The yoyo of talent changes

[Claybor]

Talents up this sim, talents down next. It's a rollercoaster ride watching players in the minors, and a bad rollercoaster at that, in my opinion. I have never liked the way ootp handled the minors, and I think it is getting worse with every version. I admit that players talents will change in the minors, and OK ootp has it setup so most will die, but should they change every other game week? Some guys go up and down so much it is just crazy.

I know just about every reply will disagree with me, and I probably wont even come back to see the fanboy brigade, I just wanted to state my opinion on a game that it totally losing it's appeal for me between the handling of prospects, and the absolutely assinine # of injuries that the designers seem to think necessary for 'realism' even when set to the lowest injury setting.

[Afino]

I'm coming into this thread to second this.

Yeah, I realize the dev engine is essentially random, even with dialing down the Talent Randomness number, but there is a LOT more ups and downs in OOTPX than prior versions, and it drives my online leagues nuts. Every sim there is development that makes ZERO sense.


And don't even get me started on actual ratings changes. This is a sample from my #1 pitching prospect. No injuries, he's not upset, this is his first professional year.

July 11, 2020
Movement versus Lefties Increases from 10 to 11
Overall Movement Increases from 10 to 11
Movement versus Righties Increases from 10 to 11
Changeup Rating Increases from 2 to 3
July 21, 2020
Stuff versus Lefties Increases from 8 to 9
September 1, 2020
Stuff versus Lefties Decreases from 9 to 7
Overall Stuff Decreases from 8 to 7
Changeup Rating Increases from 3 to 4
Stuff versus Righties Decreases from 8 to 7
September 11, 2020
Movement versus Lefties Increases from 11 to 12
September 21, 2020
Fastball Rating Increases from 11 to 12
Stuff versus Lefties Increases from 7 to 8
Overall Stuff Increases from 7 to 8
Stuff versus Righties Increases from 7 to 8
October 1, 2020
Overall Control Increases from 3 to 4
Control versus Lefties Increases from 3 to 4
Changeup Rating Increases from 4 to 5
Stuff versus Righties Decreases from 8 to 7
Control versus Righties Increases from 3 to 4
October 6, 2020
Changeup Rating Increases from 5 to 6
Stuff versus Righties Increases from 7 to 8
October 26, 2020
Movement versus Lefties Decreases from 12 to 11
Movement versus Lefties Increases from 11 to 12
November 14, 2020
Movement versus Lefties Decreases from 12 to 11

Basically, it gives me a heart attack every damn sim. But he's only gone from 8/10/3 to 8/11/4 despite ALL of that development. WHY?

[Malleus Dei]

Dial your talent development randomness down to 10 or less and dial your injuries to low (or off) and watch your complaints vanish.

[RonCo]

There is no real need for the constant talent changes. It makes no logical sense and has no gameplay value as far as I can see, and is in fact wearying to the point of being annoying at times. Opportunity for talent changes one or two times a year would seem to be the maximum required for random fluctuations, with most of these being positive in order to offset the negatives caused by injuries.

[RonCo]

Originally Posted by Malleus Dei

Dial your talent development randomness down to 10 or less and dial your injuries to low (or off) and watch your complaints vanish.

If you do this will you still see late draft picks often rise and early draft picks bust?

[Spritze]

Originally Posted by RonCo

If you do this will you still see late draft picks often rise and early draft picks bust?

What constitutes often? 1 per year, 10, 100, 1000? Enquiring minds want to know. Just by factors of 10 or so.

To me, often means 24 or so per year. To you often might mean 173 per year. The next fella might constitute often as more than once. I often find often awfully vague. So I don't often use often except when it is ofteniciously required, which is not often.

[awall5]

I have always been curious if stats drive talent changes or if talent changes drive the stats when it comes to minor leagues. Or are both completely random. What is everyone's experience?

[RonCo]

Originally Posted by Spritze

What constitutes often? 1 per year, 10, 100, 1000? Enquiring minds want to know. Just by factors of 10 or so.

To me, often means 24 or so per year. To you often might mean 173 per year. The next fella might constitute often as more than once. I often find often awfully vague. So I don't often use often except when it is ofteniciously required, which is not often.

