“It’s Still Early!” and Other Things We Tell Ourselves

It’s nearly the end of May. Roughly one third of the way through the regular season. Already, we’ve had our fair share of surprises. New main characters are scattered across your league, or all concentrated on the roster of a sure to be smug top seed.

Whether it’s the manager who added Davis Martin early or the one who was utterly convinced by Jordan Walker and his offseason with Driveline, they’re certainly sitting pretty so far. 

For every one of those managers, there’s the one who missed out on Corey Seager in the draft, bought low on the veteran slugger, watched him go hitless in the seven games after trading for him, begged for him to hit the phantom IL, and then knowingly grimaced while moving him to the actual IL at the end of last week due to back tightness. 

And another who started their draft with Tarik Skubal or Garrett Crochet instead of Paul Skenes. 

The randomness of baseball, fantasy or otherwise, is part of what makes the game so frustratingly intoxicating. 

For me, a young guy who can’t for the life of him hit with anything other than the nine-iron, it’s sort of like golf. 

You can watch a ton of video, pore over the numbers, establish a process with theoretical backing that should garner success, and then end up shanking your first tee shot with an ego-stinging, swinging bunt that mercifully rolls past the tee. 

Or you drafted Cal Raleigh instead of waiting on Drake Baldwin in one of the deepest catcher groups in recent memory. 

For all of the preparation we all do, that degree of randomness is inevitable. And for all of the bad BABIP luck, high strand rates, and flat out poor performance, there is also something to be said for accepting this randomness. Releasing yourself to it even. 

That’s what I’ll try to show you in this article. You can still tell yourself that “It’s still early!” while you watch Bo Bichette flail and miss at another hung slider. You can see your leaguemate’s smile as their season turns around thanks to the recent revitalization of Rafael Devers and wish for your own salvation. 

But at the end of the day, we may just be putting too much faith in ADP. Especially this early, it is not nearly as predictive as we may think. I came into this with the prediction that, as the collective wisdom says, pitchers are more likely to be risky and volatile assets, and therefore will be more likely to underperform as a group relative to ADP. 

The evidence seems to tell a different story. 

Let’s start with some of the biggest overperformers and underperformers to start the season, both at the dish and on the mound.

Clearly, a variety of factors are at play here, from injuries to mechanical and approach changes, to flat out batted ball variance. Over a large enough sample, this may very well flatten out. Regression to the mean is the expectation. 

But at the beginning of the ninth week of the fantasy season, this has yet to happen. Right now, for hitters, starting pitchers, and relievers alike, ADP is a remarkably poor predictor of performance. 

In the above chart, I use ADP data from the NFBC for the period of March 1 to March 26, this year’s Opening Day, and map each player’s ADP to their Fangraphs Player Rater value through Sunday, May 17. For undrafted players, I substitute an ADP of 1000 and I lump SP/RP eligible players into the SP category for cleaner analysis. 

There’s clearly a lot of variance in this data. Overall, regardless of position, ADP is at least somewhat predictive of dollars earned. But this is using the entire player pool, including hundreds of players that are largely irrelevant in most formats. I’d be willing to bet that none of you drafted Huascar Brazobán, who’s accrued $7.70 of value as a win magnet in the Mets bullpen, and despite this, he is not worth rostering. 

What happens when we remove the Brazobáns and other players who fall far outside of most reasonable draft prep lists?

At each ADP-cutoff, shown on the x-axis, I charted the correlation coefficient for the relationship between ADP and Dollars, grouped by position. For hitters, represented by the brown line, the relationship between ADP and dollars grows weaker as the ADP cutoff removes more of the player pool. 

For starting pitchers, represented by the green line, the opposite is true. As you go deeper into the player pool, the predictive power of ADP on dollars earned grows weaker. 

This suggests that the market, captured by NFBC ADP, is actually somewhat effective at pricing starting pitchers within the first 250 players drafted, despite all of the injury risk that accompanies them, and is far less effective in that same range of the draft at pricing hitters.

Here are those correlation coefficients displayed in table form. 

Admittedly, it is still early. Regression will likely come and smooth over some of this variance. But clearly, with thousands of plate appearances and pitches thrown, we can see that the market is not able to confidently predict player performance, even when using ADP from drafts that occurred less than a month before Opening Day. 

If you made it down here, past the harsh reminders of Seagers gone wrong and through the tint of the hindsight goggles we’re all provided upon entering the theater of fantasy baseball, I hope you take one lesson away from all of this. And I hope it isn’t a Camus-fueled nihilism that sees you ditch fantasy baseball for golf, or some similarly frustrating pastime. 

Instead, I hope you embrace the random. Pick up someone you’ve never heard of and don’t at all trust, and let yourself go.