By In Stuff

Burgers Sold and Pitcher Wins

When I was a kid, the McDonald’s not too far from our home used to have a sign that would brag about how many burgers the chain had sold. I’m probably remembering wrong, but I recall the first number being something like 20 million burgers sold. It was probably more than that, but that’s what I remember.* And the number would change. A month later it would be 30 million sold, then 50 million sold, then 100 million sold, then 500 million sold and so on.

*I also remember when the first number in the price of the car on “The Price Is Right” was a 3. Maybe a 2. And it was just four digits. I’m old.

I remember that Joan Rivers, back when she made a living out of making Elizabeth Taylor fat jokes, used to do one about how Taylor would go into a McDonald’s and they would have to change the number out front.

Eventually, McDonalds gave up on the counting and the signs simply read “Billions and billions of burgers sold.”

I guess the point is this: I used to LOVE that McDonald’s sign. I used to look for it every single time we went to McDonalds. I would keep a running total in my head. How many burgers sold were on the sign last month? How many burgers sold on the sign now? How many burgers did they sell in the month? Surely, I knew the numbers meant next-to-nothing.The number of burgers McDonald’s sold had no effect on my life whatsoever, and the numbers on the billboards were certainly wrong and, more than wrong, not intended to be taken seriously. It was all a marketing scheme.

But I DID care. And, I suppose, based on how long McDonald’s ran the campaign, other people must have cared too.

This brings me to the pitchers’ win. There are probably not many people in America who have railed against the pitchers’ win more than I have over the last few years. It’s a ridiculous statistic. It’s utterly illogical. As I’ve written before, if the pitchers’ win had never been invented and you suggested this goofy statistic — OK, you have to pitch at least five innings, unless you’re a relief pitcher, in which case you don’t have to pitch five innings, you can actually pitch 1/3 of an inning, depending on the circumstances, and the team has to be ahead after you come out of the game, except in certain weird times when the game can be tied, and the team must never lose the lead while you’re out of the game — people would scream at you to put on your Star Wars pajamas and go back to your mother’s basement and leave the baseball thinking to the people who have played the game or watched it for many years.

So, yeah, the win is kind of dumb and often misleading and probably the worst of the most obvious pitching statistics to determine a pitcher’s value. Here are five fairly well known pitching statistics. Which group of pitchers would you want?

Tops in wins

— Max Scherzer, 19

— Francisco Liriano, 15

— Jordan Zimemermann, 15

— Matt Moore, 15

— Chris Tillman, 15

— Adam Wainwright, 15

— Jorge De La Rosa, 15

Tops in ERA

— Clayton Kershaw

— Matt Harvey

— Jose Fernandez

— Anibal Sanchez

— Yu Darvish

Tops in strikeouts

— Yu Darvish

— Matt Scherzer

— Clayton Kershaw

— Felix HFernandez

— Chris Sale

Fewest HR allowed

— Matt Harvey

— Clayton Kershaw

— Jhoulys Chacin

— Anibal Sanchez

— A.J. Burnett


— Clayton Kershaw

— Matt Harvey

— Max Scherzer

— Jose Fernandez

— Hisashi Iwakuma

OK, what’s the best group? I used Baseball Reference WAR to measure so you can take from that what you will. Because there are so many pitchers tied at 15 wins, I took Scherzer (6.0 WAR) and four pitchers with the BEST WAR to give wins as much of an advantage as I could. Those four pitchers, in case you were wondering were Adam Wainwright (4.8), Jorge De La Rosa (4.2), Francisco Liriano (3.5) and Chris Tillman (3.4)

That gave the pitchers with the most win a combined 21.9 WAR. Not bad at all.

But the Fewest Home Runs Allowed group had a combined 24.9 WAR. And home runs allowed, let’s face it, is a pretty limited category.

The best ERA pitchers had a combined 28.7 WAR.

The strikeout leaders had a combined 29.0 WAR.

And the WHIP leaders also had a combined 29.0 WAR.

It’s not close. Wins do not do a very good job reflecting the qualify of a pitcher. As Bill James has said, wins are like watching the first 20 minutes of a movie and deciding if the movie is good or not. You may be right much of the time, but it’s no way to do movie reviews.

OK, but all that has been said in different ways a million times. The win has been back in the Internet spotlight lately mainly because my friend Brian Kenny has been leading a ferocious attack on it (Kill the win!) and some people have been fighting back and some people have been saying that this is just beating a horse that died five years ago and so on and so forth.

But here’s the thing: I don’t want the win killed. I don’t want it to go away. And, at exactly the same time, I don’t want people to use it as a factor for their Cy Young vote, and I wouldn’t want the general manager of my favorite baseball team to even look at it, and I could go the rest of my life without hearing people knock the amazing season Clayton Kershaw is having because he doesn’t have that many wins.

