Ratings vs historical stats (over time)

[macprivateer]

VERY new to the game, just been looking at the manual as well as posts here trying to understand things better before I jump into it ( which will be running historical simulations). I have some questions since I don't quite understand the manual when it talks about Ratings and how they are “created” for games such as historical simulations.

Manual states that “Players' current and potential ratings change over the course of a season or a career, according to OOTP's player development model.”

My first question is how much are ratings “built” from the actual stats players achieved? For example, are the stats for say Catfish Hunter truly representative of his 1976 stats for the Yankees in 1976? And how close to those would the game do in an accurate replay of 1976? (I come from an APBA background so replay accuracy is important to me)

My next question is about how those ratings change over time. For example, say that if I’m looking to simulate from the 1976 season onward. I WOULD like the players to perform as they have done in real life each year (i.e., they’re “real” stats each year or in OOTP universe they’re “real” ratings), should I be disabling the player development?


Let me restate this with another example, I am looking for the individual players to perform similarly to their historical stats were for each year, so again starting in say 1976 I want to see Ron Guidry become the pitcher he was in 1978 and Jim Rice to be the player he was, etc.

[Reed]

There are many ways to play and yes it is easy to set it up to play like you are talking about.
Some people like to play what if scenarios such as what if Ted Williams didn't miss seasons due to the War or what if Babe Ruth stayed a pitcher or what if a players entire career is based just on their stats/stats their first year etc..
For me that is to fictional so I play basically like you are talking about.
Hope this helps.

[Rain King]

I do think it is important to mention that OOTP is a simulation and not a replay. So, there is going to be a larger standard deviation for player performance when playing out a historical season than something like APBA.

Your overall stats are going to come up very similar to real life, but the individual players have the possibility of over/under-performing.

[Brad K]

Originally Posted by macprivateer

My next question is about how those ratings change over time. For example, say that if I’m looking to simulate from the 1976 season onward. I WOULD like the players to perform as they have done in real life each year (i.e., they’re “real” stats each year or in OOTP universe they’re “real” ratings), should I be disabling the player development?

Disable development and set ratings to be based on current year only.

Its covered in the manual that this gives very close to real life performance.

[Brad K]

Meaning performance will still vary as it will in ANY GAME that doesn't have a feedback loop on player performance.

[Mifr44]

Originally Posted by Reed

There are many ways to play and yes it is easy to set it up to play like you are talking about.
Some people like to play what if scenarios such as what if Ted Williams didn't miss seasons due to the War or what if Babe Ruth stayed a pitcher or what if a players entire career is based just on their stats/stats their first year etc..
For me that is to fictional so I play basically like you are talking about.
Hope this helps.

I'm just getting back to my baseball simulator search (sidetracked by other matters), and your post speaks to how I would like to model some all-time greats and past favorites against each other. It's a bit tougher if you want to try and average some "best" seasons on your own. No MLB simulator I know of gives you that kind of flexibility.

What if Ray Fosse didn't have Pete Rose run over him at the ASG? What if Herb Score hadn't been hit in the face with a line drive? There are several other players who would be fun to model but either aren't included in various historical packs for lack of a minimum number of PAs/IPs, or their seasons after major injuries dropped off significantly and their career stats ended up being mediocre.

[fredbeene]

Originally Posted by Mifr44

I'm just getting back to my baseball simulator search (sidetracked by other matters), and your post speaks to how I would like to model some all-time greats and past favorites against each other. It's a bit tougher if you want to try and average some "best" seasons on your own. No MLB simulator I know of gives you that kind of flexibility.

What if Ray Fosse didn't have Pete Rose run over him at the ASG? What if Herb Score hadn't been hit in the face with a line drive? There are several other players who would be fun to model but either aren't included in various historical packs for lack of a minimum number of PAs/IPs, or their seasons after major injuries dropped off significantly and their career stats ended up being mediocre.

