Hitters have been striking out more than ever before. Per NBC Sports, there was an average of 6.52 strikeouts per nine innings in 2006. That figure jumped up to 8.21 strikeouts per nine innings in 2016.
The increase is probably in part due to elevated pitch velocity, but we’ve also seen more and more “elite” hitters strike out at a high rate: Mike Trout, Kris Bryant, and Paul Goldschmidt all struck out more than 130 times in 2016.
With strikeouts occurring at an all-time high rate, FantasyLabs’ K Prediction metric is thus quite important for DFS. This prediction is based on factors such as pitcher and hitter K rates, anticipated batters faced, and each player’s spot in the order. Let’s take a look at how correlated the K Prediction metric has been to DFS production.
K Prediction Percentile
Using our Trends tool, we can see how pitchers with the best and worst K Predictions have performed over the past three seasons:
Pitchers have historically produced positive value with a K Prediction in the 61st percentile or higher. The real value has come from pitchers with a K Prediction in the top-20th percentile, as those pitchers have posted the best Plus/Minus and Consistency Ratings of the group. That said, this production has come with an ownership premium on both DraftKings and FanDuel.
Other than the heightened ownership, there has been value in targeting the upper echelon of pitchers in K Prediction. Within our Player Models, we generate a specific K Prediction for every pitcher on the day’s slate. Now that we’ve established that the K Prediction metric has historically led to value, let’s take a look at where that value has come into play.
Targeting any pitchers with a K Prediction under 5.0 has historically been a -EV move. This group’s count has made up 44 percent of the total sample size, meaning we can essentially dwindle down our pitcher pool to the top 56 percent simply by fading pitchers with a very low K Prediction.
Pitchers with a K Prediction of 6.0 or higher have consistently produced value. Of course, we see the sample sizes get smaller and smaller as the K Prediction rises. This makes sense: It’s very rare to find a pitcher with a double-digit K Prediction. In fact, the full list of last names that have had a K Prediction of 10 or higher since 2014 are Sale, Kershaw, Hernandez, Fernandez (RIP), Scherzer, Bolsinger, Hill, and Darvish.
Targeting pitchers with a high K Prediction has led to a ton of production, but also a ton of ownership. Intriguingly, this ownership jump appears to happen right around a K Prediction of 7.0 on both DraftKings and FanDuel. Pitchers with a K Prediction between 6.0 and 7.0 have still consistently produced value, and their average ownership hasn’t shot through the roof. Pitchers with a high K Prediction likely have at least one well-known factor heavily in their favor, so targeting pitchers with a slightly sub-optimal K Prediction could be a contrarian move in guaranteed prize pools (GPPs).
Targeting pitchers with high-strikeout upside isn’t imperative to DFS success, but it has historically led to some of the biggest fantasy performances. Over the past three seasons there have been 26 instances of pitchers racking up 50-plus DraftKings points. Only three of those games included a pitcher with a K Prediction less than 6.0.
Takeaways
FantasyLabs’ K Prediction metric has proven to be correlated with DFS production. Here are some specifics:
- Pitchers in the top-20th percentile of K Prediction have historically posted high Plus/Minus values with great Consistency Ratings, although this production has come with heightened ownership.
- Pitchers have started to produce significant value with a K Prediction greater than 6.0. Pitchers with a K Prediction over 8.0 have especially crushed, but those instances are rare.
- There has been an ownership premium with a high K Prediction, but this heightened ownership hasn’t historically been as much of a factor for pitchers with a K Prediction between 6.0 and 7.0.
When you’re constructing rosters with our Lineup Builder, remember to take into account pitcher K Predictions. And, of course, be sure to do your own K Prediction research with the FantasyLabs Tools.
