31 March 2016

Does Miguel Gonzalez's Orioles Career End Here?

With the regular season only a few days away, major-league teams must make the final moves to prepare their rosters for the 162-game haul. Some clubs must decide who will serve as the fourth outfielder; others will choose the backups for the starting catcher and infielders. And for teams like the Orioles, it means they must conclusively determine their starting pitching situation.

Baltimore knows that Chris Tillman, Ubaldo Jimenez, Yovani Gallardo, and Kevin Gausman will (in some order) occupy the 2016 rotation. Tillman rebounded somewhat in the second half of 2015, and his solid work prior has earned him the Opening Day start for the third straight year. Jimemez and Gallardo have solid resumes and pricey contracts, both of which will grant them some job security, deserved or not. And once Kevin Gausman returns from his shoulder injury, his flame-throwing arm will allow him to take the hill every fifth day.

The doubt comes beyond that. Mike Wright and Tyler Wilson each started some games last season, but neither has really impressed to this point. While Vance Worley has some talent, the fact that the Pirates put him on waivers tells you about his potential. Yet each of these three apparently has more potential than Miguel Gonzalez, with whom the Orioles have decided to part ways. After a solid run for the Birds, it looks as though Gonzalez has reached the end of the road.

Out of the many pitching prospects that Rick Peterson has tampered with developed, Gonzalez stands out as perhaps the biggest success. Before the Orioles signed him in the 2011 offseason, he was an oft-injured 27-year-old farmhand with 5.0 innings in Triple-A and zero innings in the majors. Then he broke out in 2012, twirling a 3.25 ERA over 105.1 innings to support the team's Cinderella run. That kicked off a three-year stretch of solid starting pitching, in which Gonzalez posted a 3.45 ERA in 75 games (69 starts — nice) and 435.2 innings. Dan Duquette and Co. seemed to have found a diamond in the massive minor-league free agent rough.

Then the wheels came off in 2015. Gonzalez deflated to a 4.91 ERA over 144.2 frames. It could have been a fluke, but since he'd outperformed his FIP in the prior years (more on that in a moment), this appeared to indicate that his luck had run out. He sealed his fate with a ghastly stint in Sarasota, stumbling his way to a 9.78 ERA in 19.1 spring training innings. That takes us to where we are now: Gonzalez on waivers and the Orioles looking elsewhere to round out their rotation.

In retrospect, we should note that Gonzalez's decline happened not in 2015, but in 2014. Over the first two years of his career, Gonzalez allowed 3.87 runs per nine innings and tallied a 3.64 DRA, both above-average marks. DRA reflected some of the more advanced areas in which Gonzalez excelled — chiefly, his ability to limit production at the plate. Despite facing difficult competition (he had an opponent TAv of .265 over those two seasons), Gonzalez held hitters to a combined TAv of just .254*. Because DRA focuses so much on plate appearance outcomes, this helped Gonzalez's case immensely.

*TAv is scaled so that the MLB average is .260.

When the success at the plate went away, Gonzalez should have given up more runs. In 2014, batters knocked Gonzalez around to the tune of a .278 TAv, driving his DRA up to 4.91. He didn't have quite as low a BABIP in that year as he did in the preceding ones, and home runs plagued him like never before. But Gonzalez got lucky: He allowed just 3.45 runs per nine innings and cruised through the season (when healthy). Gonzalez had the same problems in 2015 that he had in 2014 — too many hits and long balls leading to an inflated TAv — except his good fortune vanished. His 5.04 RA matched his 4.80 DRA and doomed him to the waiver wire.

For two or three years, depending on your preference of peripherals or results, Gonzalez was a respectable, major league-caliber starting pitcher. That's a phenomenal feat for someone who endured as many hurdles — a recurring knee ailment, Tommy John surgery, numerous trips up and down the minor leagues, and the premature death of his close friend — as he did. Still, 31-year-old pitchers who lose their touch usually won't get it back. While the memories he gave us during the 2012 and 2014 runs make this a painful decision, ultimately it'll benefit the team to run out someone such as Worley or Wright.

29 March 2016

Hyun Soo Kim And How We Got Here

Hyun Soo Kim appears to be reliving the same moment he experienced in 2006.  The surprise that past success does not guarantee that those on the next level will automatically hand you anything.  More than that, that showing oneself to be exceptional in years past means less than what you show in the past month.  When it happened before, he was a kid who had just completed high school.  Kim won the Lee Yeong-min Award, which is awarded to the best hitter in Korean high school baseball.  He submitted his name for the KBO draft and every team past him by in every round.  He wound up signing as an undrafted free agent with the Doosan Bears and was forced to reestablish himself by playing on the Bears minor league club.

In 2007, Kim became a regular, finishing as a runner-up for the Rookie of the Year.  In 2008, he finished as runner-up for the Most Valuable Player.  And, well, he never stopped being exceptional:

YearAgeTmGPA2BHRBBSOBAOBPSLG
200719Doosan993531952646.273.335.398
200820Doosan1265573498040.357.454.509
200921Doosan13357231238059.357.448.589
201022Doosan13256529247864.317.414.531
201123Doosan13056125137163.301.392.444
201224Doosan1224911774650.291.358.382
201325Doosan12251023166271.302.382.470
201426Doosan12552826175345.322.396.488
201527Doosan141630262810163.326.438.541
All Levels (10 Seasons)11314768230142597501.318.406.488
Provided by Baseball-Reference.com: View Original Table
Generated 3/29/2016.

