MATT’S BONUS LEFT CRYING
Emilio Bonifacio – 2B/3B/SS/OF
Take out a piece of paper and a writing instrument, class; it is time for a brief lesson on Small Sample Size and Ree-ee-gression!
Wait, what? Yes, I suppose the back of the test I just returned to you and your neighbor’s highlighter will do, but please come to class with all of your crap next time.
This was originally going to a bonus second Cubs “Left Crying” in our NL Central Edition of Spring Fling / Give Him A Ring / Leave Him Crying, but it took on a life of its own and became too large to be included in that post. I simply could not pass on the opportunity to briefly celebrate the cruel, heartless bitch that is regression. Of course, real statisticians (and sabremetricians) will consider this concept elementary, but it is nonetheless fun to ponder. I need to preface that I do not really have anything against Emilio Bonfacio – he is somewhat unremarkable and largely inoffensive, though I would never go out of my way to fling with him or present to him a ring (per the paradigm of the larger series of which this post is a rogue offshoot) – the opening to his 2014 campaign has so perfectly encapsulated this statistical concept in which I place a great deal of value that I simply could not pass up the opportunity to use and abuse Bonifacio’s poor little heart for my selfish purposes.
Let us begin by stating that, assuming good health (which, frankly, one probably cannot), the vast majority of batters tend to remain who they are year in and year out. Sure, a player will experience occasional positive or negative spikes in BABIP that will yield a higher or lower floors for the AVG/OBP/SLG slashes, or booms and busts in HR/FB% that yield either the magical “career year” or the disappointing “lost season” – large fluctuations in these areas (and a glut of other smaller fluctuations) are the spice of baseball numerical life. Having said that, hitters who walk a lot will mostly keep walking, hitters who swing and miss with great frequency will continue to frequently swing and miss, and hitters who scald line drives everywhere will scald more line drives everywhere. Furthermore, these characteristics tend to hover around some defining mean value regardless of whether a player is hot, cold, or somewhere in between. Improvement is certainly possible, but rarely will a player improve an aspect of his game so drastically as to render the initial baseline completely unrecognizable (i.e. Chris Carter may improve upon his catastrophically high strikeout percentage, but he will never resemble Ichiro Suzuki). By the same token, a player will typically not suddenly begin whiffing significantly more frequently in the absence of some critical physical change (injury, significant age-related decline, a dramatic change in the manner by which pitchers attack him, etc.). Regression to a player’s individual means is an overwhelmingly dominant force – this is true of both hitters and pitchers, though it needs to be noted that pitchers age quite a bit differently than do hitters.
Let us now introduce Bonafacio into these proceedings.
Bonafacio’s two calling cards are plus speed (or even plus-plus speed) and versatility – he has logged reasonably significant playing time at 2B, 3B, SS, LF, and CF. His batting average and walk percentage are both almost perfectly league-average (yielding, by definition, an almost perfectly league-average OBP). Power? LOLPower. Defense? Over the bulk of his career he has been a positive contributor at the two outfield corners and below average everywhere else. In summation, he is a speedy, unremarkable hitter whose versatility allows him to deploy his unremarkable-ness from one of six defensive positions.
Having established some measure of Bonafacio’s essence through prose, let us seek to do the same through numberz (click to enlarge this and all subsequent tables):
Again, Bonifacio is very close to league average as it pertains to AVG, BB%/OBP, K%, and LD%. His great speed makes him both more dangerous and more efficient on the basepaths than his league average peer. His slap-swing yields a ton of groundballs, little loft, and no power. The large number of groundballs yields a disproportionately high number of infield hits, and consequently, a higher baseline BABIP (which Brian correctly associates with speedy rabbits in our Carlos Gomez debate). These elements combine with the varying levels of facility Bonifacio possesses at numerous defensive positions to create a player who sits almost smack-dab-in-the-middle between the hypothetical “replacement player” and the league average one.
Mr. Bonafacio, you can step outside for a moment. Don’t wander too far, though – I’m going to need to make fun of… I mean, make an example of… er, that is, I’m going to need your help as I try to teach the other 56 kids in the class. No, you can’t have extra credit – it’s not my fault you aren’t happy with your grade after you skipped 8 classes, didn’t turn in any homework for 5 weeks, bombed 2 tests, and have posted a pitcher-like career ISO.
