2009 Minor League Stats Worth Tracking

by spurdynasty on Mar 20, 2009 8:39 AM CDT Comment 34 comments

A number of Rangers prospects who are expected to start the season in AA or AAA have a chance to impact the major league club in 2009. If injuries, ineffectiveness, or trades at the big league level do not force the Rangers to promote a prospect, then adding a minor league player to the big club will likely only happen if the Rangers believe he will improve the performance of the major league team.

Determining when a prospect is ready for a big league promotion is best accomplished by a combination of scouting and statistics. Scouts and coaches (or Jason Parks and Zywica) can determine if a player has made adequate progress with his change-up or outfield defense or pitch recognition or whatever was deemed critical for the player’s development. Statistics indicate whether a player’s skills are being efficiently converted to production. One tool that is available for fans to track whether a prospect might be ready for the majors is major league equivalents (MLE).

The MLE concept was first described by Bill James in the mid-80’s and posits that there is a strong correlation between major and minor league stats. During the past 20+ years, James and others have developed algorithms that convert minor league stats to major league stats and vice versa. MLEs have proven to be almost as predictive of future major league performance as a major league player’s prior year stats. A variety of web-sites provide custom tools for converting minor league numbers to major league numbers and vice-versa. If you are interested, Dan Szymborski provides one of the more straightforward descriptions of how to calculate MLEs at http://www.baseballthinkfactory.org/btf/scholars/czerny/articles/calculatingMLEs.htm.

Rangers prospects who might emerge as at least major league average players in 2009 are presented below. Two things worth noting:
(1) There will likely be specific things that each of these players are working on that will keep them from posting optimal minor league numbers (e.g., Feliz will likely limit the use of his fastball in 2009 to develop his off-speed pitches). This will effectively suppress the prospect’s stats.
(2) Prospects are often promoted to provide them with major league experience so that they are better prepared to contribute in future years. If that is the case for the 2009 Rangers, then it is likely that one or more players will be promoted before they can be expected to perform as a league average player (e.g. Elvis Andrus).

Presented below are the average performances of American League players by position in 2008 according to http://www.baseball-reference.com/pi/psplit.cgi?lg=AL&team=TOT&year=2008. I used the MLE calculator at http://www.minorleaguesplits.com/mlecalc.html to convert the ML average hitting numbers to AA and AAA equivalents. There is not a good calculator for major league equivalents for minor league pitchers so I reverse-engineered the calculator used by Minor League Splits (http://www.minorleaguesplits.com) to convert the American League average starter and reliever numbers to AA and AAA equivalents. Minor League Splits was the source of the career statistics that are listed after the name of each Rangers prospect. Minor League Splits is also the source of the 2008 stats and MLEs that are provided for some of the players listed below.

The combination of major league averages, MLE’s, and minor league numbers for the various prospects are used to provide some sense of what statistical weaknesses individual players can improve to enhance their chances of becoming league average or better players.

Starting Pitchers
Major League Average – 1.39 WHIP, 6 K/9IP, 3 BB/9IP, 1 HR/9IP
AAA Equivalent – 1.25 WHIP, 6.7 K/9IP, 2.3 BB/9IP, 0.7 HR/9IP
AA Equivalent – 1.25 WHIP, 7.5 K/9IP, 2.2 BB/9IP, 0.8 HR/9IP

Near-Ready Rangers Prospects
Tommy Hunter – 1.18 WHIP, 5.8 K/9IP, 1.7 BB/9IP, 0.8 HR/9IP (career minor league averages)
Hunter’s first exposure to major league hitters wasn’t pleasant, but his early AAA numbers suggest that he is close to being ready for the majors.   Hunter’s 50+ AAA innings in 2008 translated to an MLE of 1.22 WHIP, 4.2 K/9IP, 1.8 BB/9IP, and 1.5 HR/9IP. If he can decrease the homeruns and increase his strikeout rate, then Hunter should be ready to provide the Rangers with a league average pitcher.

