Forecast 2009: The 2008 Baseball Player Prognosticator of the Year

 

 

Although some of the projection services cover as many as 25 rotisserie categories, I am only interested in those covered by the most popular game in Rotisserie *5x5* (AVG, HR, RBI, Runs, Stolen Bases and W, ERA, WHIP, Strike Outs, Saves).

To develop a comparison among the services I use the commonly understood statistical method of "Correlation" to determine dependency of actual player results on the services' projections. Correlation is a popular statistical technique that is employed to show strength of relatedness of two variables. The measure of correlation is called 'Correlation Coefficient", and its value lies in the band of  (-1)-0-(+1). A value closer to -1 or +1 suggests strong inter-dependence of the variables. In this exercise, we measure the correlation of each services' 2008 preseason projections to the actual player results.

 

 

 

For those new to my analysis, there are 2 sets of projections which I compare. The batter’s projections and those of the pitchers. Among the batters I took the top 150 hitters from the 2008 season. Among the pitchers I took the top 100 pitchers from the 2008 season. Why a 150/100 split? It serves as a decent relationship to a 10-12 team rotisserie league with a 14/9 hitter/pitcher split.

Below are the correlation results for each of the major rotisserie categories that was measured (the higher the number, the more relevance):

BATTING PROJECTION RESULTS

2008 Baseball Baseball Fantistics Marcels Roto ZIPS
  HQ Prospectus     World  
ABs 0.542 0.500 0.595 0.262 0.512 0.475
             
HRs 0.740 0.765 0.741 0.742 0.715 0.749
             
RBI 0.679 0.618 0.709 0.638 0.641 0.640
             
Runs 0.551 0.570 0.576 0.427 0.554 0.557
             
BA 0.576 0.543 0.510 0.516 0.540 0.499
             
SB 0.846 0.843 0.839 0.709 0.859 0.836

Baseball HQ took the batting average (BA) category by a significant margin this year. Batting Average has historically been the toughest category to predict, as evident by the lower correlation among the prognosticators. Fantistics correlated better overall in two of the categories RBI and Runs. They were significantly better than the rest of the field in RBIs as they were the last time we ran this analysis. Baseball Prospectus took the highly correlated HR category. And the Stolen Base category, which empirically shows the highest correlation, was taken by Rotoworld. New comer Marcels and ZIPs didn't fair so well in the batting categories and part of this can be explained by the lack of correlation with projecting the number of player At Bats. Marcel and ZIPs are a free projections service and understandably don't update their projections/playing time situations as frequently as the other "pay" services listed.

When I averaged all of the correlations for 5 Rotisserie batting categories, Baseball HQ came out ahead overall. Although they only correlated the best in one category, they were close enough in the other categories to combine for the title (see averages below).

PITCHING PROJECTION RESULTS

 2008 Baseball Baseball Fantistics Marcel Roto ZIPS
  HQ Prospectus     World  
IP 0.763 0.712 0.855 0.608 0.802 0.697
             
W 0.571 0.565 0.652 0.503 0.587 0.559
             
S 0.564 0.587 0.589 0.529 0.568 -
             
ERA 0.217 0.266 0.267 0.353 0.289 0.265
             
WHIP 0.330 0.295 0.313 0.348 0.255 0.418
             
K 0.533 0.516 0.550 0.403 0.473 0.473

Similar to prior seasons, Fantistics came in very strong with their pitching prognostication. Similar to the 2007 analysis they nailed three of the categories (Wins, Saves, K) and the Innings Pitched category. Marcels took the ERA category, while ZIPs took the WHIP category for the second time since we've run this analysis. It should be noted that WHIP is historically the second least correlated or predictable category, and ZIPs was significantly better than the rest of the field. Similarly Marcels did very well with the ERA category using a simple regression model.

With 3 of the 5 categories well correlated, it's not much of a surprise to see Fantistics come up on top with the best overall correlation for 5 Rotisserie pitching categories. Similar to prior seasons they have a lock on projecting strikeouts, earning them another title win in the pitching area.

The 2008 Baseball Player Prognosticator of the Year is........

...Fantistics. For the forth year since I've been running this analysis, Fantistics has won out overall in each of these years.

  Baseball Baseball Fantistics Marcel Roto ZIPS
  HQ Prospectus     World  
Hitter 0.678 0.668 0.675 0.606 0.662 0.656
             
Pitcher 0.443 0.446 0.474 0.427 0.434 0.429
             
Total 0.561 0.557 0.575 0.517 0.548 0.542

As I've said in the past, I'm not sure what they are using for a crystal ball, but it's been an impressive run for them. If anything, Fantistics has proven that there is predictability in forecasting player projections, something beyond a random walk.

Having won 3 titles in the last 5 years in my highly competitive league (The Faux GM League), I am often asked which player projections service I use for my fantasy baseball quest. I usually rely on several sources for my fantasy information, but have relied on the Fantistics draft projections since 2001 (no need to explain why). I sprinkle in some Baseball Prospectus during the season, with a close eye on the Rotoworld web site for breaking player news. I'm also a fan of Baseball HQ and Baseball Notebook, as I subscribe to then intermittedly as well.  

With good reason, several of these sources (Baseball HQ, Baseball Prospectus, & Fantistics) have already earned the respect of some Major League Baseball organizations. They have certainly earned my respect as well. The level of insight found on the web over the last few years has increased exponentially, and so has the challenge of winning a competitive league. Fielding a championship team is a 9 month commitment that extends beyond selecting the right players come draft day. However if you're not selecting the right players on draft day, it's quite an obstructive mountain to climb.

Here's a shout out to the guys over at FantasyBaseballCafe.com and BaseballForums.com for their help in compiling the projection sources. Have a big fantasy season, see you next year.

-Dave LaDue AKA RotoDaddy

 

 

Copyright © 2009 RotoDaddy Baseball All rights reserved