We must remember that players will assemble their overall statistical profile through a composite of peaks and valleys. The larger the number of plate appearances, the more likely we will negate the anomalies brought about by that foul temptress “Small Sample Size.” Small sample sizes are often the primary undermining culprit to the integrity of any statistic; we experience its false charms whenever a broadcaster or analyst cherry picks some statistical element of a player’s performance (i.e. “Dating back to 2005, Joe Klangertang is batting .500/500/1.250 in 4 at-bats with runners in scoring position and less than two outs during day games on Thursdays during which the tide is in and the President of the United States is vacationing at Prairie Chapel Ranch in Crawford, TX”). Most every hitter will have hot stretches during which they perform at an “All-Star” level – the best players perform at such a level more frequently and for longer periods of time (perhaps you have heard of that catch-all term “consistency?”). In summation, it is dangerous to draw conclusions from statistics born of small sample size (but it’s cool – all the cool kids do it, and they’re cool!).
While this undoubtedly has most of you saying, “Um, no shit,” there is one period of each season during which most everyone forgets this principle and falls into the trap of assigning too much value to statistical anomalies destined for correction: the opening weeks. In the absence of a volume of tendency-defining statistics, a torrid hot streak in April is far more apparent than it would be over a three-week stretch in July and August, and the type of 30 PA slump through which most every batter will slog at least once a season can be seen as a prelude to a player’s grisly demise. When all is said and done, however, these season-opening super-studs and mega-duds will almost invariably experience some period of statistical correction. A 2014 example of extreme dud-turned-stud scenario can be found in the bipolar opening acts of Pablo Sandoval‘s season, though Sandoval’s struggles did persist for quite a while before he finally went super nova. Sandoval will need even further positive correction to reach or exceed his career norms – given that he wants a lot of money once this season is over, he had better hope math continues to smile upon him. Bonafacio, on the other hand, has been among the most exaggerated – and extraordinarily predictable – of 2014’s studs-turned-duds.
Let’s bring him back into the room, shall we?
So just how did Bonifacio’s season start? Well, through 8 games (Small Sample Size) 41 PA (S.S.S.), Bonifacio had done these things to baseballs, pitchers, and catchers.
Well, that’s just great (kinda). For just over a week, 57.6% of the balls Bonafacio put in play did not find fielders, he could barely miss a pitch if he tried (12.2% strikeout rate), and he could not be thrown out on the basepaths while donning the wheels of a young Eric Davis. The high BABIP lifted his AVG, which in turn lifted the portions of his OBP and SLG that are themselves built upon AVG. It all worked out to that nice .500/.537/.579 slash. Let’s briefly strip away context so as to render those numbers excellent-to-insane: a typical batting champion posts an average within 15-20 points of .350, the list of batters who have who have posted an on base percentage ≥ .537 in qualifying seasons during the live-ball era consists of “Barry Bonds, Barry Bonds, Ted Williams, and Babe Ruth,” and a .579 slugging percentage would have ranked 3rd in all of baseball in 2013. EMILIO BONIFACIO IS THE GREATEST THING SINCE BARRY BONDS 2.0!!!
Well, that was fun, now let’s view them with some context.
Because it is important, let us recall a snippet from earlier in this post (unlike a former professor of mine, I promise I am not quoting myself as an “expert source”):“…assuming good health -snip- the vast majority of batters tend to remain who they are year in and year out. Sure, one will see the occasional positive or negative spike in BABIP that will yield a higher or lower floor for the AVG/OBP/SLG slashes, or booms and busts in HR/FB% -snip- [but] hitters who walk a lot will mostly keep walking, hitters who swing and miss with great frequency will continue to frequently swing and miss, and hitters who scald line drives everywhere will scald more line drives everywhere. Furthermore, these characteristics tend to hover around some defining mean value regardless of whether a player is hot, cold, or somewhere in between.”
