Domonic Brown’s Spring Is Still No Guarantee

Domonic Brown is tied with Mike Morse for the spring training lead in home runs with six and we still have a week and a half of games left. The once top prospect had a questionable future with the team entering spring training, but he has played his way into an everyday job in the Phillies’ outfield. Aside from the home runs, he is hitting .397 and has drawn nearly as many walks as strikeouts (six to eight) in 63 at-bats. That makes this perhaps the most appropriate time to flash the reminder: spring training stats aren’t predictive of regular season success or failure.

Last year, 15 players hit six or more home runs in spring exhibition games in the range of 45-80 at-bats. There was no correlation whatsoever between their spring hitting and regular season hitting.

Note: at-bats were used instead of plate appearances because MLB.com doesn’t list PA with their spring stats.

Spring Training Regular Season
Player Team AB HR AB/HR OPS AB HR AB/HR OPS
Freeman, F ATL 78 7 11.1 .959 540 23 23.5 .796
Hague, M PIT 55 7 7.9 1.200 70 0 0.0 .527
Pujols, A LAA 60 7 8.6 1.287 607 30 20.2 .859
Boesch, B DET 73 6 12.2 .928 470 12 39.2 .659
Duncan, S CLE 56 6 9.3 .759 232 11 21.1 .676
Dunn, A CWS 57 6 9.5 1.005 539 41 13.1 .800
Gamel, M MIL 63 6 10.5 .879 69 1 69.0 .641
Guzman, J SD 65 6 10.8 .999 287 9 31.9 .737
Hughes, L MIN 66 6 11.0 .984 23 0 0.0 .255
Raburn, R DET 56 6 9.3 .994 205 1 205.0 .480
Ross, C BOS 46 6 7.7 1.257 476 22 21.6 .807
Soriano, A CHC 56 6 9.3 .963 561 32 17.5 .821
Trumbo, M LAA 58 6 9.7 1.030 544 32 17.0 .808
Uggla, D ATL 65 6 10.8 .999 523 19 27.5 .732
Young, Delm. DET 63 6 10.5 1.156 574 18 31.9 .707

Even if you remove the three players who didn’t get much playing time (Hague, Gamel, Hughes), the correlation doesn’t get any stronger. But that is a sample size of only 12. Let’s go back further in time.

2011 had only eight players hit six or more home runs during spring training:

Spring Training Regular Season
Player Team AB HR AB/HR OPS AB HR AB/HR OPS
Fox, J BAL 74 10 7.4 1.122 61 2 30.5 .756
Morse, M WSH 66 9 7.3 1.239 575 31 18.5 .910
Ka’aihue, K KC 58 7 8.3 1.306 96 2 48.0 .612
Gordon, A KC 70 6 11.7 1.187 611 23 26.6 .879
Huff, A SF 65 6 10.8 1.084 521 12 43.4 .676
Hughes, L MIN 65 6 10.8 .834 287 7 41.0 .627
Rodriguez, A NYY 49 6 8.2 1.342 373 16 23.3 .823
Trumbo, M LAA 74 6 12.3 .978 539 29 18.6 .768

2010 had eight as well:

Spring Training Regular Season
Player Team AB HR AB/HR OPS AB HR AB/HR OPS
Johnson, C HOU 62 8 7.8 1.182 341 11 31.0 .818
Bowker, J SF 77 6 12.8 1.010 151 5 30.2 .652
Hill, A TOR 48 6 8.0 1.371 528 26 20.3 .665
Napoli, M LAA 50 6 8.3 1.128 453 26 17.4 .784
Rodriguez, S TB 63 6 10.5 1.373 343 9 38.1 .705
Upton, J ARI 57 6 9.5 1.224 495 17 29.1 .799
Young, Delw. PIT 53 6 8.8 .986 191 7 27.3 .700
Zimmerman, R WSH 56 6 9.3 1.182 525 25 21.0 .899

With our sample size expanded to 28 players, the r-square between spring AB/HR and regular season AB/HR is .001. Interestingly, there is a .1 r-square between spring OPS and regular season OPS, but it is still very weak and easily explained by survivorship bias. In other words, players who perform well in spring (who, more often than not, are good players) both get more spring training playing time (and aren’t cut mid-way through March) and earn regular season playing time.

As fun as Brown has made an otherwise boring spring training, there is still no guarantee that he will enjoy tremendous success during the regular season. Hopefully, the mechanical changes Brown has made are parlayed into a breakout 2013 season.