But the win means something emotionally. It has what Bill James calls the power of language. Cy Young won 511 games. Denny McLain won 31 in 1968. Dwight Gooden went 24-4 in 1985 and Roger Clemens did it in 1986 and Ron Guidry went 25-3 in 1978. I know all these off the top of my head, and I’m guessing many of you do too. Rick Sutcliffe went 16-1 for the Cubs in 1984 after the Cleveland Indians traded him, and Doyle Alexander went 9-0 down the stretch for the Detroit Tigers as they won their division in 1987. I know Tom Seaver won 311 games — I was at his 296th victory in Cleveland when he beat Jerry Reed and the Indians 8-3. I know Whitey Ford’s career record was 236-106. I know Old Hoss Radbourn won 48 and 59 games in back-to-back seasons. I know Lefty Grove won 31 in 1931.

Does any of this information MEAN anything? No. But, in the end, do any baseball numbers mean anything? Their value is to tell baseball’s story, to help us remember, to make us feel something. I grew up with the pitcher’s win, it is engrained in me, so I do feel a jolt of excitement when i see that Max Scherzer is 19-1. It’s really cool. I get the same jolt when I see someone with 150 RBIs or a .360 batting average. There’s something fun about a 19-1 season, there’s no reason to deny that joy. Does it mean he’s the best pitcher in the American League? Well, I happen to think he has been the best pitcher in the American League so far, but it’s not because of the won-loss record.

I still want to KNOW his won-loss record, though. For posterity. For history. For the kid who grew up a baseball fan.

I wonder if the young baseball fans today — fewer and fewer, I’m told — feel differently about the win and other old fashioned statistics like batting average and RBIs and errors and the like. They’ve been given all these advanced statistics, they’ve read Moneyball and/or seen the movie, they have been told time and again that the win doesn’t measure what so many claimed through the years it measures. I don’t know what the future holds for the win.

And the win is not a good statistic for much. But it’s a good statistic for remembering. I know Bob Welch won 27 games in 1990 and won a Cy Young that should have gone to Roger Clemens. I know LaMarr Hoyt won 24 games in 1983 and won a Cy Young that probably should have gone to Dan Quisenberry. I know my South Euclid brother Steve Stone won 25 games in 1980 and won a Cy Young that probably should have gone to Mike Norris.

The fact that I remember their exact win totals probably says something. Like the McDonald’s burgers-sold signs, the push for pitcher wins have been like a marketing campaign. But, if I’m being honest, I have to admit: It’s a catchy one.

Print Friendly

94 Responses to Burgers Sold and Pitcher Wins

  1. Ryan says:

    “Why don’t they just put up a sign…”McDonalds – We’re Doing Very Well” – Seinfeld

  2. Todd Akers says:

    Instead of getting rid of the statistic, why not just *fix* it? Have the official scorer simply award the W to the pitcher they feel deserves it most? And if there isn’t anyone deserving, just credit the “W” to the team rather than an individual. At least then, earning a Win will be more “earning” and less a by-product of some convoluted rule.

  3. I view wins from season to season as a small sample size. But over a 10-15+ year career, wins do tell a story. I don’t care what your ERA is, if you can win 10+ games for 15 years, you are doing something right.

  4. I view wins from season to season as a small sample size. But over a 10-15+ year career, wins do tell a story. I don’t care what your ERA is, if you can win 10+ games for 15 years, you are doing something right.

    • Rob Smith says:

      They tell a story, unless you pitch for a team that either scores a lot of runs or doesn’t score a lot of runs & either plays great or awful defense. Then the wins are skewed.

    • Right. That’s why I said over 10-15 years and not single seasons.

    • Andrew says:

      Sure, if a pitcher wins 200 games over a career, they were probably a good pitcher. But there are other stats that do a much better job. It’s like having the cable company tell me they’ll come by on Friday between 1 and 5. That’s sort of helpful and informative, I guess. But it would be much better if they just said “We’ll come at 3:45.”

    • Jacob says:

      It’s also biased because pitchers that stick around the majors to win 200 games are going to tend to be good by any objective measure. Otherwise they would have not managed the longevity in the league. So, not really all that useful.

    • FranT says:

      This comment has been removed by the author.

    • FranT says:

      Andy Pettitte won many games fo many years for great Yankee teams. But evaluating his career based solely on his win totals would not tell you anything close to the full story.

    • How about the career of a pretty good pitcher from the 30’s and 40’s, Hugh “Losing Pitcher” Mulcahy, who had a career pitcher’s WAR of 4.6 (per Baseball Reference)?

  5. Rick Johnson says:

    Pitcher wins made much more sense in the 19th Century when games were much more likely to be a true pitchers’ duel, with each pitching complete games, one winning, one losing. Major League Baseball only began addressing pitcher wins in light of ever developing substitution tactics in 1950, when it first created a rule for assigning wins. The League then modified the rule several times over the next decade until it was refined to its current form. I think no one wanted to abandon the statistic, and so it was modified in an attempt to mold it to the modern world.

