Taking 98T1, 98T2, 99T1, 99T2, 99T3, 00T1, and R1

McGuire batted:

.256
.257
.192
.234
.211
.286
.245

His average BA was .240.

The Standard deviation which should predict where about 75% of all results will fall is .031.

In other words, based on this data, 75% of leagues McGuire would be expected to bat from .209 to .271

His actual batting average that year was .278

Now the question is, is this a reasonable amount of variation to expect in this type of simulation.

If all players exhibit this range of batting averages after around 300 at bats, then I would say no.

But, perhaps McGuire is an exception rather than a typical example.

I would think the players average BA across leagues should be within 5 points of his actual BA and that the standard deviation should be about 5% of the players actual batting average. In other words, if a player batted .400 you could expect his Standard Deviation to be 20 points or from .380 to .420, but if he only batted .200, his standard deviation would be .10 and his BA vary from .190 to .210 75% of the time.


Comments?

DaveF