Inside the Mind of Brian Sabean

As Aaron reported earlier, and as commented upon by others, the San Francisco Giants signed Omar Vizquel to a three-year deal worth more than $12 million this past Monday.

Omar Vizquel will be 38 at the beginning of his three-year deal, which means he’ll “make” more than $4 million at the advanced baseball age of 40 (more on the structure of the contract later). Want to know how many players made more than $4 million at the age of 40 this year? Just one — the player whose contract represented the best deal in baseball last year, Mr. Barry Bonds.

Now, I’m not saying Omar Vizquel will be a great value for the Giants in three years — far from it. But Brian Sabean is probably the only General Manager who would even think about offering a player in his late 30’s a three-year contract. At least, I think he is. The question is, why? Why does Brian Sabean like old guys?

Due to Sabean’s free agent signings and the phenomenal Bonds, the Giants are a pretty old team. How old are they? Well, a nifty way to measure team age is to take the average age of its players, weighted by the number of Win Shares each contributed to the team. Presented by Bill James in the original Win Shares book, this stat allows you to understand how old a team is, based on the actual ages of those who contributed the most.

Here’s a graph of each team’s Win Shares age on the horizontal axis, along with total Win Shares on the vertical axis:

I’ve added dotted lines for each of the two averages, which splits the graph into four quadrants. It’s safe to say that you want to avoid the lower right quadrant; older, losing teams don’t have much of a future. Pity the poor Dback, Rockies, Mets and Mariners fans. They’re all virtually in the same boat, though some of them have better young players than others.

You do want to be in the upper left corner — younger and winning — because young teams improve and tend to be less expensive. In fact, the future looks bright for many of these teams, such as the Twins, Rangers and Angels.

The lower left quadrant is dominated by “small market” teams struggling to find the winning formula. But how about that upper right quadrant? Winning is good no matter how old you are, right? In fact, did you notice that this year’s “final four” teams are all in the upper right quadrant? The four teams that didn’t advance to the League Championships were the four younger teams (Braves, Twins, Dodgers and Angels). Think that’s something Billy Beane would want to know? This year, at least, age seemed to help in postseason play.

There are actually five teams solidly in the upper right quadrant; San Francisco was the only one that didn’t make it to the League Championships. In fact, they didn’t even make the playoffs. San Francisco is getting older, too. In 2003, their Win Shares age was 32.5. This year it was 33.0. With Vizquel on board, no prospects in line and Bonds (and everyone else) continuing to age, their Win Shares age could creep close to 34 next year. If so, I wonder if they will be the oldest successful team in Major League history.

Other major changes from last year include the Yankees increasing from 31.9 to 33.2 and the Dodgers declining from 31.0 to 28.9 (thank you, Kevin Brown and Adrian Beltre). You can view a list of all 2003 team Win Share ages over at the Baseball Graphs website.

So Brian Sabean likes to sign old players. Really old players. So, you ask, are old players a good investment? Well, here’s another graph — this one is a bar graph of the total 2004 Net Win Shares Value per player by age bracket. Net Win Shares Value is a measure of the value of a player’s contract, as explained in this article and included in the Hardball Times Baseball Annual. This particular graph includes all players signed as free agents, and the number above each bar represents the number of players in that age category:

First of all, there’s a natural progression downwards from the age of 28 to 36. This is a reflection of a number of things, such as the effect of long-term contracts for players who age (think 36 year-old Mike Piazza or Sammy Sosa), a market that has declined in the past several years, so that younger players who have signed more recent contracts represent better deals (think 28 year-olds Vlad Guerrero and Miguel Tejada), and the structure of long-term contracts that pay out less in the beginning and more at the end (all of the above).

After age 36, things change, however, and there are a number of age brackets that represent very good value. In general, this graph resembles a “U” shape, with the bottom of the “U” occurring at age 36 and then turning positive (with a couple of exceptions) thereafter. This second part of the “U” is Sabean territory.