  6. Brad Kelley says:


    I love your writing but you need to have an “Apollo 13” moment. You know, where tragedy has struck and all the propellerheads are scurrying around trying to figure out what to do and some kid cuts through all the B.S. and says something like “Guys!. It’s all about the power. We can’t go over 15 amps. Nothing else matters.” or something to that effect. I appreciate all the advanced stats because it gives me more ways to think and talk about the great game of baseball. But despite all that, “It’s all about the W.” That is not old-fashioned. That’s baseball.

  7. What if someone found a way to normalize WAR for pitchers such at an elite pitching season came to 20 WAR. That would normalize the new Win with the old win for counting purposes. Something along the lines of every 9 innings in which 3 or fewer runs score counts as a win. I guess im talking about something like quality starts.

    Not that I like WAR all that much but it would stink to remove the win altogether

    • simon says:

      What if instead it was something like pitchers RE24. REW turns RE24 runs into wins, so maybe that’s a better start. It’s already set to workload – the more innings you pitch, the more chances you have to accumulate positive RE24. But maybe the RE24+ and RE24- could be converted into wins and losses. More RE24+ gives more wins. More RE24- gives more losses.

      To convert into a W/L record, how about this. Divide innings pitched by an innings per start number (I will use 9.0), then divide by 2. This gives a baseline .500 Wx-Lx winning percentage. Add REW to Wx and subtract REW from Lx, and round it off.

      Using fangraphs data, some selected 2013 seasons Wx-Lx (actual W-L):
      Clayton Kershaw 16-7 (14-8)
      James Shields 14-8 (10-8)
      Adam Wainwright 14-9 (15-9)
      Jhoulys Chacin 13-6 (13-8)
      Max Scherzer 13-8 (19-2)
      Matt Harvey 13-6 (9-5)
      Chris Sale 13-8 (10-12)
      Yu Darvish 13-6 (12-6)

      Now some of the bad:
      Joe Blanton 3-12 (2-14)
      Aaron Harang 3-10 (5-11)
      Lucas Harrell 4-11 (6-15)
      Joe Saunders 6-12 (11-13)
      Barry Zito 4-10 (4-11)

      It seems to pass the eye test. I think I will go write some more about this on my blog.

    • Which Hunt says:

      Stephanie, twenty wins above replacement is insane! The only person to ever do it is some guy named Tim Keefe in 1883! He never again came close.
      In fact the highest WAR after 1900 belongs to the legendary Walter Johnson in 1913. His WAR that year was a sixteen. Other notable amazing seasons according to baseball reference:
      Walter Johnson 1912: WAR 14.6
      Babe Ruth 1923: WAR 14.0
      Dwight Gooden 1985: WAR 13.2!
      Roger Clemens 1997: WAR 12.2
      Pedro Martinez 2000: 11.7

      Sandy Koufax’s best year 1963 is worth 9.9 WAR! 10.7 of that is for pitching and a -0.8 for hitting. What is clear to me, though is that a 10 WAR is a great/legendary season (does Kershaw have a shot at a 10.0 WAR this year?). I like simon’s suggestion though. Here is the list I was referencing, and it is worth a look. Ruth’s ’23 is ranked 16! Above him only Walter Johnson’s ’12 and ’13 campaigns and a bunch of seasons from the 1800’s.

    • Yes I know 20 WAR is insane. Thats why i said to normalize it so that 20 wins in an elite season would be roughly equal to what 20 wins has historically meant.

      Im not talking about WAR as it currently is. Im talking about an adjusted war that counts wins differently

    • Which Hunt says:

      You should check that link to simon’s thing. It might be just the stat for you.

  8. Jake Bucsko says:

    In 1972, Steve Carlton earned every single one of his 27 wins. The Phillies went 29-12 in games he pitched, 30-85 in the ones he didn’t. Carlton pitched 341 innings in 41 starts, an AVERAGE of 8 1/3 innings per. 30 complete games, 8 shutouts. A WAR over 12, the 2nd highest in the past 90 years (Doc Gooden in 85). In one 15 game stretch, the Phillies won all 15 of his starts, with Carlton going 13-0 with 129 IP and just 24 earned runs, for a 1.67 ERA and an 0.89 WHIP. 13 complete games, 7 shutouts, including 5 in a row. The 16th start was an 11 inning complete game, a loss.

    This doesn’t really have much to do with the topic anymore, I just love looking at Carlton’s 72 season. One of the greatest pitching performances of all time.

    • Ian R. says:

      Steve Carlton was amazing in 1972, but even in his case, the W-L record doesn’t tell the whole story. The Phillies scored 3.22 runs per game that year, which was bad – in fact, they were the second-worst offensive team in the NL. On the days Steve Carlton pitched, though, they scored 3.75 runs per game – an improvement of more than half a run. Prorate that over a full season’s worth of games, and the Phillies would have scored 585 runs, which at the time was just a tick below average.

      Carlton was a historically great pitcher, and on the days he pitched, he had essentially league-average run support. Of course he won a pile of games.

