BP Comment Quick Links
![]() | |
May 30, 2013 Dissecting the DraftExamining Kris Bryant
Since we first set out talent tiers for players in consideration for selection with the seventh overall pick in the draft, some significant events have affected the stock of two of our potential targets. We have also seen reporting that this draft class contains a unanimous “top 3”—a triumvirate that does not match up with our tiering. We’ll quickly address the shakeup in the tiers before devoting the bulk of this entry to a thorough examination of Kris Bryant—the “unanimous elite talent” that didn’t make the cut for our Tier One grouping. Tiering Updates Player: Sean Manaea, LHP, Indiana St. University Player: Ryan Boldt, OF, Red Wing HS (Red Wing, MN) Entering today’s analysis, this is where our tier currently stand: Tier One So what is our measuring stick for “elite” or Tier One talent? For my money, a selection at the top of the draft—one requiring over $4 million of upfront investment—needs to satisfy a three-pronged test: (1) impactful upside, (2) limited comparative risk, and (3) satisfactory makeup. The first prong centers on raw tool grades and the potential for the player to outdistance the raw grades. The third prong requires an examination of work ethic, on-field demeanor, and potential for off-field issues, but is focused primarily on large red flags. In other words, we are concerned with aspects of a player’s personality that could have the potential to make him ill-suited to successfully navigate the mental rigors of professional baseball. The second prong is often the trickiest, and it is the prong we are most concerned with when examining Bryant. By “comparative,” I mean risk compared to upside. So the higher the upside of the player in question, the more risk we can tolerate in the profile. Assessing Bryant as a Tier One Talent What’s not to love? Bryant easily satisfies prong one and prong three of our “elite” test. All that remains is an examination of the associated risks in his profile, and to make a determination as to how those risks might impact his ability to reach his ceiling. Concerns surrounding Bryant revolve around two primary issues: (1) he has long been viewed as a player likely to be forced off of third base, knocking down his defensive value, and (2) as a long-limbed power bat with a propensity to groove his swing, there have always been questions as to how much his raw power will ultimately play at the highest levels. Defense is the easiest question to address. Bryant has acquitted himself well in limited action in right field, making the potential shift off of third base not nearly as frightening as it was anticipated to be some 18 months ago. As a fringe-average to average defender in the nine-spot, Bryant would provide more value than he otherwise would at first base, and there is additional room for his bat to fall short of its ceiling without torpedoing his overall value. In short, Bryant’s improvements have made this risk a non-issue from a draft value standpoint. The only chief concern, then, is his future ability to implement his immense raw power. From a scouting perspective, Bryant has widened his base and simplified his barrel delivery, which has helped him to cut down on empty at-bats and improved his contact rate. Further, the fact that he has launched 31 home runs so far this spring—more than a good number of entire Division I teams—would seem to greatly alleviate concerns in this area, and on the surface it does. But if we are to group Bryant as a truly elite player—a player worthy of an up-front investment of over $4 million—we need to be as certain as possible that his carrying tool (power) will be an impact tool at the next level by looking beyond a raw collegiate stat line. To me, this closer look, reveals evidence that Bryant isn’t quite as safe an investment as many have thus far indicated. Before we take the final plunge into this analysis, I want to be clear that none of the below is intended to imply that Bryant is likely to fail, or even that he is unworthy of top-five overall consideration. The point of this exercise is to flesh out the decision-making process and, specifically, the questions as to whether Bryant should be considered an elite talent. Stat Analysis—Examining Power and Triple-Slash Here are a few subsets of Bryant’s performance this spring broken down into two tables. We are looking at total performance and then subsets of performance against starting pitchers, against relief pitchers, during midweek action (weaker subset of pitchers than weekend), and against draft-eligible arms projected to go in the top 10 rounds (or for underclassmen, that would project to the top three rounds were they eligible). Please note these totals were compiled through review of game logs and may not perfectly match the aggregate statistics posted on the University of San Diego website. Performance Against Starting Pitchers and Relievers
