Curious Case of a Minor League Player

[SunDevil]

Settings: MLB Quickstart All in One

Scouting Accuracy: Very Low
Top Prospect List: Dynamic
Lineup Selection: Sabermetric
Rating Weight: 35
Current Year Stats Weight: 45
Previous Year Stats Weight: 15
2 Years Ago Stats Weight: 5
Create and Maintain Hidden Players: Yes

This player is listed as one of the top prospects at the beginning of the 2012 season.

Ratings: (OSA)


Editor Ratings: (Commissioner Mode)


Stats: (April - June 2012)


April 6th, 2012 - Demoted from AAA to AA

He was able to play one game before being demoted from AAA. Why can the AI demote after only one game?

May 27th, 2012 - Demoted from AA to A

In AA he was allowed to play 45 games and take 189 at-bats hitting .344 before he was again demoted to Single-A. What did the AI not see or not like to demote him again?

June 27th, 2012 - Demoted from A to S(A)

Now playing on his third team in Single A after being demoted twice already, he plays in 26 games and get 100 at-bats hitting .320 before again the AI decides to demote him to Short Season Single A. Again what is missing from this player that the AI feels like he is not holding his own?

June 30th, 2012 - Injured Placed on 60 Day DL (Fractured Ankle out 4 months)

He only got to play in 3 games before fracturing his ankle and being out for 4 months. Here is a player that is considered one of the top minor prospects in the game, who has been demoted 3 times from the beginning of the season to the all star break and playing in a useless game down in Short Season Single A where he gets injured for 4 months which is pretty much the rest of the season at this point.

This is an AI controlled team handling one of the listed top minor prospects in the game.

Are the stats being considered? Even with Scouting being set to Very Low, what is it about the hidden ratings which are amazing and the stats that this player has put up on each level incur the decision from the AI to demote him?

Thanks for any feedback.

[Craig Scarborough]

Yeah, I don't have an answer for you. Only thing I can think is that the AI doesn't see him at the AAA level yet. How old is the prospect?

Do his demotions happen to coincide with injuries at the level below? I'm thinking...perhaps the AI sees a player at AAA (at league start-up) who doesn't (according to the AI anyway) belong there. He then looks for the first opportunity to send him down. The AI doesn't just send him directly to short season A (where it figures he belongs), but must go through a step-by-step process (meaning, he's only sent down one level, and only when an opening arises due to injury or some other event).

Thinking about this further, I would think this is probably a result of using the quickstart, and having an initial disparity between what the real MLB/minor leagues looks like and an AI that doesn't have any idea how to account for that initial player placement. From an AI perspective, I don't know how you'd solve this. It looks to me like it will stop occurring as you move further away from the initial season.

[BIG17EASY]

Just out of curiosity, is the player in question Bryce Harper? Sure looks like it based on ratings and the minor league teams being in the Nationals' organization.

I don't have an answer, either. But I would probably guess that it's the result of a combination of factors like the AI potentially favoring veterans over prospects, his age (Harper is 19, which falls in line with typical Class A and Short-Season A players), and the needs of each teams' rosters.

Again, that's just a random guess. Others with much more OOTP experience than myself can probably give you more educated opinions/answers.

Maybe take a look at how the game handled guys like Starlin Castro. Or Manny Machado would be a good comp -- he's also 19, one of the game's top prospects, and playing in Double A for the Orioles in real life. Look and see how the AI handled him.

I did a sim season with the MLB quickstart last night, so I can see if something similar happened with Harper when I get home tonight.

[SirMichaelJordan]

I think it's you're AI Evaulation settings.

Being that current year stats are prorated, Bryce doesn't have stats from 3 years ago and his last years stats are only a 15% weight so therefore his stats are basically being ignored and his current ratings are the deciding factor on promotion/demotion vs a player with a strong 3 year history for stats to look at.

I have AI Evaluation at 25% for all and Bryce was starting for the Nats by the All-Star break.

But I am a newb so take my post with a grain of salt.

I don't even know if minor league stats are looked at.

[SunDevil]

Settings: MLB Quickstart All in One

Scouting Accuracy: Very Low
Top Prospect List: Dynamic
Lineup Selection: Sabermetric
Rating Weight: 0
Current Year Stats Weight: 42
Previous Year Stats Weight: 40
2 Years Ago Stats Weight: 18
Create and Maintain Hidden Players: Yes

Ran the first season again, this time changing the AI Player Evaluation Options.

April 14th, 2012 Demoted from AAA to AA

He actually stayed up for 10 games this time and still held his own hitting .310 but the AI still decided to send him down to AA.

June 28th, 2012 Promoted from AA to MLB

This time around the AI left him in AA for 65 games and accumulated 262 at-bats and hitting .389. On June 28th, the AI promoted him to the Major League team.

MLB Team

So in the same season that MLB quickstart league started, one of the best minor league prospects was promoted to the major leagues. He was able to play 55 games and rack up 163 at-bats with a .331 average.

So with change AI player evaluation settings with a more stat focus setup, the AI demoted him less in the minors and eventually brought him to the Majors to help the team. This is a very small sample size and by no means do I think that I fixed anything, but it interesting how we can directly affect how the AI treats one of the best listed prospects in the game.

