Ten things I didn’t know last week

Win Shares and Loss Shares

A couple of weeks ago, I talked about the essential place wins and losses have in the sabermetric universe. There are a lot of stats that track a player’s contribution to his team’s wins (you can probably come up with a good one yourself), but Bill James’ Win Shares is perhaps the best known. It’s not necessarily the best, but it does things that few other win stats do.

Well, guess what. Bill James has changed the system. In a couple of articles on Bill James Online, he has started to roll out a new system of Win Shares and Loss Shares. Here are the career Win Shares and Loss Shares (where each figure is three times wins and losses) for a couple of players:

– Ozzie Smith: 325-231
– Alan Trammell: 282-176

Trammell’s original Win Shares total was 318, so you can see that James not only added Loss Shares to the system, he changed the overall system. Curiously, Ozzie’s original Win Share total was exactly 325 before; I have no clue why it didn’t change in the new system. Anyway, Smith and Trammell are very close in James’ new system. Smith is 94 Win Shares above average (Win Shares minus Loss Shares) while Trammell is 106 Win Shares above average.

Here at THT, we have addressed the Loss Shares issue by calculating each player’s Win Shares Above Bench. If you compare Smith’s Win Shares and Loss Shares to a .350 player, you get 130 Win Shares above replacement. In our WSAB system, Smith has 127 WSAB, which is mighty close.

It also appears that James has allowed “negative” win shares in his new system—something he avoided in the original. I wrote about the lack of negative win shares a long time ago, and I’m glad James changed his approach because it makes the system more legitimate. It’s also one of the few changes we incorporated into our Win Shares.

However, this development presents a bit of a quandary. Win Shares are popular—our survey results indicated that Win Shares and batted ball stats are the primary THT stats that you folks like to follow. However, James’ old Win Shares method is being replaced. What to do? Well, I think we’ll keep posting Win Shares this year, but we may also start to work on our own system. Yes, just what the world needs. Another baseball stat to drive Murray Chass crazy.

I’ll let you know if we make any progress.

Now you can calculate WPAB.

The only other win stat I know that exactly matches each team’s wins and losses is Win Probability Added (WPA). WPA is the difference between a player’s Win Advances and Loss Advances in each game. By subtracting one from the other, WPA compares a player’s contribution to an average (.500) player.

Well, what if you want to compare a player to another level, such as a .350 player, or .400? It’s a lot more useful to compare someone to a lower level than average, particularly when players have played different numbers of games. Unfortunately, the math for replacement-level WPA isn’t exactly straightforward, because teams compile win advancements even when they lose games. The typical team has 1.65 win advancements in a win and 1.15 win advancements in a loss.

Never fear. I’ve decided to teach myself PHP and MySQL programming this summer, so I’ve created the following little PHP script to calculate a player’s WPAB. Win and Loss Advances are available from Fangraphs and you can select your own replacement level (please use the decimal point—.350 or .400).

Win Advances (from FanGraphs):
Loss Advances (from FanGraphs):
Target Replacement Level (your choice):

You can find this form anytime you’d like on this page.

Who would have played for nothing

They Would Have Played for Nothing is a new book by Fay Vincent, the former commissioner of baseball. Vincent has done an interesting thing—he has followed the same “oral history” format of Lawrence Ritter’s classic The Glory of Their Times, but instead of interviewing players from the early part of the 1900s, the players he interviewed all played in the 1950s and 1960s.

As a result, We Would Have Played for Nothing is more nostalgic than anything else, at least for an old guy like me. Of course, the same was probably also true of Ritter’s book when it was published in 1966. Still, the amount of publicity and knowledge we have of these players is much greater than that available in Ritter’s time. As a result, reading Vincent’s book is a different kind of experience.

