PDA

View Full Version : Draft Class CSV Ratings -> FOF Ratings Translation


aston217
02-18-2012, 08:25 PM
Or, what do I gotta do to make a 100 BPR wide receiver around here?!

--

This stems out of a conversation I had with Ben E Lou a while ago at the BFL forums, about custom draft classes. Basically, where you control the ratings for every player in a draft by editing the csv file, then importing that into the game.

Ben offered the following formula for converting the 375-625 scale used in the CSV, to the 0-100 scale we see in-game:


(X-375)/2.5


CliffNotes:

1) Ben probably meant it only as a rule of thumb, but the translation does not appear to be linear like that.

2) Additionally, the translation appears to be different for different position groups.

3) There does not appear to be any noise - if you enter the same value in for a static bar, and view that pre-TC rookie static bar via the right click trick on the Trade Information screen, you will get the same number back.

4) The numbers shuffle some post-TC, but only by very small amounts outside of volatility.

5) While 375 appears to translate to 0, 625 does not appear to translate to 100.

Details

I first started with a class of 600 WRs. This means, of the 863 players in the draft class, everybody was a WR and everybody was rated 600 in every available category.

Keep in mind that not all categories are columns you can control via the CSV. The rest are generated in-game. However, I'm looking only at the BPR bar.

[i]Every single WR came out of this class with 78 BPR. OK, I didn't look through all 863. But I looked through 10-15 signed rookies, and I think it's a reasonable extrapolation here.

The linear formula has 600 => 90. So I thought, perhaps, if this is more quadratic than linear, maybe I can find a better formula here. My results:

WRs: BPR bar
625(100)->78
600(90)->78
550(70)->55
500(50)->42
400(10)->11
375(0)->0

I stopped after 625 and did a double take. I was fully expecting 100 BPR bars to come out. Instead, 78.

I tried: importing into different leagues, adjusting the vol ratings from everyone being 0 to everyone being 100, adjusting intelligence ratings for everyone, changing it so that the WRs had just 625 in the BPR column and 375 everywhere else...

Nope. The best, and only, BPR number I come out with, is 78.

Fullbacks

I did this for FBs next because thenewchuckd had also run some draft files before and thought it was odd that his 625 FBs weren't getting 100 PI ratings. He thought static bars must be masked, or something funky was going on with the draft import* (see bottom).

This is what I can say about FBs: 625 FBs have 72 power inside.

Chuck checked, and he has the same result. Mike made a comment at the USFL about how perhaps it was a relative number, to the ratings of other guys in the league. Not discounting that possibility totally, but I changed up my draft class so that there were 5 guys with 625 PI and everyone else (all 850+ of them) had 375.

So, the 375 guys were 0 and the only 625 guy I checked, had 72 PI.

Oh, and by the way, the draft file won't take values higher than 625 from the CSV. I tried.

CONCLUSIONS????

I do not know what to conclude. Clearly it is possible for fullbacks to have better than 72 PI, and WRs to have better than 78 BPR. And it is clearly possible for this to happen without volatility to kick in, the number of surefire studs at the top of the draft that come out better.

Any enlightenment here? Would be much appreciated.

* - Ben did mention that if you don't have TCY installed, there can be funky things that happen with draft class generating. I do not have TCY installed, but I understood that to mean some rare discrepancies, and not widespread trends as I am describing here.

Ben E Lou
02-18-2012, 08:38 PM
...but I understood that to mean some rare discrepancies, and not widespread trends as I am describing here....then you didn't understand what I was saying. ;) Let me be more clear: if you don't have TCY installed, it's completely useless to do any testing with the draft file editor.

aston217
02-18-2012, 08:44 PM
Gosh, well, that sucks.

That was a lot of typing! and testing! Oh well.

Ben E Lou
02-18-2012, 08:47 PM
Just took a regular draft class and put 625s everywhere. Here's one of the guys you were looking for.

http://www.fof-ihof.com/upload/Ben%20E%20Lou/hereyago.png

aston217
02-18-2012, 08:48 PM
^ Oh dang. That is pretty conclusive, as to the worthlessness of all of my results, heh!

OK, I am curious, though. Is your formula, (X-375)/2.5, meant to be exact? As in, 550 actually means 70, exactly 70. Or is it nonlinear?

Ben E Lou
02-18-2012, 08:49 PM
Oh...and it could be that it has to be not just installed, but licensed as well. I'm not really 100% sure. But I am 100% sure that without TCY, the import process isn't WYSIWYG. I have TCY both installed and licensed.

Ben E Lou
02-18-2012, 08:51 PM
Yes, it's linear. It's clear that 375-625 is the range, and that it's linear. From there, the formula was easy to derive.

aston217
02-18-2012, 08:56 PM
And I got all excited that I had found something interesting...:p

I suppose this applies to the combine-side too. I was going to say, "looks like it is even possible for 625 BPR WRs to have 4.51 dash", or "looks like 28 is a magic boundary for QB sole scores of sorts"...

Both of which, might be true, or close to it, being numbers seen in Malcpow's thread. But if WYSINWYG, that's pretty disappointing.

At any rate, thanks for setting me straight once again, Ben.

Ben E Lou
02-18-2012, 08:59 PM
And fwiw, I have long suspected that FBs are nerfed all over the place. My guess on the logic there: a fullback with 100 talent in key running ratings wouldn't exist because he would have been used as a halfback, not a fullback.

Nemesis
02-18-2012, 08:59 PM
Would be interesting to know how many historic league commishes (who do the draft files, of course) have TCY installed and licensed.

Ben E Lou
02-18-2012, 09:01 PM
(To be clear, the FB thing is pure speculation. The TCY thing I know for sure. There are patch notes out there somewhere that verify the TCY thing. Running a search now...)

Ben E Lou
02-18-2012, 09:23 PM
Couldn't find the patch notes, but found this from JG himself:

Front Office Football Central - View Single Post - The College Years? (http://www.operationsports.com/fofc/showpost.php?p=1640504&postcount=11)


The import function actually looks for and uses a piece of TCY. It works better when TCY is installed on your computer.

I had disabled the import function when TCY isn't present at one point, but reinstated it in one of the patches because that didn't seem fair to everyone. The estimates it makes when TCY is not present may not be ideal.

aston217
02-19-2012, 01:05 AM
Ah, thanks. Also, from later in the thread, Jim says this:

FOF 2007, importing a TCY draft without TCY installed on the computer, will give you a draft class that isn't completely correct experience and age-wise. FOF relies on a piece of the TCY aging algorithm when importing that, in retrospect, I should have simply duplicated in FOF rather than just throwing a quick estimate in there.


It's interesting that if it's just the aging/experience algorithm that is the missing piece here, that the results are as what they are - wholesale inaccuracy (or maybe 'unexpected modification' is better) in the translation of values from .csv to .faf.

Or maybe there are more pieces of TCY that FOF relies on, and Jim just neglected to mention when he went into detail.

Ben E Lou
02-19-2012, 05:49 AM
Or maybe there are more pieces of TCY that FOF relies on, and Jim just neglected to mention when he went into detail.That would be SHOCKING. ;)

aston217
02-19-2012, 11:40 AM
Oh, Jim :p

I checked -- it does appear you need TCY licensed, in addition to installed. Bummer.

ezpkns34
07-21-2013, 05:18 PM
Bumping to ask if this would include the draft classes found here:
Historic Draft & Roster Files (File Reupload 15-Mar-11 + Draft tweak) - Front Office Football Central (http://www.operationsports.com/fofc/showthread.php?t=55962)