[SunDevil]

Settings: MLB Quickstart All in One

Scouting Accuracy: Very Low
Top Prospect List: Dynamic
Lineup Selection: Sabermetric
Rating Weight: 25
Current Year Stats Weight: 25
Previous Year Stats Weight: 25
2 Years Ago Stats Weight: 25
Create and Maintain Hidden Players: Yes

Ran the first season again, this time changing the AI Player Evaluation Options to what SirMichaelJordan recommended in another thread.

May 15th, 2012 Promoted from AAA to MLB

This time the AI left him in AAA to start the season. He then played in 36 games collecting 145 at-bats where he hit .331 So the same player with the same abilities was left in AAA and not moved around the minors and he blossomed.

MLB May 15th, 2012 -> All Star Break

So far, he is making the most of his promotion to the big leagues. He has played in 43 games and collected 120 at-bats and is hitting .367! Not bad, especially in the first example this same player was in down Short Season A ball with a fractured ankle, which is not helping the major league club at all. In this example he is healthy of course but no where near going down to the Single A club.

MLB End of Season

So at the end of the season, he had a solid rookie year. Playing in 104 games and getting 347 at-bats, hitting a .285 and playing center field. A much better outcome, then being demoted three times to Short Season Single A ball and sitting on the 60-day DL with a fractured ankle.

Now of course this whole situation might be caused by setting the scouting accuracy to "Very Low" thus blinding the AI that originally was weighting the ratings by a value of 45. And in this example and the previous by putting more weight in the stats, the AI is able to make a smarter decision about player movement in the minors.

But for the people who like to turn off ratings and make scouts act more like real life counterparts, here are some examples of what to use for the AI to better compete in an environment when ratings are not the be all end all.

[BIG17EASY]

I checked my MLB quickstart season that I simmed last night and Harper was called up to the MLB part way through the season and did fairly well as a part-time player. But I hadn't touched the AI evaluation settings at all.

[SunDevil]

Originally Posted by BIG17EASY

I checked my MLB quickstart season that I simmed last night and Harper was called up to the MLB part way through the season and did fairly well as a part-time player. But I hadn't touched the AI evaluation settings at all.

What is your Scouting Accuracy setting? Do you have Create and Maintain Hidden Players turned on?

Now of course this whole situation might be caused by setting the scouting accuracy to "Very Low" thus blinding the AI that originally was weighting the ratings by a value of 45. And in this example and the previous by putting more weight in the stats, the AI is able to make a smarter decision about player movement in the minors.

[bwburke94]

OK, it's pretty obvious the "mystery player" is Harper. Now why didn't the OP say that at the start?

[BIG17EASY]

Originally Posted by SunDevil

What is your Scouting Accuracy setting? Do you have Create and Maintain Hidden Players turned on?

Everything is at the default setting. So scouting I believe is either normal or average, whatever that choice is, and hidden players are on.

[SunDevil]

Originally Posted by bwburke94

OK, it's pretty obvious the "mystery player" is Harper. Now why didn't the OP say that at the start?

Because it didn't matter. I wanted to focus on the AI handling one of the top rated minor prospects, and did not want to get into a discussion about a specific real life player.

[SunDevil]

Originally Posted by BIG17EASY

Everything is at the default setting. So scouting I believe is either normal or average, whatever that choice is, and hidden players are on.

So from what I have seen and what I have read before, with less accurate scouting, the AI player evaluation values should be taken into account. So since you left everything as default, I'm glad that it worked for you. But I wanted to setup a league that was a little more a challenge that uses scout accuracy as "Very Low" and the AI evaluating players was being affected with the default evaluation values. So I just wanted to find player evaluation settings that did a good job in not moving minor league players all over the minor leagues and instead promote and groom them more efficiently. From the 3 seasons I have run the 25/25/25/25 player evaluation settings in tandem with the very low scouting accuracy works for me.

[BIG17EASY]

Originally Posted by SunDevil

So from what I have seen and what I have read before, with less accurate scouting, the AI player evaluation values should be taken into account. So since you left everything as default, I'm glad that it worked for you. But I wanted to setup a league that was a little more a challenge that uses scout accuracy as "Very Low" and the AI evaluating players was being affected with the default evaluation values. So I just wanted to find player evaluation settings that did a good job in not moving minor league players all over the minor leagues and instead promote and groom them more efficiently. From the 3 seasons I have run the 25/25/25/25 player evaluation settings in tandem with the very low scouting accuracy works for me.

Good work. I'm not as seasoned at OOTP as you clearly are, which is why I left the settings on default. I'm continuing to sim with a few different settings before starting a career after the patch with updated rosters comes out this weekend. I'm not sure I'm going to go with scouting very low, but I might try it on low. I'll have to see what that does for the AI's handling of prospects. If I see anything interesting, I'll post it here.

[OutS|der]

With scouting the ratings come from the scouts so the lower the accuracy of the scouting the more the ratings will be off from what the really are.
So the lower the ratings weight the better the AI will evaluate players.



So all you showed is with Scouting Accuracy: Very Low the AI needs the rating weight lower then 35 to function properly.
Not really that interesting as that is how i assumed it would work, plus I've always played stats only 0/67/22/11 so i haven't really seen players being treated like that, though with ratigs on 35 it does seem a little excessive that he would of been sent down with those stats

[SunDevil]

Originally Posted by OutS|der

....ratigs on 35 it does seem a little excessive that he would of been sent down with those stats

Which was the exact reason why I started this thread. All my other posts in this thread were to see if I could find values for the AI Player Evaluation settings to see if the AI handled a top minor league prospect better than the initial post.