RADIO
![By Keith Allison on Flickr (Original version) UCinternational (Crop) [CC BY-SA 2.0 (http://creativecommons.org/licenses/by-sa/2.0)], via Wikimedia Commons](https://cdn.rotoballer.com/2015/04/Corey-Kluber-MLB-fantasy-baseball-analysis-e1463165512276.jpg)
The idea of the “two-start pitcher” makes an appearance on most fantasy baseball podcasts, articles, sites, and newsletters as players are ranked and recommended to owners and teams. Selecting pitchers based on match-ups, park factors, and opponents over the scoring week is a regular debate as owners plan their rosters. At the same time, other than knowing that these pitchers offer an additional start each week, often owners do not understand what these pitchers do to their line. How do they add to subtract from ratios and counting stats? What does a bad start do to the overall line?
That is where this article takes up the conversation: what exactly does a two-start pitcher do to a fantasy team? Specifically, this piece highlights three findings that can be taken from the two weeks of data studied. While this data is only generalizable to the sample size, as pitchers change week to week, it still offers some insight into the process of selecting a starting pitching strategy.
One word before walking into this piece, no matter what lessons owners can learn, all lessons should be taken within the league context of each team. While the article will try to offer some advice to both roto and points leagues, owners are the best guides to their team and strategy. With that, onto the data.
Editor's Note: Our incredible team of writers received five total writing awards and 13 award nominations by the Fantasy Sports Writers Association, tops in the industry! Congrats to all the award winners and nominees including Best MLB Series, NFL Series, NBA Writer, PGA Writer and Player Notes writer of the year. Be sure to follow their analysis, rankings and advice all year long, and win big with RotoBaller! Read More!
What to Make of Two-Start Pitchers
Before diving into the findings in this article, it is worth taking a step back to put the generic two-start pitcher in some context. Over the course of fantasy baseball’s 10th and 11th week of games, a total of 80 pitchers were expected to make two-start pitching appearances. These numbers slanted a bit to the previous week when 49 were on the docket, but overall, the numbers fluctuate based on plenty of variables and events. For example, with the new allotment of off-days this season due to the CBA, some teams play only five games in a week, whereas others, already affected by weather, might play all seven days.
Of those planned two-starts, a total of 81% ended up happening as planned. The change in schedules means that one out of every five pitchers will not make that second start. The variance will also change week to week, but in general, all two starts will not happen for the factors listed before. That being said, for the sake of argument, there are a total of 150 starters in the majors at any one point, each week approximately 27% will be scheduled to make two starts.
The other piece to add to the data and context is that not all two-start weeks are created equal, as some weeks the starting pitcher is Corey Kluber, and some weeks they are Homer Bailey. The variation means that even when comparing week to week, the fluctuations are apparent based on the match-ups. For the sake of the data used in this article, two full weeks of data offer a broad enough sample size to make general findings that should support the underlying baseline findings. At the same time, not every team has a Corey Kluber in their rotation, so knowing what the average two-starter does is, in some ways, more valuable to the fantasy owner trying to figure out who to start.
Finding #1: Two-starts are at best equal to league average one-start pitchers
Over the course of all 65 of the two-start weeks that occurred in weeks 10 and 11, the average pitcher worked for 11.21 innings, producing an ERA of 3.76, a WHIP of 1.24, and striking out 10.26 batters. At the same time, each pitcher only averaged 0.77 wins, meaning that the starter just won two out of every five starts over the sample size.
To factor in the variance in pitching quality, a second survey was run on the data during which the aces, or consensus top 20 starters, were removed from the data sample. Under this sample, the following averages appeared over the same sample timeline: 11.07 innings, an ERA of 4.19, a WHIP of 1.24, and 9.59 strikeouts.
Without the top pitchers, two-start weeks average close to the same number of innings but add close to a half earned run over that week. What does stand out from the comparison is that the WHIP stays the same, but these non-aces lost close to a K a week for their owners. When the Major League average for ERA sits at 4.06 so far this season, the average two-start starter is 3.2% worse than a league average starter. League average WHIP this season sits at 1.30 meaning that the ace-less two-starters are a bit better than that mark with their rate of 1.24. Strikeouts appear to be a bit higher, or right around average when comparing the data set to season norms.
In this way, the average two-start pitcher is worse than league average regarding runs but better or average with WHIP and K numbers.
Why might this be the case? When pitching twice over the course of the week, it makes sense that one good start and one average start ould result in closer to the average mark, whereas one bad start and one good start would equalize out to at best average and at worst, a worse average over the week. It is not uncommon for even the best starters to have a bad outing which is magnified during one week of data.
An excellent example from the data was Tyler Skaggs, who against Detriot gave up five earned runs in five innings, but rebounded again the Rangers to throw six scoreless innings. Owners would love the last start but perhaps winced when seeing the first performance. At the same time, Michael Fulmer gave up five earned in 3.1 innings, and four in six versus the Angels and Blue Jays. These results, even if not a two-start week, would not have been excellent for most teams.