Of Bryant’s 279 plate appearances this spring, 39% of them came against pitchers throwing in relief, and 61% against starters. Of his 31 homers, however, 16 (or 52%) came against relievers, with 15 (or 48%) coming against starters. We also see a dramatic delta between his triple-slash line against starters (.296/.470/.736) versus relievers (.388/.523/1.035). His strikeout and walk rates remain fairly uniform and in line with his rates on the season. Over a small sample, Bryant could be slightly lucky or unlucky against starters or relievers, and while we don’t have a large enough pool of data to make a determination, we can look at other circumstantial evidence to see if there are further indicators of troubling areas. With regards to the above table, we are presented with the possibility that Bryant’s stats are buoyed by his plate appearances against lesser relief arms, as opposed to starters. Performance Against Midweek Arms and Top Arms
College teams tend to use their best arms as weekend starters. Additionally, college teams tend to fill midweek innings with the second rung of arms on the staff, including underclassmen and pitchers otherwise not relied upon for the weekend series. Bryant has a disproportionate number of homers against these second-rung arms—32% of them over just 22% his plate appearances—versus those he faced on the weekend. He totaled 38 plate appearances against arms projected to go out in the first 10 rounds of the draft (or first three rounds for underclassmen were they eligible this year) and hit three homers—10% of his homers over 14% of his plate appearances. Again, while not definitive, the circumstantial evidence seems to indicate his home-run totals could partially be a result of his ability to take advantage of lesser pitching at the collegiate ranks. Additionally, we see his strikeout rates higher and his batting average and slugging lower versus the top arms, compared to both midweek arms and his season averages. Playing fast and loose with the stats, taking his plate appearances versus starting pitchers and extending those rates to his full 279 plate appearances would net him around 24 homers, which would still be tops at the Division I ranks. But the severe drop off in triple-slash rates is troubling, even in a small sampling. Remember, our identified risk is that Bryant will have trouble tapping into his raw power due to implementation issues. Looking at Plate Discipline and Contact Rate History tells us that as a player progresses from the low minors to the majors, we should generally expect walk rates to decrease and strikeout rates to increase. The idea is that more advanced pitchers tend to have better command of higher-quality stuff. At the major-league level, I set 20% as a loose upper limit for strikeout rate and 10% as a loose target for walk rate. Accordingly, we look for collegiate rates to offer enough cushion so that as these rates rise or fall against better competition, there is a decent chance to meet or surpass the above targets. As noted above, Bryant’s walk rates across the various subsets sit 19% to 22%, while his strikeout rates sit 12% to 14%. In each case, he looks to be in good shape to take on some negative movement as he faces better arms while still meeting or exceeding our goal of a 10% or better walk rate and 20% or better strikeout rate at the major-league level. However, because Bryant is such a huge homerun threat, and because he is the biggest bat in the San Diego lineup by a fair amount, due diligence requires we pry into the walk rate a little further, parsing intentional walks. Of Bryant’s 59 walks on the season (by my tally), 21 were intentional. If we remove those from his walk totals, his effective walk rate drops from 21.1% on the season to 14.7%. This is not a red flag, but we must remember that at least a small portion of the remaining 38 walks came when the opposition did not pitch Bryant particularly close, for fear of him punishing mistakes. As a final notation, in limited action with Team USA last summer (17 games and 68 plate appearances), Bryant posted BB/PA and SO/PA rates of 11.8% and 25%, respectively. While again a small sample size, we have to note the import of stats accrued while Bryant was using wood bats, and we have to at least acknowledge that (i) an 11.8% BB/PA rate is closer to his 14.7% effective BB/PA rate this spring than the 21.1% BB/PA including his intentional walks, and (ii) his a 25% SO/PA rate is closer to his 21.1% SO/PA against “top arms” than the cumulative 14% SO/PA rate this spring. What does it all mean? As noted in his scouting report, Bryant grades out at an Adjusted OFP of 55/61, indicating good probability, but slightly more risk and less upside than I would personally look for in a truly elite draft target. For purposes of our shadow draft, he slots in right behind Trey Ball in AOFP, but, for all intents and purposes, is on par with both Ball and Stewart. Should all three be available, factors outside of straight grades would determine our ultimate selection, with major-league ETA likely a large component of the equation. (Bryant could be ready at some point next year, as opposed to Stewart’s and Ball’s projected 2017 arrival.) Looking Ahead
Nick J. Faleris is an author of Baseball Prospectus. Follow @NickJFaleris
|
Interesting and important approach in evaluating Bryant. Have you done a similar analysis of Gray/Appel- essentially comparing performance against them by top players versus the weaker players in each lineup? Also, I would think the factors you noted being of concern would apply to any baseball player in minors or majors- certainly expect players to perform better against a 4th/5th pitcher than a 1/2 and I think that holds up under careful analysis.
