Wednesday, 12 November 2014

Should We Have the Melk-Man Delivered?

The rumour mill (or in this case The Oakland Press' Matthew Mowery) reported a 'real' interest on the part of the Tigers for free agent Melky Cabrera. Cabrera has most recently been playing in HomerDome North, where the Blue Jays take flight.

By now, the pattern should be familiar. Out comes the Brock2 projection spreadsheet, and in goes the data. However, I would like to briefly digress to take a look at the pattern of Cabrera's career.

Through age 25 (2010), Melky's career slash line was .267/.328/.379. He had mostly been a CF for the Yankees up to then, with his most recent season being at all three outfield positions for the Braves. From age 26 through age 29, he has a slash line of .309/.351/.458. It's not surprising for a player to peak at this period, but that is quite a jump, adding 42 points of BA, 23 points of OBP and 79 points of slugging. He brings doubles power, rather than home-run power. Brock2 suggests his walk rate is going to improve noticeably over the next four years, with his doubles power staying about the same. We've probably seen the best of his home-run hitting. This is a player who was suspended for violating Major League Baseball's policy on doping.

His Runs Created over the next four years are 58, 78, 66 and 72. Not the greatest for a corner outfielder, but there is definitely a spot for production like that where Tigers' outfield corner positions are concerned. He would likely be a noticeable upgrade over Torii Hunter's hitting production, and thus would improve the Tigers' lineup straightaway.

The stumbling block is money. Beyond the Box Score's free agency calculator suggests an annual average value of around $9 million, which is not a great improvement on his last contract, for which Toronto paid him $8 million.

That said, Melky potentially offers a long-term (four-year) fix of a fielding position, and is not coming off a career year. If I'm going to overpay for 274 runs created over the next four years, I think I would rather do it for a left-fielder than a DH.

Monday, 10 November 2014

Victor Martinez: Four More Years?

It seems the market for Victor Martinez' next contract has been set at four years. If I were GM of the Tigers, would I go after him? He is possibly my favourite Tiger hitter, so my initial reaction would be to get in the bidding, post-haste.

Once again, I turn to the Brock2 spreadsheet to consider this. Let's look at his Fangraphs' WAR and RC for his four seasons with the Tigers:

Season      WAR    RC
2011        2.5    97
2012         0      0
2013        0.9    93
2014        4.4   122
Right away, there is a problem. V-Mart was injured in 2012, which makes projections a little bit difficult. Then, we see that being a DH means one has to be an exceptional hitter to show any real WAR impact. (DHs don't field, which hurts them, in the WARverse.) Basically, a .301 average garnered few WAR. A DH needs to hit somewhere between .320 and .330 to start to be worth the kind of salaries a full-time DH can command. Let's also note that V-Mart's 32 home runs in 2014 is a career high. That's not likely to happen again. Here's his projections in those categories for the next four years, plus what Brock2 thinks his Runs Created will be:
Season     BA     HR    RC
2015      .297    14    67
2016      .289    19    78
2017      .314    16    82
2018      .277    14    67
That 2015 RC value is, I think, too low on account of his missing 2012. I would expect a figure more like 2017's 82 is probable. However, all those numbers are below his previous four-year established performance. That is a very large red flag.

Turning to the Beyond the Box Score free agency calculator, and adding a generous estimate of 1.7 WAR for his injured 2012 season, I get an average annual value for a 36-year-old DH of 14.9 million. But I don't think I want to use that value, because of the career-high home-run total for 2014. ZiPS, Dan Szymborski's projection system, forecast only twelve home runs for 2014, while Steamer had him with fourteen. These projections worked out to a 0.9 WAR (ZiPS) and a 1.3 WAR (Steamer). I would prefer to use a WAR number closer to these than to his actual 4.4 WAR in projecting how much I want to pay Victor Martinez.

Adding them together gives a 2.2 WAR, which leads to a far more realistic annual average value for a 36-year-old DH coming off a career year of $9.7 million. That works out to a 4/$38.8 contract, and that would be my ceiling. But we can't do that, because that is a pay cut relative to his 2014 salary of $12 million. So despite his MVP-worthy heroics for 2014, I would say goodbye to Victor Martinez, and many thanks for a great effort.

The problem for the Tigers is the uncertainty over the expensive contracts for Justin Verlander and Miguel Cabrera, both of whom had seasons below expectations in 2014. The last thing the Tigers need is to tie up $20.8 million in excess realistic-market-value for an aging designated hitter. I tend to think the new-fangled sabermetricians are laughably cheap in paying for quality players, but in this case I think they are probably right.

