clock menu more-arrow no yes mobile

Filed under:

Spring Numbers: Meaningless, or COMPLETELY Meaningless?

MLB Florida and Arizona Spring Training - SB Nation

"Mentally and physically, I feel great, and that's the important part." -- Doug Davis

It's something you hear every year around this time: someone, usually a starting pitcher after a bad outing, disparages the significance of actual numbers put up in spring training. And it's pretty much one of those unwritten rules that they are meaningless. Even the overall record posted by a team is deemed to bear no relation to what they do. When Derek Zumsteg did the math, matchng the 2004 standings for spring training and the regular season, he found a correlation of just -0.05, or as he put it, "You would get that from comparing two sets of 30 random numbers."

While writing this, I discovered that over ar Royals Review, our cousins already looked at this, and came up with some reasons for this. In summary, these are sample size; uneven competition; unreal play and rarefied air. There is some truth to all of these in varying degrees - rarefied air applies a good deal less in the Grapefruit League, for example. There are also other factors: motivation, for example - I'd expect a player whose spot on the club is assured, to be exerting much less effort than a non-roster invitee, fighting for the 24th or 25th spot. However, one would still expect talent to shine through, to a certain extent

Over at Acta Sports, John Dewan wrote about one possible metric last year: "We took all spring training hitters and found that, as expected, about half of them do better than their career norms in the upcoming season, and about half of them do worse than their career norms. However, when we chose only those players doing exceptionally well in spring training, we found that about three-fourths of them performed better than their career average during the upcoming season" They defined  "exceptionally well" as slugging 100 points higher than their previous career SLG, and provided a list of those who did so to the greatest extent in spring 2008.

There were some palpable hits. Standing out is a young prospect with a troubled past and less than 100 major-league games at that point. 32 HR, seventh-place in the MVP voting and a performance for the ages in the Home Run Derby later, Josh Hamilton is now a bona-fide star. Two Diamondbacks were also listed, though the results there were a good deal more mixed. Topping the entire list, with an SLG no less than 445 points higher, was Chris Snyder, and he certainly had a fine season, with career highs in HR, RBI and OPS. However, the other Diamondback, at +308, was... Chris Burke. Hey, no system is perfect.

My other thought was that expecting one month of spring training to foreshadow an entire 162-game season is a little much. So I decided to focus more closely on the performance in the first month of the season, as being the one most likely to reflect success or failure in the Cactus League. For the past three years, here are the batting averages and ERAs respectively, for the five most impressive AZ hitters and pitchers in spring training, followed by the same number for the player, in the regular season up until the end of April. To qualify, a player needs 25 at-bats or 5 innings pitched in spring training, and the same again by April 30th.

2008
Chris Burke: .395/.128
Justin Upton: .350/.327
Orlando Hudson: .333/.270
Chris Snyder: .323/.206
Conor Jackson: .320/.348

Edgar Gonzalez: 1.64/6.55
Yusmeiro Petit: 2.92/4.70
Chad Qualls: 3.00/0.00
Juan Cruz: 3.18/1.69
Tony Pena: 3.85/5.11

As noted above, Chris Burke would be the poster child for the meaninglessness of spring training. He roared out of the starting gate in March, smacking three homers in a mere 38 at-bats. Unfortunately, that would be more than he'd have in the entire regular season. However, Justin Upton showcased his April form early, though it didn't last much beyond the first month, and CoJack even improved upon his impressive pre-season form. Snyder suffered from a less extreme case of Burke-itis, barely hitting above the Uecker line once serious games started.

On the pitching side, most of the bullpen were lethal in spring, but the results after Opening Day were a lot more mixed. However, the number of spring innings by the top five pitched was from six to only 12.1, probably making the results even less reliable than for the hitters.Qualls and Cruz kicked butt in April; Peña and EdGon, not so much. And what of Brandon Webb, who went 6-0 with a 1.98 ERA in April? His ERA in spring was 7.90.

2007
Orlando Hudson: .438/.352
Alberto Callaspo: .434/.212
Scott Hairston: .426/.220
Miguel Montero: .419/.243
Carlos Quentin: .357/.220

Brandon Webb: 1.64/3.21
Micah Owings: 1.93/2.93
Jose Valverde: 2.70/1.64
Brandon Meddars: 2.70/4.15
Tony Peña: 2.70/1.88

Some monstrous hitting performance from the Diamondbacks that spring, most notably probably Hairston, who had six homers, 15 extra-base hits and 17 RBI in 54 at-bats. However, his batting average was almost cut in half after Opening Day, and the same goes true for almost all the top performers. Though expecting them to keep hitting .400 is probably a bit much, only O-Dawg even managed to bat above .250, following on from another excellent spring for Hudson.

Leading the chart for pitchers is our ace, a sharp contract to his pre-season struggles in 2008. In contrast to the hitters, all the pitchers did seem to experience some kind of honeymoon effect moving into the regular season, posting early ERAs better than league average - considerably better, in some cases. Overall, the top five had a collective ERA of 2.92 in their 95.2 April innings

2006
Orlando Hudson: .473/.237
Johnny Estrada: .449/.300
Jeff Davanon: .413/.311
Conor Jackson: .389/.273
Chris Snyder: .333/.300

Greg Aquino: 0.96/6.10
Jason Grimsley: 3.18/5.82
Luis Vizcaino: 3.24/2.25
Miguel Batista: 3.45/6.33
Jose Valverde: 3.86/3.24

What is it with O-Dawg? In three springs with Arizona, he hit .424 (59-for-139), but in the first month of the regular seasons, his BA drops to .288 (89-for-309). That's below his average with us, 294, and matches his career variation overall, at six points worse in the opening month (.276/.282). Outside of him, however, the other hitters did fairly well. Jackson and Snyder each make their second appearance in the list as spring blossoms.

Well, if we though we were onto something with the pitching in 2007, we're right back where we were. Grimsley must have left his steroids down in Tucson, and Greg Aquino appears to have feasted upon Double-A hitters there - major-league ones, not so much. However, Vizcaino and Valverde did well: in case you're wondering, the latter had a 4.50 ERA for Houston last spring, so didn't quite match Hudson in three-peating.

Basically, this seems to back up the assertion that you really can't predict, even the first few weeks of the regular season, based on spring performances. Some players do continue to perform well, but others don't, and there's no apparent rhyme or reason to it. Beyond the Box Score looked at this March to April link in more statistical and less anecdotal form, and found no link in BA, and only a weak correlation (no more than 0.15) for OBP and SLG. I'm not sure what those numbers would be, if you did the same comparison between April and May, say.

I can't help feeling there is some kind of correlation, if only we looked in the right place. Maybe those who played winter ball put up better spring numbers than those who didn't - having 'their eye in'. Or maybe they are worse, not being so fresh. However, as far as them being any use for the major-league season, or even the first few weeks thereof, the evidence is pretty thin. That's something to cling to, in the wake of the sound thrashing administered to the Diamondbacks by the White Sox in Tucson yesterday.