Net Win Shares Value 2006

There’s been a discussion going around a few websites recently about just how powerful rich teams really are or aren’t. It started with a book called The Wages of Wins. Wins received a favorable review from Malcolm Gladwell in the New Yorker, though not favorable enough to make me want to read it. Evidently, most of the book deals with basketball, and I didn’t think there was anything interesting about baseball in the book.

I may have been wrong about that. The authors decided to investigate Bob Costas’ stated opinion that …

The fact is, the single biggest indicator of a team’s opportunity for success from one year to the next is whether the team has a payroll among the top few teams in the league. Period.

They ran a regression analysis of yearly performance against annual payroll and found a correlation coefficient (R) of .43 and a coefficient of determination (R-squared) of .18. Using these stats, they concluded that payroll only determines 18% of a team’s success from year to year, and that Costas was wrong. You can read their reasoning on their blog.

Part of the resulting discussion has been technical (using R vs. R-squared), which I won’t get into because I don’t want to look like a fool. And part of the discussion has been related to the interpretation of the results. For instance, J.C. Bradbury, in his blog entry on the subject, interpreted the results this way: The results indicate it’s quite a stretch to say that success on the field is a product of financial determinism. I don’t know if he intended to, but I think J.C. took the results beyond the authors’ original intent.

So, allow me to express a couple of opinions. First, of course Costas is wrong. He forgot about a couple of little things called minimum salary and salary arbitration. Just remember this factoid: 37% of all 2006 Win Shares were contributed by players making the minimum last year (or close to the minimum). Ryan Howard made $355,000 last year. Joe Mauer made $400,000. Hanley Ramirez made $327,500. That’s great production from guys who weren’t paid a lot of money. If your team is able to gather a lot of these players (as were the Twins and the Marlins), then you’re going to have a great year without spending a lot of money.

I looked at last year’s results and ran a regression of WSAB from minimum-salary players and total payroll against wins to see how much those two variables could predict winning percentage. I got an R of .77 and an R-squared of .6. In other words, I explained 60% of last year’s variance in wins by adding the impact of “low-wage” players. (Yes, I’m being facetious when I label guys making six figures “low wage”.) This is not a rigorous study by any stretch of the imagination, but I think it’s a decent approximation of what a rigorous study would find.

Secondly, I do think that long term success on the field is largely a product of financial determinism. In this year’s Hardball Times Annual, Vince Gennaro (who wrote a great series on baseball economics for THT a year ago) has an article about baseball’s competitive balance. He’s taken a long, hard look at the financials of each specific major league team and has labeled six teams “haves” and nine teams “have nots.” By my calculation, the six “have” teams have won at least 90 games 26 times since 1998, while the nine “have not” teams have had 10 seasons with at least 90 wins. That’s a ratio of four to one per team, which is pretty decent financial determinism.

By the way, I chose to examine the results this way for a couple of reasons:

1. Salary isn’t the same thing as financial strength. Teams choose to spend their money in a number of different ways, including player development and management instead of payroll. I think Vince’s analysis is a better indication of which teams are truly financially strong.

2. Wins are nice, but the difference between 65 wins and 75 wins isn’t really “success on the field.” To me, success is having a strong shot at the postseason, and it’s right about the 90-win mark that separates true postseason opportunity from also-rans (You could make an argument anywhere from 88 to 92; I chose 90 because that’s the point at which more than 50% of teams have made the postseason since 1998, when we reached our current configuration of teams).

There’s a very nice mathematical approach to this discussion at The Book Blog.


Switching gears, let’s talk about the current free agent market. In 2004, players who weren’t yet eligible for arbitration (rookies and sophomores, essentially) contributed 1,720 Win Shares to their teams. In 2006, that same classification of players contributed 2,669 Win Shares. That’s 1,000 more Win Shares from players making the minimum. Thanks to arbitration, these stars are going to remain relatively inexpensive for the next several years.

At the same time, teams aren’t lowering their salary budgets. If anything, they’re increasing their salary budgets, thanks to the fine season that just concluded (record-setting attendance) and the cash inflow from Major League Baseball Advanced Media.

See what’s happening? Combine more production from inexpensive players with more money to spend, and you have teams with blow-out free agent budgets. When you see a team pay over $50 million just for the right to negotiate with a (hopefully) ace starter, you know you’re not in Kansas City anymore. Two things will result from this: free agent salaries will explode, and spending salary dollars wisely will be more important than ever. My advice to general managers is to make sure they really need that free agent they’re negotiating with.

There. Now maybe those GMs will stop calling me all the time.


This is a technical issue, but it’s something I’ve been wanting to mention for a while: the typical “replacement level” analysis isn’t appropriate for evaluating free agent salaries.

As background, most people who evaluate salaries on the web use two basic sets of numbers: salary above the minimum compared to performance above a “replacement level” player. Salary above the minimum is well defined, but replacement level isn’t. Over at Baseball Prospectus, they typically refer to “freely available talent” (or FAT) as a replacement level.

My issue is this: freely available talent isn’t the same thing as players making the minimum. I tallied up all the players who made less than $400,000 last year (the minimum was $329,000) and found that they had an average Win Shares percentage of .455. In other words, if you built a team of the average player making the minimum last year, it would have gone about 74-88, a far cry from the 30-40 wins that WARP seems to use for its replacement level. In each of the last three years, players making close to the minimum have averaged .455 ball.

The replacement level used by most systems, while fine for other types of analyses, is too low for salary analysis. As a result, these systems are all underestimating the relative cost of free agents. Have you ever noticed that most baseball critics and analysts feel that nearly every team overpays for free agents? How can it be that everyone is always overpaying for free agents? The problem is partially due to starting with an inappropriate baseline.