When Kim decided after this past season to shift over to the States to play baseball, he was considered by many in Korea to be a sure-fire success.  A great amount of hesitation was eliminated by Jung-Ho Kang's success in Pittsburgh last year.  Kang certainly had more power and played a key position in the infield, but Kim is thought of as a professional hitter.  He just hits and hits and hits.  He works at bats.  He makes contact.  Kim has effectively been a right handed left fielder version of John Olerud in a baseball environment that would greatly appreciate what John Olerud did.  In case you forgot, Olerud accrued 58 bWAR over his career, which puts him just outside the threshold of Hall of Fame consideration.  Anyway, the point is that Kim was considered exceptional and his skills were considered transferable to the Majors.

I was not as confident as my connections in Korea were.  I have not watched a lot of Kim and never saw him in person.  All I had was data, imperfect data, and that throws a great deal of uncertainty into the mix.  Still, my data, which is based on Kang as well as a handful of recent MLB exports to Korea, was not incredibly impressed with Kim.  The model I created did two things well: (1) it assessed home runs well and (2) it assessed walks and strikeouts well.  It looked at Kim's home run performance from the past few years and thought he had a 10-12 HR bat, which is what the scouts thought as well.  It looked at his walk rate and thought it would be cut in half and that his strikeouts would rise by about 10%.  In all, it thought Kim had a fringe fourth outfielder bat. 

Again, the model only hit on home runs, walks, and strikeouts.  Beyond those three true outcomes, it became far more uncertain.  What the model could not do well is figure out contact rates and gap power.  That was where the professional hitter value would be found.  What Kim specifically did well, the model shrugged and threw out weak estimates.  In that gray area, he would need to excel to show good value.

If that was the only issue, the Orioles would likely be smoothing over the difficulty Kim has had this spring.  However, Kim's value is even more reliant on that gray area because of a couple more considerations.  First, Kim is not fleet-footed.  He is smart on the base paths, but does not have the speed to turn long singles into sliding doubles.  Kim is not a first to home kind of guy.  He is not slow by any means, but it is fair to say he is below average.  Second, what Kim lacks in range in the outfield, he doubles up with poor arm strength.  If you have watched him play left field this spring, then you know the modest downplay of how well Kim might fit into right field was still an absurdity.  I also find it doubtful that he can play a league average left field, but I have found no one to comment on that.

Anyway, from this you can see why Kim was awarded a two year, seven million dollar deal from the Orioles.  It was a paltry sum for such an accomplished player because there simply are so many questions regarding his ability to succeed in MLB.  At the time, there was some discussion about an team opt out of the contract.  If he did not perform well, it was noted that he would be sent back to Korea in a manner similar to the Orioles releasing Suk-min Yoon after the 2014 season (Yoon, who was hammered in Norfolk, enjoyed a rather dominant 2015 in the KBO).  However, no firm detail of a team buyout was reported at the time. 

When spring training arrived, so did Kim.  The first rumblings were not positive.  Orioles officials noted that Kim was a little heavy and it was suggested that he was not as prepared because KBO spring training are longer with more emphasis on conditioning players.  Drills commenced and curiously little was said about Kim's performance.  The scrimmages began and Kim began his descent into an 0-for-twenty something streak.  He hit some balls hard, but it was not as if he was spraying the corners and the gaps.  His hard hit balls were basically hit at infielders.  He worked deep into counts, four-plus pitches per plate appearance, but his swing was often defensive.  He often was coming in downward.

The second half of the spring training scrimmage season was more favorable to Kim.  The hits started scooting through and falling.  Over the course of another twenty plate appearances he was flirting with .400 baseball.  Glaringly, he was also flirting with a .400 slugging percentage.  As things came together, he still was not driving the ball with authority and was unable to stretch singles into doubles or find the other side of the fence.  It is troubling that even when things appeared to be going well that his one tool that needed to carry him was simply producing singles.  Additionally, he had earned only a single walk.  This was a growing concern.

Of course, what can one say about 40 plate appearances and spring training ones at that?  Previous research has found that spring training shifts a projection around 5%, which is not enough to really emerge from the noise.  That is complicated in that Kim's projection is built upon a rather scant amount of data relating KBO performances to MLB performances.  Even if it was MLB in-season data, it takes about fifty plate appearances to say anything meaningful about walk and strikeout rates.  It takes three times as long to say much about anything else.

This likely left the team in a tough spot.  The projections are connected to a great deal of uncertainty.  The little amount of data available shows a player who is struggling mightily to do anything of much value.  And the scouting is largely unimpressed.  If Kim was a player who had stomped AAA pitching and put up this kind of spring training performance, he simply would not be considered for a roster spot.  The truth is that right now uncertainty is what both is his savior and what damns him.  It is also what makes other players look more ideal.  A known fringe player or one that performs well in a handful of play often feels more comfortable than the guy who goes out and does poorly.

