Most fantasy owners, at least to some extent, suffer from confirmation bias. They attempt to twist and use data in a way that will prove that they were (or are) correct. Their slow starting player is a historically slow starting player – all is well. Maybe. Or maybe, a player’s skills have changed and that is causing his struggles, not the time of the year. To avoid trying to prove yourself or talk yourself too far into your own beliefs, you should actively try to prove yourself wrong.
For instance, Carlos Gonzalez is a player that I own in one league that is off to a slow start. Like most owners at this point, I check his splits and uncover that the March/April is split is where Gonzalez has posted the lowest OPS of any month during the course of his career. Instead of merely brushing this off and being comforted by the fact that Carlos Gonzalez was off to a similarly terrible start in 2015 through May 15 – the year that he hit 41 homers – I dug a little deeper into what exactly is going on.
In Gonzalez’s case, the strikeout and walk rates are pretty flat year-over-year, his BABIP is low .243 (career .334) and the batted ball profile is not alarming – grounders are up a bit, hard contact down a bit, soft contact up a bit. Digging a little deeper into the numbers, I feel much better that this is just a cold streak and a few good games at Coors will boost Gonzalez back into the upper tier outfielder that I drafted.
As another example, let’s turn to Byron Buxton. I was not a Byron Buxton fan coming into the year. I thought he was entirely overdrafted as a potential starting outfielder based upon what we have seen. His early season struggles have been well-documented and validated, to some extent, my opinion. Nonetheless, his supporters have been murmuring that Buxton is beginning to come around. Let’s dig in and see if they are right…
Before we begin, I must caution that this is based upon the most minimal of samples, which gives me pause right from the start. Over the last two weeks, from April 22 through yesterday, Buxton has homered, stole 2 bases, walked 22.6% of the time, struck out just 16.1% and is triple-slashing .292/.452/.500. Those results are impressive, but do the skills support those results?
The difference that we are seeing skill-wise is that Buxton is laying off pitches outside of the zone – on the season he sports a 31% O-Swing%, but during this recent surge 20.8%. This is good because (1) pitches outside the zone are tougher to make good contact on generally, and (2) Buxton is terrible at making contact outside the zone (34% in 2017) . This is a truly positive development, but how positive and meaningful are these trends? Those questions remain. Buxton has a below contact rate inside the zone, if pitchers start coming inside the zone more will Buxton be able to make them pay? Or perhaps, as we will discuss below, maybe the pitchers that Buxton is walking against are not throwing particularly close pitches or have poor control?
Another change, which some might dispute whether it is positive, Buxton is hitting more line drives (41.2%) and groundballs (35.3%) than flyballs (23.5%). Although flyballs are good for power, if you are not a particularly good power hitter, they are going to be bad for your batting average. By mixing in a really strong, albeit unsustainable line drive rate with more groundballs, which Buxton can turn into infield hits (16.7% IFH% during this stretch), Buxton can get on base more and take advantage of his athleticism. We would much prefer a .250 10 HR 40 SB Byron Buxton to a .210 20 HR, 20 SB Byron Buxton.
However, the contact/batted ball profile is a mixed bag of sorts – Buxton has a 50% IFFB% rate during this hot stretch – hinting that the BABIP might be a little high – and his Soft % and Hard % are an identical and not ideal 26.3%. With respect to the batted ball profile, there are signs of improvement, but it’s not as encouraging as the plate discipline – the batted ball gains seem like simple variance, in my opinion.
One of the go-to Buxton arguments pre-draft was his 2016 September and specifically, the list of pitchers that he homered off during that stretch. Let’s turn that argument on this hot stretch, below is a list of his hits and walks, the hit type (if it was a hit), and the pitcher that the walk or hit came against:
|Date||Batting Line||Hit Type (If Any)||Pitcher (BB/9 or H/9)|
|4-22-17||1-3, 1K||1. Infield Hit||H1: Blaine Hardy (1.04 H/9)|
|4-24-17||0-2, 2BB, 1K||N/A||BB1: Martin Perez (5.4 BB/9)
BB2: Dario Alvarez (4.7 BB/9)
|4-25-17||2-3, 1 BB||1. Line drive single to left
2. Line drive single to right center
|BB1: Andrew Cashner (6.75 BB/9)
H1 & H2: Andrew Cashner (0.75 H/9)
|4-26-17||1-2, 2BB||1. Line drive single to right||H1: Cole Hamels (0.76 H/9)
BB1: Cole Hamels (3.31 BB/9)
BB2: Keone Kela (3.68 BB/9)
|4-28-17||0-3, 1 BB, 1K||NA||BB1: Joakim Soria (5.4 BB/9)|
|4-30-17||1-3, 1 BB||1: Bunt single||H1: Jason Hammel (1.33 H/9)
BB1: Jason Hammel (5.4 BB/9)
|5-2-17||1-4, 1 HR, 1 K||1. Home Run to Center (429 feet)||H1: Cesar Valdez (1.5 H/9)|
|5-3-17||1-4, 1K||1. Triple to Center||H1: Kendall Graveman (0.88 H/9)|
Suddenly, this hot stretch is a lot less inspiring, right? The walks make a little more sense given the quality and type of pitchers he was facing… Maybe…
This research is not meant to show that Buxton’s hot stretch is total fluke. He may have laid off pitches from better (or in Hamels’s case – healthier) pitchers during this timeframe. The pitches that he laid off may have been extremely good, close pitches. Buxton also may have gotten hits off better pitchers during the stretch and it just so happened that the collection of pitchers that he faced were mostly uninspiring. Further, there are mediocre and bad pitchers that Buxton will face during the course of the season that he can be productive against. This is a numbers game – you don’t get bonus points for hitting a home run off Chris Sale rather than Cesar Valdez.
The purpose of this exercise was to show how confirmation bias might make us short-sighted in our research or efforts to resolve an issue. It was also meant to show (via my own efforts and the research of others) how we can use data – even on the smallest sample – to try to make ourselves right. In this case to confirm or deny that Byron Buxton is a worthy fantasy baseball option. The best fantasy players will always be those striving to prove themselves wrong until they cannot anymore.
-With the rash of pitcher injuries (Hamels, Bumgarner, Syndergaard, Kluber), we discussed several pitchers on the podcast (which will be out later today) that you can target off free agency. My list included: Francisco Liriano (46% – all percentages Yahoo!), Eduardo Rodriguez (46%), Patrick Corbin (36%), Alex Wood (25%), Jesse Hahn (22%), Mike Foltynewicz (19%), Charlie Morton (15%) and Trevor Cahill (10%) with the lightly owned Cahill as my favorite of the group. Another name that I did not mention on the podcast – J.C. Ramirez (11%), the converted Angels reliever, who has struck out 31 batters in 27.2 innings with a healthy 3.44 K/BB.
-Also keep your eye on Nate Karns (2%), who had a great start last night. Over his last two starts (both against the light-hitting Chicago White Sox), Karns has struck out 12 batters and permitted just one walk. He faces his former team Tampa Bay next and could be worth a flier, as a turn around candidate, after a rough opening month.
-Lost in the long, tall shadow of Aaron Judge, is Matt Holliday (59%), who has been red hot the last two weeks hitting 3 HRs with 10 RBI, a .324 AVG 1.076 OPS. Holliday has both 1B and OF eligibility and is vastly underowned at this point. Holliday makes for an apt fill-in and bench option for those struggling with 1B and/or OF in the young season. There is no reason to own Adrian Gonzalez (74%) over Matt Holliday at this point.