Do We Need Talent Change Randomness?

[Brad K]

Seriously, do we need TCR? Isn't normal game engine randomness enough?

With ratings that match reality last season had Ron Santo in .227 with 16 HRs in a year where he really hit .300 with 31 HRs. Ron Hunt hit .317 in a year he hit .263. Frank Howard hit .302 in a year he hit .256.

This year I have Willie Stargell hitting .346 vs .237 and Ed Charles hitting .174 vx .276.

Why does there need to be any help in producing unexpected performances? It happens on its own.

[uruguru]

I think non-historical leagues need TCR. There are situations where players develop positively well out of the historical norm, and that improvement sticks.

I mean, history has shown that it's very possible for a weak power hitter (I dunno, let's say Luis Gonzalez) can suddenly blossom outside of his aging peak and become a massive slugger. And a slugger not because of RNG, but a physical improvement -- a talent change.

You can't model that with the aging curve and built-in RNG.

In fact, I suspect that much of what we generalize as real-life improvement as players age is really just talent changes. Nolan Ryan fixed his mechanics in California and went from an almost-washout to a great. That's a talent change. And then he talent changed himself again with a new pitch (the Circle Change) and added another 8 or so years to his career.

[OutS|der]

With TCR off, the player will likely keep roughly the same potential rating through his career. (With TCR on, that value will drop or change).

The development engine isn't ready to account for the talent changes that is baseball. Specially not drops in talent or potential. It can't handle a player like Adalberto Mondesí having so much potential to washing out of baseball.
There's a randomness to doing everything right to develop the player only for them to turn into barely above replacement level.
That's why we'll always need some sort of TCR

[Brad K]

Originally Posted by OutS|der

With TCR off, the player will likely keep roughly the same potential rating through his career. (With TCR on, that value will drop or change).

The development engine isn't ready to account for the talent changes that is baseball. Specially not drops in talent or potential. It can't handle a player like Adalberto Mondesí having so much potential to washing out of baseball.
There's a randomness to doing everything right to develop the player only for them to turn into barely above replacement level.
That's why we'll always need some sort of TCR

Your answer implies that there is no randomness to development. That means players will develop the same way in all saves without TCR. Is that true?

Why is there a discussion about TCR affecting potential ratings? fredbeene previously showed potential ratings don't affect future current ratings. I linked his thread recently.

Back to my original situation, why do we need TCR when normal game engine randomness already gives us performances that differ widely from real life?

[Matt Arnold]

You never "need" it, if you like how the game plays without it, you can turn it mostly off.

The main reason it's there is to have players develop different than their initial potential. When he was picked, Mike Piazza would not have had a super high potential rating. I mean, yeah, obviously if you're playing in a historical game, he will be rated based on how he actually with. But without "TCR", you'd never get someone like that, who initially starts out as essentially a nothing but develops into a superstar.

Sure, you will get players who don't develop to their full potential with or without TCR. But players will generally speaking not be able to grow past their potential without the TCR shifting that around.

[mytreds]

Call me ignorant, but I always thought TCR was used so when you play an MLB start 20 times, every time each player has a chance to develop a little different. One save Mike Yastrzemski might be a decent power hitter, in another he flames out, due to TCR. Thus, you wouldn’t be bored seeing the same players develop exactly the same way over and over.

[uruguru]

Originally Posted by Matt Arnold

The main reason it's there is to have players develop different than their initial potential. When he was picked, Mike Piazza would not have had a super high potential rating. I mean, yeah, obviously if you're playing in a historical game, he will be rated based on how he actually with. But without "TCR", you'd never get someone like that, who initially starts out as essentially a nothing but develops into a superstar.

Yeah, this is why TCR imo is essential for fictional leagues. For historical replays, having TCR make a major impact on a player that didn't occur in real life would obviously break realism for a lot of players. That's why we have recalc.

TCR does for fictional leagues does what recalc does for historical leagues... it breaks players out of that formulaic aging curve that is really just supposed to be an approximation anyway.

[Brad K]

Originally Posted by uruguru

Yeah, this is why TCR imo is essential for fictional leagues. For historical replays, having TCR make a major impact on a player that didn't occur in real life would obviously break realism for a lot of players. That's why we have recalc.

Who says people who play recalc want output to match historical? I want players to have a plausible career in OOTP, not one that matches RL.

The changes in the default settings in 25 show a trend toward helping people who want to watch and see how close the game gets to RL. Is this something that was done because the data collected by OOTP on how people play shows this was what they were doing? Or is it because someone came in and decided this is how people SHOULD want to play? I think the second.

And what does disabling TCR really do in recalc leagues? Due to the game's randomness, players already have season performances higher or lower than they ever did in any season RL. That was the point of my first post.