What this means is that in roto leagues the two starts matter much less than in points leagues, as that one bad Kluber start equals out over the 30+ starts that he will make in a season. In a points league or any weekly scoring league, the bad start is magnified. At the same time, if that Kluber start happens without a second start, then it hurts the overall line more than the averaging out or weakening of the gains from a two-start week. Owners should already be looking to both match-ups when setting line-ups, but also recognize that there is no unique benefit from having two starts in a week unless innings count in match-up specific scoring.
BALLER MOVE: Prioritize good one-start weeks over average two-start weeks in non-innings leagues
Finding #2: Road Pitchers are Better than Home Pitchers
Perhaps the most exciting piece of insight that comes from this sample of two-start pitchers was the variance in performance if the starter in question made both of their starts at home or on the road. In a vacuum, it would seem that the average pitcher at home would perform better than on the road, but that turns out not to be the case. Over the course of the two weeks of data collected, 14 pitchers made both of their starts at home and 19 pitchers who made both of their starts on the road.
For the pitchers making both of their starts at home, the gross average pitching line for both of their starts was: 10.95 innings, 4.64 ERA, 1.41 WHIP, and 9.07 Ks over that time. These numbers are much worse than league average by 0.6 earned runs and 0.5 Ks over both of those starts. Also, two-starts at home only posted 0.36 wins which is much lower than the expected total.
For pitchers making both of their starts on the road, the gross numbers were: 11.54 innings, 3.57 ERA, 1.15 WHIP, and 10.95 Ks. At the same time, the average road-only pitcher earned 0.84 wins over their two starts. Road only pitchers were more than an earned run better than the home-only set and lowed their WHIP by 0.3.
Why might this be the case? Looking to the pitchers and the match-ups there is no skew concerning top pitchers in either grouping, and the parks seem to be much the same. The road slate did feature Chris Archer and Jake Arrieta, but those two on their own, should not have affected the large sample enough. Arrieta also pitched a dud in San Francisco for his second start, hurting his case over that scoring period either way. Even if taking these pitchers out of the equation, the road starters still were a bit better than the home-group, which is still unusual based on standard fantasy ideas of park factors. The road starts also had more starts at Coors which should affect the overall line, but not in the way that was expected
BALLER MOVE: Prioritize road-only pitchers making multiple starts in a week
Finding #3: Two-start pitchers struck out more in their second start than their first start on average
Of all the factors listed, this might be the most context-dependent observation, and something that this study will return to at a later date, but also shows a clear trend over two weeks of data. For context, in both weeks there were top starters and fill-ins, and the data trends still existed with that context. The other reason this trend stands out is that it appears in both weeks with a noticeable gap, so not unique to one slate of starters.
For week 10 starters, in their first game pitchers averaged 4.48 Ks, and in the second, 5.56 Ks. For week 11 starters, in their first game, pitchers averaged 4.41 Ks, and in the second, 6.07 Ks.
The numbers are even starker when removing the aces from the data with a week 10 jump from 4.15 to 5.24, and in week 11 the increase went from 3.96 to 5.82.
Why might these numbers be the case? Typically a second start in the week occurs on a weekend date which might account for some of the increase in Ks, as some pitchers are better during afternoon games on Sundays, or even better in Saturday night games. At the same time, with days off, there is a higher chance that two-start pitchers are on their regular schedule, and are not getting an extra day of rest in between starts which might also account for the change in numbers.
Weekend games are also more likely to see reserve hitters due to wear and tear, but should not seem to account for all the difference. The best “proof” here would be starting catchers getting a day off after a night game, and the backup catcher on most teams is mostly glove and no bat. Attendance factors could mean there are more aggressive hitters at play, which would support more strikeouts across the board. While still a mystery this is one of the most actionable findings and should influence owners moving forward.
BALLER MOVE: When in doubt, two-start pitchers are most valued for high strikeout match-ups in their second game; prioritize these match-ups. Also, one clear value to two-start performances is the gross number of Ks that they can provide for teams and owners.
Next Steps
While stated in the introduction, this data should only be used to understand what happened during the 10th and 11th fantasy weeks, but this does offer a step to begin to add more context to two-starters moving forward. The plans will be to release two additional articles to support this process. The first will dig into the pitchers highlighted here, and identify who stood out and who surprised based on match-ups.
Second, the plan is to check in at least once, if not twice, over the season to see if the trends form these weeks appear to continue. While frustrating, this article leaves owners with more questions than firm answers, but if the trends in this article are accurate across multiple data sets, this could change the strategy of approaching starting pitchers based on more than just match-ups.
What can be said is that two-start pitchers might not be as valuable as they appear on the surface, and when in doubt owners should rely less on the multiple starts as opposed to the pitching pedigree itself. This means do not shoehorn a pitcher into the line-up due to two starts as the results are not much better than an average one start, but the risk is much higher.
