- FanGraphs Fantasy Baseball - http://www.fangraphs.com/fantasy -

Fantasy Hitter Evaluation: Core Talents (Part 1)

Posted By Jeff Zimmerman On October 5, 2011 @ 4:30 pm In Meta Analysis | 11 Comments

I stated a while back that I will be working on a method to predict and evaluate player’s fantasy stats. I will be using both the player’s talent and outside factors and will also combine several data points to get the player’s final value. I will try to keep the explanation as simple as possible and let me know if you have any questions. Let’s start with hitters and once done with them, move onto the pitchers.

Categories

I didn’t want the process and results to be limited to just a standard 5×5 league. If a league was 4×4, 6×6 or 8×9 , the process to evaluate the players would be the same. To find out how common each category is in a fantasy league, I used data from CBSSportsline (thanks to Marc Normandin). They allow League Commissioners to see how often certain league categories are selected. Here is a list of the categories that at least 5% of the leagues in CBS used last year (any other category available at CBSSportsline is under 5%).

Name – % of Leagues that use that stat
Runs Batted In – 99%
Home Runs – 98%
Stolen Bases -97%
Runs- 86%
Batting Average -51%
Walks – 46%
Doubles – 45%
Triples – 45%
Singles – 43%
Strikeouts – 35%
Caught Stealing – 33%
Hit by Pitch – 32%
Hitting for the Cycle – 18%
Errors – 11%
Grand Slam Home Runs – 11%
Ground Into Double Plays – 9%
On Base Pct – 7%
Hits – 6%
Sacrifice Flies – 6%
Sacrifice Hits – 6%
Intentional Walks – 5%

First, some of these stats I will not even try to examine because of their rarity (hitting for the cycle and grand slams). The standard 5 categories take the top spots. AVG is the lowest of the 5 at 51%. I figured people may have moved onto OBP, but it is only at 7%. After the top 5, Walks, Ks, 1B, 2B and 3B are all in a row. These seem to show that about 40% of the leagues were points based. I will eventually go through all the categories, but for now I would like to concentrate on the following categories because of their limited number of inputs:

Runs Batted In
Home Runs
Stolen Bases
Runs
Batting Average
Walks (Batters)
Strikeouts (Batter)
On Base Pct
Hits

Looking at the inputs to each category, I wanted a few predictable stats to generate each subsequent stat. Here are the possible inputs into the above stats:

BB%
K%
HR/PA
BABIP
PA
Surrounding Talent
Park Factors
Batting Order
SB Attempts/times on base
Manager SB Philosophy

HR, K and BB are all easy to figure out using the rate and projected PAs (PA estimation will be covered in Part 3 along with SB). The 3 rate stats of BB%, K% and HR/PA all stabilize fairly quickly in a season.

I am going to use PA, K%, BABIP and HR/PA to get AVG. Projecting a player’s BABIP is difficult. I am trying to find the best combination of seasonal BABIP and xBAPIP and career BABIP and xBABIP to get a good prediction. I will be using guesstimated values until I find a more predictable number.

The biggest leap of faith I had was taking the inputs to get an estimation for Runs and RBIs. I have a process that I find is both accurate and simple. I will cover it in my next article. Let me know if you have any questions or suggestions until then.


Article printed from FanGraphs Fantasy Baseball: http://www.fangraphs.com/fantasy

URL to article: http://www.fangraphs.com/fantasy/zimmermanfantasy-hitter-evaluation-core-talents-part-1/

Copyright © 2009 FanGraphs Fantasy Baseball. All rights reserved.