option to plug an O/D lineman in as fullback for a short yardage or goal line play.

[Ben E Lou]

Originally Posted by Jstraub

Not sure you are taking suggestions but if so... here are some other 'archtypes' that could be intriguing.

DL -> run stuffer vs. pass rusher (maybe the default engine produces enough of these already?)

OL -> road grader vs. protectionist

CB -> press specialist vs. man specialist

RB -> in addition to what you suggested... Add a Power back? No speed but great strength (would weight more) and convert short yardage?

LB - > pass rusher vs. coverage specialist vs. run stopper
WR -> Deep threat vs. possession Vs. YAC specialist

Probably important to note here in terms of setting expectations: I get zero enjoyment out of trying to build a defensive plan based on my talent like I do on offense, so there's little/no chance I'm going to be bothered with doing anything to defensive players. I created a generic 3-4 man/press plan for defense within a day or two of the release, and I just draft/acquire defensive players who fit the talent profile for that plan and never hire 4-3 coaches so I don't have to be bothered with creating a 2nd option.


On the other stuff...

OL and WR either already have that stuff in standard FOF9, or perhaps there are few enough bars that matter for those types that they just kinda happen randomly often enough that forcing them isn't worth the time. There are tons of OL with high run blocking/loss pass blocking and vice versa, for example, and

Power back isn't a "type" like the others in my code where there's a dice roll to create that specific type by increasing some ratings and decreasing others, but it's kinda baked into the player creation system I have in that heavier backs are generally stronger and slower. Plus it's kinda like the above two: really only speed/strength/elusiveness would be involved anyway, so even if Jim isn't doing it intentionally, just the dice rolls that create the bars from the overall ratings will give you plenty of guys high in strength, low in speed and elusiveness. All the stuff I'm coding involves a minimum of four bars, usually more.

[Ben E Lou]

Oh, and this is worth documenting: there's a good bit of variability in whether the AI will use the "player types" appropriately, and that may impact whether some will want to go this route.


For example, I'm fairly certain that AI teams don't built their playbooks or game plans with a behind-the-scenes "Guided" generator like some of us would. A specific example: I was testing a team with a "9" FB (86/86 overall) on it. I used the guided playbook to increase FB usage in the plays, and the guided game plan to crank up passing attempts to the FB. With a lot of FB runs in the playbook, he was getting 100-150 carries per season and 80-100 pass targets. And of course--especially if your team has an OC who is Balanced or Smashmouth, it'd be both easy and productive to get a guy like that at least 150-200 carries and over 100 targets with a custom game plan. But AI teams are going to max out at around 60 targets and 70 carries per season for a FB, even an absolute stud one. So, other things to note...

• The AI WILL crank up QB run, QB option, and RPO plays for QBs with good running skills.
• Gunslinger and Short Passer QBs do tend to have higher and lower AAYA respectively. (That's likely mainly just a function of the "Confidence" bar directly correlating to that.)
• AI teams won't throw to great receiving RBs as often as we might, nor even as often as the modern NFL. (I still have completion percentage a bit high. I might tinker with nerfing WRs a little bit to mitigate that. But the flip side might be that it throws more to the backs and gets even higher completion percentage as a result. )
• The AI will definitely target a good receiving TE a lot. Either FOF7 or FOF8 was the first version that did this really well by default, and FOF9 continues in that regard. (Prior to whichever release changed that, the default Rex game plan formation settings just didn't have TE1 on the field enough to ever get 1000 yards receiving.)

HOUSE RULE CONSIDERATIONS
One simple one that I know I'm going to use with these files: no RB-->WR or FB-->TE position switching. The AI doesn't know to do it, so I won't either, and because the player types may well create RBs and FBs with mediocre overall ratings but receiving ratings that would allow them to be converted to decent WRs and TEs, that would be too much of an extra advantage to the human.

[Ben E Lou]

Originally Posted by Ben E Lou

I *think* I'll have at least a test file for folks to play with by this evening.

Wasn't happy with the results last night, which turned out to be a good thing overall because in reviewing one section of code, I stumbled across a significant bug in an adjacent area. Squashing that bug has greatly increased my ability to control passing stats. I need to rebalance things now, because in my first run this morning they came out significantly *lower* than I want--(which is a good thing...I was having the hardest time reducing them without having to resort to extreme talent reduction, nerfing, etc.) At any rate, I just tweaked a couple of things that I suspect will get me much closer to what I'm looking for, and as such things are probably now on track for me being able to release a few test files later today.

[Harveysimpson]

It would be fantastic to be able to use such a tool.

Excited for you to share!

[DavronM]

This tool existed for FOF7/8 and I LOVED it. Unfortunately it doesn't work with the new version.

Roster/Draft Class CSV Utility - Front Office Football Central

[NawlinsFan]

Casual observation. I am running through using BLT1 right now and it seems that there are still lots of QB's running a lot and almost leading their teams in rushing. I am not controlling the game plan or calling plays so this is the AI selections.

[QuikSand]

Still looking for someone to give this a go?

[Ben E Lou]

Originally Posted by QuikSand

Still looking for someone to give this a go?

Sure! I posted files in another thread: Testing Fictional Player Pool/Draft Files - Front Office Football Central

[Harveysimpson]

Still very interested in this thread.

How is progress going?

Would you every consider sharing the utility itself?

[KingZal]

I'm interested in this too. I don't mind messing around with excel sheets but I'd love to see a full utility like FOF8 had.