Yes that is my main goal of the game, what if scenarios.
My experience has been:
if you do no recalc, then many players will be different over the years, however your ray fosse would have better change of being great

if you do a replace there is no way ray fosse will be a star. you have to manually adjust his ratings. (which to me feels like cheating). But it will work. You can try to adjust potential, but it won't work for players over 30 (this is based on 200 sim attempts for me (other players will argue that though which is fine)

[Brad K]

Originally Posted by fredbeene

Yes that is my main goal of the game, what if scenarios.
My experience has been:
if you do no recalc, then many players will be different over the years, however your ray fosse would have better change of being great

if you do a replace there is no way ray fosse will be a star. you have to manually adjust his ratings. (which to me feels like cheating). But it will work. You can try to adjust potential, but it won't work for players over 30 (this is based on 200 sim attempts for me (other players will argue that though which is fine)

Hi Fred, good to see you. Let's review.

Potential is a clue as to what future ratings might be.

Potential does not affect future ratings. Its the other way around. Current ratings affect potential.

Editing potential doesn't affect ratings now or in the future for ANY player regardless of age. It only that potential greater than current isn't shown for players age 30+.

I'm really sorry for you that Bobby Thomson fell off the cliff at age 30 giving Hank Aaron a chance to play at a young age and get a head start on breaking Ruth's HR record.

[Syd Thrift]

Originally Posted by Mifr44

I'm just getting back to my baseball simulator search (sidetracked by other matters), and your post speaks to how I would like to model some all-time greats and past favorites against each other. It's a bit tougher if you want to try and average some "best" seasons on your own. No MLB simulator I know of gives you that kind of flexibility.

What if Ray Fosse didn't have Pete Rose run over him at the ASG? What if Herb Score hadn't been hit in the face with a line drive? There are several other players who would be fun to model but either aren't included in various historical packs for lack of a minimum number of PAs/IPs, or their seasons after major injuries dropped off significantly and their career stats ended up being mediocre.

If a player keeps playing after they stop having stats IRL the development engine takes over, so this works for a few players whose careers just plain stopped cold at a certain point (Lyman Bostock, Ray Chapman, and Mickey Cochrane all come to mind) but not so much for guys who suffered their tragedy but still tried to come back with bad results (so, Bo Jackson and Tony Conigliaro for instance). For those players you'd need to remove their post injury seasons from the database you're using.

Sent from my SM-G950U using Tapatalk

[Brad K]

Originally Posted by Syd Thrift

If a player keeps playing after they stop having stats IRL the development engine takes over, so this works for a few players whose careers just plain stopped cold at a certain point

OK, so settings are three year recalc and development engine on.

Three year recalc over rides development engine for players who have stats (?) and comes into play when the player runs out of stats. Is that right?

If the development engine is off does it still come into effect for players whose careers in the game last longer than histroical?

[Charlie Hough]

Yes, development can be turned off but should hopefully take over and modify player ratings when players' historical stats run out in the database. It didn't do this properly in the past, but I believe it has been addressed in more recent versions of OOTP.

By the way, I recommend 3-year recalculation with doubling the weight of current-season stats. It's a good balance between a single-season replay approach and some smoothing of the ratings to help address anomalies such as a one-off injury or a one-year slump dramatically altering a player's historical stats and therefore his ratings for a given season.

[Brad K]

So with three year recalc on what does having development on do? (Other than taking over when players run out of stats.) This is how I've been running the game.

I don't use double for current because I'm not interested in close year by year tracking of a player's career. For peaky players I don't want to have too much of a clue of when they'll be at their best.

[Charlie Hough]

Originally Posted by Brad K

So with three year recalc on what does having development on do?

If I'm not mistaken, it means player ratings can fluctuate during the season, although it's probably rare. But I've rarely used development with historical leagues, so maybe someone else can clarify.

[macprivateer]

Originally Posted by Brad K

So with three year recalc on what does having development on do? (Other than taking over when players run out of stats.) This is how I've been running the game.

I don't use double for current because I'm not interested in close year by year tracking of a player's career. For peaky players I don't want to have too much of a clue of when they'll be at their best.

Originally Posted by Charlie Hough

If I'm not mistaken, it means player ratings can fluctuate during the season, although it's probably rare. But I've rarely used development with historical leagues, so maybe someone else can clarify.

