Archive for Trades

Who got the most out of the Shelby Miller trade?

At the moment, based on present values of prospects, players and wins? It’s the Braves. Based on the industry consensus? It’s the Braves. Based on which team will be better next season? Maybe still the Braves. How did Arizona get so comprehensively fleeced in this deal? There have been some great articles written on FanGraphs these past two weeks about the GM of the Arizona Diamondbacks. Dave Stewart seems to want to do things differently from all other teams, going with his gut when building a roster. Gut decisions can work. His gut identified Shelby Miller as a good player and so wanted him on his team. But wow did this gut decision miss something. How much something? How about 225 million dollars?

That’s right, I am suggesting that, with present values of prospects, players and wins, Dave Stewart, GM, gave away $225M more value in this trade than he got back. See below for how I got to that number.

FanGraphs’ staff have been using a great contract tool recently to estimate the value of major-league free agents. It estimates that players get better each year until aged 27 then worse each year after reaching 31. I’ll be following their method, which estimates an aging curve using WAR where players improve by 0.25 WAR each year in their 18-27 age years, keep steady WAR in their 28-30 age years and get worse by 0.5 WAR each year in their 31-37 age years. It also includes an inflation increase in the market value for each WAR, starting at $8.0M in 2016 and increasing by 5% each year going forwards.  We end up with an estimated value for players over the lifespan of their contract.

I’m going to use last season’s FanGraphs WAR as the starting value for any major-league players (I’ll discuss the minor-league prospects later). I realise I could use ZiPS or many other predictors of players performance (which offer much lower WAR values), but it seems fair as both major-league players in the deal have some issues with their peripheral numbers that seem to balance out. The tool is an estimator based on current performance, so it seems fair we start with what they achieved this last year. Anyway, let’s get started.

The Major Leaguers

Shelby Miller (worth $98.48 M) – to Arizona

Ender Inciarte (worth $172.91 M) – to Atlanta

Holy cow! Not looking good for Arizona already. Shocked how much Inciarte is worth by this model? Me too. Those wins get expensive. He’ll do well to keep his performance at this level, but I’d argue that Miller has the same issue.

The Minor Leaguers

This isn’t quite as simple to work out. Minor-league players have a habit of not making it to the majors (an average 70% bust rate of ranked players not making a significant contribution to the major league team, according to this excellent article by Scott McKinney). I have used the valuations on ranked MLB prospects from Kevin Creagh and Steve DiMiceli with a couple of modifications.

Summarised, using historical prospect rankings, they took prospects ranked between 1-100 in the top prospect rankings each year and found the average WAR produced over their first six major league seasons at different rankings. These values are in the table below:

Creagh and DiMiceli were looking at surplus value produced by a prospect. I’m more interested in total value. For each ranked prospect (Blair and Swanson) I found the average WAR produced by historical players with a similar prospect ranking (potentially flawed; I’d probably prefer median WAR for each group), then used an inflation model (again 5% per year starting at $8.0M/WAR in 2016) and an assumption that 2/3 of a prospect’s value is accrued during years 4-6 in the majors to find an estimate for total value.

Finally, some minor-league prospects don’t improve enough to reach the majors (they “bust”) so they aren’t as valuable as major-league players. This reduces the value of the prospect. Lower-ranked players are more likely to bust than higher ones so they should have a greater reduction in their value. The historical likelihoods of a prospect bust are shown in the table above (Bust % Zero WAR or less) for different ranks of prospect (again from Creagh and DiMiceli).  To account for the chance of a prospect bust I reduce their value by a factor of the bust percentage, taken from the table above (% Zero WAR or less). It isn’t perfect, but seems reasonable, I’m happy to discuss in the comments.

Dansby Swanson (worth $137.2M) – to Atlanta

• #10 prospect currently (on MLB.com)
• Hitter
• Worth 15.6 WAR on average in years 0-6
• Arrives in majors in 2017
• 33% of value in years 0-3 (at 2018 cost per WAR) – 5.2 WAR – $45.76M
• 67% of value in years 4-6 (at 2021 cost per WAR) – 10.4 WAR – $105.04M
• Bust chance of #10 hitting prospect – 9%
• Estimated total value – $137.2M

Aaron Blair (worth $14.85M) – to Atlanta

• #61 prospect currently (on MLB.com)
• Pitcher
• Worth 3.4 WAR on average in years 0-6
• Arrives in majors in 2016
• Assume only lasts 3 years in majors (due to low WAR total)
• 100% of value in years 0-3 (at 2017 cost per WAR) – 3.4 WAR – $28.56M
• Bust chance of #61 pitching prospect – 48%
• Estimated total value – $14.85M

Gabe Speier (worth $negligible) – sorry Gabe – this trade wasn’t about you.

