The Predictive Power of a Dodgers Spring Training


When spring training began earlier this month, an interesting article written by Dan Rosenheck gained some heated attention by Saber-nerds far and wide. In the article, Rosenheck took on the conventional baseball wisdom that spring training statistics don’t matter. He uses a number of theoretical arguments, statistical trends, and other big words to make his point. “The claim that spring-training numbers are useless is wrong,” states Rosenheck. “Not a little bit wrong, not debatably wrong—demonstrably and conclusively wrong.”

More from LA Dodgers News

On the surface, one would think individual performances do matter and do mean something in spring training especially for unproven commodities. Time and again we’ve heard Jerry Hairston Jr. and John Hartung on SportsNet LA talk about Los Angeles Dodgers prospect Alex Guerrero and his strong offensive spring so far. His fielding is still somewhat suspect, but as the conventional wisdom goes, “If you can hit, they’ll find a position for you.”

If Guerrero finishes the spring hitting .460 as opposed to .160 the odds of him making the team are much greater. In that case statistics mean something. If Yasiel Puig finishes the spring hitting .167 (again) there is no real fear that he will perform similarly through the regular season. That is a case where statistics mean nothing.

This makes me wonder if it’s possible to look past individual performances and see if we can make any predictions on the run production of the Dodgers this year as a collective. Throughout the spring, some players will over-perform, others will under-perform. I would wager they cancel each other out leaving us with a team result that is close to what the regular season might look like.

“You’re judging the offense already??” asks the Super Fan. “But it’s only been 14 games! It’s not enough data!”

Agreed. I also expect run disparities to adjust as batters like Puig and Adrian Gonzalez bring their averages up, while Joc Pederson, Jimmy Rollins, and Scott Van Slyke begin to regress. While some individuals shift one way and others shift the other way, the team result might give an accurate harbinger of the season to come, or so that’s the theory.

“Baseball games played yesterday, no matter how unreliable spring-training numbers may be, should offer some useful information, however modest, when all the other data at one’s disposal is at least six months old.” –Dan Rosenheck

I also get that pitchers, hitters, defenses, and managers don’t take the same approach to spring training situations as they will during the regular season. I listen closely to Rick Monday giving these insights during Dodgers broadcasts. I totally get it. Taking all that into consideration, like Rosenheck, I still believe there are enough controlling factors in baseball to allow for some reasonable insights if not apples-to-apples comparisons between spring training and the regular season.

So after just two weeks of the Cactus League, let’s take a look at the Dodgers’ offense versus opposition offense to see if we can make any predictions about the upcoming season. Surely a more cringe-inducing sentence has never before been uttered, but here we go anyway.

The grumblings about limited run production began the moment Andrew Friedman agreed to trade away Matt Kemp and refuse to extend the contract of Hanley Ramirez. To this day fans on the call-in shows still believe the Front Office should have demanded more in return for Kemp, and that the Dodgers’ firepower this year will be too limited to progress in the playoffs even if they do win another division title.

The rejoinder came in the form of Rollins, Howie Kendrick, and the Dodgers’ improved ability to generate runs by putting the ball in play, ostensibly in a wider variety of ways. Through their first 14 Cactus games, the Dodgers have scored 72 runs while surrendering 42. They’ve done this in part by hitting 17 home runs with 9 of them hit by likely first teamers.  Looks like their offense is exploding, right? What’s all the concern?

Live Feed

MLB rumors roundup: Tatis and Yankees, Dansby Swanson, Carlos Rodon
MLB rumors roundup: Tatis and Yankees, Dansby Swanson, Carlos Rodon /

Call to the Pen

  • Why Dansby Swanson makes sense for the Dodgers, explainedFanSided
  • MLB rumors: Dansby Swanson could make Braves worst nightmare a realityFanSided
  • With departures, Los Angeles Dodgers must retool for 2023Call to the Pen
  • MLB Insider: Diamondbacks showing interest in Justin TurnerFanSided
  • Latest World Series Odds for Every Team Following Big Free Agency Signings (Astros Remain Favorite After Jose Abreu Signing)FanSided
  • “Not so fast!” says Super Fan. “Good statistics, especially in such a small data set, require the removal of anomalies.”

    No problem. Taking out the anomalous 10-1 win over the Brewers, the Dodgers have outscored their opponents 62-41 in 13 games. Still not too bad.

    “Not so fast again! None of the starters play all 9 innings! There’re lots of green prospects in the lineup, and high school kids doing the pitching!”

    I’ll concede that point, but generally the first four innings of each game is when you’ll see the largest number of expected starters playing for each side. Controlling for only the first four innings, the Dodgers have still outscored their opponents 24-21. My intuition tells me this is a more accurate comparison of Dodger run production versus the opposition.

    Except we still have one more factor. Again removing the anomalous whitewashing of the Brewers where the Dodgers posted 8 runs in the first four innings to the Brewers’ 0, we are left with a grand total of 16 runs for and 21 runs against.

    Not a very heartening stat to be sure.

    Fortunately, I don’t know of any extensive studies that show this hackneyed method of team stat comparison as holding any water. If it turns out that the 31 or so Cactus League games come and go, and the Dodgers first team lineup scores fewer runs than their opposition I would still like LA’s chances of progressing to the post season. Even Rosenheck concedes that the predictive power of spring training statistics have a limit. “At best,” he says, “it might cause teams to give a handful of promising players each year a chance they wouldn’t otherwise have had.”

    So relax and enjoy the offensive explosion of this year’s Cactus League Dodgers.  It’s only spring training after all. And spring training stats don’t matter. Or do they?