Thanks! Yes, I run splits for most of the top guys I examine. As a general matter, you expect better performance against lesser competition, but I look for severe splits as a potential indicator that a player is buoying his stats against a certain player type. Again, not looking to find decision makers in the stats, but rather trying to make sure I have the right context when viewing the stats as a whole.
The pitchers are slightly different in that we have the opportunity to do a few more "fun" exercises with their numbers, and overall I think their splits tend to be more telling, overall. Most of the time, whether pitchers or position players, this exercise doesn't reveal a ton--just serves as reinforcement for the general stats. Other times, such was the case with Dustin Ackley, you get fairly convincing indicators that aspects of the player's game (in Ackley's case, "emerging power" his junior year) are not what they appear to be.
Can you help frame how unusual Bryant's splits are given that hitters as a group would exibit this same split (as Bruce points out above)? My preference would be to see a rigorous statistical study showing how many standard deviations he is from the mean in terms of difference in output vs. high and low quality arms, but that's impractical right now and if you study these breakdowns enough you should have a feel for it given your domain expertise. Would like some additional context if you can provide it.
Secondly, to me the most interesting data you highlighted was that even if you extrapolated his rates vs. only starters he would still be leading D1 in homers. Obviously there is more to production than just homers, but that is pretty remarkable and would seem to say that even ignoring his video game numbers against inferior pitching, he is getting it done better than anyone else against top arms. That would seemingly make the bat quite low risk....
Yes, your (and Bruce's) thoughts are spot on -- the risk is low in that he's a high floor/probability bat, and the existence of some discrepancies in certain splits are to be expected. In other words, it's highly likely he's a contributor in some form. The concern comes when you have an elite bat (statistically) that posts wider than expected splits, and those discrepancies dovetail with scouting concerns. Certainly, the home run production is top notch regardless of splits. But raw power isn't the concern, it's his future ability to make enough consistent contact (and hard contact) to get that raw power to play against Major League arms.
It may be the difference between averaging 35 home runs a year or 25 home runs a year. As noted in the article, either way he's an excellent draft prospect. But the only question we are setting out to answer in this examination is whether or not he is safe enough to be considered an elite selection. Opportunity cost requires we make surgical distinctions between our top draft targets--an inherently difficult task in a world that requires us to project undeveloped players one to four years into the future.
As a broader matter, we are shooting to rely on scouting for projection and overall assessment of a draft prospect's draft day value, while looking to statistics to add context. Because competition varies so much at the collegiate level, we further parse the statistics to highlight any potential problem areas.
Later this summer, once the draft is behind us, I'd be happy to explore some of these issues in more detail. The most extreme split (SP/RP) I've seen regarding a potential top 5 round guy was a report I provided to an area scout that showed a 140 bip delta. Loosely, I'd consider 0 to 25 bips noise, 25 to 60 bips "noteworthy", and 60 or greater bips potentially problematic. Even in the case of problematic splits, there has to be a corresponding scouting concern in order for me to consider strongly dinging someone. Otherwise, it's just a further data point to toss into the AOFP calculator.
Hope that helps; thanks very much for the feedback. If I can flesh things out further, please let me know.
Nick,
I'm extremely impressed with this piece and look forward to all your columns - must-read stuff. Keep up the great work.