The value of pre-arbitration players

I’ve written a ton about Wins Above Replacement (WAR), salary and when the two are used together how we can get a better picture of overall value. One of my posts looked at the percentage value (WAR percentage minus salary percentage) that each team has received from its highest-paid player so far in 2012. In that post, I showed that almost every team receives lower production from its highest-paid player than the percentage that player takes up of total payroll. This was a pretty fascinating result, but there was more to the story.

One commenter, Todd, made a very interesting point about why we may be seeing this relationship:

You can anticipate getting a certain percentage of production above the cost you pay for it from cost-controlled players. That gives teams extra money to put toward players that aren’t cost controlled

This comment gave me the idea for this post, to discover just how much a team should expect to save on average, per season, from cost-controlled players, in order know how much it can reallocate towards players who take up a larger chunk of overall payroll.

Analyses similar to this have been done before, and they tend to deal with the distinction between non-market WAR (players who aren’t eligible for free agency) and market WAR (players who can sign free-agent contracts). One particular quote from one of these analyses, a Matt Swartz’ post for FanGraphs, stands out in my mind as a good starting point:

Each year, free-agent-eligible players get about 75% of payroll, but they only produce about 30% of all WAR. The average team gets about 12 WAR from Auction-Market talent, but 26 WAR from Non-Market Talent. At the current price of free-agent talent, it’s effectively impossible to build a team out of auction-market talent alone.

So according to Matt, an average team gets 70 percent of its production from players who make up only a quarter of its payroll, an incredible 45 percent of surplus value to reinvest.

I’ve decided to take the distinction between non-market and market WAR one step further and separate pre-arbitration players from players eligible for arbitration. The savings that teams receive from arbitration players pales in comparison to the amount of money they make off pre-arbitration players. For instance, in an extreme case Clayton Kershaw is making $7.75 million in just his first year of arbitration, which is much more than many free agents make on the open market.

Teams get a ridiculous amount of value from pre-arbirtration players who, because they have less than three years of major league service, are eligible for the league minimum salary ($480k). In another extreme case, Mike Trout leads the league in WAR this season, but the Angels are only paying him the league minimum.

I only looked at players with fewer than three years of major league service who are actually pre-arb players; thus, super-2 players and free agents from Japan and Cuba were excluded.

Here’s the average percentage of WAR (and average portion of total payroll) pre-arbitration players have contributed to their teams over the last five years (2007-2011).

The average team had received between 35-45 percent of WAR from players eligible for the league minimum, which makes sense based on Swartz’ finding that 70 percent of WAR comes from pre-arb and arbitration players. Teams have spent between seven and nine percent of their payroll on pre-arb players, which also makes sense.

There is a lot more that goes into this equation, though. I don’t think it makes a whole lot of sense to say the average team gets about 35 to 45 percent of its production from players who take up just seven to nine percent of its total payroll; thus, every team get about 30 to 35 percent payroll savings that it can reinvest in more expensive free agent and arbitration talent. The problem is, the Yankees obviously are a very different organization from the Rays in terms of payroll and need fewer contributions from pre-arb players.

The spread between percentages of payroll allocated to pre-arb players for the teams in the sample was very wide, from the 2010 Yankees with 1.43 percent up to the 2008 Marlins with 53.16 percent.

To attempt to account for this issue, I broke the data into three payroll tiers, with the top ten teams in terms of payroll as “top-tier”, the next ten as “mid-tier” teams, and the last team “bottom-tier.”

Teams in the top tier, who have higher payrolls, tend to devote a lower percentage of their payrolls to pre-arb players, and in turn get lower production from those players. It seems obvious abstractly that teams who can spend more on arbitration and free agent-eligible players have less of a need for major contributions from cheaper talent in comparison to a team with a lower payroll, and the numbers back this notion.

I think this analysis gives us an even better picture of what type of surplus production a team should expect to get from league-minimum players, because a front office understands what payroll tier it is in and what portion of production it needs/will get from pre-arb players. Also, front offices get the benefit of knowing their own players well—and, therefore, have a better idea of what type of production they’re going to get from those players—than those outside the organization would.

The trend of higher-payroll teams receiving a lower level of production, but still a ton of surplus value, from pre-arb players has continued this season.

These numbers still may be too general, though. There’s a large payroll gap ($108 million) between the Yankees, who are first in payroll, and the team with the tenth-highest payroll, the Marlins. And there’s also a smaller (but, still large) payroll gap ($32 million) between the Blue Jays, who are 21st in payroll, and the Pirates, who have the lowest payroll in baseball.

I broke up the tiers from the average between ten teams down to the individual level, for 2012.

I graphed the percentage of salary that each team has spent on pre-arb players this season, broken up by tiers, with the top-tier teams on the left and bottom-tier franchises on the far right.

