Transaction Analysis: Negotiating Nick Markakis’ Defense

Reportedly signed RF-L Nick Markakis for four years and $45 million. [12/3]

Ideally, the Braves would have made their next move by now so we could talk about Markakis and Justin Upton, but we can’t stall forever, so we’ll have to talk about Markakis just.

There’s the dollars and there’s the length, so first the dollars. What’s bigger, the difference between $6 million and $11 million or the difference between $26 million and $31 million? I’m just guessing here, but I’d bet your first reaction was that the former is bigger, but that the conscious part of your brain immediately corrected yourself, so that your real response is that they’re the same. Five million is five million.

But what’s bigger, the difference between $0 and $5 million or between $26 million and $31 million? Here it seems much clearer that the former is the wider gap. If you have a player who is worth zero dollars, and you pay him $5 million, then you’re much more foolish than the GM who pays an elite player a little bit extra, right? At least the other guy got an elite player, and the difference in salary is within any reasonable margin of unpredictability. Put another way: In real life, overpaying by $500 for a stick of gum would be a bigger sign of feeblemindedness than overpaying by $500 for a house.

Put back into baseball terms: The margin of error for defensive metrics matters a little bit when we’re judging contracts offered to five- or six-win players. But the same margin of error might be almost the entire story for a player like Nick Markakis.

Over the past five years, Markakis has produced about 1.7 wins per season, according to WARP. By bWAR, 1.7; by fWAR 1.6, so you see how much agreement there is. When PECOTA comes out next month, I presume it’ll have him projected for about 1.7 WARP. For $11 million, that’s not a great return, but it’s not a terrible one—almost $7 million per win. For a four-year commitment to a player on the dark side of 30, in which the club is absorbing the possibility of collapse or injury and the probability of decline, it’s a bad starting point. This is a lot to pay for Nick Markakis, Name On A Leaderboard.

But what if Markakis is two runs better on defense than we think? Two runs is nothing in an absolute sense. I probably have more faith in defensive metrics than most people, but I have little confidence that we’re precise to within two runs. If Markakis is two runs better than metrics expect, then we’re down to $6 million a win. If it’s five runs, we’re down to $5 million per win, and suddenly we’re writing about a bargain, at least at the front of the deal. We make the best with what we have, and in almost all cases it’s just as likely that defensive metrics are too kind to a player as too cruel. But is that true in Markakis’ case? This is at the heart of it all.

Here’s a lot of what we know about Markakis’ defense. Some of these bulletpoints are worth more than others.

*He won a Gold Glove, this year.
*He was a Gold Glove finalist in 2013.
*He won a Gold Glove in 2011, too.
*He was not a finalist in 2012.
*Fielding Bible voters placed him ninth at RF this year.
*Fielding Bible voters did not place him on a single ballot in 2013. (Twenty right-fielders were named.)
*Fielding Bible voters placed him 14^th in 2012, seventh in 2011, 12^th in 2010, ninth in 2009, and second in 2008.
*In the Fans Scouting Report, he ranks third among all right-fielders.
*He is, in the Fans Scouting, average or better in every facet.
*He has been, in all six years of Fans Scouting Report voting, a plus right-fielder
*By Inside Edge’s metrics—which assign a degree of difficulty to each play—he made

• more "remote" plays than any other right-fielder in 2014*
• the second-most unlikely plays of any RF (out of 16 qualifiers)
• the seventh-most “even” plays
• the second-most “likely” plays*
• and every single routine play, one of only three RFs to do that

*DRS has him -22 runs over the past five years, and +0 in 2014
*UZR has him -20, but +6 in 2014
*The issues are with range; UZR’s assessment of his range over the past five years has been very consistent, until it wasn’t: -8.2, -9.2, -8.9, -11.5, -1.6
*FRAA has him -5, and +2 in 2014
*At least one metric has rated him positively in three of the past five seasons
*Of the 15 numbers put on him by the three metrics since 2010, 10 were negative
*The Braves, with better metrics than we have, paid him as though he’s a good defender.
*The Orioles, with better information than the Braves, did not.

If you were Markakis’ lawyer, you’d tell this story: Metrics obviously have some persistent bias toward Markakis, perhaps something having to do with his ballpark, his pitchers, his teams’ defensive alignments, or the stringers who record much of the data. People who watch him, including the experts who vote on Gold Gloves, as well as the fans who see him the most, think he’s great. You might avoid mentioning the Fielding Bible, or you might spin its swings on Markakis as correlated with nothing, or with the swings of those same biased metrics.

Believable, to a point. The Gold Glove victory this year is a weak piece of evidence, and not because Gold Gloves are an unreliable measure of player defense. (They’re actually not bad these days.) Rather, it’s weak because Markakis faced very little competition. With the outfield awards now split evenly between the three positions, he had to go up against a weak class of AL right fielders. Josh Reddick missed a third of the season and Shane Victorino missed almost the whole season; Lorenzo Cain’s case was wiped out by his multi-position usage; Daniel Nava, a DRS standout, started only 57 games in right field; same with Kevin Kiermaier; Jake Marisnick started only 29 games in right. All are arguably (probably?) better than Markakis, but all were either disqualified or underqualified because of playing time. Of the eight players with the most time spent in RF this year, seven were in the NL—only Markakis was in the AL. That meant he was in a field of basically six full-timers: Torii Hunter, Jose Bautista, Kole Calhoun, Nori Aoki, Alex Rios, and himself. Is victory in this group an overwhelming show of support from industry insiders? Not at all.