Joey Rickard was unforeseen.  I imagine the original plan was to kick him around in spring training to see if he could handle the fifth outfielder slot.  Instead, he just hits and hits and hits.  His value has not changed much.  He is a fringe outfielder, but you tend to go with the fringe player who is hot than the fringe players who are cold.  Rickard will start the season suiting up in left field until he remembers that he is Joey Rickard.

Nolan Reimold was planned.  He was going to support Kim in left field and provide insurance in case Kim needed a longer transition or just never made the needed adjustments.  He still is that comfortable sweater.  His defense can be squirrelly, but his bat can play in a pinch.  He is a solid fourth outfielder and has shown more competency at the plate than Kim has.

Pedro Alvarez was not planned.  Alvarez is really what has hurt Kim.  When the Orioles' plans for filling in right field with another left handed bat evaporated with Dexter Fowler returning to the Cubs, the team got the only meaningful left handed bat remaining.  Alvarez is a designated hitter and nothing more, which puts the right handed Mark Trumbo into the outfield.  It also creates a modest desire to have another MLB quality offensive left-handed bat on the bench, which the club does not seem to think Kim has established.

What we find now is that there is a discussion about Kim and the potential curiosity the Orioles have for the recently released and left-handed outfielder David Murphy.  Murphy used to be good.  He used to be a starter.  He is now a guy coming off a 0.0 bWAR season who gets spring training invites and contemplates retirement.  Murphy is not a great prize.  However, he is left-handed and he has experienced success in the not too distant past. 

And so we now find ourselves sorting through reported discussions between Kim's camp and the Orioles about releasing him.  The optimistic view is that this is just an errant news release that has echoed a bit.  The slightly less optimistic view is that the two sides are discussing a way to let Kim spend a month in the minors before making a full decision.  The least optimistic view is that the Orioles are following through with what they said was a possibility when Kim signed last winter: that he has not shown himself to be good enough in the opportunities that were available to him and that he will be heading back to the KBO to dominate some more.

Best And Worst Cases For The Orioles' Rotation

In case you haven't been paying close attention, or are in a constant state of denial, the Orioles have some serious starting rotation question marks. FanGraphs has been posting its 2016 positional power rankings series, and the O's starting rotation options ranked 28th in the majors.

That low ranking should not come as a surprise. With Chris Tillman, Ubaldo Jimenez, Yovani Gallardo, and Miguel Gonzalez, the O's have some decent starting pitchers, but none who should be considered superb or an ace. Then there's Kevin Gausman, who has the highest upside of anyone in the group and the potential to be a front-line arm. But he's also dealing with shoulder tendinitis, and he'll begin the year on the disabled list.

To round out the current starting options, Tyler Wilson and Mike Wright are seemingly in a constant -- but friendly! -- battle to take over for whomever in the group falters or gets injured (depending on Gausman's status in a few weeks). Then, in some order, there's Vance Worley, T.J. McFarland, Odrisamer Despaigne, Parker Bridwell, and Chris Lee. And in the impatient/pipe-dream category, there's Dylan Bundy. The Orioles could always acquire another starter, so things could change quickly.

Besides the rotation, the O's have the makings of a solid team. The offense, which doesn't have a bunch of high on-base types but does have lots of power bats, should feature plenty of homers and still score lots of runs. Defensively, the O's project to be pretty good in the infield. Mark Trumbo's defense is a serious concern in right field, but Adam Jones should at least be adequate in center. Left field is currently a question mark. Joey Rickard, the talk of O's camp, is a lock to make the opening day roster and at least comes with the reputation of a strong defensive outfielder. And despite being heavily relied on yet again, the bullpen should again be one of the best in the American League.

Things don't always go according to plan, but it sure seems like the starting rotation is the main issue holding the Orioles back from competing for the AL East crown. If you read any season preview of the Orioles, it won't take long to get to the team's starting rotation woes.

Anyway, you know what the projections say, and it isn't good. So let's take a hypothetical look at the best and worst cases for the O's projected rotation.

Chris Tillman

Tillman gets the nod on opening day for the Orioles. Because, you know, someone has to do it. The BABIP gods struck back against Tillman a bit last year (.293) after posting below average numbers the two years prior (.269 and .267, respectively). That could have something to do with him slowly becoming more of a groundball pitcher and incorporating a sinker. In his three full-time seasons, he's gone from 39% to 41% to 44%, which is an interesting shift. Still, Tillman doesn't miss enough bats and isn't efficient enough in limiting walks or generating ground balls to be considered anything better than a back-end arm.

You could see things breaking right for him and Tillman posting a similar season to 2014, when he had a 3.34 ERA despite peripheral numbers that suggested he wasn't that good. That year, he had an fWAR of 2.4. That's probably the best case. Worst case? Probably something like last year's 4.99 ERA and 1.9 fWAR, with the frightening potential to somehow underperform even those numbers.

