Overcoming Recency Bias on Draft Day

March 25, 2021

Pierre Camus addresses how recency bias affects fantasy baseball ADP and which players are overvalued or undervalued in 2021 based on the way they finished last season.

With all the draft analysis and sabermetrics being thrown at fantasy managers these days, I think it's time we take a step back and focus on the most important thing to dissect before your draft - YOU.

A few months ago, I covered some of these cognitive biases for fantasy football season. Now, I believe it's time to take a deeper dive into each by highlighting specific players that each might apply to on draft day.

Recency bias applies to us in all walks of life. It's only natural in this modern age of overstimulation that we filter out the unnecessary details of the past and retain what is most important or most fresh in our minds. This is an adaptation meant to serve us well as a species but it has its drawbacks as well, mainly the failure to see the bigger picture. This certainly applies to fantasy baseball, where we have more data than you shake a wireless mouse at. My goal is to help you take a step back and get introspective before creating a roster and then identify some players that are overvalued or undervalued based on the industry's collective recency bias.

It's Not How You Start...

...it's how you finish. At least that's the old axiom. This seems intuitively logical, especially for sporting contests, until you realize that it doesn't matter when you did what, all that matters is the end result.

If your weekly H2H team finishes with a record of 128-104-20, did it really make a difference whether your record was better in the final week as opposed to the first or somewhere in the middle? We are incapable of remembering all the significant details, especially mundane ones, of our lives. We remember what happened most recently because it's still stored in our short-term memory or if something stood out for its uniqueness. That one week where you went undefeated or had a bad beat because of that walk-off home run by an opponent, for example. That's what recency bias is all about.

The theory of recency bias is attributed to Hermann Ebbinghaus, a noted psychologist who conducted numerous memory experiments. He realized that people are more successful when recalling items from the beginning (primacy effect) or end (recency bias) of a list. More importantly, people tend to give greater importance to the most recent event.

This rears its head in the fantasy sports realm when preseason rankings remarkably resemble those of the previous season's end. Analysts essentially confirm that the way it was is the way it's going to be. As Blair Williams (@EverywhereBlair) thoroughly explained in a recent article, "...fantasy baseball players are provided with rankings that actually recreate ADP rather than attempt to dissect and overcome the weaknesses of ADP."

This speaks to another bias - herd mentality. This trend becomes more and more real throughout the preseason as rankings are compared to one another in order to form an industry consensus. We're told that we must agree that there is a clear Big Three at SP, with Jacob deGrom, Gerrit Cole, and Shane Bieber ahead of the rest of the pack. You have to attack steals early because if you don't get Trea Turner or Adalberto Mondesi, you stand no chance of winning that category (ignoring the fact you don't need to actually finish first in any individual category to win a roto league). Not true.

According to Ron Shandler, "the success rate of ADP rankings correctly identifying each season’s top 15 players (in any order) is only 33.75 percent. In fact, those top 15 players finish somewhere in the top 30 only 53 percent of the time."

Putting extra weight on the way a player finished the previous season seems like sound logic but it is often faulty. This is especially the case coming off a season whose entirety would normally account for the forgotten middle portion that we'd be prone to forget entirely. Does that mean we may as well write it off?

Should 2020 Even Count?

With the sprint of a 60-game season, there was no first half or second half, no beginning, middle, or end. Do we take the entirety of it to mean something or try to erase it from our memories? We might wish 2020 never happened for a multitude of reasons, but the question is whether we should consider these statistics for fantasy analysis since there were so many things out of the norm: universal DH, seven-inning doubleheaders, week-long cancellations, unbalanced schedules, etc.

We can't throw everything out the window, nor should we. But we also can't pretend that last season's stats will all legitimately translate over to 2021. The best course of action would be to take second-half stats of 2019 and combine them with 2020 in order to get a large enough sample that doesn't include all those variables. Of course, that means looking at something that happened 18 months ago and turning it into actionable advice. As always, the best course of action is to look at the bigger picture and the course of a player's career rather than just the last season or two. Rejecting the recency of last season is easier said than done, but it's a good exercise that should be practiced constantly.

Shooting Stars Ready to Fizzle?

If I were to make a PSA to warn against falling victim to the phenomenon of recency bias, Adalberto Mondesi would be the poster boy. So I did:

Sure, Mondesi's scorching end to the season shouldn't be dismissed. Many feel that he is worth a high draft pick even with a low batting average because he can outright win you the stolen base category by himself. But we're talking about an absurdly hot streak that salvaged an otherwise dismal 2020 season.

Mondesi is projected by ATC to hit .251 while stealing 51 bases and scoring 78 runs with modest power numbers. I'm not here to dispute whether those are realistic expectations. I'm merely pointing out that midway through 2020, managers were ready to drop Mondesi from their rosters.

