Avoiding Information Bias

A funny thing happens with the sequence in which we are given information to process.  In psychology, the “anchoring” effect is a cognitive bias in which we rely too much on the first piece of information offered, and not enough on subsequent information that should otherwise be just as valuable.  When it comes to baseball analysis, the way I see this frequently applied is when analysts attribute a “hot start” in April to some underlying fundamental change that will potentially be sustainable throughout the season, whereas a similarly torrid pace that occurs in June may be deemed as an unsustainable anomaly and is explained away as a function of opponents or small sample size fluctuation.  It’s maddening! 

We should look at month long (or whatever interval you deem convenient and relevant) samples equally whether they are the first such sample you encounter or the fourth.  Some examples…

Xander Bogaerts:

For simplicity’s sake, we’ll use WRC+ to show the differences as the year has gone on for Bogaerts.

April: 145

May: 158

June: 154

July: 93

Bogaerts incurred a pretty big drop off in July.  The power has still been there somewhat, with a .182 ISO that isn’t such a dramatic drop off, but the batting average has dropped to only .234 and the OPS to only .730.  Not good!  So is he hurt?  Does he not like hot weather?  Is it just a random fluctuation like maybe his May and/or June also were and the true Bogaerts is somewhere in between?  Only time will tell, but we should be paying just as much attention to a steep drop off in July as we did to a hot start at the beginning of the season. 

Isiah Kiner-Falefa:

If we are worried about Bogaerts, we should be absolutely pulling our hair out and reaching for the anti-anxiety meds when it comes to Kiner-Falefa.  He absolutely fell off a cliff in July, hitting only .188 with a .213 BABIP and an absolutely atrocious .035 ISO.  He also stole zero bases since it’s tough to steal bases when you aren’t on the bases.  We also know what he’s capable of when he’s right, and in leagues where he has catcher eligibility especially, he’s a tough guy to move on from.

For what it’s worth, he hasn’t really done anything differently under the hood in July to achieve the worse results.  The K% is still about the same, he still doesn’t walk, and he is hitting the ball roughly as hard and at the same trajectory as he previously had.  It’s hard to explain such a massive drop-off, perhaps there’s more of a “book” on him now with him getting regular plate appearances atop the Rangers lineup as opposed to sporadic and strategic starts in years past.  Whatever the case may be, if Kiner-Falefa’s downward spiral continues well into August, we may need to look for another contributor to salvage our fantasy season. 

These sorts of changes don’t have to be negative, though.  Just like Bogaerts and Kiner-Falefa have been in poor form lately, there are other players who have turned it on as of late, although none of them quite as much as…

Joey Votto

In a season where Vlad Guerrero Jr. and Shohei Ohtani are doing such incredible things, you might have missed what Joey Votto did in July.  If you did, it’s worth taking a look.  After a rough start to the year and being injured or ineffective for the vast majority of May, Votto really started to turn it on in late June and July.  In July in particular, without really changing his approach (walking and striking out at about the same rates as the rest of the season), Votto started absolutely crushing the ball.  Slashing .319/.440/.734 good for an OPS of 1.174 and a wRC+ of 198 (remember Xander Bogaerts from the first paragraph- his best month was 158!). 

The biggest shift for Votto was to outcomes that lend themselves to better production- lots more fly balls, and lots more hard contact.  Whether he can continue this torrid run remains to be seen- he almost certainly cannot, and he certainly won’t hit home runs in seven straight like he did to close out July, but if this approach of hitting the ball extremely hard and in the air is something he maintains, well then he just better and more useful as a fantasy producer. 

It will be interesting to see how pitchers adjust to Votto- he is typically thought of as having a discerning batting eye and excellent bat to ball skills, but not a behemoth slugger that is trying to go yard at every opportunity.  July Joey Votto however was absolutely that- and if you pitch to a guy with that approach the way you would pitch to Adam Frazier or Nick Madrigal, you’re going to get a guy that hits home runs in seven straight games.  August will be very interesting when it comes to how successful Votto can be as pitchers make adjustments to his new and more aggressive approach. 

Luis Castillo

If you are a regular reader in these parts (which, thanks!), you might recall that in early June I wrote about how awful Luis Castillo had been and that I expected him to improve dramatically.  The analysis was simple, nothing under the hood had changed, the velocity was where it always had been, he was starting to miss more bats, and was likely the victim of some bad BABIP luck.

Here’s how Castillo has performed by month:

April: 17. 1 K%, 6.3 BB%, .317 BAA, 1.60 WHIP, .346 BABIP

May: 21 K%, 11.6 BB%, .325 BAA, 1.93 WHIP, .393 BABIP

June: 25.6 K%, 9.6 BB%, .173 BAA, .98 WHIP, .228 BABIP

July: 26.8 K%, 11.8 BB%, .209 BAA, 1.22 WHIP, .278 BABIP

So- not to toot my own horn too much, but that is pretty much exactly what was happening.  If you were able to buy low on Luis Castillo in June as I recommended, it likely worked out very well for you!  He is very much more the guy that we have seen in June and July than he was early in the season, and looking ahead is much more likely to produce at that level in August and September than the way he performed early on. 

With all of these guys- Bogaerts, Kiner-Falefa, Votto, Castillo, the takeaway is that different information can present itself at different times.  Performance in the middle of the season counts just as much as performance in the beginning of the season, and should be evaluated as such.  We should not dismiss things because they don’t fit our early season narratives, or worse, find explanations to excuse production that does not fit with our narratives.  In being truly objective, we have to look at all trend information objectively and determine its usefulness independently from our biases. 

Leave a Reply

Your email address will not be published. Required fields are marked *