All the talk about individual players... it doesn't matter. They already have seasons where they perform well outside plausibility.

[Brad K]

Originally Posted by mytreds

Call me ignorant, but I always thought TCR was used so when you play an MLB start 20 times, every time each player has a chance to develop a little different. One save Mike Yastrzemski might be a decent power hitter, in another he flames out, due to TCR. Thus, you wouldn’t be bored seeing the same players develop exactly the same way over and over.

People think different things about TCR. That's because there is no official explanation about what it does. And its flawed in concept. There's and adjustment. What does it adjust? Frequency or severity? Actually there should be a setting for both.

And what about the R? Random? Not when it drives Jim Bunning into a bull pen mop up man five consecutive years.

BIG question. How can people insist TCR is necessary when they don't know what it does?

[uruguru]

Originally Posted by Brad K

BIG question. How can people insist TCR is necessary when they don't know what it does?

Why do you think no one knows what it does?

I have literally created fictional players in earlier versions of the game (23, I think) within a historical league and closely monitored the changes in their potential ratings over the course of a season at both low and high TCR ratings -- just so I could understand what the setting did and what I preferred

Do I know exactly how it works? No. But do I have a good idea of what it does? I believe so.

[Brad K]

Originally Posted by uruguru

Why do you think no one knows what it does?

I have literally created fictional players in earlier versions of the game (23, I think) within a historical league and closely monitored the changes in their potential ratings over the course of a season at both low and high TCR ratings -- just so I could understand what the setting did and what I preferred

Do I know exactly how it works? No. But do I have a good idea of what it does? I believe so.

You snipped the background information in the quote for your response. What you snipped is relevant and not addressed by your response.

[thenewchuckd]

Originally Posted by Matt Arnold

When he was picked, Mike Piazza would not have had a super high potential rating. I mean, yeah, obviously if you're playing in a historical game, he will be rated based on how he actually with. But without "TCR", you'd never get someone like that, who initially starts out as essentially a nothing but develops into a superstar.

The issue is, under the standard game ratings, you've gone to a system where the most likely scenario for superstar development is the Mike Piazza scenario.

This is especially problematic right now for high school and international players.

I get that Mike Piazzas happen - guys projected with lower talent end up being superstars. I get the most guys selected in the 1st round won't make it to the majors. On the other hand, the superstar pool should more often than not come from the high projected players.

I just did a sim with 100% scouting on standard settings otherwise, drafting and tracking pitchers and their ratings. I liken the 1st look at ratings similar to the first look from the development lab. Gives you a feel, a starting point but 99.9% chance his final rating will be way different. Guys drop from top 10 prospects all the way out of the top 200 without an injury or anything meaningful to derail them. It could happen in a very short timeframe as well.

You're never going to find a can't miss prospect like a Jackson Holliday.

[fredbeene]

Still waiting for a manual to describe the actual changes.
Seriously, we have seasoned players here that cannot clearly define how it is to be used. From what i can tell the mechanism of what is supposed to do has changed over versions.
Are there overrides so certain players can avoid me affected my TCR?
I would like my leagues to ensure Ron Santo plays close to real life and not take the chance that TCR will drop him.
1.25 setting would mean 25% of the players, we dont know who, will be affected. However because Santo is override then I know for sure he won't be a player affected.

In another sim, I just want fidrich and rico carty to play injury free. There is no recalc that will allow fidrich to continue on.

[thenewchuckd]

I think this article gives some ideas of how to make the game more realistic:

https://content.iospress.com/article...tics/jsa200586

Lots of interesting stuff there, although a few takeaways & really in my mind this game is not even close to realistic until it achieves this. I feel like previous versions of the game got this to a degree, I'm not sure how we got to this version and it got so bad:

-a solid majority of signed 1st round picks make it to the majors
-the average career WAR of these players is a respectable 6.93
-the drop-off after the first round (for both making it to the majors and career war) is massive. It scales for every round there-after as well
-somewhere between the 6th-10th round it becomes a complete crap shoot

I'm sure if you took the top 10 and top 5, you'd get similar scalable findings. Like, most superstars are detected and found in this range.

Really, I'm not sure if you can do that in this version, although eventually you need to get to a more probabilistic TCR. Instead of a complete crap shoot TCR, that treats all players like they're 10th rounders or later.

Edit: reading the article further, it has a chart of career war for groupings in the 1st round. Exactly as I said, career WAR for picks 1-5 is much higher (13.29), diminishing until pick 16, where it levels off (although still generally decreasing as the picks advance)..