Ubaldo Jimenez

You never know precisely what you're getting with Jimenez. And that should be well documented by now. He's consistently inconsistent (another way of saying he's somewhere between average and above average), and from start to start can look like a completely different pitcher. O's fans have already seen the worst-case version of Jimenez: the 2014 edition, with a 4.81 ERA that essentially matched his performance, a ridiculous amount of walks (5.53 BB/9), and many fewer generated ground balls. That year, he posted an fWAR of 0.4, and it's hard to imagine him being worse.

Jimenez's best-case scenario is something close to 3 fWAR -- so a bit better than last year, when he cut his walks by more than two per nine innings and posted his highest groundball percentage since 2009.

Yovani Gallardo

Gallardo is a good pitcher with a proven track record, and he makes the Orioles slightly better. But looking beyond ERA, his declining strikeout rates are noteworthy, his decreasing velocity is concerning, and a concern with his shoulder that was revealed during the O's physical process led to a restructured two-year deal (with an option).

Gallardo hasn't posted a 3+ fWAR season since 2011, but the Orioles would be thrilled to get anything within the 2.5-3 fWAR range in 2016. But there's always the chance that this is the year when he's unable to limit the amount of damage and his low strikeout totals start to haunt him. In that case, something closer to an fWAR of 1 seems possible, and would be problematic.

Kevin Gausman

Gausman is the youngest of the group, and he also has the biggest chance to be a difference maker. Unfortunately, it's getting dangerously close to that time of worrying if he'll ever live up to his draft position and hype. He'll start the year on the disabled list, but there's optimism that he will be able to pitch pain-free soon. The Orioles are more or less counting on him to take a huge step forward, even if everything they've done with him the past few seasons wasn't exactly in his best interest. It's not like the Orioles deserve the benefit of the doubt when it comes to grooming pitching prospects.

A full season building off what he did last year, while being better at keeping the ball in the ballpark, would surely be welcomed. Gausman's has top-of-the-rotation upside, so an fWAR above 3 is certainly within reason if he's able to come back healthy and avoid any injury concerns. Then again, injury issues could plague him the entire year, or perhaps he starts the year and is ineffective enough to warrant another ride on the Norfolk shuttle. If Gausman's not able to put everything together this year, it may be time to wonder if the O's eventually deal him away.

Miguel Gonzalez

Gonzalez has been awful this spring, and he was some combination of bad and injured last season. Depending on what you believe (and because he can be optioned), he's in a close battle with Tyler Wilson and Mike Wright for fifth starter duties. It's been a marvel to watch Gonzalez outperform his peripherals every year -- until last year, at least. If you want to give him the benefit of the doubt for pitching through pain, that's fine. It's not like his peripherals were all that different from years past, though his inability to limit the damage with runners on base cost him dearly.

Because he never rates well in fielding independent metrics, Gonzalez has never posted an fWAR over 1.6. So that would likely be his best-case scenario: something close to 175 innings and an ERA in the 3.5-3.75 range. The worst case is that he's ineffective early on, quickly loses his rotation spot, and is unable to rekindle what worked for him from 2012-2014. Even if that ends up being the case, he'll be regarded as one of the best finds and surprises of the Buck Showalter/Dan Duquette era.

-----

It's not a good sign when a large portion of a team's starting rotation needs to rebound considerably. And it's tough to overlook the starting rotation woes. But it is possible to still be positive about this team, and that's because starting rotation questions are nothing new for the Orioles under Buck Showalter.

Go back and look at other rotation options during Showalter's tenure. Rotation mainstay Wei-Yin Chen is gone, but there really is not much of substance there. There are names like Jason Hammel, Tillman, Gonzalez, Tommy Hunter, Joe Saunders, McFarland, Jake Arrieta (before he became Jake Arrieta), Scott Feldman, Bud Norris, Freddy Garcia, Jimenez, Gausman, and several others (none of them amazing).

Frustratingly, things have not changed much. It's an uphill battle to overcome a lack of quality starting pitching. But it's hardly impossible, even if it's occasionally hard to watch. And I can guarantee you that it will be very tough to watch at times this season.

28 March 2016

How MLB Expansion Makes Sense

As Commissioner Manfred said in an interview with Joe Passan, baseball is a growth business and sooner or later, growth businesses expand. Manfred told reporters that he’d be interested in expanding from 30 to 32 franchises and that baseball would have a better structure as two 16-team leagues split into four divisions of four. It seems that the addition of two new franchises is almost inevitable. Many articles have been written discussing the merits of adding franchises in cities such as Montreal, Vancouver, Portland, Austin, Las Vegas, Nashville, Sacramento, Indianapolis, San Jose, Memphis, Charlotte, Virginia Beach and San Antonio.

But there are significant challenges to expanding the league. Large-market clubs are upset about the amount of money that goes towards revenue sharing. Adding two small-market clubs would force large-market teams to devote a higher percentage of their budget towards revenue sharing while reducing revenue they receive from league merchandise, the national TV deal and MLBAM. If MLB is going to expand, they’ll need to choose markets that can support teams without requiring significant revenue sharing.

