I tested my offensive WAR projection on Derek Jeter. Using just the stats from the first 9 seasons of his career I projected the rest of his career.
No projection is perfect. With most careers my projections there is a margin of error that is dependent on the number of seasons away from the current one and how many seasons were used to project, and unless the player is injured and misses the majority of the season his actual WAR should fit.
The graph above shows the absolute value of the difference of the actual WAR and my projection. As you can see the only year in which he was passed the margin of error was when he only had 17 at bats in 2013.
I tested my ptching WAR projection on John Smoltz. Using just the stats from the first 8 seasons of his career I projected the rest of his career.
Due to the fact that pitchers only play in 25-35 (starters) or 60-90 (relievers) games per season and each game has a chance to be an inaccurate measure of the pitchers worth, a pitcher projection is much harder and less accurate. For this reason I do more offensive WAR projections. When I do project pitchers there is a larger margin of error, but it is still somewhat accurate. John Smoltz's average difference was 0.4. more than twice as high as Jeter's.