Derek Holland – 1.03 WHIP, 9.7 K/9IP, 2.5 BB/9IP, 0.4 HR/9IP
Holland’s minor league numbers indicate that he is on-track to be an above average major league starter. If Holland can hit his career numbers while pitching in AA or AAA this year, then he would project to post numbers akin to the 2008 version of James Shields (~3.5 ERA, ~1.15 WHIP, ~8 K/9IP, ~3 BB/9IP).

Neftali Feliz – 1.13 WHIP, 11.2 K/9IP, 4 BB/9IP, 0.2 HR/9IP
In 45 AA innings in 2008, Feliz’ MLE consisted of a 3.16 FIP, 1.26 WHIP, 9.3 K/9IP, 4.6 BB/9IP, and 0.2 HR/9IP. Except for the walks, those translated stats suggest that Feliz is ready for a spot in a major league rotation. If Feliz can add a second plus offering to his repertoire while continuing to dominate AA hitters, then he will be ready for a starting role in the major leagues by 2010 at the latest.

Thomas Diamond – 1.35 WHIP, 9.5 K/9IP, 4.8 BB/9IP, 0.74 HR/9IP
Diamond’s career numbers, which are admittedly buoyed by his first two years in the minors, suggest that he is on-track to carve out a major league career. The stat to watch in 2009 will be walks – if Diamond can drop his walk-rate to below 3 per 9 innings, then he has a legitimate shot to become at least an average major league pitcher.

Relief Pitchers
Major League Average – 1.39 WHIP, 7.5 K/9IP, 3.9 BB/9IP, 0.9 HR/9IP
AAA Equivalent – 1.25 WHIP, 8.3 K/9IP, 3 BB/9IP, 0.6 HR/9IP
AA Equivalent – 1.25 WHIP, 9.4 K/9IP, 2.8 BB/9IP, 0.7 HR/9IP

Near-Ready Rangers Prospects
Pedro Strop – 1.23 WHIP, 11.7 K/9IP, 4.1 BB/9IP, 0.8 HR/9IP (career minor league averages)
Strop’s career walk-rate is high but the rest of his career numbers suggest that he can be an average to above average relief pitcher. If he can remain healthy and drop his walk rate to ~3 per 9 innings, look for Strop to spend time in the Rangers bullpen in 2009.

Guillermo Moscoso – 1.08 WHIP, 9.5 K/9IP, 1.9 BB/9IP, 0.7 HR/9IP
Moscoso’s stats scream above-average starter. His repertoire and age relative to league suggest bullpen. If Moscoso can remain healthy and post the same stats as a reliever in AA or AAA that he’s accumulated as a starter in the lower minors, then look for him to become an above average major league reliever in 2009 or 2010.

Beau Jones – 1.42 WHIP, 8.5 K/9IP, 5.0 BB/9IP, 0.5 HR/9IP
Jones’ career numbers include time as both a starter and reliever. His 2008 numbers, which came primarily as a reliever, included a lower WHIP (1.30) and slightly lower walk-rate (4.3 BB/9IP). If he can continue to push those two stat lines down in 2009, then it is likely that Jones will be challenging for a spot in the Rangers’ pen in 2010.

Brennan Garr – 1.36 WHIP, 10.2 K/9IP, 4.8 BB/9IP, 0.4 HR/9IP
Garr’s walk rate is the difference between being a candidate for the major league pen and being an after-thought in AA or AAA. As with many of the pitchers on this list, Garr needs to improve his control in 2009 without sacrificing strikeouts or homeruns.

Center Fielders
Major League Average – 269/334/412/746; 27 2B, 2 3B, 13 HR, 57 BB, 121 K
AAA Equivalent – 323/398/514/912; 39 2B, 7 3B, 21 HR, 77 BB, 112 K
AA Equivalent – 341/421/545/966; 39 2B, 7 3B, 23 HR, 84 BB, 110 K

Near Ready Rangers Prospects
Julio Borbon – 321/363/425/788; 32 2B, 2 3B, 7 HR, 29 BB, 62 K (2008 minor league averages)
Borbon’s numbers in Frisco (336/380/457/837 in 256 at bats) translate to an MLE of 272/307/352/659. The MLE batting average would have placed Borbon as above average but the 307 OBP would have been almost 30 points below average in 2008. If Borbon can improve his walk rate by 3-5% in 2009 while maintaining the high BA, look for him to be patrolling centerfield at RBiA by September.