MLB hasn’t had such good luck with expansion recently. In 1998, MLB expanded to include Tampa Bay and Arizona. Tampa Bay has financially been a complete failure since its creation. In their inaugural year, they were able to attract over 30,000 fans per game. Since then, despite having a successful period from 2008-2013, they’ve attracted no more than roughly 23k fans a year and have ranked last or second last in attendance in the American League for 13 of the 18 years they’ve been in existence.  Their TV deal is considered to be poor and isn’t eligible for renewal until after 2018. They typically have one of the lowest payrolls in the league and are able to stay above water solely due to revenue sharing and other league shared revenues.

Arizona has been more successful than Tampa Bay. They ranked in the top half of the National League in attendance from 1998 to 2004, while winning the division three times during that period including a World Series victory. However, since 2004, the Diamondbacks attendance has been ranked between 11th and 14th in the National League. As part of their deal to get a stadium, the Diamondbacks agreed to have low ticket prices so they receive minimal revenue from attendance. As a result, the Diamondbacks have struggled to spend enough money to field a competitive roster and have only once spent more than the $102M that they spent in 2002 on payroll. Jayson Stark wrote that he was told that the Diamondbacks received almost $80M from revenue-sharing in the past three years. Their new TV deal will help, but it pays less than large-market TV deals (my article). This may be better than Tampa Bay, but it isn’t a very good result.

In 1993, the MLB expanded to add the Marlins and Rockies. Jon Heyman reported that the Marlins’ revenues are lower than the Rays and Athletics and that the Marlins received $50 million in revenue sharing this past year. Despite this, the Marlins are still not making a profit. The Rockies have been more successful. They had a stretch from 1993-1999 where they led the NL in attendance. They’ve leveled off since then, but are ranked about at the midpoint in attendance. However, the Rockies still receive revenue sharing funds each year and aren’t known for having high payrolls. All four of these teams are annually eligible for the competitive balance lottery, which is eligible only to clubs in one of the ten smallest markets or has one of the ten smallest revenue pools. It’s safe to say that these four clubs are some of the weakest in baseball.

The American City Business Journals recently did a study to determine whether any market in North America had the economic capacity to support an MLB franchise. They didn’t find a single market strong enough to support an MLB team. Even worse, they felt that Montreal was the best candidate in North America for a new MLB team, but that it only had a limited chance of success. Montreal already failed to support one team and it’s questionable that they’d do better with a second chance. If Montreal could support another team, there are plenty of franchises that could benefit from relocation.

The arguable failure of previous expansion attempts combined with the lack of any remaining large markets available illustrates why further expansion in the United States will be a significant challenge. Therefore, it should come as no surprise that Commissioner Manfred has stated that he feels that the best available markets in North America aren’t United States markets but rather Canadian or Mexican.  Canada only has two possible remaining markets for an MLB team, Montreal and Vancouver. Metro Vancouver has fewer than 3 million residents and would be a small market at best while Montreal has been discussed above. Mexico, on the other hand, is interesting.

Mexico has three metropolitan areas that could potentially support an MLB team; Mexico City, Guadalajara and Monterrey. According to the OECD in 2014, Greater Mexico City had an estimated population of 20.4 million people and a GDP of $421 billion ($21,000 GDP per capita). Metro Guadalajara had a population of 4.9 million people and a GDP of $70.87 billion (14,463 GDP per capita). Metro Monterrey had a population of 4.8 million people and a GDP of $117 billion ($24,400 GDP per capita). In contrast, Baltimore has 2 million people and a GDP of $114 billion ($57,000 GDP per capita). In my opinion, Mexico City and Monterrey are the two best Mexican candidates for an MLB team.

A team in these locations will still have significant challenges. These cities have a low GDP per capita and therefore teams will need to be able to attract the upper class. Oscar Suarez, an MLB player agent born in Cuba that represents multiple Mexican players, stated that Mexico is a big country and parts are heavy into baseball, but that he isn’t sure whether the general population can afford sustaining a big-league team. The Boston Globe reported that other challenges include high altitude, the lack of new stadiums and safety issues. In addition, Mexico City is 650 miles away from Houston, 930 miles away from Dallas and over 2,000 miles away from New York. Not only this, but politics will be a potential concern as there has been significant tension between Mexico and presidential candidate Donald Trump.

But Manfred does have a significant interest in having a team in Mexico. Manfred has stated that a team in Mexico could help grow the Hispanic market in the United States. As of mid-2014, Mexico had an estimated 15.4 million cable and satellite subscribers and Mexico has a robust $1.5 billion ad market. If a cable station broadcasting Mexican MLB games could charge a subscriber fee of $1.50 per month, then it could earn over $300 million in revenue. If so, one would expect these teams to earn slightly less than $100 million in rights fees starting in 2016 with significant room for growth as Mexico’s broadcast market grows and their populace becomes more prosperous, which would go a long way towards ensuring MLBs financial viability in Mexico. Best of all, unlike a team in Canada or the United States, a team in Mexico wouldn’t cannibalize another teams’ existing market. For all of the potential challenges, Mexico has a significant chance of helping the league financially rather than being a burden.

