Nolan Ryan rated a real 1.5 stars in 1982, how does this happen?

[Brad K]

In my America's team save Nolan Ryan has a real 1.5 star rating in 1982. I looked under the hood and this is not a scouting error. Settings are three year recal dev on TCR 100 historical minors default make bad for 23, whatever that actually is. (This is on 23 because I do not change horses mid stream. Whatever version a save starts on is where it stays.)

OOTP lists his WAR for 81, 82, and 83 as 4.7, 5.4, 4.7 for an average of 4.9. How is this a 1.5 star SP?

Note on the screen shot of the editor his current ratings support the 1.5 star rating and his potential ratings are much higher and in line with a really good pitcher.

Predicted ERA in editor is 5.42. Historical ERA for 84, 82, and 83 1.69, 3.16, and 2.98.

I suppose this is TCR. But I have to ask, does setting the TCR number affect how much a player changes or how often players change? Is this effect of TCR, if that's what it is, at all reasonable?

[Brad K]

I received a PM suggesting I run a rescout with accuracy set to 100%. I did. Ryan is now rating 3.0 stars.

However, his actual real used by the engine ratings have changed. A rescout should only result in the scout changing his ratings not in a change in the number the engine uses.

But here it is.

FB from 61 to 152
Curve from 50 to 122
Circle curve from 41 to 93
ERA from 5.42 to 3.40

How can a rescout result in a player's real ratings changing?

See first post for original figure. Rescout figures below.