The Price of a win: Part 2

On Monday, we looked at the most recent free agent market, and how teams spent overall. Taking every deal into account, and each player’s projection, age, and aging them at the average rate, we came to a nice round number of $9M per Win Above Replacement (WAR) for the off-season. Last year, the number floated by analysts was around $8M per win, so this represents a decent jump, 12.5% in just a year. While the market looked tame, it seems that it had more to do with the fact that there just wasn’t all that many good players out there to choose from. Just one player, Yoenis Cespedes, reached the 9 figure mark, securing a $100M+ deal. Seven players (Jason Heyward, David Price, Justin Upton, Jordan Zimmermann, Zack Greinke, Chris Davis, and Johnny Cueto) reached that plateau last year.

First, let’s break down these deals into position players, starting pitchers, and relievers:

Starting pitching offered the best deals, but as we’ll see later, that’s probably just a function of the fact that no starting pitcher signed a four year or longer contract, which also probably just a function of the fact that this was a historically weak pitching market.

What probably isn’t just a coincidence is the much higher cost for relievers. Relievers being over-paid is one of the earliest findings of sabermetrics, and that market inefficiency appears to continue to this day. That’s particularly interesting because Kenley Jansen, Aroldis Chapman, and the Cardinal’s own Brett Cecil all signed deals that give the team some projected surplus value.

Here’s a breakdown of single year deals compared to multiple:

Almost two-thirds of all signings were one year deals, but those deals only represented about a quarter of the money spent. The large majority of the money went to multiple year deals. That makes sense, as the players good enough to score multiple year deals are also going to secure higher Average Annual Values (AAV) to go along with it. Players that signed one year deals averaged 0.7 projected WAR in 2017; those that scored multiple-years deals averaged twice as much. Single-year deals averaged quite a bit less $/WAR than multiple year deals, and may just be a function of Supply and Demand. Maybe there was just a lot of mediocre players available, and the few teams that just needed to fill a hole was able to take advantage. Let's dig deeper to see if that was the case.

On Monday, I showed you a scatter plot of every free agent signing. It was difficult to see most the one year deals though, as they were bunched up in the bottom left-hand corner of the graph. Here’s a scatter plot of just the one-year deals:

So the player’s salaries weren’t exactly determined completely by their WAR, but we see a pretty good correlation. That makes sense, because teams are getting smarter and smarter. The stat guys have won the war in MLB front offices. Of course, those teams aren’t checking out Fangraphs’ leaderboards to find out what to offer players, they surely have their own evaluation systems they’re relying on. It does, however, speak to the fact that many people working for teams don’t see a large difference between the public analysis and a team’s internal analysis.

Let’s break down the multiple year deals by year:

Interestingly, Those who scored two year deals on average were projected nearly the same as those that scored one year. Those two year deals seemingly didn’t come with any discount for the team, as on average the $/WAR paid on two-year deals exceeded the rate paid on one year deals. Maybe teams were just higher than the projections on those players, or perhaps those players just had more leverage than their similarly projected one-year counterparts.

We’ll start off by looking at the ten best one-year deals for the team, according to the projections:

Chacin has some injury history, and has been a below average pitcher for most of his career. He heads the list. That type of profile lends itself to a good “buy low” for the team. Carlos Gomez made positive comments about the Rangers’ coaching staff and helping him get through his hitting issues, so it’s not surprising that he ended up taking a discount to stay with them. He’s also a risky asset after his struggles the last couple of years and is hoping a strong season that allows him to score a multi-year deal next year.

There were five average or above average players (2+ WAR) that settled for one year deals. Gomez is one. Neil Walker and Jeremy Hellickson both accepted the Qualifying Offer, and Jose Bautista ended up returning to his team for the same amount as the Qualifying offer. The other was Bartolo Colon, who is entering his age 44 season. Bautista is entering his age 36 season, so it seems easy to explain both away as deals where the market was concerned about age.

And for the ten worst:

The first thing you should notice is that the bad deals aren’t near as bad as the good deals are good. The ten best deals averaged a surplus of $7.25M, the ten worst averaged a deficit of $3.25M. When you add on the fact that a one-year deal by definition won’t affect future budgets, this lends a lot of credence to the common rule of thumb that a one-year deal can’t ever be really bad for the team.

For comparison, we’ll look at all the two year deals from this winter:

Steve Pearce’s deal always looked like a bargain, and this only confirms it. He signed by far the most team-friendly two-year deal according to the projections. The Royals and Rays got good deals for Hammel and Ramos as well. On the other side, Brandon Moss’ deal grades out really bad, while every reliever on a two-year deal looks over-paid according to the projections.

The two-year group doesn’t uncover a single above-average player, but the teams’ actually paid out more per WAR then the single-year deal players. The lower $/WAR for single-year deals seems to be just be a result of the fact that five average or above players settled for one year deals, rather than an ability to pay below-average players less for their production. One year deals may limit the upside for a team, but getting an average or better player on one is a pretty great scenario.

Now to three year deals:

Eric Thames heads the list for position players, thanks to getting the smallest guarantee of any other three year deal. There’s probably more risk for Thames than anyone else on this list, as his projection is based on numbers he put up in Japan. The Brewers were exactly the team to take on that risk when considering their current rebuilding status, and I commend them for doing so.

All anyone can do is talk about the huge discount that the Indians got for Edwin Encarnacion, but according to the projections he received right at what he was worth. Kendry Morales and Mark Trumbo also back up the idea that maybe sluggers are still overpaid.

Ivan Nova’s deal looked like a huge discount when he signed it, this analysis just boils down to what extent it was. Rich Hill might continue the trend of older players getting under-paid because of the increase in risk. The Twins perhaps got a discount for Castro by being the only team to offer him three years. The Rockies’ deal with Mike Dunn continues to look very strange.

And finally, four and five year deals:

Well, hopefully you already saw this coming because you read my article on Monday. The Dexter Fowler deal looks like a disaster according to the projections, though in my previous post on the subject I make the case that he’s about a win better than the projections say, and that would put his deal around break-even.

The Justin Turner and Cespedes deals look like they should be flip-flopped. I predicted that Tuner would be underpaid and that Cespedes would be overpaid, though I won’t toot my horn too much because many others did as well.

As we talked about on Monday, despite being long-term deals for relievers, Chapman, Turner, and Cecil all signed deals that project as a surplus for the team. Mark Melancon helps balance things out, though his high ground-ball percentage could continue to allow for him to produce better results than his FIP-based WAR implies. Three wins is a lot to make up though.

So what have we learned? Well, I think number one is that if you can get an average or better player on a one-year deal, you should do it. Of those five, only Bautista signed a deal that wasn’t a surplus for his team, and that’s because I decided to penalize the Jays a draft pick for it. They didn’t forfeit a draft pick, but they would have eventually gained one if they hadn’t signed him, assuming Bautista would have otherwise signed with a team before the 2017 trade deadline.

I think another take away is that high-end relievers are worth spending a bunch of money on. As a whole, the group received $20M/WAR, but when you take out Jansen, Chapman, and Cecil, the cost is astounding. When Melancon signed, I saw it as a victory for the Cardinals, who signed Cecil at a much more appropriate rate, and wondered if it meant that Jansen and Chapman would get $100M+ guarantees. They didn’t end up getting it, but this analysis shows that it wouldn’t have been outrageous if they would have.

Overall, we can’t come to too many firm conclusions from one season. However, we can keep these things in mind, and see what happens in the next off-season. Before that happens though, we got a lot of baseball to watch, and that’s just fine by me.