The Fans Scouting Report is a bit more convincing, but, as with any metric, it might not necessarily get better in large samples; its biases might just get calcified.

I’m not planning to answer this, by the way. Well, fine, I’ll give you my estimate: I’d estimate than Markakis has not been about five runs worse than average per season over the past half-decade. I’d say most of the evidence suggests he is a skilled right fielder who makes the plays he gets to, that he has a strong arm, and that his range isn’t good; considering his experience, arm, and reputation—his steady competence—I’d be fine assuming little tiny bits of excess value here and there, backing up a play or hitting the cutoff man just right or some such. If I were paying him, I’d probably estimate he’s average defensively. Maybe a run above.

But the point isn’t the number, so much as how much our assessment of his value depends on the number. Remember when I said up there that I have more faith in defensive metrics than most people? It’s true. I believe that they generally point toward something true, that they correlate with wins because they’re picking up real information and that over the course of hundreds of players the two or three or ten runs they’re sometimes off by get smoothed out. But for a single player and a single transaction, where the margin of error might be the difference between below-average defender and above-average defender, and where that difference might account for half the player's salary, the lack of precision can make for very difficult transaction analysis.

So, the bottom line: The difference between this price for Markakis and the “right” price for Markakis looks a lot bigger than it really is. You only have to budge by three or four runs for $11 million to look reasonable. I’ll budge. Seems right.

**

Oh, but the length. There’s no talking me into the length. It’s very rare for a player of Markakis’ level to get such a long contract, and I can say that based not on what I think about Markakis, but about what the Braves think of him—based on what they’re paying him.

Going back to 2005, there have been around 50 free agents who were given four-year contracts or longer. (This excludes players who signed extensions, or international free agents.) Based on their salaries, these players are expected to be three- or four-win players. By average annual value, and adjusted for the leaguewide spending levels, Brian McCann’s contract from last winter is our median. To get a four-year contract (or longer), a player typically has to be a McCann-type.

Markakis, who will make about two-thirds of what McCann will, is nearly at the bottom. Just three of the free agents signed deals as long as his with an average annual value that was (adjusted) lower than Markakis: Omar Infante, Jason Vargas, and Luis Castillo. Long contracts in general have a way of going sour by the end, but the bottom of this list in particular—the guys who were second-tier players getting first-tier contract stability—are a particularly grisly group. Immediately above the four players I named are Ricky Nolasco, Ubaldo Jimenez, Matt Garza, Michael Bourn, Jhonny Peralta, Jarrod Washburn, B.J. Ryan, Gary Matthews, Edwin Jackson. What a list! Between them, they have produced, to date, 22.5 WARP in 28 player seasons covered by these contracts, and of course most of them still have the sad, saggy back-ends of their deals ahead of them. You could make the case that maybe three of the 28 seasons so far have been in line with what the club was expecting—the first seasons of Peralta, Vargas and Ryan's contracts. So the lesson here might be that, if you’re going to go big, go big. Otherwise, there’s just no reason to sink so much into an ordinary player.

But the Braves don’t have an outfielder among their top 10 prospects, unless you count Braxton Davidson, who is 18 years old and already described as “on the clumsy side” with “below-average speed.” Markakis provides certainty in a sense—he gives the Braves a corner outfielder for this year, freeing them to shop Upton or Gattis; he gives them a corner outfielder for the next four years, filling in where a top prospect won’t. He’s disaster insurance. It’s just that, as teams have been finding out with multi-year contracts forever, there’s no guarantee that the disaster isn’t the one you sign. —Sam Miller

*The Inside Edge stuff comes with big sample-size caveats, as all of the groups of plays except "routine" include small numbers of attempts and catches

Fantasy Impact

Unfortunately for fantasy owners, Markakis’ value peaked back in 2008—the last season in which he hit at least 20 home runs or above .300. Since returning from an injury-plagued 2012, which saw him break his right hamate bone and left thumb, his power has almost completely evaporated. Markakis was able to muster 14 home runs this year, but his ISO was a mere .111. Moving from Camden Yards, which had a park factor of 114 for home runs for left-handed hitters this year, to Turner Field isn’t going to help. It wouldn’t be surprising if he failed to reach 10 home runs next year.

Despite the lack of power and speed, Markakis still carries value in fantasy because he’s a leadoff man, as he’s scored 81 and 89 runs in the past two seasons respectively. He’s nothing special anymore, but he’ll make contact (11.8 percent strikeout rate), take his share of walks (8.7 percent walk rate), and get the most at-bats on the team. Ideally, it’ll all add up to another healthy runs scored total. Markakis was still worth $14 in 12-team mixed leagues this year, according to BP’s valuation expert Mike Gianella, but it’s starting to feel like more of a deep league profile. The arrow points down because he’s leaving Baltimore and playing like he’s much older than 31.—Nick Shlain