You can't use a specific number, because the raw number would have to vary based on the size of your league. I use percentages of players drafted in any particular round--focusing heavily on the first ten rounds because I think that's the most common range people use while playing OOTP.

I can't speak for X, but in v9 and v2007 using the default settings late picks (after round 1) develop into major league players at about half the rate they do in real life (using two real life draft years in the late 80s as I remember when I tested the snot out of the game). This is a better representation of real life than any other version, but still is too low. When the random talent change level was reduced in v2007 or v9, this rate fell even further--meaning draft value was much more predictable. This is not desirable for most online leagues.

What this meant is that an online league was put in a design box such that if they wanted to remove the annoying trickle-down/trickle-up development wandering, they also needed to remove the variance in draft pick development. There really is no need for this design approach, but it is what it is.

I asked MD this because it's possible that Markus recoded the less-than-optimal approach in v9 with one that provided variable amplitudes...in other words, provide the option for fewer random changes, but make them of higher amplitude (similar to the old v5 approach) instead of the constant flow of talent changes that are 1-5 points as a general rule. This is probably a better design, anyway, because fewer random changes that are larger actually provides a better ability (IMHO) to control average talents, anyway...as well as making the game a bit more fun (again, IMO only).

But, anyway, to answer your question about what "often" is, I suggest Markus use some form of real life data to answer what "often" ought to be. My baseline, as stated above is a pair of real life draft classes that I tabulated back when I was a beta.

[RonCo]

Originally Posted by awall5

I have always been curious if stats drive talent changes or if talent changes drive the stats when it comes to minor leagues. Or are both completely random. What is everyone's experience?

Through v9, stats could drive talent changes (as could just random events and injuries). This is a big flaw, because it's a reversal of the cause and effect relationship of skills -> stats. My understanding is that the relationship between stats and talent changes has been removed in X, but that's just based on a comment from Markus at one point.

[edddgar]

I don't know, talent does randomly fluctuate quite a bit in real life as well. The sim by sim changes don't exactly make sense, but maybe it's the only way to ensure the talent progression acts as it does it real life?

Take a look at the past BA top 100 prospect lists here. Guys fall off the face of the map and guys come out of nowhere. We see it in real life and we see it in OOTP.

[RonCo]

Originally Posted by edddgar

I don't know, talent does randomly fluctuate quite a bit in real life as well. The sim by sim changes don't exactly make sense, but maybe it's the only way to ensure the talent progression acts as it does it real life?

Take a look at the past BA top 100 prospect lists here. Guys fall off the face of the map and guys come out of nowhere. We see it in real life and we see it in OOTP.

You can argue that performance changes in real life...which is modeled in OOTP as ratings (not talents/potential). But (1) there is enough random variation due to sample size to negate the need for lots of small ratings changes, and (2) there is no need at all for such an aggressive talent engine since talent/potential don't affect the results on the field.

My opinion of the root cause of why this design is in place is the misconstruing of talent/potential changes as "scouting errors." Scouting errors would be modeled in the scouting algorithm, not the development algorithm. But that's getting into some deep OOTP neep.

[RonCo]

Originally Posted by edddgar

Take a look at the past BA top 100 prospect lists here. Guys fall off the face of the map and guys come out of nowhere. We see it in real life and we see it in OOTP.

Realize that this list is a scouting projection of what the real life development engine is going to create. To get "development" right, you need to ignore scouting. Then once the development algorithm is in place, scouting becomes a "guessing" algorithm that projects what fake human beings say is going to happen.

[kq76]

Originally Posted by RonCo

My opinion of the root cause of why this design is in place is the misconstruing of talent/potential changes as "scouting errors." Scouting errors would be modeled in the scouting algorithm, not the development algorithm. But that's getting into some deep OOTP neep.

This depends on whether you view potential as an absolute max of what the player can achieve if he does absolutely everything right or if you view it as the likely future ability if the player continues on the path he is currently on. Personally, I view it as the latter in that if the player listens more to other players, eats better, trains better, parties less, etc then he could very well increase his potential or vice versa if he does the wrong things. The other is just way too set in stone for my liking. For example, you'd have no hope for those late rounders to turn into really good players because even if they reached their potential they wouldn't be good enough.

