Periodically, I have a serious problem with an aspect of “predictive baseball.” When I have a problem with something baseball-related, it bleeds into my writing. This time, it’s bleeding into a prospect piece of a player who hasn’t played a game in the USA with a box score. My problem with predictive baseball seems as applicable now as ever, so today is as good of a day as any. My Raino Coran perspective is as good of a Main Street at OK Corral as any.
Many of you are employed by predictive baseball, in one form or another. More of you subscribe to the basic tenets. When taken as a loose guideline, they generally make sense. A person with a career .800 OPS ought to have an .800 or so OPS given a full season, give or take. Good players generally do well. Historic strugglers generally struggle. However, two problems override the general norms. One is, nobody has any seriously valid evidence about how a player will do when he gets his first legitimate look as a starting regular.
Three Cubs were forced, effectively, into regular roles this season. If a wealthy man would have bet a third of his fortune on the going line (failure) for Rafael Ortega, Patrick Wisdom, and Frank Schwindel, he’d have been penniless by October. Ortega outplayed Joc Pederson. Schwindel outplayed Anthony Rizzo. Wisdom set a rookie home run record for the Cubs. But the models said they would be bad.
When the Astros received no MLB value out of Mark Appel, they scrapped their prior predictive model as worthless. They needed to find something more predictive. I haven’t heard a thing about the predictive model of fringe-to-regular players. The major outfits don’t care about them. (And, no. I don’t believe in “three outliers on the field at the same time” for a speck of a second.) My predictive model is: “I’m not going to act like I can know what I can’t know.”
Raino Coran, outfield
Born May 13, 2002, Willemsted, Curacao
Signed internationally by the Cubs in February 2021
Each cycle when I do a full-scale writing barrage, I look for a few specific types of players. This cycle, Coran is one of them. Generally a starting outfielder in the Dominican League in 2021, his slash line was a bit odd. Despite a .179 batting average in 106 at-bats, three homers and 36 (!!!!!) walks bumped his OPS up to .710 in a league where the average OPS was .682. Coran should play next season, likely getting near everyday chances. The question is, where?
This leads to my other problem with the predictive model, which seems to like to believe 2022 will flow naturally from 2021. Which I consider garbage. In one of the Nelson Velázquez articles I’ve read over the last few months or so, his development spurred from a more serious level of commitment from the player. Which the predictive model doesn’t assess until it resonates in games, which seems as valid at wagering on a game after it takes place.
If a predictive model is going to predict, it ought to predict before the early election results start reaching the “30 percent of votes are tabulated” mark. If you can’t predict, admit your system wasn’t good enough to assess that Kirby Yates was going to be an elite reliever for two of three years after his waiver claim by the Padres.
Coran should get four or five starts a week in 2022. The questions are, where? And how could a system be put in place to figure out where? Leave him in the Dominican, and he could be far better than the league, and waste a bit of his time. Move him to the States for 2022, and he could be “Ed Howard in May and June” underwater. If only there were a sort of intermediary step to trial him against better pitchers to see if he represents he can hit against said pitchers, to aid in the decision process. Either way.
Alas, there is. Or, at least, there was, at one point. Extended Spring Training takes place shortly after the full-season players have been sent off to their destinations. Teams have leftover players who won’t initially make the full-season squads. (That would be roughly 100, depending on if the lockout continues into the minor league season.) About 80 players are on the stateside XST roster. Some are injured, and the rest ought to be angling for spots on the (yet to be determined) one to three other rosters.
Cristian Hernandez figures to be stateside, as will one of my personal favorites Pedro Ramirez, who closed with an OPS of .919. However, how many arms will be able to gets outs in Mesa is a perennial question. Playing as many Extended Spring Training games and innings as possible should be an organizational goal, along with paying the players for participating. The best way to find out if Coran is ready for the USA ought to be to get him as many looks as realistic against pitchers with college experience, in the extended spring training experience. In the states, in XST.
Maybe he’s proven not quite good enough. And his experience helps him in the DSL. Or, maybe, he is proven good enough, by the extra looks. The seemingly best predictive method would be more along the lines of monitoring his actual at-bats in actual games (scrimmage or not) after his off-season adjustments. That seems far better than assuming that past results will guarantee future success, which is a disclaimer on every CNBC look at past results.
I have no delusions about Coran’s baseball future. His commitment to adjusting to pitches that have given him trouble in the past will be a key, and I have no idea how he’s doing at that. I’m completely willing to laugh off anyone valuing my predictive abilities for future production. For any player. That’s why games are scheduled. Track a game, and see if anything has changed. If it has, adjust your expectations off of what’s happening. Predictive models that had the Giants as a third-place team had serious flaws in them. Yet, the same model is largely cuddled and valued as meaningful. Even though it will likely be bunk on things that are unknown. Which are the only things worth speculating about, anyway.