Greg Golson – 265/310/411/721 (career); 286/337/440/777; 27 2B, 6 3B, 19 HR, 51 BB, 192 K (’08 est for 700 PA)
Golson’s 2008 OPS in 420 AA ABs was the highest of his career. His walk-rate and power numbers don’t jump off the page, but they do suggest that he could be an average major leaguer. As has been noted by others, Golson’s statistical flaw is his strike-out rate. Assuming Golson maintains his career .350 BABIP, he could hit .310 and put himself on the prospect map in 2009 by simply reducing his strike-out rate to 20%.

Middle Infielders
Major League Average (SS and 2nd base) – 274/329/393/722; 32 2B, 3 3B, 9 HR,
AAA Equivalent – 325/389/473/862; 42 2B, 4 3B, 13 HR, 64 BB, 87 K
AA Equivalent – 346/414/503/917; 43 2B, 4 3B, 14 HR, 70 BB, 85 K

Near-Ready Rangers Prospects
Joaquin Arias – 293/319/401/720 (career minor league averages)
    ’08 est for 700 PA – 296/329/421/750; 22 2B, 13 3B, 10 HR, 30 BB, 79 K
Arias is a .290-.300 hitter with a good glove and not much else. His slugging and OBP will likely keep him from being more than a bottom third starting SS or average utility infielder. Assuming he begins the year in AAA, a sudden spike in his walk-rate could indicate that Arias is finally ready to become a major league shortstop.

Jose Vallejo – 292/345/425/760; 29 2B, 4 3B, 11 HR, 41 BB, 91 K (2008 averages)
Vallejo could take another big step up the prospect ladder if he proves that he can play shortstop and he increases his BA to over 310 without sacrificing his walk rate or power. If successful, the resulting 310/365/450/815 line would look like what you would expect from an average major league shortstop playing in Frisco or Oklahoma.

Catchers
Major League Average – 258/322/393/715; 28 2B, 1 3B, 12 HR
AAA Equivalent – 306/379/484/863; 39 2B, 1 3B, 20 HR, 67 BB, 108 K
AA Equivalent – 327/403/517/920; 39 2B, 1 3B, 22 HR, 72 BB, 106 K

Near-Ready Rangers Prospect
Max Ramirez – 317/424/521/945 (career minor league averages)
’08 est for 600 PA – 347/439/628/1067; 38 2B, 4 3B, 38 HR, 84 BB, 138 K
Ramirez’ MLE in 2008 (269/339/442/781) would have placed him third in hitting among AL catchers with at least 300 ABs. If he can catch, Ramirez will likely be one of the best hitting catchers in the majors.

Corner Outfielders
Major League Average – 275/344/445/789; 32 2B, 4 3B, 20 HR
AAA Equivalent – 328/409/555/964; 42 2B, 5 3B, 28 HR, 112 K, 83 BB
AA Equivalent – 350/435/594/1029; 43 2B, 6 3B, 30 HR, 110 K, 90 BB

Near-Ready Rangers Prospect
Brandon Boggs – 257/362/463/825 (career minor league averages)
In 68 AAA at bats, Boggs’ 309/368/456/824 batting line was solid though probably not yet major league worthy. As expected, his time in the major leagues was solid but unspectacular. If Boggs is relegated to AAA in 2009, look for an improved slugging percentage to precede a mid-season call-up.

Corner Infielders
Major League Average – 267/341/438/780; 30 2B, 2 3B, 19 HR
AAA Equivalent – 320/409/545/954; 41 2B, 3 3B, 28 HR, 87 BB, 110 K
AA Equivalent – 342/435/585/1020; 42 2B, 3 3B, 30 HR, 94 BB, 108 K

Near-Ready Rangers Prospect
Justin Smoak
Smoak’s minor league resume is too sparse to predict what he should improve prior to a promotion to the big leagues. Assuming he begins the year in Frisco, an OPS above 1000 would suggest that his time as a prospect will be short-lived.

21 recs | Comment 34 comments

Story-email Email Printer Print

Comments

One hour, 4 recs.

