My last name is Calle, which means ‘Street’ in Spanish. This is a piece about narrative street. Ergo, ‘Narrative Calle.’
Home Run Derby Curse
There is a belief that participating in the Home Run Derby can ruin a hitter’s swing and preclude him from replicating or maintaining his first-half prowess after the All-Star break. That idea may hold merit in general, but mainly because derby participants are discernible regression candidates due to their impressive first-half performances. Many outperform their career power numbers and parlay that success into a derby invitation.
According to research conducted by FiveThirtyEight, Home Run Derby participants incur a noticeable dip in home runs per at-bat during the second half of the season, but so does the rest of the league. Scott Lindholm discovered that 152 of the 238 participants from 1985 to 2014 suffered a second-half drop in OPS+. Neil Greenberg concluded that contestants tend to carry similar batted-ball profiles throughout the season with the noticeable exception of Home Run/Fly-Ball Ratio, which cumulatively drops for participants. A study at FanGraphs affirms Greenberg’s results and adds:
“Although data suggests that Home Run Derby participants do indeed regress in the second half of the season, the derby is not to blame.”
While there have been plenty of studies attempting to debunk the myth tying second-half regression to participation in the derby, all of these studies are incomplete from a DFS perspective in that they do not take into account the salaries DFS platforms assign to these batters.
What are the DFS implications of participation in the Home Run Derby? Does the pool of eight participants each year suffer drastic reduction in the ability to meet salary-based expectations after the derby? How do these batters compare to similar players who don’t participate in the derby?
May our free Trends tool be of service in this endeavor.
Home Run Derby Participants
| Home Run Derby Participants (2012-2015) | |||
| Plus/Minus | Consistency | Count | |
| Pre-Derby | 1.38 | 43.2% | 2813 |
| Post-Derby | 0.42 | 40.2% | 2014 |
The data was collected using the FanDuel database, which dates back to 2012. All derby participants are accounted for in the dataset. Removing contestants with limited time on either side of the derby didn’t affect the final overall trend.
Since 2012, we’ve seen 27 individuals participate in the Home Run Derby — six of whom have competed more than once. The total number of instances settles at 34 if duplicates are separated into multiple data points. We’re dealing with a small sampling of mainly power hitters.
Only six participants improved their Plus/Minus values after the All-Star break, and two of those data points belong to Yoenis Cespedes (after both of his derby victories). Five of the aforementioned six advanced past the first round. For the most part, the derby batters not to suffer second-half DFS slumps are those who did at least moderately well in the event.
For this group, the expected average points decreased by 0.5 points after the derby. Only six times since 2012 has a participant averaged more expected points after the All-Star break — an indicator of increased salary. We can’t argue that a price hike impacted the decline in Plus/Minus.
Seven players registered a negative Plus/Minus after the All-Star break, a departure from the first half, in which all participants individually sported a positive Plus/Minus.
Overall, the evidence is damning for a four-year span of Home Run Derby participants. Nonetheless, if non-participants faced a similar fate — if they also face a second-half decline — then maybe the second-half slump of derby batters is just a noisy myth.
Open-and-Shut Case?
Home Run Derby participants experience a sizable dip in Plus/Minus production during the second half of the season. Whether that’s due to participating in the derby or other factors such as age or fatigue remains unclear and outside our scope.
Rather than extract data from every non-participant, I limited the pool to players with similar games played to weed out injury-impacted players and shortlived callups. Choosing from everyday starters may illuminate fatigue’s fingerprints.
To qualify for the comparison cohort, players must have played at least 125 games during the season and at least 75 percent of the first half of the season — which varied each year based on how many games were played prior to the All-Star break. The data was pulled straight from our Trends tool, and (because there is no All-Star Break filter) the line of demarcation was set by the team that played the fewest games heading into the break. All non-participant data was collected up until that game. The same strategy applies after the All-Star break. Between 110 and 120 players were culled for each season.
| Non Home Run Derby Participants (2012-2015) | |||
| Plus/Minus | Consistency | Count | |
| Pre-Derby | 0.45 | 41% | 33,853 |
| Post-Derby | 0.49 | 40.5% | 19,819 |
Non-participants witnessed an increase in Plus/Minus after the All-Star break, one that marginally trumps participants over the same space. Case closed. Ignore context. Don’t question my methods.
Not so fast.
Last year was the outlier that tipped the scales. From 2012 to 2014, non-derby participants registered a better Plus/Minus prior to the All-Star break. In 2015, non-participants who matched the games-played criteria recorded a 0.21 Plus/Minus before the break and a four-year high of 0.69 Plus/Minus after the derby.
Was 2015 the beginning of a new trend in second-half production? Was it merely a one-year blip on the radar? Or was it the year that in this small sample makes the short-term trend match what would be a longer-term trend?
Per the data, offensive production spiked during the second half of last season, yet only one Home Run Derby contestant followed suit: AL MVP Josh Donaldson. During the second-half revival, runs increased by nine percent and home runs improved by nearly 16 percent. A simple stat like batting average has declined league-wide eight straight seasons until last year.
2016 Crop
Of the eight participants this year, four have competed previously: Giancarlo Stanton, Todd Frazier, Carlos Gonzalez, and Mark Trumbo.
Frazier has lasted until the final round in three straight competitions. However, his production after the All-Star break has always declined significantly.
|
Todd Frazier (2014-2015) |
|||
|
Plus/Minus |
Consistency | Count | |
|
Pre-Derby |
2.04 | 44.9% |
176 |
| Post-Derby | -0.005 | 39.7% |
131 |
I urge you to be wary of Home Run Derby participants who overperformed in the first half, especially those who outpaced their career HR/FB rates: Frazier, Trumbo, and Wil Myers.
Trumbo is six homers shy of his career high, and he leads the majors in total bases and home runs. Back in 2012 when Trumbo participated in the derby, he posted a negative 2.72 Plus/Minus in the second half after being eliminated in the semifinals.
Myers’ 19 home runs is already a career high, 13 of which were hammered in Petco Park. His 3.4 Plus/Minus and 47.6 percent Consistency lead all participants this year, and he (somehow) has found the spacious ballpark in San Diego befitting:
Frazier has two years that prove he’s likely to regress, but his career- and league-worst .202 BAbip provides plenty of room for second-half improvement. In case you’re wondering, that’s the sixth-lowest BAbip since 2010 among qualified first-half hitters, per FanGraphs.
Conclusion
Home Run Derby participants from 2012 to 2015 suffered a cumulative drop in Plus/Minus after the derby yet still returned value, given their salary-based expectations. Non-participants accrued a slight uptick in Plus/Minus during the second half of the season, but the offensive renaissance in the latter half of 2015 skews the data, most likely making it unrepresentative.
In the DFS landscape, the Home Run Derby curse has endured over the past four years. Overall and on a yearly basis, those competing in the derby largely fail to recreate their first-half success.
I hesitate to provide a definitive conclusion based on the cumulative Plus/Minus of Home Run Derby participants compared to similar non-participants. Save for one abnormal second half, both groups have submitted better first halves.
What is certain, though, is that throughout the data set derby participants have seen their Plus/Minus values decline in the second half much more than non-participants have. If you are entering the second half of the 2016 season thinking that you can find some relative hidden value by paying up for Home Run Derby participants, you should reexamine your position.