Will Smith, the catcher not the relief pitcher, is the third player at his position flying off draft boards. This is more than post-World Series hype from Dodger fans too. In The Great Fantasy Baseball Invitational (TGFBI), an analyst-only tournament comprised of nearly 400 industry folks, Smith was taken 111th overall on average and was indeed the third catcher drafted. He went as high as the sixth round in one league at the 75th pick, ahead of Salvador Perez. I get that he's the shiny new toy at the catcher position and it's a highly-competitive two-catcher league, but still... I had this belief even before recent comments by manager Dave Roberts stating that Smith will only catch about 90 games this season.

Will Smith Expected To Catch Around 90 Games In 2021 https://t.co/ZUtqyKjyus

— RotoBaller MLB (@RotoBallerMLB) March 20, 2021

Take this into account when deciding if the sample size matters. His 100-PA rolling xwOBA shows an increase of .080, 18th-highest of all batters to end the season. His 50-PA rolling xwOBA shows a decrease of .115, which is 26th-worst of all batters. Selective memory tells us he is the next big thing at catcher but the statistics tell a different story depending on how you look at them.

Byron Buxton continues to tantalize in spurts, showcasing his elite sprint speed and power upside. It looked like he was putting it all together in the second half of 2017 when he hit .300 with 11 HR, 35 RBI, 40 R, and 13 SB in 56 games after the All-Star break. That elevated his ADP all the way to 61 overall in 2018 and then... bust city. A combination of injuries conspired to limit him to 94 plate appearances over which he batted .156 with zero home runs and four RBI.

What's changed since then? Well, not too much. He had an IL stint in 2020, batted just over his career average at .254 and his walk rate was worse than ever at 1.5%. Why are fantasy managers still willing to take a chance on him within the top 125 picks? The answer may shock you. But it shouldn't - it's recency bias.

Buxton ended the short season strong by hitting .290 and jacking eight homers in 16 games started after returning from the injured list in September. The upward tilt is undeniable.

What's missing from all that is the one thing that is supposed to make him most valuable - speed. He stole all of two bases last season. It seems we're destined to never have it all with Buxton, so don't let a hot finish lure you in again.

Fast Faders Ready to Rebound

Austin Meadows looked like a blossoming star after being freed from the Pirates to thrive in Tampa. Unfortunately, he was one of the players affected early on by COVID and then suffered an oblique strain after struggling through 2020. It's no surprise his ADP has fallen by 50 spots from a year ago.

It's understandable to some extent since we only have one standout season from him with two mediocre, albeit short, seasons that could suggest 2019 was an outlier. This could be labeled normalcy bias but we don't really have a long enough set of data to determine Meadows' true ability. Given the combination of injury and virus-related issues he experienced, along with his prior prospect pedigree, I'm leaning toward him being the real deal. As a potential five-category contributor, he could be a screaming value.

Max Kepler slashed a disappointing .228/.321/.439 and wasn't nearly as valuable as his 121 ADP would have suggested. Coming off 2019 where he slugged 36 homers and crossed the plate 98 times, he should have been a solid OF3 but his early struggles along with an abductor strain that cost him 12 games eliminated those hopes. With more time to recover, Kepler could have been a second-half bounce-back candidate but alas, 2020. His .256 xBA was just one point lower than 2019 and his plate discipline was still excellent as he posted a 7.2% K-BB%, marking the third straight season with a K-rate under 20%.

The best part is that he's gone 2-for-30 so far in the spring. That should be discouraging but it should only serve to drive his draft cost down. Kepler isn't in a battle for playing time and should have no problem straightening things out at the plate once the regular season gets underway. At 28 years of age, he is still in his prime and slotted to leadoff for one of the most potent lineups in the American League. Steamer projects him to reach 31 HR, 82 RBI, and score 96 runs while chipping in seven steals.

Jesse Winker started out like a ball of fire last season. In August alone, he hit .369 with 10 HR and 16 RBI before slumping in a big way in September. Tightness in his lower back clearly played a part in his late struggles but it didn't stop him from finishing with a hard-hit rate in the 90th percentile, barrel rate in the 88th, and xwOBA in the 93rd. His walk rate was better than ever at 15.3% and he's already showing his trademark plate discipline in spring training, taking more walks (eight) than strikeouts (six). With an ADP of 222, it's hard to believe he's become an afterthought in 2021 drafts but given how his best showing was sandwiched between the first and final months of last season, it's pretty obvious that primacy bias is at play.

Download Our Free News & Alerts Mobile App

Like what you see? Download our updated fantasy baseball app for iPhone and Android with 24x7 player news, injury alerts, sleepers, prospects & more. All free!

More Fantasy Baseball Advice