A Final Look at the Market

A timeline:

November 14th: Omar Vizquel signs with San Francisco for $12.25m/3yr.
Commentary: "Wow, Sabean must really hate draft picks."

November 15th: Cristian Guzman and Vinny Castilla sign with Washington for $16.8m/4yr and $6.2m/2yr, respectively.
Commentary: "Bowden's a real tool."

November 17th: Troy Percival signs with Detroit for $12m/2yr.
Commentary: "Will the Tigers ever stop overpaying for stuff?"

November 23rd: Juan Castro signs with Minnesota for $2.05m/2yr.
Commentary: "Wait, hold on a second..."

December 9th: Troy Glaus signs with Arizona for $45m/4yr.
Commentary: "-how...I don't..."

December 10th: Russ Ortiz signs with Arizona for $33m/4yr.
Commentary: "Am I asleep? Is this a nightmare?"

December 13th: Mike Matheny signs with San Francisco for $10.5m/3yr.
Commentary:

...and it went on like that. GM's liked to blame the Kris Benson ($22.5m/3yr) and Troy Glaus deals for the sudden inflation, but several deplorable contracts had been handed out before those two went to press. Instead of settling down, the market only got worse after the initial spending rush, and by the end nearly $600 million of per-year salary was given to free agents (as opposed to just under $400m in per-year salary in last year's market).

In a way, the Tigers bookended the offseason with two context-appropriate deals - the Percival contract came at a time when useful veterans were getting a few too many million or a few too many years, while the Ordonez deal is the ideal example of how teams threw caution to the wind over the course of the market, ignoring projections and red flags in order to sign familiar names.

By mid-December, everyone could tell that the market had reverted to pre-2003 form. But what was the actual extent of this inflation? How many steps backwards did front offices take this year, after spending money in an entirely reasonable fashion a year ago?

I bring you the 2003 and 2004 free agent markets:

[It may not look like the curves match the data very well, but those of you who are familiar with Excel will note the good correlation (r = .78 for each chart).]

In case you've missed the previous editions of this post, I'll provide a brief explanation. Studies have shown that performance in the most recent year has the strongest influence on contract size (think Jaret Wright), so for the sake of simplicity I used a player's 2004 VORP and put it against the average per-year salary of his contract. For example, Troy Percival got $12m/2yr after posting a 15.7 VORP last year, so you match the 15.7 point on the x axis with 6 on the y axis (shouldn't be too hard to spot). Buyouts were included in the average salary calculation instead of option years, and incentive clauses were carefully included.

You should notice that the "best fit lines" aren't linear - they're curvy, and for pretty good reason. The more productive a player, the more his additional productivity is worth. For example, a guy worth five wins a year might get X on the market, and a guy worth six wins a year might get X+1, but a guy worth seven wins might get X+3, and so forth. The basic point that you should understand is that the reationship between performance and salary is curvilinear.

One other point to mention: players who missed a significant chunk of time last year (Sexson, Ordonez, etc.) had their previous full season VORP figures added to their 2004 numbers and then averaged over however many years it covered - so, for Sexson, it was (58.7+7.3) / 2 = 33. This provides a quick and dirty way to account for injury risk while not skewing the data to say that the Mariners will be paying $12.5m a year to a 7.3 VORP first baseman.

Because you can't get a real good idea of how the two markets compare when you look at them individually, let's put them on top of each other (Black = 2003, red = 2004):

Now you see it. The two curves barely even come close to intercepting, with the gap widening as you reach higher levels of productivity. That apparent pink anomaly you see to the left of the graph? Derek Lowe, who registed a -11.5 VORP in 2004. By comparison, Cory Lidle was the highest paid negative VORP player a year ago, earning himself $2.75m after a -2.5 VORP season. As a group study, 14 players who were below replacement value signed as free agents this year for a combined $23.1m of average per-year salary; last year, those numbers were 10 and $11.975m.

For the sake of further comparison, let's take a few player examples and see what the market equations say they deserve in salary:

Adrian Beltre: $14.4m (2003), $18.2m (2004)
Jaret Wright: $5.4m (2003), $7.2m (2004)
Carlos Beltran: $11.4m (2003), $14.5m (2004)
Edgar Renteria: $3.6m (2003), $4.94m (2004)
Pedro Martinez: $7.2m (2003), $9.4m (2004)

(You can see why we consider Adrian Beltre such a spectacular deal.)

One of the other things I wanted to look at was how the market reacted to early spending; that is, if GM's were paying too much in the early going, did that trend continue, or did the market correct itself later on? Let's see (blue = 2003, pink = 2004):

I broke the offseason up into five sections - the dates included in each of those sections are shown by the x axis of that chart. The first data point of each line shows the average VORP per $1m spent for all players signed through November 30th. The second data point shows the average VORP per $1m spent for all players signed through December 15th (including those signed in November). And so on.

What you notice: a year ago, early market inflation was corrected in the first half of December, as teams began getting better deals. By comparison, early market inflation this year only got worse in December, as the Winter Meetings brought about a truckload of regrettable contracts. There was some correction in late December, thanks in large part to Beltre's contract, but the market just flat-out sucked for the duration of the winter. The highest VORP/$1m for any of the five 2004 sections - 6.07, in the second half of December - was still 10.6% below the 2003 average.

Now on to my favorite chart. Blue = 2003, Purple in 2004:

This is how I explained it before:

This year's contracts have been worse across the board, from stars to scrubs. Even where the two columns are closest - that is, in the 40-60 VORP range - the 2004 value is still down 11% from where it was a year ago.

Take a look over at the last category - the overall average VORP per $1m spent (including negative VORP players).

2003. 6.79
      
2004. 4.90

Contract value decreased by 28% in one calendar year. I shouldn't need to tell you how significant that is.

What conclusions are there to draw from this? Well, aside from the most obvious (that the market was inflated), none, really. It's also worth noting that it wasn't just average salary but also length of commitment that seemed to increase this year.

If I weren't in a rush then I'd think of a snappy way to finish this post, but I don't have that kind of time. So, Mariners fans, just be thankful that Bavasi was able to bring in the best deal of the offseason in a dramatically inflated market environment. It could have been a lot worse.