[uruguru]

Originally Posted by Brad K

You snipped the background information in the quote for your response. What you snipped is relevant and not addressed by your response.

ok, well first of all the Jim Bunning issue almost certainly seems related to stamina (a low stamina would disqualify him from starting) or possibly a negative TCR that made him a mediocre pitcher. None of that would happen if you are playing with recalc but, if you are not, that's the risk you choose to take.

The other issue was about wanting detailed precision to adjust TCR or even to know about it. That is irrelevant to the "BIG QUESTION" you asked at the end of your post that I answered. There are lots of valid reasons why developers don't specify exactly how underlying mechanics work in strategy games. One of them is specifically every time you explain how something works in a video game, you open yourself up to criticism from armchair quarterbacks who insist it should be done differently, and there are precisely zero ways to implement any feature that will shield you from that criticism.

If you don't like how something works, explain why it feels wrong. When possible, provide evidence of the behavior you think is aberrant. But you can't ever expect devs to give you detailed information on how an underlying mechanic works, especially one that they may tweak from time to time. Nobody owes you a screenshot of their source code.

[uruguru]

Originally Posted by fredbeene

In another sim, I just want fidrich and rico carty to play injury free. There is no recalc that will allow fidrich to continue on.

Fidrych played through 1983 (age 28) and was washed out by that point. If you use recalc and turn off retirement, that is the baseline for his age-29 season in 1984 and it's not good.

If you want Fidrych to play well after his rookie season as if he had never been injured (which the recalc will reflect), you can do that easily if you are simply trying to make a special case for him. Here's how:

1) Set recalc to whatever setting you like (1-year, 3-year, 5-year, w/e) and set the TCR to its lowest setting (1).

2) In Fidrych's rookie season (74 or 76, depending on if you have minors), CLONE Fidrych. Then give the original Fidrych a career-ending injury.

3) Put the Fidrych clone in the Detroit system. Since he is technically a fictional player, not a historical one, the OOTP development engine will develop him according to his potential. He will be untouched by the recalc.

There. You are done.

[Brad K]

Originally Posted by uruguru

ok, well first of all the Jim Bunning issue almost certainly seems related to stamina (a low stamina would disqualify him from starting) or possibly a negative TCR that made him a mediocre pitcher. None of that would happen if you are playing with recalc but, if you are not, that's the risk you choose to take.

I'm playing with recalc. His WAR for 1964 - 1967 is 5.5, 8.0, 9.0 and 7.8. Each year TCR drove him to a 2.0 or 2.5 star rating as a SP and he didn't bounce back higher than 3 stars before falling again. This is inconsistent with several things you have said about how TCR works.

[Brad K]

Originally Posted by uruguru

If you don't like how something works, explain why it feels wrong. When possible, provide evidence of the behavior you think is aberrant. But you can't ever expect devs to give you detailed information on how an underlying mechanic works, especially one that they may tweak from time to time. Nobody owes you a screenshot of their source code.

Where have I asked for that? No where. But perhaps a better phrase for what I am asking would be what does it do.

[uruguru]

Originally Posted by Brad K

I'm playing with recalc. His WAR for 1964 - 1967 is 5.5, 8.0, 9.0 and 7.8. Each year TCR drove him to a 2.0 or 2.5 star rating as a SP and he didn't bounce back higher than 3 stars before falling again. This is inconsistent with several things you have said about how TCR works.

My first question would be.... if you are using recalc, then why do you have TCR also on? What is your TCR setting at? (from 1 to 200)

What is your recalc range? 1, 3 or 5 years?

Also what is your AI evaluation set at? When you are talking about star ratings for a player you have to remember that you are not necessarily talking about the internal ratings of the player.

Jim Bunning from 1964 to 1967 was aged 32 to 35. It is possible that what you are seeing might not be TCR but some sort of age-related decline by the OOTP engine (which you have turned on if you are using TCR).

Is Bunning on the same team in OOTP as in real life? The effect of defense on pitching is significant and you have to rule that out as well.

Is your league setting set to 5-man rotations or 4-man rotations? Strict or Start Highest Rested? Because Bunning in that period was pitched in a 4-man strict rotation, which would give him about 5-7 more starts than OOTP in a 5-man (7 if strict, 5-6 if start highest rested). That league setting alone would give his real-life WAR a 20-25% advantage over his OOTP WAR.

And keep in mind that WAR is calculated a bit differently in different places, so I'm not sure how comparable OOTP WAR is to whatever WAR you are referencing. FIP might be a better comparison.

There are a lot of things to consider before jumping to the conclusion that the fault lies in TCR. It might be!

[OutS|der]

I've always assumed TCR is a dice roll for +/- of any players ratings and potential. Sliding scale based on age, so bigger rolls on potential when young and actual ratings when older.
With other every other factor you can think of added to the formula.