Baseball HQ vs. PECOTA

Posting has been a bit fluffy lately, so here's one for the stat geeks. Before the season, Doc Gram did an analysis of the projections provided by the creators of two of the most popular annual baseball books for fantasy leaguers- Baseball Forecaster and Baseball Prospectus. The purpose was not only to get two perspectives on each player, but also to determine market opportunities, based on knowing which managers in his league rely on which sources more. In the graph here, the lines are actually made up of points, one for each player, along the x-axis. The red lines refer to projections made by the folks at BaseballHQ.com. The blue lines refer to the projections Baseball Prospectus provides based on PECOTA. To normalize the two projection sets for comparison, Doc Gram kept it to those players for whom both sources projected stats, and used a third source for estimating 2007 MLB ABs. Then, a secret formula was applied the same way to both sets of projections, to create a meta score that we'll call mgSCORE, which represents the y-axis. The yellow line is a simple average of the two projections for each player. The player all the way to the left is Albert Pujols, as he achieved the highest scores for both sources; the player all the way to the right is some scrub you don't give a shit about. Anyway. When you look at the graph, the places where the blue points and red points are furthest away from the yellow line essentially represent the largest disagreements between the two expert sources. The first big blip from the left is Derrek Lee, who Ron Shandler loves and the Prospectus people only like. Players like these warranted the most attention from Doc Gram during pre-draft preparations. A few more where BP and BHQ are pretty far apart include David Dellucci, Gary Sheffield, Hideki Matsui, and Bartholemew Bonds. Doc Gram will provide some more details on some of these types players over the course of the season, using 20-20 hindsight to rip on whoever's conveniently wrong as of that day. But for now, it's interesting to note one trend in the graph. PECOTA projections seem to be more conservative than Shandler for the top 250 players or so, and more aggressive for the rest. This would indicate that PECOTA is predicting that more overall production may come this year from players who go undrafted by the fantasy world. This seems realistic given the fact that the average major leaguer is now nearly 29 - older than the average has been in years. Maybe the folks at Baseball Prospectus are thinking that this is the year some of those younger guys will start to get their chances.