    • Ian R. says:

      I should acknowledge that some of the difference in the Phillies offense can be attributed to Carlton himself (he posted a .501 OPS on the year, which was pretty darn good for a pitcher). Still, that’s not worth half a run per game.

    • Devon Young says:

      Carlton definitely had a great season in 1972. I still wouldn’t want to credit him with team victories though. Carlton faced 1351 batters. He struck out 310 of them & fielded 47 batted balls. That left the responsibility for the other 994 batters faced (73.57% of Carlton’s BF), to the Phillies defense. They had to turn most of those batted balls into outs, or else a lot more runs would’ve scored. So the defense did more for Carlton’s W-L record than Carlton did.

    • Rob Smith says:

      You miss the point. Pitchers performance determines whether the batter hits routine ground and fly balls… Or rockets in the hole or in the gaps. Obviously the former makes it a lot easier on the fielders. So saying all but strikeouts and come backers are up to the fielders doesn’t even begin to grasp what good pitching is all about. Greg Maddux said that he wanted the batters to swing at his late breaking pitches and ground them to short or second for an easy out with a minimal required pitch count. His formula was to throw strikes, get ahead in the count and make the hitter hit a pitchers pitch. So, tell me again how balls in play are merely up to the fielders.

    • Devon Young says:

      Rob, I believed that up ’til 2-3 years ago myself. You might be interested in DIPS (Defense Independent Pitching Stats). If you look into that, you’ll see there’s plenty of statistical evidence that pitchers don’t really have control over where the ball is hit to.

    • Rob Smith says:

      If that means that all balls in play are equal, and I don’t think that it does, then Maddux must have been really lucky since his strikeout count was pretty average. Pitchers may not totally control where the ball is hit, but they do have an element of control on how hard it’s hit. A big element. Which, I think, is what DIPS tried to do…. Takeaway the defensive advantage or disadvantage to see the pure impact of their pitching. Pitching is certainly not all about strikeouts. Strikeouts are important, but if strikeouts were everything then Maddux and Glavine would have been journeymen, at best.

    • invitro says:

      I can believe that pitchers don’t really have control over where the ball is hit… but do they have any impact on how hard the ball is hit? I would think they do, and I would think a harder-hit ball is more likely to become a hit.

      I won’t believe that fielding is a larger factor than the pitcher for defense until I see some really good and convincing research.

    • Rob, why are you saying Maddux and Glavine would be journeymen if strikeouts were everything? Maddux is 10th in K’s and Glavine 24th. They weren’t flamethrowers like Randy Johnson but they had more than their fair share of strikeouts.

    • B Zman says:

      Especially with Maddux, the key to his success was not his BAA (or his DIPS) which both varied within normal statistical fluctuations, but it was his impeccable control that made him elite. He could not control what happened to the ball at all. He could, however, limit the number of walks to extremely low totals while still striking out a fair share of people.

    • Rob Smith says:

      Of course Maddux’s legendary control was a huge factor. Glavine’s numbers, however are quite different. His “never give in” approach created a fair number of walks (3 per 9 innings). And, he allowed a fair number of hits (8.8 per 9). He had a 1.31 WHIP, which is not elite. But he refused to put the ball over the middle of the plate & somehow that translated into getting out of a lot of jams…. and eventually will land him in the HOF.

  9. Derek Legler says:

    Instead of WAR maybe we should look at win shares or something similar.
    How much did the pitcher contribute to the win or loss and measure his record that way.

  10. Derek Legler says:

    This comment has been removed by the author.

  11. James J. says:

    It was probably 1000 times more than 20 million. 1 billion served was hit in 1963, and 20 billion served was hit in 1976.

  12. marshall says:

    I like the win as a statistic for the same reasons Joe mentions.

    Here’s an idea for a way to evaluate starting pitchers (as if there are not already enough): When the pitcher exits the game, look at the number of innings he went and the number of runs allowed, and calculate the probability (based on league-wide averages) a team wins having given up that many runs through that many innings. Then, subtract off .5. Just keep a running total for the season.

    This is pretty similar to WPA (I think), but does not take into account the runs scored by your own team. If I understand correctly, if your team scores 10 runs in the first inning, there is very little you can do to impact WPA for the rest of the game.

    • Why would you subtract the .5?

      I actually like the WPA type of measurement, which would account for the 1 relevant quality of the Win, which is the idea of pitching to the score. A WPA type stat would penalize less for giving up a run with a 10-0 lead than it would for giving up a run with a 1-0 lead, which I think makes some sense. You can accrue more positive WPA for tight games, but don’t get penalized as much in blowouts, which would normalize it for strength of team/opponent.

  13. BobDD says:

    Movie review after only 20 minutes? Is that the baseball equivalent of an Early Wynn?

  14. Devon Young says:

    Why can’t batter who smacked a walk-off HR, be given a W? Seems he would deserve it more than any pitcher who was sitting in the dugout when the team won.

  15. Did the McDonald’s sign ever say “burgers sold?” I always thought it said Over __ Billion Served and that the number referred to the number of customers the franchise had served, not the number of burgers sold.