Final value totals

Arizona:

• Shelby Miller – $98.5M

Atlanta:

• Ender Inciarte – $172.9M
• Aaron Blair – $14.9M
• Dansby Swanson – $137.2M
• TOTAL – $325M

DIFFERENCE IN VALUE – $226.5M

There are a number of obvious caveats here. Miller could be better than this, Inciarte may not be that good, Swanson may never make it, Blair may never make it. However, at this moment, these are some of the values that you could reasonably ascribe to these assets. This is a staggering loss for Arizona. In what business can you lose $225M dollars in one transaction and keep your job? By this rather flawed measure, the Braves have just increased the value of their organisation by $225M. That pays for half a new stadium. Or Jason Heyward’s recent contract. Or the next 3 years of performance of Mike Trout (Trout is seriously valuable). I realise that the money can’t be accessed like that, but still, wow. Dave Stewart might be using his gut feeling when making deals, but he’d better start listening more to his analytics department or he’s liable to get robbed again.

A lot of the inspiration (it wasn’t plagiarism, honest) for this article came from Craig Edwards and his piece on “Attempting to rationalise the Shelby Miller Trade”. I just took it a different way. Thanks to Craig though! You should read it – http://www.fangraphs.com/blogs/attempting-to-rationalize-the-shelby-miller-trade/

With the trade deadline approaching (ed. note: so long!), we have players donning their current affiliation’s uniforms for a final game.  Aside from a few tears shed by baseball’s infantile devotees (ages 3 and up), these will be business-like transactions; it’s the buying and selling of goods.

For this article, we’ll be looking at pitcher performances in their final game before being traded.   The last game a pitcher pitches for a team is a relatively inconsequential event.  What I mean is that teams searching for that piece to propel their momentum in the efforts of a playoff push know what Cole Hamels is worth.  His value did not take a significant ding after two abysmal starts when he totaled 6.1 IP, 20 H, 14 ER, and 5 K (July 10th and 19th), and neither did it skyrocket after his performance on Saturday.  You know what happened, but let me recap: In his last outing for the Philadelphia Phillies, Hamels threw a no-hitter against the Cubs – the very team who he could have ended up with.  That’s probably already a better idea for an article – pitchers that were traded to the team they just faced – but I’ve started the research for the article I’m currently writing, an article that, much like my night, is, for lack of a better word, aimless.  My one year old son is asleep, my wife went to a birthday dinner, I’m in my underwear watching X-Men 2: X-Men United (for the umpteenth time), drinking Diet Hansen’s Tangerine Lime, and thinking about Hamels’ almost 100-point game score in what was his last game as a Phillie.

For the purpose of this article, but mainly to eventually get some sleep tonight, we’re going to limit the research to this current decade (2010 being the starting point), and for pitchers that went to teams that made the playoffs only.  Cole Hamels’ no-hitter scored a 98 on Bill James’ game score calculation, with the only two blemishes keeping it from being a nice even 100, being the two walks he issued over the 9 IP.  There have been two pitchers traded already, Scott Kazmir and Johnny Cueto, and they each posted game scores north of 70 (73 and 78 respectively).  The starting pitcher talent level that could swap hands this year is pretty special.  It’s hard to remember the last time there was a talent pool this deep or when there were this many teams in viable contention to make playoff pushes, so it’s no surprise that we’ve seen a couple of really, really well-pitched games.  But in terms of single-game performances, how good is a game score of even Kazmir’s 73 in the final game before a player gets swapped?  You’d have to go back to 2010 to find a game score as high as 70, which was what Cliff Lee scored for the Mariners after dominating the Tigers over 8 IP.  (Zack Greinke had a 72 before he was traded to the Angels in 2012, but they didn’t make the playoffs.)

Below you’ll find a table with players that were traded close to the deadline to teams that made the playoffs from 2010 – 2014.  Next to their name you’ll find the game score of their last game pitched with the team that traded them (organized by game score) compared to the average game scores they posted with the team that traded them and the team they were traded to.  It’s highly likely that, with so much going on tonight – X-Men just ended – I missed one or two or three starting pitchers.  ((Also spoiler alert (Not about X-Men which I’ve seen close to 100 times now): There are no correlations between that one game score and their value for the rest of the season because a pitcher is the pitcher they’ve been for the body of their work)).