Leave a Reply

*

36 comments

  1. LarryM

    March 20, 2013 09:59 AM

    This is a helpful corrective, and sure there’s no guarantee, but it’s not JUST the home runs. You have the home runs, the K/BB ratio (which has always been his strength), the minor league performance, and the better mechanics. And a couple other things going right for him. Of course it’s impossible to do a study where you examine whether all those factors TOGETHER are correlated with regular season success, but I suspect that they will be.

    If healthy, Brown is almost certain to break out this season. IMO the only question is just how far the break out season takes him – average to above average regular, or stardom.

  2. Bill Baer

    March 20, 2013 10:12 AM

    When you add too many parameters, you limit yourself to an awfully small sample size. (Think of how Luddites mock Sabermetrics by saying, “What did Player X hit on Sunday afternoons when the wind is blowing in from right field at 7 MPH?”)

    I’m not sure the Minor League stats are meaningful in any way here, either.

  3. TomG

    March 20, 2013 10:22 AM

    Morse has hit his six HRs in only 39 at-bats. Imagine how disappointed the Seattle faithful are bound to be when he doesn’t hit a home run every 6.5 at-bats.

    And yes, I am boldly predicting he will not do so. You heard it here first.

    I hope Dom breaks out this year. This great ST should at least afford him that chance.

  4. Evan

    March 20, 2013 10:33 AM

    I’m going to go out on a limb and predict that Dom Brown has a better season than Delmon Young. This is the minimum level of performance Mr. Brown requires to achieve the rank of “Evan is Pleased With You”.

  5. Pencilfish

    March 20, 2013 10:56 AM

    While ST stats may have little correlation with regular season stats, ST has already served its purpose for Dom Brown: he will get his chance to show what he can do come opening day.

  6. LTG

    March 20, 2013 11:37 AM

    FWIW, there is a correlation between large gains in slugging percentage during spring training and increases in power output during the regular season. Brown qualifies for this correlation, as things stand right now. Of course, even the correlation is no guarantee.

  7. SJHaack

    March 20, 2013 11:43 AM

    But it means there’s a chance he could be really good, right?

    I mean, Dom might totally be the best. Right guys?

  8. LTG

    March 20, 2013 11:50 AM

    That all depends on what ‘might’ means in that sentence.

  9. Jonny5

    March 20, 2013 12:14 PM

    I hope this is no ST mirage, I really really do. What can Brown do for you?

  10. schmenkman

    March 20, 2013 01:23 PM

    I try not to get too excited about ST stats one way or the other. My favorite indicator (aside from his very good plate discipline for a young player, as mentioned earlier) is that Brown has been a better major league hitter through age 24 than any current or recent Phillie (wRC+ with 150+ PAs), with the lone exception of Delmon. Young has gone downhill in the two years since his age 24 season (and Brown has hit better than him over those last two years), but if Dom improves after age 24 like we typically expect players to do, he should be a pretty good hitter.

  11. Marty

    March 20, 2013 01:52 PM

    I have to disagree with your conclusion. While we shouldn’t look at spring training stats for much predictive value, generally, this set of data is actually pretty substantial. If you get rid of the two guys who hit 0 HRs (AB/HR is nonsense for them) and the obvious outlier in Raburn, the r^2 value (preseason vs. regular season, AB/HR) is .07. Take out the other three guys without 100 AB (both 0-homer guys had <100 AB), and the r^2 jumps all the way to .21.

    Qualitatively speaking, it's very unusual for someone to hit 6+ HRs in the spring and then not hit for at least average power. Going back to the full dataset, only 4 out of the 32 hit less than 10 HR/550AB (550 ABs is a bit less than a full season); 22 of them hit at a rate better than 17 HR/550AB.

  12. EDGE

    March 20, 2013 02:29 PM

    So, he may or may not have a good season.

    One positive thing is he didn’t croak under the pressure put on him this ST. If he had the start the Ruf had, that might have been it for him mentally.

  13. LarryM

    March 20, 2013 03:49 PM

    First of all, in terms of your response to my original post: I’m aware that it would be essentially impossible to design a study to test all of the relevant parameters because of sample size. But that misses my point – YOUR study, limited to HR (and another critique of it follows), doesn’t tell us much given Brown’s other positive indicators.

    Second of all, OF COURSE minor league performance matters. Properly evaluated e.g., considering age/level and looking at the most important predictive indicators), minor league stats are fairly highly correlated with major league performance. There are exceptions, but less than 500 major league AB has not yet established Brown as one of them. That’s why the various purely statistical projection systems are fairly optimistic about Brown (albeit IMO not quite optimistic enough at this point). They do take minor league performance into account.

    Finally, there is a serious flaw in the study which I didn’t notice at first. It doesn’t even address the question you’re asking. You want to know “is hitting a lot of HR in spring training correlated with regular season success?” Your study doesn’t address that, at all. The question that your study addresses, OTOH, is the uninteresting question of “whether differences in performance among a small group of hitters who hit a lot of HR in spring training is correlated with regular season success.”

