Feedback Requested: Tweaked Settings for Aging Curves

[Seviien]

!!!Updates!!!
July 25,2006: Settings below have been updated for use with patch 1.02 build 12262
July 26,2006: Pitching settings tweaked & modded Data expanded to show 1916-1979
Sep 2, 2006: Settings adjusted for patch 1.03; updated charts and summary stats to follow.
Sep 2, 2006: Charts and summary stats updated


Below are charts that shows averages for recent MLB data alongside both the default OOTP settings and my modded settings. The tweaks aren't perfect, but I think they're a nice improvement over default.

Keep in mind that OOTP is a complex game, and the number of factors that seem to influence the aging curves are many. One of the prices for having such a complex and customizable game is that the interaction of customized options can produce results slightly different than what the customizer may desire. In fact, given all of the options in the game, the accuracy of the default settings is really a testament to the quality of the product and the people behind it.

Note that I made all my test reads after letting the game run for 15 years worth of sims. By using data from the 16th-20th years, I felt that I would be capturing the tendancies of AI player development -- which is crucial to long-term single player sims.
To allay concerns about the modifiers becoming dated through time, I used my suggested settings to sim from 1901-1979. That sim is represented below.

The settings I've used are depicted below. Nothing is ironclad, so feel free to experiment. That said, if you change the "Things That I Think Matter" ... well, I wouldnt be surprised if your results are different than mine


!!!SETTINGS!!!
(Set in-game values equal to values below)
~~~~~~~~~~~~~~~~~~~~~
Things That I Think Matter:

10 round draft

3 Tier Minor System with the following MLEs:
```````````````AAA````AA``````A
Average```````0.850```0.775```0.700
Extra Base Hits`0.800```0.700```0.600
Homeruns`````0.800```0.650```0.500
Walks````````0.900```0.763```0.625
Strikeouts`````0.900```0.763```0.625

Batter Aging = 0.900
Batter Development = 0.700

Pitcher Aging = 4.000
Pitcher Development = 3.000

Coaching system is turned off

Injury frequency is "normal"

162 game schedule (Anything between 140-180 should be fine)

~~~~~~~~~~~~~~~~~~~~~
Settings That Probably Don't Matter:
League totals and player creation modifiers are set to track historical data.

~~~~~~~~~~~~~~~~~~~~~
Suggestions:
I found the accuracy of the curves to improve over time, so for best results I would suggest running your sim for 20-30 years before digging in and starting to play.


!!!SUMMARY DATA!!!
~~~~~~~~~~~~~~~~~~~~~
Summary of Average Player Ages:

MLB Actuals: 28.7
Modded OOTP Batters: 28.9
Default OOTP Batters: 29.1

MLB Actuals: 28.5
Modded OOTP Pitchers: 28.9
Default OOTP Pitchers: 30.4

[Seviien]

Info consolidated into above post

[jbergey22]

fabulous work, thanks for the info

[lighthousekeeper]

Thanks for this information - very useful. I just can't believe how much you need to adjust the aging factors! What would you have done if 0.001 didn't do the trick?!

Question: with the 0.001 and 0.002 aging factors , are you seeing some outlying players still hanging around at 50?

[turdfurgeson]

Ummmm, nevermind

[Seviien]

Either lower the number of rounds in the draft or reduce the development modifers. But yeah, part of the reason it was 15+ tests was that I didn't think I'd have to go that low

Indicentally, the game does work with negative values for the aging modifers, but it didn't seem to help...

[jarmenia]

Good work Seviien. I've plugged your modifiers into a test league and will play around with them. I'll let you know if I find any problems.

[scefalu]

Excellent stuff. I always enjoy looking at this type of information because I just don't have the time to generate it myself.

Thanks for taking the time and sharing the information with the rest of us.

[Carplos]

Have you tried these numbers on non-MLB sized leagues at all?

[lighthousekeeper]

Im curious, what was your reasoning for tweaking the MLEs?

[jarmenia]

Originally Posted by lighthousekeeper

Im curious, what was your reasoning for tweaking the MLEs?

I'm guessing because he wanted his minor leagues to be of a lesser quality than the Major Leagues.

[Seviien]

Carplos:
All my tests were with a 24-team league using a 162-game schedule. So technically, I haven't tried it with a MLB-sized league
My guess is that there wouldn't be any real change in a league as small as 16 teams or as big as 32. Beyond that, there might be some changes, but I wouldn't think they'd be too big.

Lighthouse:
I tweaked the MLEs because I wanted a three-tier minors system; I know it's more realistic, but 5 levels of minors felt a little like bloat. That being the case, I didn't think it would be best to plug the default single-A settings from the game into my lowest minor league, since Markus had made the rookie and Short-Season-A settings lower than A ball. I was concerned that it would make development occur too soon. So I just averaged the default single-A settings with the default rookie settings and used those for my single A settings. Given that I was using custom settings for the low minors, I didn't want the jump to AA to be too severe, so I used the default AAA settings for AAA and then just averaged the default AAA and my custom A settings for AA. That way the jump from AAA to the majors was the same as Markus intended and the jumps up the minor leagues were all smooth and even.

It's probably worth noting that I used these MLEs in the leagues that became both the default and recommended lines. As such it taints my comparison a bit, since "default" isn't truly default. Still... I don't think that it was my MLEs that threw the baseline off of MLB actuals, so the tweaks should still constitute an improvement.

[The Wolf]

Originally Posted by jarmenia

Good work Seviien. I've plugged your modifiers into a test league and will play around with them. I'll let you know if I find any problems.

"

[Elendil]

What were injuries set to?

[CaLíKrAzY]

Originally Posted by Elendil

What were injuries set to?

normal

[Seviien]

yup. injuries were at normal.

[CaLíKrAzY]

BTW,

Just wanted to say thanks for taking time out to do this...

[Seviien]

You're welcome
Smoothing the arcs added to my sense of enjoyment from the game, and I figured it had a chance of doing the same for some other people. I'm just glad I could be a help. I'm really, really proud of the batting arc.

Also... is Skydog around anywhere? He did something like this for 6.5 and I was curious if he had any thoughts or ideas for improvements.

[Carplos]

How do you check the average age of your league? I'm testing these numbers out with some my leagues--which aren't nearly as deep as your test league--and also Tiger Fan's 1930 Fictional Template, which I noticed after doing an initial 50 year sim that players fell flat consistently at 33.

[Seviien]

Um...
I'm sure there's an easier way, but I first did a data dump to csv files. Then I looked at players.csv and the carrer batting and pitching files. I imported players.csv into Excel and did a text-to-columns to distill the players' date of bith to just the year.

I imported the batting and pitching files into separate tables in MS Access. I then filtered out all the unwanted splits and the minor league data before exporting to text files. Those text files were then imported into Excel. Using playerID as a unique key, I used the "vlookup" function to interject the year of birth data into the career batting and career pitching worksheets. I then subtracted year of birth from the season year for a given stat line. Then it was just a matter of creating a pivot table, plopping age into the data field, changing the view option from "sum" to "average," and finally filtering out the seasons I didn't want to look at.

Like I said... I'm sure there was an easier and more direct way. The main issues are that (1) player age in current season is not included in the career batting and pitching files and (2) after 25 seasons, the career batting and pitching files were so large as to choke excel if they were imported directly without filtering unnecessary rows in Access.