- FanGraphs Sabermetrics Library - http://www.fangraphs.com/library -
Posted By Steve Slowinski On February 18, 2010 @ 12:35 am In | 5 Comments
So we have all of these statistics, but when can we use them? Suppose a player goes three for three in their first game in the big leagues. Should we expect this player to continue batting 1.000 for the rest of the season? Of course not, that’d be silly. Three at-bats is way too small a sample to draw conclusions about a player, but then we’re left with the question: at what point do statistics become reliable?
There has been a lot of research done in this area by Russell Carleton (AKA: the artist formerly known as Pizza Cutter). For his most recent work, you can find his full research at Baseball Prospectus. We’ve included links and a summary below:
In case it’s not obvious, you can tell a lot more about a hitter from one year of data than you can about a pitcher. If a statistic is not included, the means it did not stabilize over the intervals that Russell Carleton tested.
Also, a quote worth remembering: “In small sample sizes, a good scout is ALWAYS better than stats.”
Article printed from FanGraphs Sabermetrics Library: http://www.fangraphs.com/library
URL to article: http://www.fangraphs.com/library/principles/sample-size/
Copyright © 2011 FanGraphs Sabermetrics Library. All rights reserved.