    Bill, I love your stuff and agree with you 95% of the time at least, but this study is … not one of your better efforts.

  14. Cheesecrop

    March 20, 2013 05:04 PM

    Have a little faith in Brown. He’s drawing walks, he looks patient at the plate, & if he’s surrounded in the lineup by Utley, Howard, maybe Mike Young, he might be able to at least give us something resembling Jayson Werth’s 2009 season (w/obvious consideration being given to the fact that the hitting environment is slightly diff. now – his #’s would fall in line w/the best of the N.L. now).

  15. LTG

    March 20, 2013 06:01 PM

    “Have a little faith in Brown.”

    Right, faith. Not a rational argument based on the new evidence produced by spring training stats. At least, this is how I interpret BB’s post.

    However, given the correlation I cited above, Brown’s spring training is enough to raise our probabilistic confidence that he will improve his power production this year compared to his career so far.

  16. Joecatz

    March 20, 2013 06:25 PM

    Bill I’m as sabermetrically inclined as anyone and I think that part of my big problem with this piece and hopefully I’m not speaking out of turn with the other folks who feel this way by speaking for them, is that domonic brown has been a statistical enigma for three years.

    And people REALLY want him to break out and turn into what they were hyped out on that day he took the field in July of 2010 for an injured Shane Victorino.

    And for the first time since then this spring he’s doing it. And it makes people happy and excited and no one wants to be poo pooed with correlations here.

    We get it. We know. But this is about the human element now. It’s a narrative and a story. And he’s the hero and we want to root for the hero

    Anyone who objects in anyway is the villain now.

  17. dave

    March 20, 2013 07:31 PM

    I think it is important to distinguish between everyday palyers versus bench/role players. Less than 300 AB’s for the entire season would not constitute an everyday player. So, therefore if you do your correlation between everyday players versus bench/role players the stats do seem to suggest there is a correlation. Of the 9 obvious everyday fulltime players, they averaged 25 HRs on 537 ABs – not too shabby.

  18. dave

    March 20, 2013 08:22 PM

    Bill – Hardly – you have to compare apples to apples. And if we/you project DBrown as an everyday player than the stats (in a small sample) suggest that there is a correlation. But if we are unsure of whether DBrown will be one of the 6 scrubs or 9 everyday players – then yes your argument would make sense.

  19. Rei De Bastoni

    March 20, 2013 08:33 PM

    Per Zo Zone:

    Baseball statistician and author John Dewan found that players who beat their career slugging percentage by more than 200 points in Spring Training have more than a 60 percent chance at beating their career slugging percentage during the regular season (minimum 200 regular season at-bats and 40 Spring Training at-bats). Brown’s regular-season slugging percentage entering this season is a modest .388, so he will have a chance to prove Dewan right once again.

  20. jtj06

    March 20, 2013 10:26 PM

    Doesn’t this hold for any month of a season and not just March and Spring training? I’m not ready to crunch the numbers, but eyeballing splits and just watching many, many seasons player performance varies greatly from month-to-month. If you tried to predict what kind of player Jimmy Rollins is based on July or August 2012 you’d say Mario Mendoza. If you based it on September 2012 it would be Alex Rodriguez.

    It seems to me the only difference between spring training vs. any other month number is that spring numbers won’t end up on the back of Brown’s baseball card or earn him any money during contract negotiations.

  21. Richard

    March 21, 2013 07:54 AM

    “It seems to me the only difference between spring training vs. any other month number is that spring numbers won’t end up on the back of Brown’s baseball card or earn him any money during contract negotiations.”

    Well, the main difference is the highly variable quality of pitching, in terms of talent, MLB-readiness, and simply pitchers trying to find their form before the season starts.

  22. Evan

    March 21, 2013 01:53 PM

    Hasn’t anyone debunked that article on SLG in spring training yet. The more I think about it the more it seems like bull. Why using SLG, why not use ISO, or OPS, or any better statistic to suggest a real improvement. SLG has a BABIP component that has a lot to do with luck over a small sample.

    Spring training has so many variables like bad defensive players and crappy pitchers over a very small sample size that any offensive stats are of less value than a qualitative measurement like, hey good news Utley and Howard are healthy in March.

    That said, qualitatively Brown has beat the crap out of the ball.