The smart bet would be to expect MLB to ultimately expand into Mexico City and Monterrey while potentially relocating a team to Montreal. Despite all of the challenges, there’s significant growth potential in Mexico and there’s a plausible path for franchises in those cities to be net assets for MLB that don’t require revenue from other markets. Even if things don’t work out, it’s still good PR. But if I was a consultant for MLB, I’d recommend looking at a different solution. I’d look into adding MLB teams in Japan.

I’d add two new teams in Japan, while also relocating the Oakland Athletics and Tampa Bay Rays to Japan, resulting in MLB having four teams in Japan with MLB having a total of thirty-two teams. These four teams would be placed in Tokyo, Osaka, Nagoya and Yokohama. The OECD states that Tokyo has a population of 35.9 million people and a GDP of $1.475 trillion ($41,000 GDP per Capita), Osaka has a population of 17.3 million people and a GDP of $597 billion ($34,508 GDP per Capita), and Nagoya has a population of 6.5 million people and a GDP of $255.7 billion ($39,300 GDP per Capita). Other sources suggest that the Nagoya metro area consists of 9.1 million people and has an average GDP per Capita PPP of $40,000. The OECD doesn’t measure the population or GDP of Yokohoma, but others rank its population at about 3.7 million with a GDP per capita of roughly $30,000. Realistically, Tokyo and Osaka should be considered large markets, Nagoya should be considered a strong mid-sized market and Yokohama would be a strong small-market.

Japan has a strong broadcast market that was projected to earn $12 billion in revenue. The NPB draws over 20 million fans to its games, proving that there is high demand for baseball. In 2014, Japan’s government suggested that baseball expansion could help it get out of its recession. If so, it’s possible that Japan would be willing to pay for necessary items such as stadiums. In short, Japan is a rich country and could potentially make current MLB owners a huge amount of money. It's worth devoting time and making sacrifices to potentially expand to this market.

There are a number of challenges that this expansion would face. By far, the largest challenge would be the distance between Japan and the US. In addition, MLB would need to come to an agreement with Japan’s baseball league, the NPB. But NPB franchises have historically struggled and it should be plausible to come to a mutually beneficial arrangement.

Solving the distance challenge would be difficult, but not impossible. A 152 game season compared to a 162 game season might cost MLB as much as $500 million in revenue per year, but I’d expect the actual losses to be half of that number. Meanwhile, a successful entry into the Japanese market would be worth billions of dollars of revenue per year not to mention that the extra rest could potentially improve player health. MLB would have to be willing to significantly restructure the schedule and cut between 6-10 games out of the season in order to make this plan work. Still, better schedule optimizing techniques and different scheduling structures will offset MLB travel mileage.

In fact, I developed a proof of concept schedule for a potential expanded league in which every team traveled far less than 50,000 miles. If I spent the time to use better optimization techniques, I’m reasonably confident I could ensure that all teams travel fewer than 45,000 miles. To put that in context, the Mariners are expected to travel 46,000 miles in 2016 and the Angels are expected to travel nearly 45,000. In the second installment, I’ll discuss how a potential division realignment could look, how the schedule would be required to change, and what a possible schedule might look like if MLB did expand to Japan. Based on my initial results, I’m convinced that expansion into Japan could work if MLB decides to do so.

24 March 2016

The 2012 Zach Phillips



With left-handed spot relief pitcher Brian Matusz questionable for opening day, the Orioles signed free agent left-handed relief pitcher Zach Phillips to a spot on the 40-man roster. Phillips had been in the Orioles organization before, pitching in 16 big-league and 56 AAA games in 2011-2012. He was one of the 75 players (including 6 named "Zach") who played for the 2012 Norfolk Tides; he left the organization as a free agent and has since been with the Marlins and White Sox organizations. He also spent 2014 playing in Japan.

Although Phillips may have evolved as a pitcher since leaving the Orioles - his AAA strikeout rates have been significantly higher in his 2013 and 2015 AAA seasons than before - I thought we could look Phillips' performance, especially his 2012 performance, to see what we could expect if Phillips does replace Matusz. The data comes from the scoresheets I maintained as a MiLB.com datacaster and as a Baseball Info Solutions scorer. Some of the hit trajectories I recorded are imprecise and subjective. I tried to observe the BIS convention that a ground ball is a batted ball that an infielder didn't or wouldn't have caught in the air.

I saw Zach Phillips phase 72 batters in 2012. 25 of those batters hit left-handed; 47 hit right-handed. Some of the numbers will have apparent discrepancies because two of the batters - one left-handed, one right-handed - executed successfully sacrifice bunts. So, for on-base percentage purposes, Phillips faced 24 left-handed and 46 right-handed batters.

Phillips generally wasn't used exclusively as a left-handed specialist in Norfolk. As you might expect when a team uses 40 pitchers, pitching roles aren't well-defined; Phillips was used as a set-up man and part-time closer. Left-handed batters produced a slash line of .261/.280/.391; right-handed batters produced a slash line of .238/.304/.262.