Do you see the columns I’ve conveniently marked in green? Of course you do – you are clever, you are pretty, and you did stay at a Holiday Inn Express last night. If one can look beyond the eeeeeNORmous BABIP disparity, one will find a hitter who: 1.) walks at a roughly average rate (perhaps a tick below average); 2.) hits with the power of a pitcher (as measured by ISO, which strips AVG from SLG); and 3.) hits a ton of groundballs (which, again, will produce a lot of extra infield hits given his speed). As for things we might expect to correct themselves (pink and darker pink)? Certainly the strikeout and flyball rates (both exhibit far too dramatic a change to be taken seriously, and because S.S.S.); probably the percentage of line drives (though that is less of an outlier); and, in all likelihood, the success while stealing bases (because S.S.S.). As a matter of technicality, however, there was an actual day on which people could open their morning iPads while sipping their morning Venti Quad 9-Pump White Mochas and find that something called an “Emilio Bonficacio” was batting .500/.537/.576 for the season. Small sample size? Meet the opening act of a baseball season.
Man, those 8 games were so unbelievably insane that it would take… what… 2 months of hitting like Daniel Hudson to bring his numbers back into career norms? Now, Daniel Hudson is actually a fairly capable hitting pitcher (career .229/.268/.305, 52 wRC+), but the big picture version of the sentence would nonetheless read something like: “After the first 8 games of the season, Emilio Bonifacio would need to hit like a pitcher for 2 months in order to sport numbers that look like numbers Emilio Bonifacio would sport.”
Emilio Bonifacio has hit like a pitcher for 2 months.
Wow, that’s bad. Funny thing though – Bonifacio still profiles as a hitter who 1.) walks a tick below the average hitter; 2.) hits with the power of a pitcher; and 3.) hits a ton of groundballs. Why does that sound vaguely familiar? As could have been predicted, he did begin to swing and miss a little more (though he has settled in at a lower strikeout rate than his previous career norms), and both his line drive and flyball rates regressed almost perfectly to his career averages (the latter of which is more or less useless for Bonifacio given that he has more of a chance of legging out a hit on an infield grounder than he does accruing extra bases via a deep flyball). His batted ball luck and his success on the basepaths severely over-corrected, but then those were the two areas that were most egregiously out of line with his career norms.
So… what would happen if we combined the 8 games of “surface-temperature-of-the-sun” production with the 8 weeks he has spent in Siberia?
Huh… where have I seen THAT before?
It should not surprise us that Emilio Bonifacio was made to look like… well, himself… when math was sufficient enough time to perform its relentless work. As a hitter, Bonifacio is who we thought he was; the disparity in the value he is providing this season can actually be found in his defense. To date, he has actually provided positive value from every position from which the Cubs have deployed him; perhaps he has benefited from exposure to the Cubs’ coaching staff, though math tells me that is probably going to change, too (ahem… SMALL SAMPLE SIZES).“Bonifacio is very close to league average as it pertains to AVG, BB%/OBP, K%, and LD%. His great speed makes him much more dangerous (and efficient) on the basepaths than the average big leaguer. His slap-swing yields a ton of groundballs, little loft, and no power. The large number of groundballs yields a disproportionately high number of infield hits, and consequently, a higher baseline BABIP.”
There is nothing about the above numbers or prose that screams “excellent big league player.” Sure, there is undoubtely a place for such a player on the bench of a good team – he’d be a valuable utility infielder, pinch-runner, and pinch-hitter in a pinch (though he doesn’t possess an especially useful platoon split and would certainly never scare an opposing pitcher or manager). To give Bonifacio everyday or close-to-everyday playing time (as the Cubs have this season, as the Royals did after acquiring him at the trading deadline last season, and as the Marlins did in 2009 and 2011) is to overexpose him. He could be a valuable tool in the hands of a smart manager, but frequent use is inefficient use, and proclaiming him a “bona fide leadoff hitter” is an absolute absurdity. As we have already discussed, however, early-season production is greatly magnified in the absence of stabilizing data, and I suppose there is something to be said for the power of a first impression, as well. Can you imagine this article being written today if the author had to stare Bonifacio’s current .269/.319/.340 slashes in the face? I suppose it is possible if the author was still seduced by Bonifacio’s opening week, and I will say he correctly pointed out that the Cubs are getting good return on their $2.5 million investment relative to the market prices of the past offseason, but I can say with certainty that no such article would be penned if after 16 games Bonifacio was slashing .206/.254/.317 (as he has in has over his past 16 games). Do yourself a favor: don’t be fooled by obvious disguises.
Alright, class. No homework. Extra credit for everyone. Take next Tuesday off. Extra credit for everyone. Just remember that a player will always be made to be whomever he really is if math is given enough time. If you can remember that, you should ace the next test (which will count for 3 tests to save ALL of your grades). Get out of my sight.