Regardless, there are some great stories in this book…

• Ralph Branca talks about Bobby Thomson hitting his shot heard round the world while the Giants were stealing signs (something that Branca heard about in 1954). He is bitter about it, but he never complained publicly.
• In the very next interview, Bill Rigney talks about 1951 and tells a tale of the Dodgers taunting them in August (Branca was singing “roll out the barrel, the Giants are on the run…”). Rigney claims that incident helped fuel their drive to beat the bums.
• Brooks Robinson has a great story about how the players were upset with the fact that Earl Weaver didn’t like to play “small ball.” So they met on their own and decided to ignore Earl and bunt when they felt they should.

Jackie Robinson wasn’t interviewed, of course, but his presence runs throughout the book. He is mentioned by nearly every ballplayer, and several of them have excellent stories about Jackie. My favorite is Lew Burdette’s tale of a misunderstanding between the two of them that ran for most of Burdette’s career.

My favorite interview might have been Burdette. In addition to the Robinson story, he’s got a great one about Orlando Cepeda, who had always hit him hard. Consequently, Burdette and his catcher (Sammy White) decided to tell Cepeda what Burdette was going to throw on each pitch (I guess they figured they had nothing to lose). The Baby Bull didn’t like that, and he popped out weakly on a fastball down the middle. It worked so well that Burdette called Cepeda’s pitches for the next five years.

As you can probably guess from the title, salaries are discussed in every interview. One of the players literally says “we would have played for nothing,” but I took something else away from the interviews: These guys were proud of the advances they made for player’s rights and salaries. Carl Erskine feels he helped pave the way for a new pension plan. Brooks Robinson talks admiringly of Andy Messersmith and Marvin Miller.

There was also some complaining about the attitude of today’s players, from Branca and Frank Robinson in particular, but the complaining wasn’t as intense as I expected.

They Would Have Played for Nothing could have used a little less New York focus (Ralph Branca, Duke Snider, Carl Erskine and Whitey Ford were all interviewed). But overall this was a great read and I highly recommend it.

1976: the year the rotation changed

In my last column, I showed a great graph of the yearly number of days of rest between starts. The data indicated that 1976 might have been a critical year in the movement from three days of rest to four, so I decided to research the subject a little further.

Here’s a little table that shows how dramatically the number of rest days changed between the two years:

  Days      1975    1976    Diff
   0           2              -2
   1           4       6       2
   2          30      25      -5
   3       1,194     860    -334
   4       1,368   1,613     245
   5         439     527      88
   6         166     188      22
   7         105     106       1
   8          69      50     -19
  Tot      3,377   3,375      --

334 fewer starts with three days of rest? That’s what you would call a sea change. The next obvious question is: What happened?

Well, as someone pointed out to me, Wilbur Wood’s kneecap was shattered by a line drive in early 1976, and knuckleballer Wood was the king of three days of rest. But this change was deeper than that. Take a look at the patterns of change for some specific pitchers.

Doc Medich      2  3  4  5  6+   Tot
          1975  0 15 16  4  2    37
          1976  0  0 11  8  7    26

Pete Falcone    2  3  4  5  6+   Tot
          1975  0 14  7  7  4    32
          1976  0  1 14 14  3    32

Phil Niekro     2  3  4  5  6+   Tot
          1975  0 20  9  5  3    37
          1976  0  3 20  7  4    34

Frank Tanana    2  3  4  5  6+   Tot
          1975  1 16  9  3  4    33
          1976  0  6 23  1  4    34

Bert Blyleven   2  3  4  5  6+   Tot
          1975  0 24  6  1  4    35
          1976  0  7 22  4  3    36

Steve Carlton   2  3  4  5  6+   Tot
          1975  0 15 18  1  3    37
          1976  0  4 23  5  3    35

Don Sutton      2  3  4  5  6+   Tot
          1975  0 18 15  0  2    35
          1976  0  3 20 10  1    34

As you can see, there was a widespread spread in pitching patterns across many teams. I looked through my Sporting News Baseball Guide of 1976, and the breadth of this change wasn’t mentioned anywhere. But it happened, and it was significant.