  23. Richard

    March 21, 2013 02:24 PM

    Evan, I do believe that there’ve been articles making your very point (that SLG contains batting avg, and thus a BABIP small-sample component), etc

  24. Jonny5

    March 21, 2013 02:48 PM

    Evan has it right. Bill has it right. We all have to keep in mind the quality of the opponent as well here. Sure his numbers are going to be inflated in ST. He’s seeing scrub pitchers, and pitchers working on their stuff trying to hone their craft, rather than make outs. We should always be skeptical of ST stats, ALWAYS. Very skeptical.

  25. Richard

    March 21, 2013 03:04 PM

    “Sure his numbers are going to be inflated in ST.”

    well, let’s not over-state the case – obviously not everyone has good ST numbers… still the overall point is valid: Brown’s stats, as such, don’t mean much of anything, aren’t predictive, etc etc, the works.

    But they don’t hurt! The main thing about this spring has been: he’s been able to play, looks comfortable, has been decent in the field, and has shown power. The latter is important – whether 6 homeruns or not – simply because of his hamate injury. The upshot is we have every reason to be optimistic about Brown going into the regular season. Which doesn’t mean we can expect any given level of production.

  26. LTG

    March 21, 2013 07:26 PM

    The objection from ISO establishes that we should distrust jumps in SLG based primarily on singles. Brown is not Ben Revere.

    By the by, if the correlation holds for SLG, wouldn’t the correlation most likely be stronger run on ISO?

  27. LTG

    March 21, 2013 07:30 PM

    Let me just register my confusion that those who are fine making arguments based on correlations in some cases are quick to dismiss this one. Shouldn’t we treat it as evidence, indeed, better evidence than we allow for other ST stats?

  28. LarryM

    March 21, 2013 09:48 PM

    I’m really not on board with EITHER Bill’s original analysis, for reason already stated, OR the “SLG%” study. My problem with the latter is that the effect shown is pretty small really – 60% increased their SLG%, but that means that for 40% it just stayed the same or even declined. Moreover, saying 60% increased their SLG% doesn’t tell you how MUCH they improved.

    Depending upon a number of characteristics of the study which are, to me at least, unknown, that might not even be statistically significant. And, of course, in this particular case exceeding his career slugging percentage is no great feat – IMO even BEFORE spring training the chance that Brown was going to exceed his meager career SLG% was well over 60%. But if that means (say) a .400 SLG%, well, that wouldn’t be terribly impressive.

    (FWIW, various statistical projection systems, none of which consider his spring performance, project Brown between .427 and .461.)

    We’re just in a area where the numbers can only tell us so much. It’s true, trivially true (you don’t need a somewhat poorly designed study to prove it) that Spring training success doesn’t guarantee regular season success. But the sum total of the evidence – and the statistical evidence is a part of that, but only a part – seems to me indicates a guy who has turned a corner. What that ends up meaning – a solid regular, or a star, or somewhere in between – is still a very open question.

  29. KH

    March 22, 2013 01:14 AM

    It’s good he is having a strong spring if only because the Phillies will be more inclined to give him an extended tryout as the starter because of it. If he bombed this spring they would be looking for every excuse to give up on him. I’d just like to see Dom Brown get at least a full half season to sink or swim.

  30. Jonny5

    March 22, 2013 09:03 AM

    @ Richard

    Right. Small sample size in ST = meaningless stats. But, fans should definitely have a very positive outlook due to his play as of late. It shows promise. He just looks better, more comfortable. I know it’s Faux pas to mention Fantasy baseball, but you’re looking at a guy who owns Howard, Brown, And B Belt. All were dirt cheap picks for me and they allowed me to focus more $ to top points earners in pitching and other positions. Point being, if ST numbers meant anything to me I’d be pretty smug about my team and it’s chances, but tis not so my friend…

  31. GB

    March 23, 2013 08:09 AM

    KH has got it…the key to Bill’s point is PLAYING TIME. The more consistent playing time Brown gets, the better the chance he will produce at a level all would like to see granted, knock on wood, he stays healthy.

    OTOH, if the Phils yank him around like they have the last several years, replace him for “defense” in late innings, platoon him with Young, replace him if he struggles initially or in short spurts throughout the season etc. then the chances he will produce drop off…

    Like in all sports, opportunity = production potential and one cannot get opportunities if they are stuck to the bench or down in AAA for the 59th time…if Manuel/Amaro do not give him 500+ AB this year to show what he can do, they need to trade him since obviously they do not believe in him and do not think he can help the team win…

  32. Joecatz

    March 24, 2013 08:43 AM

    The reason you DON’T put stock in hitters SP statistics is because pitchers aren’t using scouting reports video etc. same reason not to put stock in pitchers lines.

    But that doesn’t mean you disregard them completely. What I get from his spring is that he’s focused and his hamate is healed and he’s gonn get his shot.

Next ArticlePhillies Pitchers Are Good at Hitting