I’d like to take this a step further with an add on question…


With your guys help I am beginning to understand the recalc settings (1, 3, 5 years) and how it relates to and uses the player’s stats to form a rating. Or if using multi year recalc settings takes several years and averages them (this “average” can then also can be modified, amazing and amazingly complex game, so happy I’ve been sucked in).

Anyway to the question, which stats are used in this rating? The actual stats only? What if a player had no stats? Again this is with development OFF. For example, and again development is off, those players who are missing data due to the military service/injury situations, how do these players get ratings?

Let’s take DiMaggio for 1942), what is the basis for his ratings?

I see why it makes more sense (if I want to have Joltin’ Joe play in 1942) to instead have the recalc on 3 so that at least his rating will be based on 1940+1941+1942? But again, what about 1942? Does the recalc then really only “recalc” on 2 years instead of 3?

And to follow on with that concept…If I then play DiMaggio for a full 1942 season, whatever his stats end up being, would they ever be part of the new 3 year recalc for 1943? Even though they are not real stats? In other words, his ratings would be based on 1941(REAL)+1942(SIM)+1943(nothing yet so not included)? Or does it use only real stats so by 1944 Joe is hosed since his ratings will be based on 1942(nothing)+1943(nothing)+1944(nothing).

So do I "have" to have development on to have players get reasonable ratings for years missed during their careers?

[Garlon]

I suggest using 5yr recalc because that eliminates the problems with the WWII seasons mostly.

When you use 5yr recalc change the import values for batters to Adjust if Below 167 and Weaken if Below 67

For pitchers use Adjust if Below 19 and Weaken if Below 14.

Remember that these are averages so, over 5 seasons 167*5 = 835 AB and 67*5 = 335 AB so you are still using a good sample for creating the ratings.

For SP this works out to at least 19*4*5 = 380 IP and for RP 14*5 = 70

I suggest that you always play with development disabled when using recalc so that the player ratings only change once per season. The players will still have their ratings u[dated each season, but with development on their ratings will change each month of the season. I suggest turning off the "Storylines" in the global settings as that will cause some players to retire earlier, and use Retire According to History to keep the talent available in the league accurate.

For the 1942-1945 seasons the default file in the game uses the real league totals for those seasons. If you want players like Ted Williams to remain in the league they will be playing in that environment. To get around this I made a custom file that sets the 1942-1945 seasons to the average of the 4 seasons before and after the war so that performance works out better.

To use this file you need to rename it era_stats and replace the default era_stats file in the game folder. From the main menu go to Settings>Troubleshooting>Open Folder Containing Customizable Data>database and you will see the original era_stats file. I suggest renaming the original file to something like era_stats original so that the game will not use it and put this new file in that folder.

If you bought the game through Steam I suggest going to the Steam Workshop and subscribe to "Major League Stadiums 1901-2019 by " so that you get models of every stadium since 1901 and the "Major League Jerseys and Caps by No Pepper and Reds1" to give you uniforms. If you want fantasy teams of the best players for each franchise subscribe to "Baseball Greats".

There is also a historicals forum here if you have ore questions about playing historical games.

I suggest using these strategy settings too. Make sure you turn off the option for having OOTP automatically update the strategy settings each season.

[thehef]

Originally Posted by Garlon

I suggest that you always play with development disabled when using recalc so that the player ratings only change once per season. The players will still have their ratings u[dated each season, but with development on their ratings will change each month of the season. I suggest turning off the "Storylines" in the global settings as that will cause some players to retire earlier, and use Retire According to History to keep the talent available in the league accurate.

For the 1942-1945 seasons the default file in the game uses the real league totals for those seasons. If you want players like Ted Williams to remain in the league they will be playing in that environment.

So in the above scenario, if you are also using historical transactions, then regardless of whether you check the Miss Seasons Accdg to History box, Ted Williams we be will be "loaned" to a "team" called "military" and will miss the war seasons (and then later be "returned" to Boston). Manual edits would be required to keep him and others active through the war years.

Similarly, if historical transactions are enabled, the Retire Accdg to History checkbox will be ignored, and players will retire according to history.

Both of these seem to be, IMO, questionable design decisions that were implemented a few versions ago. Basically taking away functionality/options. (IOW, it used to be that you could play with historical txns but introduce some variables such as players not missing seasons or retiring according to history.