Overperformance Metric: Who’s Most Likely to Breakout
Breakouts and busts. If there was a set procedure for finding both, it would have been found years ago and incorporated into projections. For now, all we have is the overall chances of either happening. Over the past few weeks, I’ve been trying to put a simple value on these chances. I’ve completed the underperforming calculations and will now finish the overperforming metric. Additionally, I will compare both metrics to get an overall idea of the projection’s volatility.
In my last article, I found the breakout thresholds for plate appearances (222 PA) and wOBA (.040) and won’t change these values. Besides these two values, I determined who had both thresholds crossed and when both were partially achieved. The overperformance needed to increase near to the threshold values.
Then I found when these breakouts occurred using historical Steamer projection. With the hitter’s age and projected wOBA and PA, I fed the numbers through a logistic calculator to get the chances for each event one occurring.
I did make one major change with the overperformance metrics. When I found the underperforming metric equations, I determined the percentage change values made more sense. Not so with the overperforming values. The percentages changes, especially with the PA, found too many non-breakouts in my opinion. The final values didn’t find the players I was hoping to find. For that reason, I removed those chances.
Here are the final overperformance results (including the underperformance numbers) with my thoughts on each.
- Plate appearances: This value is straightforward with young hitters having the best chance to see improvement and those projected for more than 600 PA seeing none.
- wOBA: These values are consistent with the values I found in my last article, there is no consistency. While there is more room to move up from the bottom, any hitter can see a major uptick like Joey Votto and Jose Altuve did. There is only a 9%-point range for these values while spread was double for plate appearances.
- Minor increase: Just like with underperformance, this value lines up with my overall expectations. It is about half the value of underperformance metrics. Injuries increase the breakdown rate.
- Major increase: This value almost mirrors the minor increase rate with the best fit curve having an r-squared of .94.
For now, I’m just going with the Minor increase value. Additionally, I’m going to just use the Minor value for underperformers. With these formulas, I hope to be able to identify the up and downside quickly during a draft or in an auction’s dollar days. It could be a tiebreaker between similarly valued players or a chance to increase or decrease risk. A few percentage points could add up to getting one of the unknown breakouts or not draftng a bust.
Another feature is to combine the under and overperformance values to get the total volatility of the players. Here are the top and bottom ten hitters in the chance they’ll diverge into the breakout or bust territory.
| NAME | Age | PA | wOBA | Drop% | Increase% | Total |
|---|---|---|---|---|---|---|
| Phillip Ervin | 25 | 208 | 0.298 | 55.1% | 25.7% | 80.8% |
| Travis Jankowski | 27 | 204 | 0.281 | 54.2% | 25.3% | 79.5% |
| Jesus Aguilar | 28 | 201 | 0.312 | 57.1% | 22.2% | 79.3% |
| Drew Robinson | 26 | 216 | 0.307 | 55.6% | 23.7% | 79.2% |
| Dixon Machado | 26 | 216 | 0.297 | 54.7% | 24.4% | 79.1% |
| Taylor Motter | 28 | 201 | 0.294 | 55.5% | 23.6% | 79.1% |
| Tyler White | 27 | 214 | 0.315 | 56.5% | 22.3% | 78.8% |
| Lane Adams | 28 | 205 | 0.283 | 54.4% | 24.2% | 78.6% |
| Adam Frazier | 26 | 224 | 0.316 | 55.8% | 22.7% | 78.5% |
| Mac Williamson | 27 | 220 | 0.281 | 53.2% | 24.5% | 77.7% |
| Shin-Soo Choo | 35 | 645 | 0.342 | 34.4% | 5.3% | 39.7% |
| Dee Gordon | 30 | 630 | 0.301 | 31.5% | 8.2% | 39.7% |
| Odubel Herrera | 26 | 661 | 0.324 | 31.3% | 8.3% | 39.6% |
| Kole Calhoun | 30 | 651 | 0.328 | 32.5% | 7.0% | 39.4% |
| Dustin Pedroia | 34 | 649 | 0.335 | 33.5% | 5.7% | 39.2% |
| Ian Kinsler | 36 | 641 | 0.324 | 33.3% | 5.5% | 38.8% |
| Kevin Pillar | 29 | 654 | 0.310 | 30.9% | 7.8% | 38.7% |
| Ender Inciarte | 27 | 667 | 0.314 | 30.3% | 8.2% | 38.4% |
| Brian Dozier | 31 | 678 | 0.347 | 32.6% | 5.8% | 38.4% |
| Nick Markakis | 34 | 646 | 0.313 | 32.0% | 6.2% | 38.2% |
The unpredictable list contains young bad hitters with little-projected playing time. These guys rarely perform like their projections suggest. Most drop out of the league while a few rise to the top. These are the guys to take a chance on.
The bottom list is old, blah hitters with full-time roles. They are likely to end the season with OK stats. Nothing better, nothing worse. While some gain fantasy relevance because of their speed (Herrera and Gordon), most are the “reliable” guys near the draft’s end.
I’m want to take the values a step further and add them to some fantasy values. With the help of @smartfantasybb, I hope to have the 2017 SGP values and some initial values for NFBC style leagues. Then, I can begin to compare hitters with similar values to start finding some targets.
While these procedures and results are far from perfect, they provide a baseline for player variability. An owner may want to take a chance on the upside while another owner may be needing a safer pick. These values are just another tool for an owner to utilize.
Jeff, one of the authors of the fantasy baseball guide,The Process, writes for RotoGraphs, The Hardball Times, Rotowire, Baseball America, and BaseballHQ. He has been nominated for two SABR Analytics Research Award for Contemporary Analysis and won it in 2013 in tandem with Bill Petti. He has won four FSWA Awards including on for his Mining the News series. He's won Tout Wars three times, LABR twice, and got his first NFBC Main Event win in 2021. Follow him on Twitter @jeffwzimmerman.
It took a little bit of thinking and a couple of times reading through the article and list to actually come away with some appreciation of this post. My gut reaction was, “this (top part of the second list) is a list of nobodies dominated by speed guys.” I then thought about it more, and realized that my own biases and lack of knowledge about these players were playing a factor in my judgement. I then took time to click on the player page link for a bunch of the players, and this caused me to look at it differently. There are some speed guys, but also former first round picks and some guys with a good amount of power in their MiLB profiles. There are players that I wouldn’t have paid any attention to that I will now have an understanding of going into my auction in the spring, and that’s all I can really ask for in a “potential breakouts” post. Thanks!
Thanks. I found a few players also and it really helps when the values are in a draft list. Villar is a perfect example the speed guy with a breakout season. The speed helps to give the player some floor.