  16. I am surprised the NFL (or followers of the NFL) did not more publicly track W-L records for QBs. They are hung up on Super Bowl wins, but why not trumpet the regular season wins!? (Not that I agree with this, of course.)

    Ah, I found QB wins and losses tracked here for active players:

    • Rob Smith says:

      While the W-L record of QBs is not an official NFL or college stat, it is nonetheless frequently used on broadcasts and on fan shows to rate the QB. They even have W-L vs. Top 10 or playoff teams… And other permutations. And there are the same arguments about the validity of W-L records…. And in football, there are 22 guys on the field, so while the QB is very important, it’s even more silly to credit a QB for a win…. I.e. the credit Tebow gets for winning a playoff game that was won by the team, especially the defense while Tebow was passing horribly. Meanwhile QBs like Matt Ryan get downgraded by some because their teams haven’t won enough in the playoffs. Oh yeah, it’s out their in football, and it’s really dumb.

    • Which Hunt says:

      Doesn’t Ryan get downgraded for playing his home games in a dome too? At least with football they acknowledge that different fields produce different numbers. Splitting hairs, I know.

  17. Brandon Gray says:

    We have a much simpler stat at our hands that people don’t use nearly enough; outs, or innings/3. Just count the outs.

    Sure, you can’t compare outs in the 1920s to outs in 2013, but when looking at specific eras, it’s simple and extremely significant.

    • Rob Smith says:

      I presume you mean innings X 3 = outs, not innings / 3. What does that give you that’s different than innings pitched? It’s just innings X a fixed number. And if you compare hits to outs, you’re essentially getting their opponents batting average against a pitcher…. Which is commonly used these days. But outs without the context of hits, walks and runs is as meaningless a stat as is innings pitched.

    • Brandon Gray says:

      Yes, obviously meant innings*3, oops. I was trying to point out a simple statistic for those against formulas and fractions to use, one that is more telling of performance than wins. Innings pitched is absolutely not meaningless, unless you’re a Craig Kimbrel for Cy Young or Trevor Hoffman for HOF supporter, which would just make you wrong. All the stuff about hits, walks, runs wasn’t mentioned because it’s irrelevant to what I was suggesting. If you give up hits and walks, you depart earlier, and thus produce less outs.

    • Rob Smith says:

      There are plenty of “innings eaters” out there that pitch well enough not to get taken out of the game… While not really pitching all that well. The main purpose they serve is keeping their team close and pitch long enough so that the bullpen doesn’t get in the game too soon. Paul Maholm comes to mind. A losing record and sub average ERA+. But generally, outside of injuries, pitches a lot of innings. He serves his purpose, but his innings pitched (multiply by 3 if you must)don’t make him a good pitcher.

    • Brandon Gray says:

      Paul Maholm has pitched over 200 innings once in his 8 year career.

      On the other hand, James Shields will throw over 200 innings for the 7th consecutive year.

      In it’s simplest form, which was the goal here, you can imply that Shields is a better pitcher than Maholm.

    • Brandon Gray says:

      You might be missing the point of my original post. I’m suggesting for a simple counting stat, innings is superior to wins.

  18. Aaron Reese says:

    The problem is never ever ever the statistic, any statistic, only how it is used.

  19. Chad says:

    Let’s kill the save first.

  20. MtheL says:

    Joe – I agree with you for purposes of relievers that wins are meaningless (although, I think the loss is very meaningful for a reliever). But judging a pitcher by ERA can be just as flawed as judging by wins. To reword Bill James a bit, “ERA is like watching an entire movie and determining the whole movie is a failure because of 10 really bad minutes when the rest of the movie is brilliant.” Liriano and Wainwright are great examples – up until recently, they were both brilliant. Both have been rocked by a couple of recent REALLY BAD starts that have caused their ERA’s to take a significant climb. Absent those starts, both pitchers have been fantastic, but they have bloated ERA’s because of those starts. Is it fair to judge a pitcher by his ERA when it was severely bloated by a few REALLY bad games when the rest of his starts have been great? Now, don’t get me wrong, I realize the shortcomings of pitcher wins, but to get a full picture of a pitcher’s season, you really need to consider all of the stats including wins. (I won’t even get into a discussion about pitching to the score, because that’s another big rabbit hole where wins could be more important than ERA.)

    • Chad says:

      Rick Porcello gave up 9 ER in 2/3 of an inning early in the year against Anaheim.

      His ERA with that start: 4.44
      ERA without that start: 3.91

    • MtheL says:

      Exactly! I’m starting to wonder though if instead of killing pitcher wins and losses, maybe we should just kill pitcher wins. Pitcher losses actually tell much more than wins.

    • Rick Johnson says:

      The Quality Starts statistic attempts to address this problem by allowing one to see the number of games in which the pitcher pitched relatively well, I.e. roughly average or better. This then gives context to ERA in a stat sheet.

