Fantasy Freestyle: Explaining the Pie

Over the past few weeks, I examined retrospective fantasy valuation, along with how past earnings may impact future earnings trends. One of the things that I discovered is that the 10 best National League pitchers in 2015 earned $320. This is more play money than the top 10 earned in a season since 2010 in either the American League or the National League.

Table 1: Top 10 National League and American League Pitchers, 2010-2015

When there is a significant change in earnings—whether it is for an individual player or within a bloc of players—readers frequently want to know if it is a trend or an anomaly. In 2015, the earnings spike in the National League was driven mostly by Jake Arrieta ($44), Clayton Kershaw ($42), and Zack Greinke ($41). These three pitchers combined to earn 12.5 percent—or one-eighth of the pitching earnings—available in 12-team, 5×5, NL-only leagues.

The initial reaction many have to an elite season in fantasy baseball is to focus on the individual performance and wonder if that performance is repeatable. This type of analysis is certainly useful. However, one of the challenges with this type of analysis where fantasy baseball is concerned is that while a great pitcher is likely to put up great statistics, wins, earned run average, and ratios are all subject to fluctuation thanks to circumstances completely or somewhat outside of a pitcher’s control. A team’s bullpen or offense can impact a pitcher’s win totals in either direction and the defense can impact either earned run average or WHIP. (This argument certainly can be made about strikeouts in regard to umpire tendencies and catcher framing, but it is safe to say that a pitcher still exhibits a great deal of control over this category).

This is all known and considered when it comes to determining valuation baseline for pitchers. However, something that is seldom if ever considered is what impact league context can have on player value.

Table 2: NL Pitcher Baselines per MLB Team, 2013-2015

With very few exceptions, contextual changes on a year-to-year basis do little to alter the fantasy landscape. The 1993 season is one of the most significant historic example of a seismic shift in valuation, due to the expansion of the major leagues from 26 to 28 teams. The average ERA in the National League jumped from 3.50 in 1992 to 4.04 in 1993 with the average WHIP increasing from 1.287 to 1.357. Jose Rijo’s 1993 season would not nearly have been worth as much in 1992’s context, while Greg Maddux’s 1992 campaign would have been worth far more in 1993.

My player valuation model does not use major league baselines to determine player value. Instead, the earnings baseline is derived using the 168 most commonly purchased hitters and the 108 most commonly purchased pitchers from the CBS, LABR, and Tout Wars expert leagues. Using these numbers instead of the overall major league average paints quite a different portrait.

Table 3: NL Pitcher Baselines per Fantasy Team, 2013-2015

The numbers on the left-hand side of Table 3 take the statistics purchased in the three expert leagues and distribute them equally among 12 teams in each season from 2013-2015. Regardless of how many wins or saves were actually bought, the fantasy dollars allocated to the 108 most expensive pitchers in these categories stay flat. ERA and WHIP earnings fluctuate based on performance, but the baseline for earnings in the category is $0 (some disagree with this in theory, but this discussion is beyond the scope of this article). Strikeout earnings do fluctuate from season to season, although as you can see in Table 3, the fluctuation is not significant.

For the most part, ERA and WHIP are what drive earnings for pitchers. The curious difference between Table 2 and Table 3 is that while the National League ERA on the whole jumped from 3.66 in 2014 to 3.90 in 2015, the baseline ERA for fantasy baseball only jumped slightly, from 3.49 to 3.53. And WHIP, which went up from 1.267 in 2014 to 1.299 in 2015 in the NL, dropped in the expert auctions from 1.226 in 2014 to 1.215 in 2015.

This is a trend I did not notice while compiling player earnings and writing about retrospective valuation last month. While I certainly thought that Arrieta, Kershaw, and Greinke were amazing earners compared to their recent historic counterparts, I realize now that their earnings could have potentially been even higher:

Table 4: Top 10 2015 NL Pitcher Earnings with “Modified” ERA, WHIP

In order to achieve these modified earnings, I took the average difference between the 2013-2014 NL ERA/WHIP and fantasy ERA/WHIP and applied them to the 2015 formulas. So instead of using a 3.53 ERA as the NL baseline, I used a 3.765 ERA. For WHIP, I switched from a 1.215 to a 1.256.

The difference is not as dramatic as some would expect. A $46 season from Arrieta certainly would have been amazing, but it would not have come close to matching Greg Maddux’s $53 campaign in 1995 (for example). Moving the ERA and WHIP needle up has an impact, but the league context still remains important. In 1995, the average league ERA was 4.18 and the average league WHIP was 1.376.

However, valuation formulas using fantasy league averages as opposed to true league averages typically aren’t as disparate as they were in 2015. This is because the expert market did a significantly better job predicting who the best 108 pitchers were in 2015 than it had in the two years prior.

• In 2013, the top 20 undrafted NL pitchers earned a combined $267. Jose Fernandez was the top free agent earner at $29. Five free agent pitchers earned $15 or more.
• In 2014, the top 20 undrafted NL pitchers earned a combined $266. Jake Arrieta was the top free agent earner at $23. Eight free agent pitchers earned $15 or more.
• In 2015, the top 20 undrafted NL pitchers earned a combined $188. A.J. Ramos was the top free agent earner at $19. One free agent pitcher earned $15 or more.

Was 2015 the beginning of a new era when fantasy experts finally saw the light and have become better at predicting who the best pitchers will be?

I doubt it.

NL closers in 2015 were the most stable commodities they had been in years, with only Steve Cishek, Bobby Parnell, Jenrry Mejia, and Addison Reed losing their fantasy owners $5 or more. While there was a decent rookie class of hitters, there wasn’t nearly the same level of depth on the pitching side. Noah Syndergaard, Anthony DeSclafani, and Ken Giles were all part of the draft pool and didn’t contribute to the free agent bounty that typically comes later in the season.

The biggest reason for the disparity, though, was that the best starting pitchers were almost all snagged at NL auctions in 2015. J.A. Happ was the only starting pitcher who earned $10 or more who wasn’t purchased in an NL-only auction. In 2014 there were 10 starters who earned $10 or more. In 2013, there were seven starting pitchers who accomplished this.

We don’t know which pitchers we are going to swing and miss on in fantasy auctions or drafts. But we almost always know that we are going to miss on more than a few. This didn’t happen very often in 2015, and as a result while the top pitchers’ value was very high, it could have been a couple of ticks higher on the elite arms had events shook out as they typically do. This is something to consider next year when constructing your roster in deeper formats. The separation between Arrieta, Kershaw, and Greinke was fairly significant in NL-only, but it could have been even greater. If you believe that these pitchers are candidates to repeat in 2016, it would behoove you to spend a little more money than usual on these arms in your auctions.