I believe it was even you who had the great line about why playing with scouting isn't necessary because there's already so much uncertainty with potential. Before that I felt like I was really missing out on not playing with scouting, but after that not so much. If Markus went the other way with certain potential then that'd be a real shame. The uncertainty, even without considering scouting, is good.

I don't really have an opinion on the original topic other than that I agree with JAttractive's response to the OP in our online league's version of this thread in that we're most likely making more out of these changes than they really are, especially in comparison to past versions.

[RonCo]

Originally Posted by kq76

This depends on whether you view potential as an absolute max of what the player can achieve if he does absolutely everything right or if you view it as the likely future ability if the player continues on the path he is currently on. Personally, I view it as the latter in that if the player listens more to other players, eats better, trains better, parties less, etc then he could very well increase his potential or vice versa if he does the wrong things. The other is just way too set in stone for my liking. For example, you'd have no hope for those late rounders to turn into really good players because even if they reached their potential they wouldn't be good enough.

Since OOTP no longer uses potential/talent as a ceiling neither of your proposed views is right. The gap between potential and rating is simply a parameter used to determine the most likely direction and rate of the player's development right now. That said, though, the potential rating is obviously very important to the engine.

I believe it was even you who had the great line about why playing with scouting isn't necessary because there's already so much uncertainty with potential. Before that I felt like I was really missing out on not playing with scouting, but after that not so much. If Markus went the other way with certain potential then that'd be a real shame. The uncertainty, even without considering scouting, is good.

Yes, I said that. And I agree with it. I don't find much need for the fake fog of war that scouts give, since there is always going to be be randomness merely due to development. That said, a well-designed development algorithm could make scouting much more fun by actually making it do what it does in real life--predict at some rate of success what is actually going to happen.

Actually, just to characterize my entire thougth pattern here, I argue that even having a "potential" rating is unnecessary...potential could be determined based on age and current skills alone (and the scouting fudge-factor put in play for those who like the veneer of scouting). This means rating progression--not some weird software parameter called "talent" would drive the engine. This is too radical for most, I guess, but I figure the real problem is either (1) I'm not good enough at describing it, or (2) people are so tied to the rating/potential concept that they can't break that mindset enough to see how simple, workable, and elegant this idea actually is.

I don't really have an opinion on the original topic other than that I agree with JAttractive's response to the OP in our online league's version of this thread in that we're most likely making more out of these changes than they really are, especially in comparison to past versions.

I agree that since the development engine is more aggressive day-to-day that there's limited value in looking at it every sim, especially if you're at a resolution higher than a 1-10 scale (and even then it's probably too much...or at least not too fun to watch).

[PSUColonel]

Originally Posted by Afino

I'm coming into this thread to second this.

Yeah, I realize the dev engine is essentially random, even with dialing down the Talent Randomness number, but there is a LOT more ups and downs in OOTPX than prior versions, and it drives my online leagues nuts. Every sim there is development that makes ZERO sense.


And don't even get me started on actual ratings changes. This is a sample from my #1 pitching prospect. No injuries, he's not upset, this is his first professional year.



Basically, it gives me a heart attack every damn sim. But he's only gone from 8/10/3 to 8/11/4 despite ALL of that development. WHY?

where do you find this screen that tracks changes? I thought there was only the player development report

[damientheomen3]

Originally Posted by PSUColonel

where do you find this screen that tracks changes? I thought there was only the player development report

neags23 made a talent development mod here

[Afino]

Originally Posted by PSUColonel

where do you find this screen that tracks changes? I thought there was only the player development report

It's a Getch mod.

Look at this player for example:

Player Report for #26 Jeff Bradford

On the right side, underneath all of his vitals, info, etc, is a link to his development history.

[PSUColonel]

Originally Posted by damientheomen3

neags23 made a talent development mod here

oh...I use a mac

[damientheomen3]

Originally Posted by Afino

It's a Getch mod.

Look at this player for example:

Player Report for #26 Jeff Bradford

On the right side, underneath all of his vitals, info, etc, is a link to his development history.

is this mod only available for online leagues?

[Cooleyvol]

StatsLab has it as well. There's a Dev History link next to every guy on the report.


Development Tracker - StatsLab for OOTPX