Stats messed up in Dynasty Mode - possible cause

[babouisi]

Hi all,

I've noticed some players like Kobe or T-Mac will lower their scoring averages in 2nd and over Dynasty seasons. However, others like Ray Allen will still score close to reality.

As I edited the roster file, I looked for Kobe, T-Mac and Allen to compare. Guess what? While Kobe and T-Mac have SCORING AREA 5 (which states for a big man with outside shooting skill), Allen has SCORING AREA 6 (which states for a swingman).

I checked all the starters SG/SF (I'm using a season roster file -not edited-, so some players could appear out of their correct teams) that's what I saw:

Player, scorearea (meaning according toolkit editor)
------------------------------------------------------------
Glover, 2 (shooting player)
Stephen Jackson, 5 (big man with outside shooting)
Pierce, 4 (unskilled big man)
Kedrick Brown, 4
Rose, 5
Pippen, 4
Ricky Davis, 4
Darius Miles, 4
Finley, 5
Jamison, 9 (serviciable big man)
Jon Barry, 2
Carmelo Anthony, 4
Richard Hamilton, 4
Prince, 5
Jason Richardson, 5
Dunleavy, 5
Mobley, 6 (swingman)
Jimmy Jackson, 4
Reggie Miller, 5
Artest, 4
Quentin Richardson, 5
Magette, 4
Bryant, 5
George, 4
Mike Miller, 5
Posey, 4
Eddie Jones, 5
Odom, 9
Redd, 5
Tim Thomas, 9
Sprewell, 6
Garnett, 9
Kittles, 2
Jefferson, 4
Wesley, 6
Lynch, 4
Houston, 5
Van Horn, 4
McGrady, 5
Gooden, 9
Iverson, 6
Glenn Robinson, 9
Joe Johnson, 4
Marion, 5
Wells, 6
Rasheed Wallace, 10 (big man with mid-range)
Christie, 4
Stojakovic, 10
Ginobili, 4
Walt Williams, 5
Ronald Murray, 2
Rashard Lewis, 4
Ray Allen, 6
Carter, 5
Stevenson, 2
Harpring, 4
Hughes, 6
Hayes, 4

It's clear that some players have a strange scorearea field (at least, much different than other players with same role), and I've confirmed those that have scorearea 6 (Allen, Iverson, Mobley) do achieve much better stats in the 2nd dynasty season than ones with scorearea 4-5.

Maybe it can be the key for the messed stats. I'll try to test it later.

Regards,
Felipe

[Metsis]

You should definitely do that...

I think we could all appreciate a patch that places all the people into their correct shooting forms.

You do the research and tell us all what you found out. Since when has Pierce been a unskilled big man! I really don't matter that much for his style for game etc. but he's everything but unskilled. And I wouldn't put him in the big man category either.

[babouisi]

Maybe it isn't as easy as the previous message say...

When playing a season/dynasty, a lot of files are created. Some of them are really suspicious and could have the cause for the messed stats:

SIMOOP.DBF
SIMBENCH.DBF
SIMENGINE.DBF
SIMMOD.DBF
SIMSTATGEN.DBF

I can't edit them now as I'm at work, but surely you tweakers will do it

Regards,
Felipe

[Metsis]

Another thing just popped up into my mind...

Do these change from year to year? I mean can someone become a big man with mid-range from a unskilled big man?

Could you please check that out too. Cause if it does not change, well then this is just as rigid as the dstats system we had before.

[babouisi]

Changing scorearea didn't work

[JonathanAu]

I think scorearea is only for gameplay.

[Mark.]

I think i have the reason. In 2003, there were coloums in the DBF called D-Stats, which created the averages for players when the games were simmed, i remember some one said that when they deleted these stats for all players, Matt Harpring led the league in scoring. Well to get to the point, 2004 doesnt have D Stats for some screwed up reason, it does however have C Stats which i think are for the first season, hence the realistic stats in the first season of a dynasty, so when it comes to the second season there are no D Stats for the computer to go by so it just uses the players ratings, you shud see that Shaq and Ray Allen still average about 25ppg cos they have a high feild goal rating, but i think the max is about 25ppg because Shaq has a 99FG rating.

So the only way to fix this would be to try adding in coloums for D-Stats, but this still may not work if the game doesnt recognise them. If we cold find out from EA, if it was an accident they were left out, it will be easy just to add them, but if the game was never intended to have them ( i wouldnt know why) then the only cure would be for EA to make a patch, did they make a patch last year?

[JonathanAu]

the problem with Dstats was Shaq would score 25ppg even if he's already 40 years old and 60+ rating.

I think EA did the right thing dropping DStats. Now, if someone will only be kind enough to review all the ratings and adjust them accordingly.