    • Rick Johnson says:

      The Quality Starts statistic attempts to address this problem by allowing one to see the number of games in which the pitcher pitched relatively well, I.e. roughly average or better. This then gives context to ERA in a stat sheet.

    • Rick Johnson says:

      The Quality Starts statistic attempts to address this problem by allowing one to see the number of games in which the pitcher pitched relatively well, I.e. roughly average or better. This then gives context to ERA in a stat sheet.

    • Rob Smith says:

      Losses? Cliff Lee last year: 6-9 3.16 in 211 innings. Nolan Ryan had 292 career losses. Although there are no decisions, pitcher losses are just the other side of the pitcher wins coin. Bad teams = more losses and less wins and vice versa.

    • Which Hunt says:

      A pitcher loses a complete game 1-0 and that loss is somehow “meaningful”? How?

  21. tomrigid says:

    ​Joe, I know what you’re getting at when you say that wins have the power of language, but I think you’re misusing Bill’s point. Wins do not have the power of language, because as a statistic its meaning is muddied. “Language” in this form is another way of saying “meaning” or “informational clarity” or perhaps “utility,” which are absent from wins as a stat and therefore from the word in this context.

    Wins may have an emotional meaning — kids on little league teams got wins in 1950 and 1980 and 2010 and they all wanted to be the horse on the mound because of that word, with its implication of personal victory in the midst of a team sport. In youth baseball we often see the big kid with “years ahead” stuff who can dominate his peers, and I’m sure that wins were originally a way to keep that kid’s self-interest aligned with the team goal of victory. You can still feel the resonance of it everytime some MLB hurler talks about the value of wins: “You take the ball and go out there and compete, every single time,” they say, and every time I hear it I think no, you can’t believe it matters.

    ​​I always liked what Bill said about stats and language. For me it was a touchstone for clarity and precision, reminding me always to ask myself if the thing I’d said or written actually meant something or if it was just a bunch of words strung together without a point. Like innings, I guess: strung together between a lead change — meaningless.

  22. Dinky says:

    Here’s why wins are a bad statistic. Monday, Clayton Kershaw had perhaps the worst start of his career. Gave up 11 hits, most in his career. Gave up 5 runs in 5 innings. He was better at bat than on the mound (2 hits, 2 RBI). LOST 0.4 value off his WAR at b-ref. But the Dodgers scored 8 in 5.5 innings so he got a precious win. If it had been a home game, no decision (he would have left the game before the Dodgers scored three to make it 8-5). If the Rockies had scored their last 3 before the Dodgers scored their last 2, also no decision. But Kershaw gets a win this time, to make up some for the games he left with leads the bullpen blew, or allowed no runs but didn’t get a win because the Dodgers didn’t score any or score in time. There’s a place for wins, but it is far far down the list.

  23. Phil says:

    The trick is to separate pitcher wins from team performance, or in other words try to determine the pitcher’s “true” or “pragmatic” record — which would keep the expression of value (wins & losses) in terms everyone is familiar with. I focus on runs, using ERA+ to find the pitcher’s annual runs surrendered in comparison to a league-average pitcher. For example, Nolan Ryan in his infamous 1987 season: 24 decisions (8-16), 65 earned runs, league-leading ERA+ of 142. So an average pitcher would have yielded 92.3 runs. Take those runs and combine them however you like (Pythagorean, Negative Binomial, etc.) and spread them over 24 decisions. You get something like 16-8, which is a much better assessment of Ryan’s pitching prowess. Do it for every year, and you find Ryan might have a “pragmatic” record of 344-272 — which everyone can easily get a handle on. Do it for almost all the pitcher every year, and you have a pet project of mine! 🙂

  24. invitro says:

    I became a baseball nut a few years after Joe. I don’t remember baseball statistics well, but the few I remember from the ’80s are exactly the ones Joe mentioned… the famous pitcher’s W-L totals. They definitely tell a story, and I want to keep them because it’s a worthwhile story.

    Nolan Ryan was my favorite player for most of my childhood, and I remember reeling at the injustice I felt when he didn’t win the Cy in 1981 and 1987 when he led the NL in ERA. I knew W-L didn’t tell the whole story, or really even a true story. 8-16 tells a story about Ryan’s 1987, but it’s mostly about his lack of run support and luck. (Although I knew the Astrodome was great for pitchers, I didn’t know how to quantify that until later.)

    I wonder if there’s a stat that could tell the story that W-L does, or at least the true part of it, and tell even more. A better story. OBP tells a better story than BA, and tells almost all the story that BA tells, so we don’t really need BA.

    It was very popular in the ’80s and I think ’90s to turn pitchers’ records into “correct” W-L records, like some people in these comments have done. It was also popular to do the same for batters’ seasons… that is one of my favorite things that Bill James did. I wonder if anyone is doing that now, or if there’s an easy way to do it with WAR and Clutch and WPA and etc. I know Win Shares… are there Loss Shares too?

