The Z Files: Hey Todd!

Last week's review of my new Top-20 hitters generated some interesting discussion concerning both the method and players conspicuous by their absence. Some were addressed in the comments section of the piece as well as on podcasts, Sirius XM and social media. But since there are probably many that had the same questions and didn't see the replies, let's address a few of the more compelling queries now.

Hey Todd, I heard you say the ranks are so close that a gain or loss of 15 at-bats can really change things. How much of a difference will that really make?

Let's look at a couple of top-20 hitters from my May 14 update, Chris Davis and Francisco Lindor. Both landed just inside the Top 20, Davis at 17 while Lindor checked in at 19. For the record, Eric Hosmer fell in between but Davis and Lindor were chosen since one relies predominantly on power while the other has a large speed component to his production. I added and subtracted 15 at-bats from each, then prorated their stats accordingly.

With an additional 15 ABs under their belts, Davis jumps from 17th to 13th in the rankings and his 15-team mixed league value rises from $26 to $28, while Lindor moves from 19th to 14th and also sees his value rise from $26 to $28. Take away 15 ABs, however, and Davis falls to 21st and $24, while Lindor plummets all the way to 25th and $23.

So 15 at-bats can raise or lower a hitter in this range anywhere between four and six places, with the speedy Lindor on the high end and the slugger Davis on the low end. Fifteen at-bats is sitting out a series after tweaking an ankle or getting hit on the hand with a pitch. As of Tuesday night, the San Francisco Giants played five more games than the Tampa Bay Rays, meaning Evan Longoria could be projected for 15 more at-bats than Matt Duffy just based on games left. Also keep in mind the further we progress into the season, the less 15 at-bats will impact the run. The difference between 635 and 650 at-bats isn't as telling as the difference between 435 and 450, which isn't as influential as that between 235 and 250.

Hey Todd, don't you trust your own ranks? On a podcast you said you wouldn't always trade for a better player according to your rankings.

While I'm not positive, I think this was in reference to Lindor and Carlos Correa and saying I'm not sure I'd deal Correa for Lindor even though the rest-of-season outlook was actually better for Lindor. While I absolutely trust my own projections and rankings, I'm not naïve. They're not going to be 100 percent accurate, and conventional wisdom estimates the best anyone can do is around 70 percent. Rankings are simply a starting point. Upside, track record, injury history and roster construction all shape a draft list. Some of this is baked into a projection but not all, and as we progress further into the season, team needs become increasingly crucial. If we were doing a draft right now, just counting stats going forward, I'm honestly not sure which shortstop I'd prefer. Lindor is safer and looks like he'll be a solid contributor to the precious stolen base category while Correa has the talent to be the best player in the game. Chances are I'd actually opt for Lindor, in part on my rankings but moreso because I like his floor better than Correa's. Others will call me crazy and grab Correa's ceiling.

Hey Todd, no Mookie in your Top 20? Are you freaking nuts?

I took the liberty of editing that one just a bit. Keeping in mind the update was based on stats through May 12, in which Mookie Betts placed 26th. The reason at the time was a small dip in contact rate as well as a slightly depressed hard contact mark. One of the complications in projecting Betts, as well as other talented young players with a limited track record, is that we don't know their baseline. I may have been too light on Betts coming into the season, still drawing from his MLEs (Major League Equivalencies) and dragging his baseline down. At the time of the update, his rest-of-season batting average was .277. Admittedly, this didn't pass the sniff test even though he was hitting just .253 at the time. My original expectation of .289 was lower than most but still, I'd have taken the over on .277. If someone asked what player outside of the Top 20 would be in it come season's end, the answer would have been Betts without hesitation. For the record, after my latest update, Betts climbed to 12th.

Hey Todd, if a guy can jump up that much in just a couple of weeks, doesn't that mean the system isn't very good? Should a player really move that much based on a couple weeks of games?

Confession time: I asked myself that very question, though I don't talk to myself in third person. The answer is while all projection engines need constant tweaking based on new data and research, a player can indeed move that much. The influence on playing time was discussed above. Preseason projections, of which playing time is an integral component, are based on initial expectations and I doubt anyone would argue the Boston Red Sox offense is better than expected. This turns the order over more which yields Betts more plate appearances, not to mention boosts his run production. The team improvement is slowly captured by the engine, keeping in mind the team will slump like all MLB squads do at some point. Still, some of the improvement has been accounted for both in terms of a change in playing time for Betts as well as the potential for more runs and RBI.

There's another factor that feeds into the 70 percent cited above. We're not always right. The underlying principle of this process is a player's over/under line is altered. Regardless of where that falls, the implication is there's an equal chance the player performs better or worse than that mark. Then, that new mark gets baked into the ensuing update. The concept may be hard to grasp, but the same player can bounce up and down all season but that doesn't invalidate or discredit the method. In fact, I prefer a process that ignores bias and just goes by the numbers and then it's up to all of us to season to taste.

Hey Todd, what does your little black box say for Joey Gallo? What's his projection after striking out less in the minors?

Screw my little black box, I need a freaking crystal ball for this one. Yeah, I took the liberty of editing my original response a little here too. Sure, I can drop Gallo's 2016 numbers at Triple-A Round Rock into my spreadsheet, do the MLE adjustment and apply an age factor which will inch his contact rate upwards. But is that the real scenario? If Gallo's 71 percent contact rate is a true indication of a skill's boost, the impact will be more than the couple of percentage points improvement from an original expectation in the mid-fifties. Yes, that's how poor it translated coming into the season. Are you really interested in Gallo if my little black box ups his contact rate from 57 to 59 percent based on 106 Triple-A plate appearances?

Either he's better or he's not. If it's the former, he still needs to show he can carry the gains over to The Show as well as continue his growth. Maybe you had the luxury of seeing him play and can add a little scouting to your evaluation. Even then, we're all just guessing. Or maybe hoping, which is fine, as that's now an element of the game we play. "Betting on the come" is the new "practicing excruciating patience". So yeah, when I do my next update there will be a projection for Gallo. But my level of trust is about the same as the chance I'll completely eliminate pizza from my diet even though it's on the banned list from my doctor.

Hey Todd, do you update pitching ranks too? Any chance you can write about those?

Yeah, I'd say there's an excellent chance. That's the plan for next time. And here's a spoiler: Clayton Kershaw is on top.