St. Louis Cardinals postseason roster: Pete Kozma might make the NLDS roster because of Mike Matheny's small-minded devotion to small sample sizes

The St. Louis Cardinals drafted Pete Kozma out of high school in the first round of the 2007 amateur draft. Shortly thereafter, Jeff Luhnow (who was working in the Cardinals front office at the time) compared Kozma to Nomar Garciaparra. Nowadays baseball fans might nod because they remember Garciaparra but the boldness of such a comparison would be lost on them because of the way one’s memory fades with time. In 2007, Garciaparra was still playing. In fact, he had been a National League All-Star the year before with the Dodgers, finishing the season with a .303/.367/.505 slash line that was strikingly on par with the .313/.361/.521 career line he would have after his last game in 2009. By comparing Kozma to Garciaparra in 2007, Luhnow made clear that the Cardinals brass projected Kozma to be very, very good.

The Garciaparra comparison Luhnow dropped in 2007 was never fair. Nomars come around once maybe twice in a lifetime (although there sure were a lot of them in the mid-to-late 90s and early aughts). But Kozma’s minor-league batting performance makes it seem absurd if not cruel. Kozma toiled in the minors for five years, fading from top-prospect lists as his inability to hit demonstrated itself at each stop. Sure, he hit .284/.363/.398 for the Quad Cities over 434 plate appearances in 2008. But after that, if Kozma notched more than 100 PAs, he never hit for an average higher than .231, got on-base at a rate higher than .318, or slugged at a clip above .384. In 2012, Kozma hit .232/.292/.355 with Memphis—by then, he was playing all over the infield in the hopes of ascending to the majors as a glove-first utility man.

But late in 2012, Rafael Furcal suffered a season-ending injury. With no organizational depth at shortstop, the Cardinals were effectively forced to not only promote Kozma to St. Louis but to play him every day during the home stretch of a pennant race (though the Cardinals were chasing one of two wild card play-in game berths as opposed to a pennant). For a month, Kozma hit like Garciaparra only better. Kozma notched just 64 plate appearances during the regular season. During them, he posted a .333/.383/.569 line. There’s an argument that he carried the team to the wild-card game—or that he at least shouldered the biggest load for the Cards down the stretch.

Kozma went hitless in the play-in game and was a participant—yielding to the lumbering Matt Holliday so that a pop fell to the grass in somewhat shallow center field—in the infield fly call that has come to define that game. But he re-conjured his September magic during the NLDS, hitting .250/.450/.500 with one of the biggest hits of the year during the team’s amazing Game 5 comeback. During the NLCS, though, Kozma looked like the Kozma his minor-league batting line indicated he was.

Even after Kozma’s excellent fall, the Cards were understandably reluctant to project any sort of major-league role for him in 2013. Furcal had rested and rehabbed his stretched UCL, so he sat atop the shortstop depth chart. The front office preferred a different Plan B to Kozma, so it explored signing free agent Stephen Drew but those talks didn’t get off the ground due to the Cards’ inability to guarantee Drew playing time with Furcal the prioritized incumbent. St. Louis was stuck with Kozma as its shortstop backup. And so the club was stuck with Kozma backing up Furcal.

When Furcal’s previously strained UCL snapped and ended the veteran’s 2013 season before it began, Kozma became Plan A at shortstop. Throughout spring camp, Cardinals front office members and coaches talked about how much Kozma had grown as a hitter. It was as if the front office had decided to see if they repeated this incantation over and over enough in the media, it might cause reality to shift and Kozma to hit all season like he had done the autumn before.

But the real Kozma was revealed over the 162-game grind. His glovework was excellent, but his batting stunk. It was so bad in fact that manager Mike Matheny started sending Daniel Descalso (who finished the year with a .238/.290/.366 line) out to stand with a glove on his hand where shortstops typically stand on the infield. Given the lack of range Descalso displayed, it’s unclear whether Matheny ordered him to do anything other than stand at the position. Kozma ended the season with a .217/.275/.273 slash line. His .548 OPS equaled a 53 OPS+.

Understandably, the Cardinals prioritized shortstop as the position at which to upgrade during last offseason and signed veteran Jhonny Peralta, who has been a revelation. With Peralta in the mix, the Cards relegated Kozma back to Triple-A—his hitting so bad that even with his excellent defensive capabilities, he was unable to unseat the no-hit, no-field Descalso as the apple of Matheny’s utility-man eye. During his third tour of the Pacific Coast League with Memphis, Kozma batted a not half-bad .248/.330/.372 in 437 PAs over 117 games.

The Cardinals recalled Kozma to the majors in June, but didn’t play him and placed Kozma on revocable waivers in order to clear a roster spot for rookie Marco Gonzales, so that the lefty could make his MLB debut. Kozma passed through waivers and went back down to Memphis, where he stayed until late August. This season, Kozma played in six St. Louis games with four starts. He’s tallied just 26 PAs.

Despite all of this, Kozma appears likely to not only make the Cardinals’ 25-man NLDS roster, but start Game 1 at second base. Why? Because he has four hits off Clayton Kershaw in eight at-bats. Three of those hits were doubles. Kozma’s line vs. Kershaw in nine PAs: .500/.556/.875. That makes for a 1.431 OPS.

This is just the latest exhibit in the case of Matheny’s small-sample-size myopia. Earlier this year during an episode of The Best Podcast in Baseball (a podcast to which all Cardinals fans should listen because it’s excellent), St. Louis Post-Dispatch beat writer Derrick Goold shared a story of Matheny making a joke about his lineup and small sample sizes. But self-deprecation doesn’t excuse the manager consistent and foolhardy devotion to results over a number of PAs so small as to be meaningless.

This type of foolish Matheny decision-making is not new.

Entering 2014, Descalso posted a .375/.500/.500 line against Volquez over ten PAs. That’s right, successful results over ten PAs (eight at-bats) led Matheny to conclude Descalso merited playing against Volquez. Descalso went 0-for-5 against Volquez in 2014. Now Descalso is batting .231/.333/.308 against Volquez.

More recently, there was the decision to send Mark Ellis in to pinch-hit against Aroldis Chapman. I don’t necessarily mind the decision from a straight skillset analysis. Ellis is a righthander who is good at making contact, which makes him an okay choice to face the flame-throwing, strikeout-inducing southpaw. But it emerged that another considered was Ellis’s past success vs. Chapman. He was 3-for-3 against Chapman entering the series. Today, he’s 3-for-6. One wonders whether Ellis’s .500 average against Chapman might, in Matheny’s eyes, have merited a PA against the lefty next year if Ellis were still under contract with Cards. Thankfully, we’ll never find out.

Matheny should not start Kozma against Kershaw because results over nine PAs should never outweigh thousands of PAs worth of performance. Kozma is as bad a batsman as Kershaw is an excellent pitcher. Kozma is a career .236/.297/.320 big-league hitter with a .617 OPS that works out to a 71 OPS+. To expect a player as unskilled at hitting as Kozma to continue to experience successful results against a pitcher as good as Kershaw is as unfair as comparing an 19-year-old high school draft pick to Nomar Garciaparra.

Correction: This post incorrectly stated that Kozma was 18 years old when the Cardinals drafted him. He was 19.