  25. Larry Jackson, Mudcat Grant, Wilbur Wood, Ron Bryant, Russ Ortiz, Jimmy Key, Joaquin Andujar, Mike Hampton, Mike Boddicker, Andy Messersmith. They all led the league in wins, and none of them won the Cy Young Award. Because people have always known that pitcher wins don’t tell the whole story. Juan Marichal won 26 games in 1968 and lost the Award, because Bob Gibson—who only won 22—posted an absurd 1.12 ERA. Of course Gibson was the best pitcher. No one disputed it. Which is why this argument over pitcher wins is so aggravating. Pitcher wins is an important statistic, but no one has ever treated it like the Holy Grail of pitching. Stat geeks who would die defending misleading crap like WAR would have you believe that wins are completely meaningless. No. Games are won and lost—that is their point after all—and pitcher wins mean that you were durable, that you pitched deep into games, and that you pitched well in high leverage situations. Jose DeLeon had a great arm and good peripherals and found a way to lose all the time. Sure, wins mean that you had run support, but guys like Walter Johnson and Tom Seaver found a way to win for low scoring teams, and Steve Carlton famously won 27 for the worst team in the league. Wins are a terrific stat that go back to the dawn of baseball, illuminating the ways pitchers have been used and misused, telling the story of baseball itself—providing you use a little common sense in interpreting them. To junk one of the game’s great numbers in favor of some inscrutable algorithm is more proof that the savants who want to turn baseball into a branch of mathematics are the most clueless fans of all.

    • Grant says:

      I was with you until you called WAR “misleading crap.” Do explain, please.

      It’s not everything. But I fail to see how it’s ‘misleading crap.’

    • According to WAR, Mike Trout of the woefully under-performing LA Angels is the runaway MVP in the American League over Miguel Cabrera (currently 8.3 to 6.9), even though Cabrera is having one of the greatest seasons at the plate in history, leading the league in batting average/on-base/slugging (a ridiculous .355/.446/.676), plus runs and rbis, with countless clutch hits for a first place team. That alone is enough to flush WAR down the toilet for me.

    • Rob Smith says:

      So Rick. Fielding, throwing and baserunning shouldn’t be considered? And we shouldn’t try to balance out the fact that Trout plays in a tough hitters park? Sure, Cabrera is a legendary hitter. The rest of his game, not so much. Trout is not as good of a hitter as Cabrera. Nobody disputes that. Cabrera is putting up an offensive season that will be rated among the best in years. But Trout’s a better fielder and a much better baserunner (30 steals, I believe, and scores runs because of his legs) playing in a really tough hitters park. That’s all WAR is reflecting. That said, voters can still say Cabrera’s hitting is so awesome that he deserves the MVP…. and most of us who like WAR are still OK with that.

    • Andrew says:

      The whole “pitcher wins will be important as long as the aim of the game is to win” is a silly argument that conflates a team winning with a stat that credits that win to one random player (as someone else mentioned, why can’t the second basemen be given a win?).

      And Joe laid out the whole Trout vs. Cabrera debate in an earlier post. WAR doesn’t take team performance into account, as it’s trying to isolate the performance of a player (unlike misleading team-based stats like pitching wins or rbis). And while Cabrera is having a better season offensively, Trout is pretty close, and is much better at other aspects of the game (i.e. defense and base running, which do matter). Give Cabrera the Hank Aaron Award; that’s what it’s there for.

    • Rob Smith says:

      Rick: of those you listed:

      Messersmith came in 2nd on the CY to Mike Marshall’s historic season for a reliever. So, Messersmith was first in voting among starters. Incidentally, they were both on the Dodgers at the time. BTW: had Messersmith not gotten injured, he may well have had a HOF career. He was an excellent pitcher with a career ERA well under 3.00.

      Larry Jackson came in 2nd to Dean Chance, who had a 1.64 ERA & won 20 games. Koufax came in 3rd with 1.74 ERA, but “only” 19 wins. So Jackson’s wins go him rated ahead of Koufax.

      Mudcat Grant, I’m not sure where he came in for the CY since Koufax “hogged” every first place vote (voting was still for the entire MLB). But Grant did come in 6th in the MVP voting.

      Wilbur Wood led the league twice in Wins (although that was based on his being in a 3 man rotation which gave him more about 10 extra starts) though he had 20 and 18 losses. Still, one year he came in 2nd to Gaylord Perry, who tied him in wins. The other year he came in 5th. It was won by Jim Palmer who had 22 wins.

      Ortiz came in 4th in voting despite a 3.81 ERA. Eric Gagne won with his “perfect” saves season, and each of the Top 3 had ERAs below 2.50.

      Jimmy Key came in 2nd to David Cone. He won 17, Cone won 16.

      Andujar came in 4th to Rick Sutcliffe’s 17-1 year that got the Cubs in the playoffs.

      Hampton came in 2nd to Randy Johnson, who struck out 364 batters

      Boddicker came in 4th. Willie Hernandez, Dan Quisenberry, and Bert Blyleven came in front of him. This one is probably your best example. With a mediocre field typically the most wins would have carried the day.