I give you spurdynasty.

"[Font} doesn't turn 19 until the end of May and his heater can already hit 99 on the gun. That's baseball porn." - Jason Parks

by hightowersmith on Mar 20, 2009 9:43 AM CDT reply actions 0 recs

you might add

Corey Young to the list of relievers to watch. I know it is probably too early, but the buzz around him is too strong to ignore, and he fits into a definite area of need for the big league club. I am starting to wonder if he starts the year in Frisco.

by clark on Mar 20, 2009 9:59 AM CDT reply actions 0 recs

+1 on this

corey young is someone i would like to see included/added

i have been waiting all offseason for a spur dynasty post!

Scout: He was a first-round pick right? Got a huge bonus?
KG: Oh yeah.
Scout: Well, he spent a lot of it on milkshakes.

by knockoutking on Mar 20, 2009 10:03 AM CDT up reply actions 0 recs

Good idea on Young

The problem with including him is that, like Smoak, his stats are so limited that it is hard to get a sense of where his limitations might currently be.

In 31 innings last year in Spokane, Young had a 1.10 WHIP, 9.9 K/9IP, 4.1 BB/9IP, and 0.3 HR/9IP. In 12 innings vs lefties, his numbers were 0.75 WHIP, 8.3 K/9IP, 3 BB/9IP, and 0 HR/9IP. His numbers are superior to Beau Jones but the standard deviation for the stats are huge given that the averages are based on so few innings. If Young can come close to replicating his short-season numbers in 50 AA innings this year, then it is very likely that he will bypass Beau Jones as the Rangers’ top minor league LOOGY candidate.

by spurdynasty on Mar 20, 2009 10:36 AM CDT up reply actions 0 recs

also

and you can correct me if I am wrong, but in many (perhaps most) of these minor league stat lines the numbers are weighted in favor of their A ball numbers and below. MLEs are really only relevant at AA or above, correct? So doesn’t it only make sense to look at their numbers at AA or above?

by clark on Mar 20, 2009 10:04 AM CDT reply actions 0 recs

The predictive power of MLEs

In the brief time that I spent researching MLEs, it appears that equivalents are most predictive when you compare one level to the next higher level and vice versa. That would mean that MLEs based on AAA numbers are more predictive of future ML performance that AA numbers. The lower the level that you use to predict ML performance, the less accurate you are likely to be. That’s why I restricted the Equivalents to AA and AAA.

I used career minor league numbers as a default so that I could have enough numbers to get a sense of where a prospect’s primary weaknesses lay. With the exception of Diamond, I tried to also include AA and/or AAA numbers as a carve-out for those players that had accumulated a reasonable number of innings or ABs. As you suggest, the upper level stats provide a much better sense of how a player will perform in the majors.

by spurdynasty on Mar 20, 2009 10:50 AM CDT up reply actions 0 recs

Excellent read spurdy.

"...my balls are really like a veiny flesh color" blueballlefty on Jun 4, 2008 7:44 PM EDT
"you gonna lose your horse. seriously." FX2
Yes we can! November 04, 2008

by Rodney on Mar 20, 2009 10:11 AM CDT reply actions 0 recs

Statistics indicate whether a player’s skills are being efficiently converted to production.

I like the way you put that. In your format description, I think you meant the average ML starter has 3 BB’s per 9.

The MLE’s for some guys look great, but we all need to remember that number suggests the level at which they are capable of performing at, not what they will actually do right away. Feliz, for example, is better off spending all or nearly all of this year in the minors polishing his secondary stuff and the rest of his game. Even then he’s likely to struggle when he first comes up. I think most of us here are pretty realistic about that, but casual fans will still be disappointed in guys like Feliz and Andrus early on.

Thoughts on specific players:
-I was surprised to see Moscoso’s career BB rate, thought it was higher than that.
-Those MLE’s for Borbon should be sobering. Hope that his walk rate bumps up and that his defense is as good as advertised.
-I think Holland’s control means he will adjust a bit more quickly to ML than Feliz does, but Feliz will top out higher. Hardly an earth shattering view, I know.

Great post as usual.