[chipper]

I also agree that Dstats are better off being removed.

scorearea (meaning according toolkit editor)

Maybe those scorearea meaning are changed for 2004?

[babouisi]

I think if EA removed D-Stats they should have implemented another way for development of players. I think so since some players like Shaq and Iverson will score 25ppg in the 2nd Dynasty season, and others like Yao will increase his average up to 20.

I couldn't find any new database field for the players which could state for a development ratio or something similar. In fact, players.dbf file structure is almost identical to the 2003 one (excluding the absence of D-Stats). Even still remains the development ratio (A to F).

Maybe the new way EA implemented is wrong (and we have to wait for a patch from them). However, I think hard patchers could focus into the new .dbf files created when in season/dynasty mode, which I edited and seem like they contain some kind of modifiers and correction percentages. Maybe (or as I think, surely) the game do apply them to generate sim stats.

These files are (mentioned above):
SIMOOP.DBF
SIMBENCH.DBF
SIMENGINE.DBF
SIMMOD.DBF
SIMSTATGEN.DBF <--- very suspicious, don't you think?

Another issue is Shooting Guards averaging 8-9 rebounds, despite his rebounding rating is low. This should be affected by an error in the sim stats generation.

I'm pretty sure we are able to fix these issues by ourself. We'll have to find what and where are the sim stats generators within these .dbf files and correct them.

Regards,
Felipe

[chipper]

Nice work, babouisi.
Keep us update as you find out more about them.

I am currently busy doing exams and working on my 90-91 season roster but if I find anything related to that, i sure will let you know about it.

[smithy_17]

I am currently busy doing exams and working on my 90-91 season roster

glad to hear that. can you say how long it will approx. take you until the 90-91 roster will be ready. will it be playable in franchise mode? -thnx

[EGarrett]

Maybe when you find out how to fix that...you'll find a way to fix all the generated players averaging 13 points, 4 rebounds and 4 assists...

[JonathanAu]

there are some tables in there that seem related to player development.
C_TT, PF_TT, SF_TT, SG_TT, PG_TT

Don't know what TT stands for, but inside you will see the fields PCT_0, PCT_5..... PCT_100 which looks like the Increase/decrease value for each range of ratings.

I'd like to check it out but I'm at work right now and I don't have the game here (just the DB).

Also, can someone verify if the ratings are significantly changed between the first and 2nd seasons? (for McGrady, Allen, Bryant, and Iverson)

[chipper]

there are some tables in there that seem related to player development.
C_TT, PF_TT, SF_TT, SG_TT, PG_TT

Don't know what TT stands for, but inside you will see the fields PCT_0, PCT_5..... PCT_100 which looks like the Increase/decrease value for each range of ratings.

I'd like to check it out but I'm at work right now and I don't have the game here (just the DB).

Also, can someone verify if the ratings are significantly changed between the first and 2nd seasons? (for McGrady, Allen, Bryant, and Iverson)

Hey Jonathan , good to see you again on the forum.
I was actually looking around that file too and I thought they could be training files.

I will change all the numbers to 10 (just to be extreme) and test the traning.

[chipper]

I am currently busy doing exams and working on my 90-91 season roster

glad to hear that. can you say how long it will approx. take you until the 90-91 roster will be ready. will it be playable in franchise mode? -thnx

It will take a while. I have pretty much done the roster, I need to do lots and lots of cyberfaces. Thankfully, some people are helping me out so it will be a good one

If you are interested, please check this thread over at nbalive.org
http://nbalive.org/forums/viewtopic.php?t=1631

[JonathanAu]

it's nice to be back here. I didn't enjoy Live 2003 much so I went inactive for a while.

Anyway, update us on your experiments.

[babouisi]

I think c_tt.dbf, pf_tt.dbf and the other files are related to the training effect. These files contains a positive or negative number for all the ratings, which matches with the training camp effect screen.

SIMOOP.DBF may be a position checker correction rating; i.e., a PG who plays at C will receive: -7FG, -5Three, -5Inside, -5OfReb and so on.

I'm checking CAREERPKG and I discovered something VERY interesting.
I simmed 7 complete seasons in dynasty mode, letting the CPU to autoplay the offseasons. I put all the stats into Excel and got those players with significant improvements or decreasing along the years, then went to the editor to check things. These are the results:

The vast majority of key players which did decrease dramatically his average scoring (Iverson, Wells, Kidd, J.O'Neal, Richardson, Rose, Stack, Kobe, K.G., Pierce, Hamilton, Nash, Marbury, T-Mac, Carter) have CAREERPKG = D. Also decreased a lot: Allen (C), Francis (C)
Others decreased slightly but not dramatically: Baron Davis, Gasol, Ricky Davis. They have CAREERPKG=E (Davis has F).

