Major Issues with Conversion from 25 to 26

[TheSpeakerfortheDead]

After checking what my online league looked like on 26 and noticing some major issues with player splits I decided to do some testing. I started a live start on 25 and converted to 26 and these are the results.

For this test, I am analyzing the sum change in a split of editor ratings. (sum of Gap, Power, Eye, Ks and BABIP for hitters / Movement and Control for pitchers) I am not combining both splits. Out of the 13,000 players 461 had significant movement of 100 points. This is over 3% of players becoming a completely different player. Over 99% of hitters showed rating movement. While over 40% of hitters received significant ratings movement. Pitchers do not seem to be as extreme of an issue. 10% of pitchers are receiving moderate change but still over 85% of pitchers are displaying ratings changes. Most of these examples seem to be massively exaggerated splits that cannot be justified due to the minor changes of the 1-550/1-600 ratings scale. None of the 1-200 or 1-250 ratings show any changes.

[TheSpeakerfortheDead]

I should add to this as well that it is consistent. The variance isn't random, it's the same players by the same margins every time.

[LansdowneSt]

I saw this in the Master Change List:

Considerations when importing from OOTP 25

Ratings:
Player splits may shift slightly to adjust to engine balance changes
Notably, relative ratings
Player overall movement rating has reduced importance of PBABIP
Overall Movement rating should be closer to a players’ HR Allowed rating
Contact rating calculation has shifted slightly
Player Overall Ratings calculations have shifted
Scouting Task List will be deleted upon import due to a change in how it is handled internally
IAFA Invitees are reset upon import
League engine modifiers may shift on import, to balance new default values. On import, leagues will have their modifier multiplied by 3 as the game engine is balanced around 1.0 as the default value now.
Note that some values we do not shift on import, like errors. If your modifiers previously increased errors, you may want to shift

[Lukas Berger]

Adjusting the splits on some players is intentional as in previous versions they were not really distinct enough in many cases.

Now, that being said we can always take a look through the logic to see if we need to tone this down a bit on import. It might even be possible to provide an option to turn this off when you import. Not a promise, but it's something we can consider.

[kidd_05_u2]

The player in the screenshots is a real player. Real players already had good splits in OOTP 25. The problem was with fictional players.

If the import process is adding more split to real players, then that is the problem.

[TheSpeakerfortheDead]

Originally Posted by Lukas Berger

Adjusting the splits on some players is intentional as in previous versions they were not really distinct enough in many cases.

Now, that being said we can always take a look through the logic to see if we need to tone this down a bit on import. It might even be possible to provide an option to turn this off when you import. Not a promise, but it's something we can consider.

Honestly Lukas if it was maybe at most 5 one way and 5 another that's still a massive shift in talent. But when there's a bunch of examples of 15 one way 10 another, it's way too extreme. It's also just not consistent with players splitting up strong side and down weakside either. There is a large number of players that get worse against their strong side and improve versus their weakside. It seems like whatever was changed is not doing at all what was intended to amplify splits.

I'm kinda confused why it needs to happen on importing a league at all? You're screwing with people's set rosters. Why not just have players generate on 26 to be much more splitty and it'll fix itself anyway overtime? When we're talking a player across the board losing or gaining 20% of their ratings on a split a lot of times as a player with visually no splits. It just doesn't seem right at all.

[TheSpeakerfortheDead]

Originally Posted by LansdowneSt

I saw this in the Master Change List:

Considerations when importing from OOTP 25

Ratings:
Player splits may shift slightly to adjust to engine balance changes
Notably, relative ratings
Player overall movement rating has reduced importance of PBABIP
Overall Movement rating should be closer to a players’ HR Allowed rating
Contact rating calculation has shifted slightly
Player Overall Ratings calculations have shifted
Scouting Task List will be deleted upon import due to a change in how it is handled internally
IAFA Invitees are reset upon import
League engine modifiers may shift on import, to balance new default values. On import, leagues will have their modifier multiplied by 3 as the game engine is balanced around 1.0 as the default value now.
Note that some values we do not shift on import, like errors. If your modifiers previously increased errors, you may want to shift

I've seen this as well. But these shifts are not slight and they are in the editor, nothing to do with relativity. They also are not anywhere near a pattern of amplifying platoon splits as a lot of times it has the opposite effect.

[Xtoforas]

Originally Posted by Lukas Berger

Adjusting the splits on some players is intentional as in previous versions they were not really distinct enough in many cases.

Now, that being said we can always take a look through the logic to see if we need to tone this down a bit on import. It might even be possible to provide an option to turn this off when you import. Not a promise, but it's something we can consider.

If we want to make new fictional players splittier to support the generally historical trend I think that's great. If it's toned down to reduce the impact on existing players that's great too. If we're rejiggering an entire online league, then it's sad.


The same thing happened when STU got rebaselined on the 24 to 25 conversion. It'd be better if the impact to please the engine and design goals were reduced on import to allow for organic growth of online leagues over time rather than a tumultuous reorg every time an online league upgrades to a new version.

[Lukas Berger]

Originally Posted by Xtoforas

If we want to make new fictional players splittier to support the generally historical trend I think that's great. If it's toned down to reduce the impact on existing players that's great too. If we're rejiggering an entire online league, then it's sad.


The same thing happened when STU got rebaselined on the 24 to 25 conversion. It'd be better if the impact to please the engine and design goals were reduced on import to allow for organic growth of online leagues over time rather than a tumultuous reorg every time an online league upgrades to a new version.

We're discussing if it might be possible to give you the option to choose adjust splits etc on import or not. Not sure we're at the point of actually being sure we'll be able to do this, but it's definitely under consideration.

[TheSpeakerfortheDead]

Tested conversion on the beta patch. Pitchers look good but there is still a pretty significant change on 2.4% of hitters and a whopping 35% of hitters are experiencing significant change to one split or another.

All of the players that have a significant change have it to their vL split which ranges from -87 to +111. The vR split has a much more reasonable range from -35 to +31. I think if the vL range was similar to the vR range this would solve the problem. (I still like the toggle solution a lot more.)

[TheSpeakerfortheDead]

Any updates on this? Still not sure why we need to rewrite players and don't just let the development engine add more splits and solve not having enough splitty players that way.

It just really sucks for any competitive online league or any long term save to rewrite a bunch of players.

[FlexD]

And that's the biggest beef I have when converting from version to version, player ratings change, not sure why a new update can't be pushed out without the change in plyer ratings. The switch from 24 to 25 was a treat to player values, dont even want to know what the jump to 26 will bring