I’m going to show you two more graphs to try and explain why this happens. First, here’s a graph of average salary by age:

See that jump in salary at the age of 36? 36-year-olds made $5 million a player last year, which was higher than most of the age brackets on either side of it. Those are some old, bad contracts that owners now regret. Salaries dropped 50% at age 37, to $2.5 million, which is the key reason that Net Win Shares Value improved so much at that age.

Some of the key value players at age 37, such as Tom Gordon, Reggie Sanders and the Giants’ Marquis Grissom, signed deals in last year’s bear market. But the Value “U” continues past the age of 37, into 38 and, after a big drop at 39, into the 40’s.

Some of the valuable 38-year-olds were guys who had signed longer term contracts and were subsequently paid more in 2004 ($4.4 million per player). But these guys, including Moises Alou and Curt Schilling, delivered the Win Shares and Net Win Shares Value. Age 39, however, was a completely different story. Several of these guys were also signed to long-term contracts a long time ago (Kevin Brown, Al Leiter) but they just weren’t that good in 2004, as measured by Win Shares Above Replacement. To illustrate, here’s a graph (the last one, I promise) of Win Shares Above Replacement by age:

See, there’s a general “U” shape in this graph, just like the Net Win Shares Value graph. Players hit bottom around age 35 and then, as the talent pool weeds itself out in the mid 30’s, the older players are literally more valuable. Perhaps the average good player in his late 30’s loses it more slowly than the average good player in his early 30’s; I don’t know if that’s true, but it makes intuitive sense to me. That, plus the short-term contracts many of these players sign, make players over the age of 36 generally good values.

Who were the stars at age 40? Well, Bonds of course, but also B.J. Surhoff, Barry Larkin and Kenny Rogers. 41-year-olds included Randy Johnson and David Wells, and the key 42-year-old was Roger Clemens, whose value more than offset Jamie Moyer’s terrible year.

Those two one-player samples at ages 44 and 46 are the Francos. John Franco, at the end of a fine career, was not really worth the contract he was given in 2004. But Julio Franco, who had probably the second-greatest season of anyone at 46, gives a nice flair to the end of the “U.”

Give Brian Sabean a little credit. He has a feel for this phenomenon, and he’s taken advantage of it. In a crazy way, this is the “Moneyball” philosophy. Approach the market in a different manner, and take advantage of values that others don’t recognize. Billy Beane and Paul DePodesta valued OBP before others did; Brian Sabean values really old guys.

Now, this business of signing free agents before the deadline on purpose to give up a draft choice is simply suicidal. And I’m still not saying the Vizquel contract is a good one — two years should have been enough — but it isn’t as bad as it appears on the surface. Vizquel will be paid a base salary of $2 million next year, $2.5 million in 2006 and $4 million in 2007. His bonus will be paid in four installments from 2005 to 2009 (skipping 2007), which lowers the overall cost of the contract. And Vizquel has an excellent chance to be better than Cristian Guzman for at least the next two years.

So maybe there’s a method to Sabean’s madness. Just for fun, try this thought exercise: If Bonds and Randy Johnson continue their elderly brilliance and the Giants continue to contend in 2005, what will Sabean offer Johnson when he becomes a free agent next year? Your answer will indicate just how “mad” you think Sabean is.

References & Resources
When computing age, I subtracted the player’s birth year from the current year (2004), which differs from the typical baseball age (which is computed as of July 1st of the year). By doing this, I essentially round the player’s age, measured in years and months as of July 1st, to the nearest age, measured in years.

As an example, Barry Bonds was born July 24, 1964, so his “baseball age” is 39. But, actually, he was 39 years and 11 months old on July 1st, so my method rounds his age up to 40, which seems more appropriate for an analysis like this. Thanks to Tangotiger for pointing this out (and I highly recommend you read his aging analysis).

I apologize for any confusion.