Washington Nationals 2014 Batted Ball Analysis

My batting reviews use types of batted balls to essay an assessment of how much a hitter's statistics might depart from their 'true talent level'. Another way to think about it is to see who might be hitting over their heads, getting that dying quail, or who might be suffering from an excess of 'at 'em' balls and may be likely to improve.

As anyone familiar with sabermetrics knows, one can evaluate batting events by means of linear weights. What this means is that a single is worth about two-fifths of a run while a home run, because it can drive in the men on base, is worth over three times a single. Research has revealed that types of batted balls can assign similar values. Line drives are worth a lot, while infield flies are almost as good as strike-outs. There is a problem in deciding what category to place a batted ball sometimes, especially the difference between a fly ball and a liner, as symbolised by the neologism 'fliner'. So one needs to treat these numbers with a degree of circumspection.

What this chart shows is the difference between a batted-ball linear weight and Fangraphs' wRC. wRC gives a supposed aggregate number of runs that should have been scored based on hitting events. Some people have flares falling in, while others hit the ball hard, but see it caught. The batted ball number also includes Ultimate Base Running, to make it more compatible with wRC. Note that the chart excludes pitchers' hitting. The first column is wRC, the second the batted ball expected runs.

Ian  Desmond       79    61   +18
Bryce  Harper      51    41   +10
Jayson  Werth     100    94   + 6
Danny  Espinosa    31    26   + 5
Adam  LaRoche      84    80   + 4
Zach  Walters       5     4   + 1
Michael  Taylor     4     3   + 1
Tyler  Moore       11    11     0
Nate  Schierholtz   4     4     0
Ryan  Zimmerman    32    33   - 1
Steven  Souza       2     3   - 1
Anthony  Rendon   100   102   - 2
Greg  Dobbs         1     3   - 2
Jeff  Kobernus      0     2   - 2
Scott  Hairston     5     8   - 3
Sandy  Leon         2     6   - 4
Wilson  Ramos      38    43   - 5
Jose  Lobaton      17    24   - 7
Nate  McLouth       9    17   - 8
Asdrubal  Cabrera  21    30   -11
Kevin  Frandsen    19    32   -13
Denard  Span       88   102   -14

I only did this analysis once for the Nationals in 2014 and, just like last time, Ian Desmond has been the luckiest batter, while poor Denard Span has been poorly rewarded for his efforts at the plate. Although I haven't don't a study of this, I use a rule of thumb that +/-5 is within reasonable expectations. So the main candidates to expect a decline from next year are Bryce Harper and possibly Jayson Werth (and Desmond), while we should expect more from Jose Lobaton, Nate McLouth, Asdrubal Cabrera and Kevin Frandsen (and Span). So it's just as well the Nationals picked up Span's 2015 option.

Saturday, 8 November 2014

Washington Nationals Fielding in Review 2014

Here is an update to last time's fielding numbers, one that closes the book on the 2014 Nationals fielding as monitored on this blog. ('Last time' was a l-o-o-o-n-n-g time ago, the end of July.) My source for this is Fangraphs, which includes all the main metrics that interest me except for Michael Humphreys' Defensive Regression Analysis.* From Fangraphs, I've used Mitchel Lichtman's Ultimate Zone Rating, my personal first choice of the 'converted-to-runs-play-by-play' metrics, and my preferred measure of RZR. RZR is Revised Zone Rating, which is like a fielding average but counts balls hit into a zone, rather than those the fielder actually reached. I have included the MLB positional averages for RZR, to help give the players' numbers some context. Note that catchers do not have a Zone Rating. Instead, I have used the runs saved by framing, supplied by

Player              UZR    Change    RZR   Change    LgAverage      DRA    Change   PFr
Lobaton (C)         n/a      --      n/a     ---         n/a        8.8     +1.8    0.2
Ramos (C)           n/a      --      n/a     ---         n/a        2.8     -2.9   -2.6
Span (CF)          -4.7     -4.3    .924    -.010       .919        8.2     +3.6    n/a
Espinosa (2B)       2.3     +2.2    .784    -.004       .787        5.0     -0.7    n/a
LaRoche (1B)       -5.0     -3.1    .793    -.007       .796      -13.2     -5.2    n/a
Harper (LF)         1.5     +2.8    .886    +.014       .881        1.1     +6.6    n/a
Desmond (SS)        0.1     +2.4    .822    +.019       .799      -10.3     +0.5    n/a
Werth (RF)         -1.6     +2.9    .931    -.002       .906      -10.0     -1.4    n/a
Rendon (3B)         4.2     +3.8    .683    -.016       .709        4.2     +1.8    n/a
Zimmerman (LF)      0.4     ---     .907    +.002       .881      - 3.0     -3.6    n/a
A. Cabrera (2B)    -2.2     ---     .758     ---        .787      - 1.5     ---     n/a 
minimum 220 innings

I have been tracking fielding in this manner for a few years now, and my only solid conclusion is that by and large outfielders tend to find their RZR level for the season quite quickly, while infielders can be more mercurial. A metric like UZR, by contrast, swings quite wildly over the course of the season at all positions. This is the fundamental problem with metrics that apply a relative standard, which is the approach adopted the most widely accepted fielding metrics. This is to say that most fielding metrics measure a player's performance against an average for the position throughout the league. I find this deeply unsatisfactory. I have been thinking about this problem, and have a solution that I'd like to test; but I make no promises about posting it here at any time.