G G G E-flat_______ F F F D__________....

by t ball on Mar 20, 2009 10:37 AM CDT reply actions 0 recs

Thanks for catching the mistake on the average walk-rate for ml starters; it has been corrected.

MLE’s are designed to directly translate a player’s minor league stats into major league stats. Bill James et al would say that if a given hitter had been in the majors instead of AA or AAA, that his ML stats would have been within 25-50 OPS points of his MLE. The studies that I’ve seen would support James’ contention. The two primary exceptions to this rule seem to be players who have:
(1) repeated AA or AAA several times and developed inflated numbers due to their familiarity with the league
(2) a exploitable flaw (think Jason Botts). Feliz’s lack of a consistent secondary pitch might be such a flaw. Even if Feliz could post league average ML numbers right, I agree with you that he should spend a year or more in the minors to complete his development.

Borbon’s MLE OBP jumped off the screen for me as well. I know that Zywica’s talked about it since the dawn of time, but I hadn’t appreciated just how big a deal Borbon’s minor league walk rate was until I saw that 307 OBP translated from his AA stats.

by spurdynasty on Mar 20, 2009 11:11 AM CDT up reply actions 0 recs

Borbon's glove

if it’s as good as we hope, he can be a 3-4 win player just by bumping his walk rate from about 5% to between 6-7%, which would mean Ellsbury’s bat with as good or better defense.

Is there a way to calculate MLE’s using wOBA? That just occured to me, and I wonder how Borbon’s .370 AA wOBA would translate.

G G G E-flat_______ F F F D__________....

by t ball on Mar 20, 2009 11:20 AM CDT up reply actions 0 recs

I’m not aware of any free sites that convert minor league wOBA to a major league equivalent. STATS Inc is the most likely group to have it.

by spurdynasty on Mar 20, 2009 2:29 PM CDT up reply actions 0 recs

Greatness.

Now blogging again (temporarily) at "Hello Win Column".

by lonestarJon on Mar 20, 2009 3:16 PM CDT reply actions 0 recs

I hope everyone will keep in mind

that a player will tend to improve his major league performance level over time, if that player is called up to the major leagues at a young age (18-25 or so). So Nelson Cruz most likely won’t be increasing his stat line over the next 5 years since he is close to the prime age for a major leaguer. I wonder if Josh Hamilton would be an exception to this rule because of the time off? Andrus (or Smoak, or Saltalamacchia, or anybody very young), on the other hand, will likely increase his stat line in the majors over a period of several years. You can’t expect him to have his best year in 2009. Prime age of productivity for a baseball player tends to be between 26-28 years old.

Rec’d the article of course!

"Evolution happened, now get over it." Michael Shermer

by rodcarew on Mar 20, 2009 5:01 PM CDT reply actions 0 recs

Apologies to tball ahead of time.

I promise that I read this article before the one you posted on Salty.

"Evolution happened, now get over it." Michael Shermer

by rodcarew on Mar 20, 2009 5:16 PM CDT up reply actions 0 recs

None needed

I expect Salty to improve on the projection I made over the next few years. Salty was called up young, especially for a catcher, so I think we should be optimistic about him. The same with Andrus if he manages to not fall on his face this year.

G G G E-flat_______ F F F D__________....

by t ball on Mar 20, 2009 5:46 PM CDT up reply actions 0 recs

Age-dependent improvement

I couldn’t agree more with the idea that many of the prospects listed will develop into above average major leaguers during their careers even if they fail to post league average ML numbers or MLEs in 2009. Along those lines, I recently completed a statistical comparison of who are young for their league to those who are age-appropriate. Interestingly, the younger players tend to improve as they move up through the minor leagues whereas age-appropriate players tend to stagnate and then regress. I’ll finish and submit the FanPost when minor league rosters are settled in a couple of weeks, since I want to recognize the Rangers prospects whose developmental time lines have been accelerated.

by spurdynasty on Mar 20, 2009 9:06 PM CDT up reply actions 0 recs

That premise

doesn’t surprise, really. I would expect guys who were young for their league to be better, their teams were probably aggressive with them because they saw something special. In that fanpost, are you looking at different positions relative to age? I mean, I’d expect catcher, for instance, to be less likely to be very young relative to level overall, but I’ve never tried to examine that premise. Sounds like it might fit right in with your post, though.