Know what? My system does the same thing. I’ve set the bench level for Win Shares Above Bench at a .350 winning percentage (.250 for starting pitchers). Someday I’ve got to fix that.

By the way, I ran a regression analysis of all free agent hitters’ and pitchers’ performance last year vs. their salaries. Specifically, I regressed their salary above the minimum to their Win Shares Above Bench to see how well teams had spent their free agent payroll. Most of the best projection systems achieve a correlation of .6 to .7 when comparing projected to actual performance in a given year (for batters only), so I figure that is a pretty good standard for evaluating free agent spending. If the best projection systems can’t beat .7, then it doesn’t seem likely that general managers (who are also dealing with long-term contracts and other negotiating demands) can do better.

For hitters, I found a correlation of .67, which is about equal to the best projection systems. General managers don’t appear to be doing too badly. Running the same analysis for pitchers, I found a correlation of .54. Considering what a risk pitchers are, that’s not too shabby, either.


In this year’s Hardball Times Annual I’ve got an article that evaluates the effectiveness of 2006 contracts. I call the system “Net Win Shares Value” and I’ve been employing it for several years now, though I’ve modified it a bit each year. The key thing to remember is this:

  • For every win above a bench player from a player making the minimum (essentially, first- and second-year players) teams pay essentially nothing (because the system assumes that an average bench player makes the minimum salary. Yes, definite flaw in the system).
  • For every extra win from an arbitration-eligible player, teams pay $2.4 million.
  • For every extra win from a free agent, teams pay $4.4 million.

Putting this in perspective, Mark DeRosa just signed a three-year deal with the Cubs for a little over $4 million a year for the next three years. If he contributes just one win per year above a bench player, his contract will have produced an “average” free agent value for the Cubs.

And remember, these figures are based on last year’s free agent contract results—they don’t incorporate this year’s (or the next two years’) inflation at all. I don’t know if the Cubs really needed to sign DeRosa, but his contract wasn’t out of line with the free agent market at all. In fact, it may have been a steal.

Speaking of the Cubs, did you notice that they gave Aramis Ramirez another opportunity to walk away in his latest contract? He has the right to void the contract after the fourth year (which isn’t as bad as his last contract, which allowed him to walk after two years), plus he has a no-trade clause for four years. I really believe that contractual clauses like this often hurt teams more than the dollars involved. When you give up your right to flexibility (which no-trade clauses do) or you give players an out, you are hamstringing your ability to manage your player roster in the future. Don’t do it!

Which reminds me: A-Rod has an “out” clause too, effective at the end of next season, in which his salary is raised if he’s not the highest-paid player in the major leagues. At $27 million, that doesn’t seem likely, but in this free agent market, who knows?


Net Win Shares Value essentially estimates the “expected” production from a player based on how he was signed (as a free agent, arbitration-eligible or not eligible for arbitration) and then compares that to how he actually did. The difference is multiplied by the average amount teams paid for each WSAB last year. If the number is positive, the player was a relatively good deal for the team; if not, not. The best Net Win Shares Value of all last year belonged to the Marlins and Miguel Cabrera, who was paid slightly above the minimum but had an MVP-type year.

Carlos Beltran, my personal MVP choice, was paid $13 million, yet was also one of the 10 best values last year because free agents are paid a lot to begin with. He outstripped his expected production by about 18 WSAB, which made his Net Win Shares Value about $14.5 million.

To determine the bench level of WSAB, I use a .350 Win Shares percentage for everyday players and relievers and a .250 Win Shares percentage for starting pitchers (due to the market’s love of starting pitching and Win Shares’ apparent undervaluation of starters). As I said before, these bench levels are too low; that means that my salary figures are relatively too low. Teams pay relatively more for free agents than even Net Win Shares Value indicates.

Anyway, you can now calculate last year’s Net Win Shares Value for each player yourself. Just use the following calculator to input a few key stats, and you’ll see how much a player’s contract was worth last year, relative to other contracts. Or, you can just wait until the Annual arrives in the mail and get access to the Net Win Shares Value of all 2006 major league players.

The Net Win Shares Value Calculator

To calculate a player’s Net Win Shares Value, you need just a little bit of information. Specifically, you’ll need his Win Shares and Expected Win Shares, as well as his contractual status (free agent, arbitration-eligible or not arbitration-eligible) and his salary. For salary information we recommend the USA Today Salaries database. For a player who is not eligible for arbitration, the calculator will compare his contract to all contracts.

Player’s Name:
Contractual Status:
WSAB:
Salary:

References & Resources
There are two differences between this year’s Net Win Shares Value approach and last year’s. For one, I lowered the bench level of starting pitchers from 60% of expected Win Shares to 50%. Without that adjustment, pitchers are horrifically undervalued and overpaid. Secondly, I removed the constraint by which a player’s Net Win Shares Value can’t be lower than the negative value of his salary. Angel Berroa convinved me to do it.

J.C. Bradbury and Vince Gennaro both have books coming out next year. I’ve seen early copies of both books, and I think they’re going to be fine additions to baseball business analysis and economics.

Print Friendly
 Share on Facebook0Tweet about this on Twitter0Share on Google+0Share on Reddit0Email this to someone
« Previous: How much is Matsuzaka worth?
Next: The Most Democratic Pitcher »

Leave a Reply

Your email address will not be published. Required fields are marked *

You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <strike> <strong>

Current ye@r *