Some interesting notes:

-The obvious takeaway from this graph is the expected (fairly) distinct trend of increasing percentage of payroll made up by pre-arb players going from left to right.

-The outlier in the top-tier of payroll is the Chicago Cubs, because Jeff Samardzija makes a salary ($2.64 million) that is much closer to a salary of an arbitration-eligible player than the league minimum.

-The outlier in the mid-tier is the Washington Nationals, who have invested a great deal of money into pre-arb players, such as Stephen Strasburg, Bryce Harper and Anthony Rendon.

I also graphed the WAR percentages of pre-arb players for each team, broken up in the same manner:

Some interesting notes:

-This graph is much less skewed to the right than the payroll graph, which is interesting but also makes some sense. The payroll graph almost has to be very skewed to the right, because the percentage of total payroll is a component of total payroll, so if teams with lower payrolls are on the right, then their percentages should be higher. The same is not true for percentages of total WAR, because where a team’s WAR comes from is not a component of, nor dependent on, payroll.

-For example, the Angels, who have baseball’s fourth-highest payroll (top-tier), have received over 34 percent more WAR from their pre-arb players than the Blue Jays, who have the 21st highest-payroll (bottom-tier). The Angels have received large contributions from Trout, Mark Trumbo, Peter Bourjos and Ernesto Frieri, who all make league-minimum salaries, while the Blue Jays’ pre-arb players have not contributed nearly as much.

-The Astros’ percentage of WAR (91.67%) from pre-arb players jumps off the graph but really isn’t impressive, even in the slightest. The Astros’ team WAR is 4.8 (almost three wins fewer than Trout’s WAR), so any percentage of their payroll isn’t a whole lot. Also, they’ve traded away all of their highest-paid players, so now Jed Lowrie is their only player who makes over $1 million (his salary is $1.15 million); thus, it’s pretty much necessary that their WAR should come from pre-arb players, because that’s about all they have left.

The goal of this post, originally, was to look at what type of savings (or surplus value) a team should expect to get from pre-arb players, so its knows how much it can invest in higher-salary players. So here’s the graph of each teams’ surplus production (WAR percentage minus payroll percentage from pre-arb players), so far in 2012.

Some interesting notes:

-The Phillies are the only team in baseball this season that hasn’t been receiving surplus production (-1.31 percent) from its pre-arb players.

-Teams without double-digit surplus production aren’t successful, other than the Yankees. In Swartz’ original quote, he said it’s pretty much impossible to build a team without cost-controlled talent, and only the 2009 Yankees had enough free-agent (market) WAR to be a playoff team. The teams who have received fewer than 10 percent of surplus production from pre-arb players other than the Yankees are: the Phillies, Marlins, Rockies and Blue Jays, all teams who have records well below .500.

-Teams who receive very high surplus production aren’t necessarily having very good seasons. Here is the top five:
{exp:list_maker}Houston Astros
Pittsburgh Pirates
Chicago Cubs
Atlanta Braves
Kansas City Royals {/exp:list_maker}-The Pirates are a team who has been making tons of headlines, as they are in thick of the playoff race for the first time in decades. They also have been receiving a ton of production from pre-arb players, particularly Andrew McCutchen. I wrote about McCutchen’s incredible production for so little cost a few weeks ago and how that production gave the Pirates room to spend and attempt to improve at the trading deadline.

What does it all mean?

The original goal of this post was to find out what type of savings teams get each season from pre-arb players. From 2007 through mid-August of 2012, the average percentage of production from pre-arb players is 38.66 percent at only the cost of 8.03 percent of payroll, good for a savings of 30.63 percent.

The fact of the matter though, is that we can’t necessarily assume that every team is going to get 30.63 percent production above cost from pre-arb players to use to overpay for free-agent talent. We can’t even assume that a high-payroll organization will receive less surplus production from pre-arb players than a low-payroll organization. It is more likely, because on average this is true, but it’s not automatic.

The obvious reaction to this is that, on an individual team level, an organization that is more adept at drafting and developing young talent will receive a higher percentage of production from pre-arb talent than an organization that is less adept in those facets of team building. While I think drafting/player development plays a large role here, it isn’t everything.

There are many ways to acquire cost-controlled talent (trade, free agency, waiver wire, Rule 5 draft) other than through drafting and having a good farm system. The main issue here is to get a solid estimate of what type of surplus value an individual team is going to get in the coming season from pre-arb talent in order to reinvest that money into players who take up larger chunks of payroll.

At the time when teams have an opportunity to acquire high-cost talent on the open market (the winter), developing young talent is less important than just knowing what pre-arb talent an organization already has, because farm systems can’t be changed overnight.

References & Resources
Special thanks to Matt Swartz for going above and beyond with research assistance, especially with this article and providing the data behind it.

All WAR data comes courtesy of Baseball-Reference and is through Sunday August 12.

All payroll percentages come from Baseball Prospectus’ Compensation Tables.