How do I STOP this loop of daily Sign/Release?

[StLee]

This is driving me crazy! I tried to limit minor league roster sizes. Then I made them unlimited. Then, after this team continued over signing, I went back to roster limits.

I have one level of minors, but I'd consider adding a second level if that is needed to stop the loop.

For now, I have a story to write about how Sunset Sarsaparilla is trying out multiple players to improve its franchise.

[endgame]

Yeah, you'll need to add more levels to provide room for movement or increase the roster limit for your single level. What's it set to now?

[StLee]

Originally Posted by endgame

Yeah, you'll need to add more levels to provide room for movement or increase the roster limit for your single level. What's it set to now?

At first I had an unlimited number. Then that franchise placed 71(!!!) players on its minor league team. Last season, it was Ultra-Luxe that did something similar.

After, I limited the number to 35 I believe. Now SS has dropped and added players every stinking day.

I'll look into adding a second minor league, but I don't really want it.

[Fyrestorm3]

Obligatory Curtis Borkholder reference.

[The Wolf]

Originally Posted by Fyrestorm3

Obligatory Curtis Borkholder reference.

Somebody had to do it.

[Ts-Rock]

Originally Posted by Fyrestorm3

Obligatory Curtis Borkholder reference.

THANK YOU!

that was a funny read! Great post!

[Leo_The_Lip]

Actually it should be an Adam Rosales reference:

Fixing the Waiver System | FanGraphs Baseball

Creating an accurate model of a broken function will create weird doings.

Markus got it right, but it is wrong to begin with, so don't moan about reality.

[mpejkrm]

"Mountainside Richards" is an amazing name.

[endgame]

Originally Posted by Leo_The_Lip

Actually it should be an Adam Rosales reference:

Fixing the Waiver System | FanGraphs Baseball

Creating an accurate model of a broken function will create weird doings.

Markus got it right, but it is wrong to begin with, so don't moan about reality.

Interesting take on things, Leo, and great article. Thanks for the providing that link.

[Le Grande Orange]

Originally Posted by Leo_The_Lip

Markus got it right, but it is wrong to begin with, so don't moan about reality.

While basically correct, notice what StLee showed in his screenshot was not waiver claims but signings and releases. That's a different situation from player movement due to waiver claims.

[Leo_The_Lip]

Originally Posted by Le Grande Orange

While basically correct, notice what StLee showed in his screenshot was not waiver claims but signings and releases. That's a different situation from player movement due to waiver claims.

I think it is caused by the same issue, however. That is, the AI player evaluation function. The AI does a calculation and thinks it can improve, so it signs a guy. However, other teams also make moves and then a new calculation is made. This new evaluation shows the new signing is now a bit below average, so he is cut or traded again.

There is an excellent thread of recent vintage that talked about the star system and how the overall changes in the league cause the star ratings to change. I think that also impacts the AI evaluations 'on the margin'. So when a few big stars come on or go off the DL, the resulting recalculation of what helps a team changes.

[RchW]

Originally Posted by Leo_The_Lip

I think it is caused by the same issue, however. That is, the AI player evaluation function. The AI does a calculation and thinks it can improve, so it signs a guy. However, other teams also make moves and then a new calculation is made. This new evaluation shows the new signing is now a bit below average, so he is cut or traded again.

There is an excellent thread of recent vintage that talked about the star system and how the overall changes in the league cause the star ratings to change. I think that also impacts the AI evaluations 'on the margin'. So when a few big stars come on or go off the DL, the resulting recalculation of what helps a team changes.

Then it's too simplistic.

There should be logic in the evaluation that attenuates the response ie doing nothing is a good response unless the value exceeds a threshold and applies to the top level team. The criteria for minor leagues should not be the same as for major leagues. This could explain why some prospects get bounced around too much. The minor league logic should be to get playing time for the identified top prospects, not signing a non-prospect who is temporarily better because of current ratings.

I'm glad I control my minor leagues.

[TDO]

I saw the title of this thread and was hoping in it would contain the linke of the legend, and sure enough!

[Leo_The_Lip]

Originally Posted by RchW

Then it's too simplistic.

There should be logic in the evaluation that attenuates the response ie doing nothing is a good response unless the value exceeds a threshold and applies to the top level team. The criteria for minor leagues should not be the same as for major leagues. This could explain why some prospects get bounced around too much. The minor league logic should be to get playing time for the identified top prospects, not signing a non-prospect who is temporarily better because of current ratings.

I'm glad I control my minor leagues.

I agree, however, there will always be a few cases where 'on the margin' changes cause a move to be chosen. There is no way to code a program and not have that occur, I think.

One of the things I've been wondering about lately is when scouting ratings are applied and their effect on moves. Does my scout use his (in this case average) ability to scout the minors on my minor leaguers, but then his (in this case very good) ability to scout the majors when the player is promoted? Does this change in scouting ability account for some of the behavior we are seeing? That is, does this change in scouting ability cause the overall AI to promote a poorly scouted prospect, but then see a more accurate picture in the big leagues, thus triggering a move to improve the big club?