Fantasy Draft Strategy: Searching for the K

One of my main goals in writing fantasy content is to give you real-life examples of how to put fantasy analysis into practice. As a few of us here at Fake Teams are currently in the midst of a 20-team dynasty draft that's been covered here and here already. There has been an incredible run on starting pitching thus far, whereby 19 pitchers have been selected in the first 69 picks (27.5% of the picks), including our first closer, Craig Kimbrel at #69.

I will typically try and grab an ace within the first 4-5 rounds (let's say 3-4 rounds since this is a deep league), and then wait until my offense is in place. Obviously that strategy has gone out the window as all the legitimate #1 SPs are gone. However, I've held my ground so far and have shifted my thinking now to round up all the best strikeout pitchers I can when they are of good value. Now this may be a bit of a rustic approach to looking for pitchers, as there are more sophisticated ways of looking at pitcher success these days in the sabermetric community; namely, there is FIP, xFIP, tERA, and the list goes on with all the various ERA predictors.

However, the strikeout still remains for me (as well as for the aforementioned metrics) one of the primary components of a good pitcher. If a pitcher can strike out a batter, then fantasy categories like ERA, WHIP, and K (of course), move in the right direction and there is a better possibility of a pitcher throwing a quality start (QS) or getting the win.

In this two-part post I wanted to look at what pitching stats are significant predictive indicators of an increase in K/9 rate from one season to the next. After identifying these stats, I wanted to analyze what pitchers have the best indicators from 2013 of a bump in their K/9 rate so that I could target them later in my (current) draft. In the following post I will do the same for BB/9 rates. Strikeouts are nice, but when combined with an increase in control (i.e., fewer walks), we could have something really nice going here.

Methodology

First, hats off to Steve Staude over at Fangraphs, who recently published this awesome tool that helps us look at all sorts of pitching correlations. What I wanted to know was using the default filters, which pitching statistics best predicted an increase in K/9 rate from the current season to the subsequent season, using all available pitching data from 2007-13.

First, I wanted to focus only on statistics that had medium or strong correlation, for our purposes, numbers that were either >0.3 to 1.0 or <-0.3 to -1.0. This would give us the most important characteristics that influence K/9 rate. The following statistics met those requirements:

Taking out any statistics that already incorporate strikeouts themselves (e.g., K%, swinging-strike%, FIP, xFIP, BERA, etc.), we find that there are four primary factors that correlate with predicting a K/9 rate in subsequent years:

To make this exercise more practical I thought it best to focus on the two factors that have the biggest predicative impact; namely, contact%, "the overall percentage of a batter makes contact with when swinging the bat" and z-contact%, "the percentage of pitches a batter makes contact with inside the strike zone when swinging the bat." It's not like these are rocket-science discoveries here; both make sense. If the batter is making less contact in general, and especially if he is making less contact in the strike zone, a pitcher is more likely to strike him out, and the inverse relationship is also true-the more contact a batter makes the less chances of a strikeout.

Results

I found the difference between the 2013 and 2012 for both contact% and zcontact% for each pitcher (min. 130 IPs), and then added them together to get a "total contact %" (TotCont%). The results for the best 15 starting pitchers (indicated by negative values), and the worst 15 starting pitchers (indicated by positive values):

These are the pitching statistics that most strongly correlate with K/9 that we have. My goal was to identify pitchers who's contact% and zcontact% changed most drastically from 2012 to 2013, in order to try and predict who would have a bump in K/9 rate for 2014. As the season begins next year you can also keep an eye on pitchers who just began pitching last year (e.g., Jose Fernandez, Tony Cingrani, etc.) and monitor their plate discipline statistics to see where they might be trending-in order to trade away or try to acquire talent before the trade deadline. So, now we have a better idea about K/9 prediction. It's no slam dunk, as correlation does not mean causation, but come draft day and I've got this list in front of me, it will most definitely help me break some ties when I'm on the clock.

Follow me on Twitter @agape4argentina and opine in the comments below as to if this will be helpful for you and/or what other things you may be wondering about for the upcoming season.

More from Fake Teams

Follow @faketeams Follow @SBNationMLB

• Top 10 Fantasy Prospects for 2014: Team by Team
• AL Only: The Koji Uehara Tier
• Fantasy Baseball Draft Strategy: Deep Leagues
• Dynasty League Draft Recap: Rookie Drafts
• Oscar Taveras: Starting right fielder out of spring training?
• Playing with stats: What was your Fantasy Average?
• The 2014 All Breakout Team: AL-Only leagues
• 3 UP / 3 DOWN: 2014 Outfield Position
• Early Rankings Love for 2014