What caused the change? Well, I can tell you what else happened in 1975 and 1976: Peter Seitz struck down the reserve clause, the owners lost their appeal and then locked players out of spring training for 17 days while negotiating a new player contract, Catfish Hunter signed a $3.3 million contract and a couple of dozen players became free agents at the end of 1976. Baseball evolved from a pastoral game to a business nearly overnight; the Dodgers raised their ticket prices for the first time since 1958.

And pitchers started resting a lot more between starts. Was it a coincidence that teams started using their pitchers more cautiously when they suddenly had to pay a lot more for them? We may want to believe that money doesn’t impact what happens on the field, but that is far from the reality.

By the way, many thanks to good buddy Brandon Isleib for the pitching data.

I hate early season graphs

I take a lot of pride in the graphs that we post on our Teams page, but they sure can look ugly early in the season. Take a look at the runs scored/allowed graph above (created last Saturday). The Tigers got off to such a horrid start (see them in the lower left?) that they forced all the other teams to group together on the graph. As a result, you can’t really tell the difference between those other teams. Just another reason to lament Detroit’s poor start.

I’ll be redrawing the graphs about once a week during the season—less often as the season progresses.

Protest the major league blackout!

Does the major league blackout policy for MLB.tv drive you nuts? Me too. Now, we have a place to register our dismay: Check out the MLB Blackout blog, sign the petition and express your disapproval.

And get this. MLBAM has stopped producing condensed games and their archives have evidently disappeared from the site. And they didn’t bother to announce these changes before most of us subscribed to the service.

Seriously, is there anybody who understands customer service at MLBAM?

Marcel is the most dependable system

A lot of people pay a lot of attention to projections systems. Here at THT, we list the Marcel projections in our player pages and we’ve developed our own special projection system for the THT Season Preview.

Imagine my surprise, then, when a fairly in-depth review by Tangotiger found that good ol’ simple Marcel is the most accurate system for established major leaguers.

Now, I’m someone who happens to like simple, explainable approaches (which is why I like stats like FIP and WPA) so this is wonderful news to me. In this case, simple has beat complex. Still, measuring the accuracy of projections is a tricky subject, and this wasn’t a definitive study. We’ll continue to research and find ways to improve our projections. Why? Because it’s fun.

Historic DER graph

Four years ago, I wrote an article on my Baseball Graphs blog that I’m fond of. I looked at FIP and DER (two basic pitching and fielding stats) over the course of baseball history to see how the game had progressed. At the end of the article, I posted a graph of DER by year, and noticed that there was an uptick in the most recent years. I wondered, Is fielding on the upswing?

I was reminded of that article recently, and decided to update the graph. Here it is:

—

The DER numbers are a little different because I calculated them differently, but the trend is the same. And, as you can see, fielding hasn’t gotten “better” in the past three years; in fact, last year’s DER was the third-lowest since 1910 (1930 was the lowest). Do I really think fielding is worse now than it used to be? No. These figures are more likely the result of changes in the ball, turf, ballparks and maybe even a dash of artificially enhanced swings.

Nice start

Twins infielder Matt Tolbert is off to a nice start, batting .435/.500/.478 as of Monday. In fact, as Craig Wright recently pointed out (subscription required), Tolbert started hot and got hotter. After going 1-for-3 (.333) in his first few games, his batting line actually improved each of the next five games:

1-for-2 (average up to .400)
2-for-4 (.444)
2-for-4 again (.462)
1-for-1 (.500)
3-for-4 (.556)

Since his nice little run, Tolbert has gone 0-for-5, but that must have been fun while it lasted.

Please contribute to the sabermetric wiki

During the offseason, Tango started something I had always wanted to do: a sabermetric wiki. A wiki is a great idea for sabermetricians, because it can serve as a local “encyclopedia” of the latest sabermetric insights, with links to all the resources on the Internet.

Unfortunately, the inevitable has happened. After an initial burst of activity, new entries and edits have virtually disappeared. This is too bad, because a wiki like this would be a tremendous resource for baseball bloggers and fans. So if you’ve got the writing or editing bug and would like to contribute, please lend a hand.