Moving on to the actual metrics, the most important thing for Nationals fans to note, I think, is just how much Bryce Harper improved as a left-fielder over the course of the season. Nonetheless, he is not much more than an average fielder overall. The defensive star for the Nationals seems to be Anthony Rendon, if one believes UZR, but RZR paints a very different picture. What is strange is that UZR gives him a high range value. One can only throw up one's hands in despair at such a divergent picture. DRA likes Denard Span best, but UZR seems to think he was one of the worse fielders. RZR suggests he was about average. Again, we see a slight divergence between RZR and UZR on Span's ability to range the outfield. Well, at least we can all agree that Adam LaRoche was a bit of a defensive liability, surely. Nope — though UZR and DRA see him as the worst regular on the team, RZR thinks his range was a little bit below league average.

The only person on whom the different metrics reach a consensus is Asdrubal Cabrera, who was got for his bat. Just as well, too. He is seen as a subpar with the glove.

Let me conclude with a note about Jose Lobaton. He was picked up to provide some defensive help, and to be honest he did exactly that. DRA likes his fielding, and his pitch-framing, which last time I checked was -1.7, rose into positive territory. A good acquisition, Mr Rizzo!

Friday, 7 November 2014

Stranding Inherited Runners 2014: Highlighting Tigers

Matt Snyder at CBS' Eye on Baseball (the Home for All Baseball Fans, as I think Dayn Perry puts it) has once again generated a list using Baseball Reference's Play Index of relievers performance in 2014 at stranding runners on base. Because the Tigers' relief corps has been a matter of some controversy this season, I thought I'd highlight their performance. These are drawn from a list of 110 relievers who over the season added up to having at least twenty-five runners on base when they came into the game.
Player        IRS%     Rank
Hardy         27.6      62
Alburquerque  27.7      63
Coke          36.5      92
Krol          42.1     103
The Tigers' relievers were not particularly good at keeping runners from scoring. None of them are in the top half of the ranking, let alone in an elite zone like the top third or quarter. This isn't really a surprise to those who followed the games. We can surmise from this that Brad Ausmus' main options in a jam during the season were Phil Coke and Al Alburquerque, although Blaine Hardy seems to have taken over from Coke after he arrived to stay.

It does raise the question of why Ausmus changed his bullpen set-up in the playoffs, though. Alburquerque had the second-best percentage by not very much. But, then again, the 2013 table shows he was a much riskier proposition the year before, with a IRS% of 35%.

Even more astonishingly, that 2013 performance was still good enough for second-best on the team. (Phil Coke's 33.3% was the best.) The Tigers may have come in for a lot of criticism for their bullpen this season, but last season was worse. Dombrowski's efforts to improve that sector show he achieved something.

What to offer Giancarlo Stanton?

Yesterday, one of my favourite sites,, mentioned that contract talks between Giancarlo Stanton and the Miami Marlins were underway.

Earlier this year, I wrote a post about how much Ian Desmond might actually be worth on the market, based on what the Washington Nationals suggested he might be worth in their contract offer to him. I am going to use some of the information discussed there, combined with Beyond the Box Score's recent post on estimating free-agent contracts, to see just what kind of price the Marlins might need to pay. Stanton, who won a Silver Slugger award for his performance in 2014, is going to be an expensive proposition, but signing him would do a lot to challenge the 'fire-sale cheapskate ownership' reputation of the Marlins franchise that began after 1997.

Projecting careers is a fool's errand, especially after the kind of traumatic injury experienced by Stanton. Some comparisons have been made to Tony Conigliaro, making projecting just next season somewhat problematic. How much will that experience affect Stanton's ability to stand in the box and attack the ball? Worse still, I don't have a super-sophisticated projection model like ZiPS or Steamer. I do, however, have a Brock2 spreadsheet, based on a system that Bill James devised about thirty years ago. This lets me work out a basic Runs Created value for each of Stanton's seasons, which we can compare to his FanGraphs WAR for the same season:

Season      WAR      RC
2010        2.3       59
2011        3.2       97
2012        5.6       97
2013        2.3       74
2014        6.1      118

There is a big difference in relating WAR to RC in 2010 and 2013 and in relating 2011 to 2012. Examining FanGraphs' breakdown of WAR values between batting, fielding and base-running in more detail, what one sees is that 2011 is suppressed by poor base-running and a lower fielding score. In 2013, Stanton fielded a lot worse than he did in 2010. So we have to be a little bit careful here. Stanton's 6.1 in 2014 for a 118 RunsCreated is probably the benchmark to use, as his base-running and fielding numbers fall around the medians for his career.