(Game Score Equation courtesy of Bill James: 50 + (outs recorded) + (2*IP after 4th inning) + (1*K) – (2*H) – (4*ER) – (2*Unearned Runs) – (1*BB) = Game Score)

**GmScA = Game Score Average

I know that 18 games is a small sample size, but 47.2 is a pretty sizable drop from averages of 54.6 and 54.5.  Perhaps it’s the uncertainty that these players are facing with the looming trade deadline that causes a dip in performance, or perhaps this is a silly, SILLY thing to look into and it means absolutely nothing!!!

But back to this year’s impressive pool of starting pitchers that were rumored to be available.  Let’s say all the top pitchers that are thought to be moving have thrown their last game for their respective teams.  Here’s a look at a table just like the one above but for the pitchers thought to be moving this year.  (This is assuming the caveats that all the players have pitched their final game for their respective teams and that they will all be traded to or have been traded to teams that will make the playoffs).

That’s a lot of 70s.  There are also three pitchers in this group with average game scores of over 60, which is one more than the number of pitchers that went to playoff bound teams from 2010 – 2014.  After all this, I wish there was a way to make this post more interesting, or show some correlation between any of these numbers, but there’s simply not!  It was a thought and I ran with it.  One game does not make a player, but I think in some TINY way, it is another example of what we know to be true:  there is some serious talent that’s about to switch hands – get excited.

Wow. Craig Kimbrel traded right before the start of the season. I have to admit to being rather shocked. I know the Braves are rebuilding this off-season and it made sense to trade him. He is very highly valued for a player who only pitches 60 innings a season, perhaps over-valued. If the Braves are going to be hopeless this year then who needs a dominant single-inning pitcher?

The trouble is I love watching Kimbrel pitch, no matter the situation. I live in London, in the UK, so a lot of Braves games happen from 1-4am and I don’t get to watch them live. Every morning I use my MLB.com subscription to check the last night’s action. If I don’t have the time to watch the whole game, which is common, I skip to the innings where the Braves scored plus any inning Kimbrel pitches. Pace, a banana curveball and strikeouts, Kimbrel is one of those rare players who is worth watching every minute he plays. Even when he is (rarely) hit you feel a strikeout is coming next. So emotionally, I hate to see him traded (just like I hated seeing Heyward traded). Lots of reporters are saying the trade is a good deal for both sides or an outright win for the Braves, so in emotional despair, I thought I’d have a proper look into it.

The facts of the trade

To the San Diego Padres:

• Craig Kimbrel – 3 years at $34.75m (includes option buyout) or 4 years at $46.75m
• Melvin Upton Jr – 3 years at $48.15m

To the Atlanta Braves:

• Carlos Quentin – 1 year at $11m (includes option buyout) or 2 years at $18m
• Cameron Maybin – 2 years at $16.2m (includes option buyout) or 3 years at $24.2m
• 2 prospects and 41^st pick 2015 draft

N.B. Bold text highlights the likely choices.

I’ll not be analysing the prospects in much detail, instead ignoring the less relevant trade pieces and looking at the end outcomes. My method is below, but if you like, skip to the summary, that’s the important bit.

Methods

From the Padres POV

• Upton not wanted/needed. Treat him as a league-minimum replacement-level 5^th outfielder for 3 years (cost $1.5m). Add the rest of his salary to the Kimbrel contract.
• Dumped 2 unneeded players and $27.2m in contracts off the books. Remove these values as savings for the Kimbrel contract
• Gained Craig Kimbrel. Assume option taken (it is great value – see later*). Contract for 4 years at $46.75m – $27.2m (from Quentin and Maybin savings) + $46.65m (Upton cost)
• Given up 3 prospects (effectively); 1 good (Wisler), 1 risk (Paroubeck) and 1 draft pick

I feel these are all reasonable assumption/treatments. The Padres want Kimbrel, don’t care much about what they get from Upton (assuming he continues as in 2013-14) and used the Quentin and Maybin savings to pay for it all.

From the Braves POV

• Quentin not wanted/needed (not sure why – seems a better bench bat than most and nobody will trade for him as they know they can get him for minimum once the Braves cut him). Add his contract to the 2015 payroll – $11m
• Maybin – Assume continues poor health/form and option buyout is taken. Treat as decent defensive replacement OF (23 career DRS in 8 years). Possibly gets 75 games a season but produces nothing more than T.Cunningham in AAA so set effective salary to league minimum – $1m for 2 years. Add rest of his contract to 2015-16 payroll ($15.2m over 2 years)
• Payroll changes:
  • Savings – Kimbrel ($46.75m – 4 years), Upton ($48.15m – 3 years)
  • Wastings – Quentin ($11m – 1 year), Maybin ($15.2m – 2 years) – both include buyouts
• Receive 3 prospects (effectively); 1 good (Wisler), 1 risky (Paroubeck) and 1 draft pick

Again, I feel these assumptions/treatments are reasonable. Maybin may produce better than this, but his batting numbers were as bad as M. Upton the last few years (70-80 wRC+) so I don’t think we can expect much more of him than Melvin (apart from his defence being better).

