Every year around this time, I get the phone call. And even though it's an annual occurrence, it always seems to take me by surprise.
"AJ, it's me, Tristan. I'm on vacation and I need you to take over my top 250 rankings. Hold on. Room service is here with my chicken parm. Gotta run. Thanks!"
Even if you play a lot of fantasy baseball and watch a lot of games on a regular basis, this is no easy task. It's made even harder when you realize you have to work off a list created by someone else; after all, we don't all value players in exactly the same way. It's not that simple to predict who might be the best player going forward when we may not even agree who has been the best player thus far. That said, I will not let Mr. Cockcroft down! Let's dive in.
The season is about one-third of the way finished, as the end of May will be upon us at the end of this week's games. There's always a lot of debate about exactly when the sample size is large enough to be of any use to fantasy players in terms of predicting future performance -- or when it's still just background noise. But even if we were to concede that two months of a season might well be a large enough number of at-bats or innings pitched to be considered meaningful, exactly how do we use this information? Even if we agree it does mean something, what is that something?
For example, let's take a look at Dallas Keuchel of the Houston Astros. He's currently ranked as the No. 4 starting pitcher on the ESPN Player Rater, and No. 17 overall. Where should a pitcher like him rank in the Top 250 for the remainder of the season? How much weight should we give to his overwhelmingly strong start to 2015? How much should we hold firm to the original projection we gave him before the season started? Where do we draw the line?
Let's take a quick look at four different projections for Keuchel for the rest of the year, each derived from a different set of assumptions:
At first glance, which one of these seems to be the most likely outcome for Keuchel? Obviously, the numbers get a lot less optimistic as we move down the chart -- but exactly where should our expectations land if we're striving to have the most accurate outlook? Allow me to explain where we got these numbers, and perhaps we'll have our answer.
Assumption Set A: This projection was generated by simply using Keuchel's first two months and doubling them up the rest of the way. Certainly, the likelihood of this outcome is extremely low, but it can easily be used as one of our data points, albeit (in this case) on the very high end of the scale. This projection assumes that the Keuchel of the first two months of the season is a completely new pitcher, and takes no notice of any historical performances to the contrary.
Assumption Set D: Set A was the ceiling, and this is the floor. This projection derives from the assumption that ESPN's preseason projection for Keuchel was completely accurate. In that case, his final season numbers will ultimately match what we thought going into the season, and the stats you see here are simply how the pitcher would have to perform over the next four months in order for him to do so. Another case of extremely low likelihood, but it gives us the proper frame.
Assumption Set C: Keuchel was No. 46 among starting pitchers in our preseason projections, and if we were happy with that basic assessment of his ranking at the time, we're probably thinking his first two months are merely an outlier, with a lot of regression to come. This line is predicated on the assumption that our original projection was, overall, an accurate assessment of Keuchel's true abilities all along, and these stats represent his expected rest-of-season numbers without being influenced by his hot start.
That brings us to Assumption Set B, which lies in between believing in "Keuchel the Cy Young contender" -- who we've seen so far in 2015 -- and the "who's this Keuchel guy" we had seen in the past. Given this projection, we factor in both the fact that he has never performed at this level before and that he's made huge improvements that may stick and propel him to the elite level.
That makes Set B the most likely outcome, and if we were to add those numbers to his current stats, it would make him a 16-game winner with a 2.49 ERA, 146 strikeouts for the season, and a WHIP of about 1.00. That seems pretty reasonable to me. And if we were to follow a similar kind of procedure to figure out where the Set B numbers were to lie for the rest of the player pool, we'd paint a picture that looks an awful lot like the Top 250 ranks that we'll see below.
It's not the only method you can use, and it's by no means a guarantee of predictive success. However, it does a good job of attempting to remove any confirmation bias that we may unintentionally allow to influence our opinions of players whose to-date performances have not been in lockstep with what we thought would be the current reality.
New ESPN position eligibility
The following players added new position eligibility within the past week. As a reminder, position players need to appear 10 times at a new position to gain in-season eligibility, while pitchers need to make three starts to earn starting-pitching eligibility or five relief appearances to earn relief-pitching eligibility:
Daniel Descalso (SS), Adam Rosales (2B), Jonathan Herrera (2B), Elian Herrera (3B), Danny Valencia (OF), Alex Guerrero (OF)
The following players are within two games of earning new eligibility, with their total games played noted at the listed position:
Garrett Jones (OF, 8 games), Tim Beckham (SS, 9 games), Danny Espinosa (3B, 8 games), Marwin Gonzalez (1B, 8 games), Rey Navarro (2B, 9 games), Cory Spangenberg (3B, 8 games), Kennys Vargas (1B, 8 games), Alex Guerrero (3B, 9 games), Max Muncy (3B, 9 games), Pedro Ciriaco (3B, 8 games), Tyler Moore (OF, 8 games)
Going-forward rankings: Week 8
Listed below are my updated, going-forward rankings. Don't worry! I've tried not to upset Tristan's apple cart with too many big swings this week. These are based upon an ESPN standard league of 10 teams and Rotisserie 5x5 scoring. Click here to see these rankings sorted by position.