Players that improved the scoring: D.George, Nowitzki, Arenas, Tskitishvili, LaFrentz, Dalembert, Hudson. All of them have CAREERPKG=C
Other significant improvements: Chandler (A), Tim Thomas (D), Yao, Brad Miller (both E).

A big number of players (like Shaq) are PKG=D and they had a normal development: Some went a bit better, some went a bit worse. Carmelo and LeBron, who also are PKG=D, kept scoring 13-16 points every year; they didn't become superstars but they didn't fall into the crap.

Conclusion:
CAREERPKG=A: 1 improved, 0 decreased
CAREERPKG=B: (nothing tested)
CAREERPKG=C: 7 improved, 2 decreased
CAREERPKG=D: 1 improved, 17 decreased
CAREERPKG=E: 2 improved, 2 decreased
CAREERPKG=F: 2 decreased

It may be the key: Seems like this year the CAREERPKG is highly important, but it should be kind of randomizer. To have a best letter (A,B,C) seems to be a pretty good chance of improvement year by year.
Letter D may be considered as an average chance of development, but some superstars rated D will go down in scoring (some others didn't).
To have a letter worse than D gives a VERY high chance of decreasing.

So, I think superstars should have a better letter, from A to C. This may help to correct the messed stats.

I'll test it in-depth during the weekend.

[chipper]

Nice work.
We should have a dedicated space for our findings.

I will just add mine to this. XX_tt files are indeed the training effect files. Change all the numbers to 10 and there wil be significant increase (7,5,3.. all over the place) as a result of traning. Now I am not quite sure but i think traning becomes more effect as seasons go by? Will have to do more experiment but it will take time... because my brother is doing all the tests for me... he lives in LA.

Also, babouisi, you might want to take players' ages into consideration. Andrew said old players with high careerpkg will decline slowy, and young players with so-so careerpkg still have good chances of improving.

Finally, I thought Lebron has A potential rating

[The Duke]

yo chip i wouldnt mind testing those edited training files if u could send me the dbf's savenger@aol.com i would also like to see the result from the editing, thats if u dont mind.

[Little_Big_Man]

this probably is a little off topic though it is related to Dynasty stats. I am in a Dynasty with the Nugg's and for Rookie Player of the Month of West it was Nick Collison who averaged 10ppg,4rpg,3asp,1spg while with Melo I averaged 24ppg,3rpg,3asp yet he got the award, plus he led my team to a 7-1 record so far, what gives??

[chipper]

yo chip i wouldnt mind testing those edited training files if u could send me the dbf's savenger@aol.com i would also like to see the result from the editing, thats if u dont mind.

I deleted that save file because my players all improved too much.
But I will soon make a new one which is more reasonable (just removing the negative effects maybe) and post it up in a couple of days.

And Little_Big_Man, this is my bro's idea and he thinks human controlled teams are somewhat discriminated when they pick various awards.

Try to change the human controlled team (which i believe is possible by modifying teams.dbf file) and see if Anthony gets more votes.

[babouisi]

I discovered the players affected by the bug suffer a dramatic decrease of two ratings: FG and PRIMACY. The FG went down by a little (2-5 points), but primacy seems to fall a lot, 20-30 points.

[Mark.]

I am in my second season and t-mac's stats are ok the reason being that b4 i started the dynasty i edited his FG rating he also im proved from a 96 to a 99 so i dont think any of his ratings have gone down, this is the same for PP and B-Diddy, who i changed a little. It helps a bit but still T-mac is averaging only 26ppg instead of his 34ppg in my first season.[/i]

[fgrep15]

for guys like Kobe, Carter and Mcgrady just give them A for Career PKG

i also figured out how to finally get a players to get double digit assists

for all the stats your guys needs a certain combination of ratings

for assists high : pass, ball handle, off awareness, and i think primacy i had a guy on the bench as the 7th man playing 15min per game and avging 10.8 dimes b/c i made those 99 except for primacy he had 80 already......so if only your pass is 99 you'll only get about 5-7 assists per game at the most

rebounds: strength, reb ones of course, jump my player avg'd 17.8 rebs

block: block, jump, def aware (if these are at 99 the guy wll average about 6-7 blocks a game)

steals - def aware, steal, quick,speed i think , my player had 2.8 steals per game

so basically the player can't just be good in the one thing he has to be at least 80 in the others also

i also gave lebron, melo, wade, ford and bosh A's for career pkg, then when melo was 33 i made all his thing's 99 and he average 29.7ppg 16.7boards, 12.7 assts, 6.5blocks, 2.9assists got 3 mvp's in a row then retired (only got one defensive player, my guy on new york got the others) but his team only won one championship and lost twice in the finals