Stanton has two arbitration years remaining. What we need is to estimate the value of his next two seasons by WAR, and then guess what they might lead an arbitrator to award. Using that 118 RC = 6.1 WAR value for 2014, for 2015 Brock2 projects Stanton at 121 RC and for 2016 a projection of 126 RC. These work out to 6.3 WAR and 6.5 WAR respectively. MLB Trade Rumors has already projected a $13 million salary for 2015, a doubling of his 2014 salary. Despite the injury, Stanton played quite a bit, making 639 Plate Appearances, and playing time is an important component of arbitration raises. One probably ought to anticipate a similar doubling of salary based on his 2015 projected Runs Created, so let's say Stanton can expect something like $35 million over the next two seasons, allowing him to underperform his projection a little, a possibility after such a horrific injury.

When one gets to Stanton's free-agency years, the problems begin because at this moment we have almost no idea how many years the Marlins might offer, nor how many Stanton might be looking for. There are effectively three broad-based options here. The obvious first two are a long contract that basically ties up Stanton's most productive years, or a short contract that allows Stanton to return to free agency with a prime year or two left giving him a chance at a second monster contract. Based on his Brock2 projection, Stanton would probably want a short contract to end after his age 29 season (2019), the Marlins might want to keep him through age 31 (2021). A longer-term contract would probably see the Marlins wanting him through age 38 (2028), although they might want to curtail it at age 35 (2025) in order to reduce the risk of carrying a high-salaried underperforming player. Stanton probably wants to go right to age 40 (2030), which would still give him a chance to sign a one-year deal with a contender needing veteran leadership from a proven winner.

A third option could be a longish deal with an opt-out. This has the advantage from Stanton's point of view of putting pressure to contend on the Marlins or whatever team they may trade him to after agreeing to an opt-out. Stanton has made it plain he wants to get to the post-season, and he wants to see Marlins' ownership show a similar commitment.

The Beyond the Box Score free agency calculator suggests that, after three 6+ WAR seasons, a free agent right-fielder can expect an annual average value salary of $39.2 million. Since this is based on a 2014 season, we can only expect this sum to go up. I suspect the actual figure might be between $42-45 million, given that MLB player wages rise somewhere between 5 and 8 per cent per annum. That kind of annual salary equals the Marlins' entire estimated payroll for 2014, according to Baseball-Reference. It is also double what the Marlins were estimated to have offered Albert Pujols in 2011.

The only way this deal is going to work for the Marlins, I believe, is with a substantial chunk of deferred money, the technique Jerry Colangelo and Joe Garagiola, Jr used to build the 2001 World Champion Diamondbacks. I suspect Stanton's willingness to sign with the Marlins is going to depend on his willingness to accept a substantial number of deferred payments.

What kind of proposal might open the negotiations? I think I would copy Alex Rodriguez' contract with the Texas Rangers, signed during the 2000/01 off-season. That was a ten-year deal with an opt-out after the seventh year, with an average annual value of $25.2 million and about 20 per cent of the pre-opt-out salaries deferred. I would add together two arbitration years at a value of $35 million, and eight years at $40 million to give a total of $355 million. That gives an annual average value of $35.5 million, and an estimated actual in-season value during the first seven years of about $30 million after the deferrals. (I actually would prefer to defer a lot more, ideally as much as 50 per cent.) I'm honestly not sure I'd go much above $355 million simply because I'm not convinced the South Florida market can sustain a large payroll for an extended period. The evidence is that Floridians are not even particularly good about supporting even winners over the long haul. Sometimes it is better to work around one's shortcomings, than to pretend they don't exist.

The immediate problem, though, is that such an offer means finding an additional $17 million for 2015, before adding any other payroll. At the moment Cot's says that the Marlins have about $18 million in commitments for 2015. Adding $30 million for Stanton will take that up to $48 million, before anyone else's arbitration raise or salaries for pre-arbitration players. For 2015, the Marlins might be looking at a payroll in the $70-80 million range to be as competitive as Stanton would like. That's almost a doubling of payroll in one season. Is that really affordable right now? It's hard to say given Mr Loria's history of tight control on the team's spending.