G G G E-flat_______ F F F D__________....

by t ball on Mar 20, 2009 9:25 PM CDT up reply actions 0 recs

I haven't looked by position,

but that’s a nice thought. Like you, I wasn’t surprised to see that young prospects catch and, in many cases, surpass the older players in performance as they progress through the minors, but I was surprised by the disparity in the performance gains as they advance.

by spurdynasty on Mar 20, 2009 11:02 PM CDT up reply actions 0 recs

imho

Golson is not near ready unless he’s fast enough to cover all 3 outfield positions at once.

"Evolution happened, now get over it." Michael Shermer

by rodcarew on Mar 20, 2009 5:23 PM CDT reply actions 0 recs

Golson = New starting WR for the cowboys :-)

by Michael Cave on Mar 22, 2009 1:45 PM CDT up reply actions 0 recs

imho

Golson is not Near Ready unless he’s able to cover all 3 outfield positions at once.

"Evolution happened, now get over it." Michael Shermer

by rodcarew on Mar 20, 2009 5:29 PM CDT reply actions 0 recs

Golson

I’m not optimistic, but the numbers suggest that if Golson can increase his contact rate, he can put up a league average hitting line for center fielders. Unfortunately, reducing his K-rate from 28% in 2008 to 20% in 2009 might be a bit of a stretch.

by spurdynasty on Mar 20, 2009 8:52 PM CDT up reply actions 0 recs

MLEs

I should note at this point that I wrote that primer on how to calculate the James MLE like a dozen years ago and that translation method has become somewhat obsolete. The difference between AAA and MLB has widened a bit over the last 15 years, in part because a lot of players that wouldn’t have gotten a chance in 1990 are getting real chances in the majors and Japan is skimming off most of the guys like Greg LaRocca that don’t.

You also get more accurate results with a non-linear model of translating minor league statistics. One of the problems with the linear models is that you either end up overestimating the stars or underestimating the merely good, simply because stars lose a greater percentage of their numbers than lesser players.

Just to respond to multiple points at once, the age issue surprisingly turns out to mostly be a BABIP issue – the reason that MLEs had trouble with repeaters and the like is that a repeater is more likely to have an extremely high BABIP and the older MLE methods didn’t deal with that. So older minor league translation will occasionally have guys like Dave McCarty translating with a BABIP in the .350-.380 range when it’s really hard to sustain anything north of .340.

--
Dan Szymborski
dan@baseballprimer.com

by D.Szymborski on Mar 21, 2009 2:17 AM CDT reply actions 1 recs

Thanks for the commentary

One thing that I noticed in researching the topic is that there is a lot more MLE literature for offensive stats than for pitching. Is that because pitching stats are less consistent when you move from one level to another?

by spurdynasty on Mar 21, 2009 11:03 AM CDT up reply actions 0 recs

That’s mostly because Bill James didn’t do a pitching MLE at the time and I don’t believe anyone was doing much (in public) with pitching translations until the mid-90s when Davenport and I (and later, McCracken) were working with those numbers. I could be wrong there, but pretty much all the public baseball stat action was on usenet (without which, there’s no BP, Primer/BTF, or Baseball-Reference) during that period.

--
Dan Szymborski
dan@baseballprimer.com

by D.Szymborski on Mar 21, 2009 2:36 PM CDT up reply actions 0 recs

If the MLE metric is in fact one that's been scrapped for awhile now, how much stock do you put in the projections listed in this fanpost?

Are they credible? Or something that requires better research?

by oc on Mar 21, 2009 6:11 PM CDT up reply actions 0 recs

Clarification

Dan was saying that the methods used to MLEs have become more sophisticated since the description that I linked in the FanPost. The hitting MLEs that I used in the FanPost were calculated using what I presume to be the latest and greatest that Minor League Splits has to offer. The pitching MLEs were calculated using a simple estimate of the algorithm that Minor League Splits uses for pitchers.

by spurdynasty on Mar 21, 2009 6:27 PM CDT up reply actions 0 recs