[Mick]

Actually, I think I might put more stock in a 25 year old's recent plate appearances than year old or two year old MLB efforts.  His prior best AAA stat was .266/313/.452 -- and now he's hitting 115 points higher.  Maybe he improved, maybe the training/coaching got better, maybe he benefitted from a change of scenery.  Or maybe not.  We'll find out.  Seems as if you have him at a ceiling of .239.

That last sentence is a deliberately unfair summary of what I said.

A projection is not a ceiling. A projection is a projection, meaning the median projected outcome. His ceiling could be as much as 1-2 standard deviations above his projection, just as his floor may be 1-2 standard deviations below his projection. And over a short period the variance could be even larger.

Well, two standard deviations above or below the projection would cover 95% of the players in the league, so the projection doesn't seem very useful to me.  You stated they were six different "sophisticated" projections.  By your use of the term, I inferred that you considered them dispositive and therefore stood behind them.  So it seemed to me, my read on your comments was that the optimal end of the six sophisticated projections was .239 but you're telling me it's the mean (incidentally, the current NL leaguewide batting average is .235, same as SFG). So two standard deviations above the .239 would put him at about .288, right?  And two standard deviations below would put him at .205?  Assuming 425 +/- players, top 2.5% or about top 10 hitters and bottom ten hitters among position players.

[Mick]

Actually, I think I might put more stock in a 25 year old's recent plate appearances than year old or two year old MLB efforts.  His prior best AAA stat was .266/313/.452 -- and now he's hitting 115 points higher.  Maybe he improved, maybe the training/coaching got better, maybe he benefitted from a change of scenery.  Or maybe not.  We'll find out.  Seems as if you have him at a ceiling of .239.

That last sentence is a deliberately unfair summary of what I said.

A projection is not a ceiling. A projection is a projection, meaning the median projected outcome. His ceiling could be as much as 1-2 standard deviations above his projection, just as his floor may be 1-2 standard deviations below his projection. And over a short period the variance could be even larger.

Well, two standard deviations above or below the projection would cover 95% of the players in the league, so the projection doesn't seem very useful to me.  You stated they were six different "sophisticated" projections.  By your use of the term, I inferred that you considered them dispositive and therefore stood behind them.  So it seemed to me, my read on your comments was that the optimal end of the six sophisticated projections was .239 but you're telling me it's the mean (incidentally, the current NL leaguewide batting average is .235, same as SFG). So two standard deviations above the .239 would put him at about .288, right?  And two standard deviations below would put him at .205?  Assuming 425 +/- players, top 2.5% or about top 10 hitters and bottom ten hitters among position players.