Three things are apparent from this, although one of those three isn't really true. First, Phillips wasn't more effective against left-handed batters than right-handed batters, so he doesn't fit the classic lefty-specialist profile. On the other hand, that makes him less vulnerable to opposite-side substitution. Second, Phillips isn't vulnerable to the home run. Although the slugging percentage against left-handed pitchers looks bad, he didn't allow a home run; the isolated power results from three doubles (one of which came when Phillips was gassed as the end of a 4-inning stint in an extra-inning game.) Third, he appears to have excellent control - but that's not really true. He didn't walk batters in these games but did walk them more frequently over the course of the season.

More interesting is the trajectories of balls put in play against Phillips, as documented in the below table. The fields should be self-explanatory; the totals combine both hits and outs.

 
Vs. Left
Vs. Right
Ground Ball
15
18
Fly Ball
4
8
Line Drive
3
5

At least in the 2012 games I saw, Phillips was a ground-ball pitcher, especially against left-handed batters. This supports the idea that Phillips is not vulnerable to the home run. The Orioles infield defense is good, and Camden Yards is a home-run park, so there is an intuitive logic that the Orioles should acquire ground-ball pitchers. If Phillips continues to get ground balls, he could be effective and useful to the Orioles.

So, based on Phillips' 2012 patterns, there's reason to believe that he can be an effective piece of the Orioles bullpen. They don't need him to be the closer or top set-up man. He's probably going to be a lefty specialist, but he should be a lefty specialist who can face the occasional right-handed batter (as, for example, if a righty is followed by another lefty or two.) Of course, he could also start his Orioles career with a couple of bad outings and find himself back in Norfolk or as a free agent.

23 March 2016

Southpaws Without Options: Orioles 5th Starter?

Throughout the offseason, several drums beat for the Orioles unendingly.  One of those drums was a search for a viable southpaw to diversify their starting rotation.  Conventional wisdom suggests that a rotation performs better when there is a lefty in the mix as it makes clubs less succeptible to teams with a high density of exceptional left handed bats.  Currently, the left handed starters the Orioles have in their system are simply not ready for prime time and with the trials and tribulations of Kevin Gausman and Miguel Gonzalez, a new starting option might be preferable.  Around the league, though, several left handers are battling it out for fifth starter gigs or are looking in from outside the expected rotation.  Perhaps one of these players could be acquired.

Tommy Milone
Minnesota Twins
Year Age IP ERA+ FIP BB9 SO9 GB%
2013 26 156.1 94 4.30 2.2 7.3 35
2014 27 118.0 90 4.69 2.8 5.7 39
2015 28 128.2 106 4.30 2.5 6.4 42
5 Yrs 619.0 99 4.23 2.2 6.5 38
Provided by Baseball-Reference.com: View Original Table
Generated 3/22/2016.

Milone is not an exceptional pitcher.  He is a good command lefty with a fastball that tops out 89 mph if the breeze blows just right.  His ability to be a back end southpaw is based on his mix of changeups, curve balls, and cutters.  Milone's success last season was derailed for a bit with a forearm strain and some time rehabbing in the minors.  Upon his return, the cutter disappeared and his starts were less useful.  Removing the cutter from his repertoire is a concern as it suggests that the injury issue was still present or, at least, he was concerned the cutter would hurt his ability to improve arm strength.  That said, he only threw about a mile per hour slower upon his return.

In Twins camp, Milone is trying his best to be the only lefty in the Twins rotation.  He is currently facing off with the very well paid (but so far disastrous) Ricky Nolasco for that fifth starter spot.  One could imagine Milone being available from the Twins, but what exactly could the Orioles offer in return?  Also keep in mind that the Twins also have Danny Santana and Oswaldo Arcia who could provide some left handed outfielder options for the Orioles.

Chris Rusin
Colorado Rockies
Year Age IP ERA+ FIP BB9 SO9 GB%
2013 26 66.1 98 4.75 3.3 4.9 48
2014 27 12.2 55 4.08 3.6 5.7 48
2015 28 131.2 88 4.71 2.8 5.9 52
4 Yrs 240.1 84 4.70 3.0 5.7 50
Provided by Baseball-Reference.com: View Original Table
Generated 3/22/2016.

Where Milone still stands in perceived competition to get into the rotation, Chris Rusin instead looks toward sticking on as a longman who will spot start.  The Rockies appear to be shifting toward collecting pitchers who light up the guns and away from those that depend highly on off speed pitches.  Rusin finds himself chucking a 90 mph fastball occasionally, but that is not where he lives.  He lives with his 89 mph sinker and induces a great deal of groundball contact.  A curveball and change up round up his pitches.

His troubles in Colorado largely dealt with poor defense, his sinker not sinking as much, and poor receivers.  In Baltimore, he would not have to deal with many of those things.  He would not be great, but I would think Rusin's 50% groundball rate would play better than Millone's 40% rate.  Rusin is a player who I think is quite likely to find himself pitching for another team by the end of Spring Training.  Maybe the Orioles, he certainly does a few things Baltimore likes.