      So, wins were still extremely important in the CY voting. Five of the examples you cited came in 2nd in the CY voting. All the others were no worse than 5th…. except Grant who came in 6th in MVP voting. Ortiz probably had the “worst” winning season of all noted, but still came in 4th. As an aside, I saw Ortiz pitch that season and it was almost weird science. He did not have good command, walked a lot of hitters (more than 4 per game) Had a WHIP above 1.3 and seemed to always be in a jam. But, he ended up winning a lot, so it felt like a successful season.

    • invitro says:

      Assume players A and B come to the plate with bases loaded an equal number of times in a season, and play in parks with equal Park Factors. Can you give me an argument why in MVP voting, a grand slam for A shouldn’t be worth more than a bases-empty HR for B?

    • Wilbur says:

      I agree. What else do you do with misleading crap except flush it down the toilet?

    • Ian R. says:

      “..guys like Walter Johnson and Tom Seaver found a way to win for low scoring teams, and Steve Carlton famously won 27 for the worst team in the league.”

      Walter Johnson was the greatest pitcher of all time, and Tom Seaver is easily top 10 if not top 5. Of course both of those guys won a lot of games. If they’d played on better teams, they would have won even more. Why is this difficult to understand?

      As for Carlton in ’72, I pointed out above that the Phillies basically gave him league-average run support. Yes, they were a terrible team, but for whatever reason they hit much better when he was pitching.

    • Rob: You left out my favorite on that list—Ron Bryant, who won a robust 24 games but came in 3rd in the Cy Young voting, behind Tom Seaver, with “only” 19 wins, and Mike Marshall, whose value came in appearances, not starts.

      If you led the league in wins, by definition you had a pretty good year, and should be involved in the discussion of the top pitchers in the league. But wins did not automatically equal best, then or now.

      Which is the biggest problem I have with WAR. Wins, though perhaps overrated, were not the only stat that counted. You couldn’t just say, “Ron Bryant won the most games therefore he was the best pitcher case closed.” With WAR, however, that’s exactly what you do. “Mike Trout has the highest WAR therefore he is the best player case closed.” You can’t go to Tom Tango and say, “Get a grip, Tom. Miguel Cabrera is clearly the most valuable player in the American League this year.” WAR doesn’t measure the way a game-tying home run in the ninth inning of a pennant race is worth 20 bases empty doubles in the middle innings for a last place team. WAR measures WAR, the way IQ measures IQ, not necessarily intelligence. But if you go to an expert and say that there is a lot more to value than your metrics can measure, and by the amazo-meter, Miguel Cabrera is having a jaw-dropping, eye-popping, once in a lifetime season, he will point to his spreadsheet and say “Computer says Trout.” At which point, why bother watching the games at all?

    • jkak says:

      Well, you watch the games because that’s what matters. And you count wins, team wins, because team wins matter. And the arguing about how WAR matters beyond its ability to calculate WAR is peripheral to the game.

      WAR is interesting, at least if being discussed with someone who does not view it as an absolute. But it is wrong to say WAR is “objective” or tries to isolate performance from context or, as someone above put it, “it’s trying to isolate the performance of a player (unlike misleading team-based stats like wins or rbis)”. WAR chooses the contexts and team-based stats to which it gives value, such as scoring runs.

      For example, someone else said Trout has increased value because he “scores runs with his legs”. But like all other players, Trout, unless he homers or steals home after tripling, scores runs because of something that someone else did. If he scores from second on a single, it’s because someone else hit the single, not because of his legs.

      If it is random and misleading for a batter to come to the plate with a runner on third, it’s just as random and misleading to be the runner on third when a batter hits a single. If it is random and misleading to give credit to a player whose hit (or ball-in-play out) brings home a runner from third, it is just as random and misleading to give credit to a player who reaches first on an error, takes second and third on ground balls hit by two teammates, and scores on a hit by yet another teammate.

  26. Grant says:

    I dunno if I count as a young baseball fan. I’m 27, which I suspect counts as old for these sorts of things (but I hope not!). It’s my peak season if I’m a hitter.

    I’m also an Orioles fan. And I see that Chris Tillman has 15 wins and I roll my eyes. He’s had a pretty good year, I guess. But he’s no Cy Young candidate. Anyone who can’t see that isn’t looking, as far as I’m concerned.

  27. j-vt says:

    About the McDonalds sign. Do you remember Woody Allen’s movie, Sleeper, set 200 years in the future. There was a scene with a McDonald’s sign in the background that said “Over 795,000,000,000,000,000,000,000,000,000,000,000,000,000,000,000,000,000″ Served.”

  28. Yes! Mike Norris! The best of the pitchers Billy Martin ruined for the 1980-82 A’s teams of my youth.
    1980 A’s rotation complete games.
    Langford, 28; Norris, 24; Keough, 20; McCatty, 11; and Kingman, 10. Heck, even reliever Bob Lacey threw a complete game in his one start that year.
    Norris was cool itself, and burned out far too soon.

Leave a Reply

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