Summary

Padres POV

• Get Craig Kimbrel – effectively 4 years for $66.2m ($16.55m/year)
• Get spare replacement-level 5^th OF at minimum salary for 3 years
• Lose 3 prospects; 1 good, 1 risky, 1 draft pick

Braves POV

• Lose Kimbrel (and M.Upton)
• Get spare replacement-level 4^th OF at minimum salary for 2 years
• Payroll savings $67.7m over 4 years ($16.9m/year average)
• Get 3 prospects; 1 good, 1 risky, 1 draft pick

Analysis

Lots of contract money going back and forth, but the end result is that the Braves get payroll savings of around $17m a year for 4 years and 3 prospects and the Padres give up 3 prospects to get Kimbrel at a reasonable free agent price* of around $17m a year for 4 years.

If you consider that the Padres would have lost that 3^rd prospect (the draft pick) if they signed Kimbrel as a free agent, the deal starts to look pretty good for San Diego and AJ Preller. The Padres almost certainly wouldn’t have been able to sign Kimbrel as a free agent with other teams competing (everyone needs a Kimbrel and the Dodgers/Yankees/Tigers/Red Sox etc all have the money for him). The contract would certainly have been longer as well (see footnote on Kimbrel’s historic value*). The Padres are paying Kimbrel a lot, but the amount is fair and they didn’t give up much.

The Braves had signed Kimbrel to a much friendlier contract than he would have got as a free agent (he’s homegrown and a Braves fan so gave a large discount – again see footnote*). Kimbrel gets $13m /year for his free-agent years, when he could have had much more. John Hart effectively used Kimbrel’s generosity to swap the spare value for 3 prospects, one of whom is extremely risky (Paroubeck) and one who is completely unknown (the draft pick). The Braves have rid themselves of Upton, but in taking back other contracts they have effectively only saved the money they should have been paying Kimbrel (had he not given a home discount).

In conclusion, John Hart basically declared he didn’t want a well-paid but high-value closer and swapped it for one good (but not great) prospect and two unknown prospects. So how do I feel now? I would have preferred to watch Kimbrel play for my team every week… Enjoy it San Diego.

*A footnote on Kimbrel’s free agent value

Craig Kimbrel is currently 26 years old and 10 months. Below is a summary list of contracts for comparable relievers and their ages when signing.

You’ll notice that Kimbrel is younger than them all and although the average yearly value is not as high as Kimbrel’s $13.0m 2016 salary, the elite arbitration-eligible relievers are likely to beat them all (apart from maybe Rivera). If he were a free agent this last winter, you can assume that he would have been offered 5-year (and possibly longer) contracts.

Kimbrel’s career numbers are also historically unprecedented at his age. This has been said many times before, but my favourite Kimbrel stat is the WAR leaders for relievers over the last 10 years. Kimbrel has the 5^th highest WAR from 2005-2014. He entered the league at the end of 2010. Since entering the league in 2010 he leads reliever WAR by 1.5 over Holland and Chapman (who have comparable service time). Before signing his (very team friendly) extension Matt Swartz estimated his first year of arbitration salary should be $10.2m. For a detailed analysis of how much Kimbrel is worth I recommend you read his article (http://bit.ly/1GEjKyT). The point being, he is probably worth at least a $17m/year, 4 year contract.

References

http://www.spotrac.com/mlb/san-diego-padres/melvin-upton/

http://www.spotrac.com/mlb/atlanta-braves/cameron-maybin/

http://www.spotrac.com/mlb/san-diego-padres/craig-kimbrel/

http://www.spotrac.com/mlb/atlanta-braves/carlos-quentin/

http://www.fangraphs.com/statss.aspx?playerid=5015&position=OF

http://www.fangraphs.com/statss.aspx?playerid=5223&position=OF

http://www.fangraphs.com/blogs/evaluating-the-prospects-san-diego-padres/

http://www.baseball-reference.com/players/r/riverma01.shtml

http://www.fangraphs.com/leaders.aspx?pos=all&stats=rel&lg=all&qual=y&type=8&season=2014&month=0&season1=2005&ind=0&team=0&rost=0&age=0&filter=&players=0

http://www.mlbtraderumors.com/2013/10/arbitration-breakdown-craig-kimbrel.html