Brett Oberholtzer
Philadelphia Phillies
Year Age IP ERA+ FIP BB9 SO9 GB%
2013 23 71.2 147 3.65 1.6 5.7 36
2014 24 143.2 88 3.56 1.8 5.9 37
2015 25 38.1 91 4.49 4.0 6.3 49
3 Yrs 253.2 100 3.72 2.1 5.9 39
Provided by Baseball-Reference.com: View Original Table
Generated 3/22/2016.

Like Milone, Oberholtzer is trying to insert himself as the lone left hander in a rotation, but appears to be seen more as a long relief left handed option.  He, too, is a lefty who tops out in the upper 80s.  He has gotten by with a two seamer, a change up, and a knuckle curve.  Last season saw an uptick in groundball rates from his historical 40% to 50%, which also coincided with him having trouble with blisters.  That trouble led to issues throwing his knuckle curve.  I would put his profile between Rusin and Milone, but Oberholtzer's past prospect status was slightly more rosier.  However, the bloom is off that rose.

It is questionable whether the Orioles would be able to get Oberholtzer from the Phillies without giving up something of some significance.  Phillies are going nowhere and it helps having a guy who can log innings.  They have long had a shine for Parker Bridwell, so a one for one could make sense.

Felix Doubront
Oakland Athletics
Year Age IP ERA+ FIP BB9 SO9 GB%
2013 25 162.1 95 3.78 3.9 7.7 46
2014 26 79.2 72 5.13 3.7 5.8 38
2015 27 75.1 73 4.45 3.1 6.7 47
6 Yrs 513.2 85 4.32 3.8 7.7 44
Provided by Baseball-Reference.com: View Original Table
Generated 3/22/2016.

Doubront use to have a low 90s fastball and was a promising back end arm in the Boston rotation.  In the last three years though, a shoulder injury was followed with loss of velocity and poor command of his pitches.  To make amends, he has embraced the cutter.  However, the new proportional pitch mix has given little indication he may be able to replicate what he did in his early to mid 20s that made him somewhat interesting.

The Athletics appear to be eyeing a long relief role for him, but would probably deal him for an interesting fringe prospect.  So what is a left handed pitcher worth who strikes out few batters, walks too many, and gets tagged hard?  Eh.

Brad Hand
Miami Marlins
Year Age IP ERA+ FIP BB9 SO9 GB%
2013 23 20.2 129 4.02 3.5 6.5 42
2014 24 111.0 85 4.20 3.2 5.4 50
2015 25 93.1 73 4.08 3.1 6.5 46
5 Yrs 288.2 82 4.54 3.7 5.9 44
Provided by Baseball-Reference.com: View Original Table
Generated 3/22/2016.

Hand is another fringe 5th starter who gets the ball on the ground and has decent velocity for a left hander.  He works both a four and two seam fastball, an inconsistent curve, and a decent change.  His main problems are that he has difficulty hitting his spots and difficulty missing bats.  Hand has performed better as a reliever than he has as a starter and it would not be surprising for the Marlins to value him as the long man.  He is still fighting for that 5th slot though.

Sean Nolin
Milwaukee Brewers
Year Age IP ERA+ FIP BB9 SO9 GB%
2013 23 1.1 13 15.05 6.8 0.0 30
2014 24 1.0 59 16.13 0.0 0.0 25
2015 25 29.0 77 5.13 3.7 4.7 42
3 Yrs 31.1 60 5.91 3.7 4.3 41
Provided by Baseball-Reference.com: View Original Table
Generated 3/22/2016.

Sean Nolin feels like a left handed version of Miguel Gonzalez.  He gets by with an above average changeup and needs good command of a lackluster 90 mph fastball.  He also mixed in plenty of cutters and a show me breaking ball.  There are question marks, of course.  In 2014, his fastball rode around 93-94 mph.  A hernia surgery, groin issues, and shoulder soreness saw him working for the Athletics in the upper 80s.  Now with the Brewers on a waiver wire pickup, he is a low value, high risk kind of pitcher.  Again, this is MiGo though MiGo was never supposed to work out.  Nolin is a longshot as well.

Tyler Lyons
St. Louis Cardinals
Year Age IP ERA+ FIP BB9 SO9 GB%
2013 25 53.0 79 3.73 2.7 7.3 47
2014 26 36.2 83 3.65 2.7 8.8 43
2015 27 60.0 106 4.53 2.3 9.0 39
3 Yrs 149.2 90 4.03 2.5 8.4 43
Provided by Baseball-Reference.com: View Original Table
Generated 3/22/2016.

While we are on comparisons, Lyons makes me think of Brian Matusz.  Lyons was a somewhat promising southpaw (though nowhere near as promising as Matusz) and has been roughed up in his stints starting for the Cardinals.  At best, they see him as a swingman, but probably have him slated on the edge of the bullpen as a lefty one out guy.  He works with a fastball, sinker, slider, and changeup which happens to be what Matusz works with.  The main difference is that Matusz' slider is a little harder than Lyons.

Lyons would give the Orioles a cheaper left handed option in the pen for this year and beyond.  He will provide a potential solution to Matusz' exit after this season.  Of course, one does not plan much to ensure there is a situational lefty on the squad and Lyons' days starting may be as wishful as Brian Matusz still having that option.