Skip to content

Kill the win? Nah, let’s just reduce its importance

Aug 27, 2013, 9:45 AM EDT

Max Scherzer AP

So I’m reading the latest Jon Heyman column and I’m finding myself agreeing with him more than I am with people coming at the issue he’s discussing from a sabermetric point of view. And I wasn’t even hit on the head this morning or anything.

The topic: the Kill The Win campaign Brian Kenny has been waging at MLB Network and on his NBC Sports Radio show. You’ve probably seen or heard some of it. Basically Kenny is crusading to kill pitcher wins as a stat. Now, I’m not sure that he really wants to eliminate the pitcher win. I suspect this is more about conscious rhetorical overstatement in order to shock people into looking at the issue more thoughtfully. Like, Kenny is at 11 in order to get people up to a 5. It makes sense, as do Kenny’s underlying arguments about why the win stat is misleading and way less useful than people tend to think. We’ve been talking about that for years here, of course.

But to the extent this is actually about killing the win — or to the extent people take the baton from Kenny and make extreme arguments about wins being utterly meaningless — I’m more on Heyman’s side of things. Heyman, you may not realize, is actually pretty sensible about pitcher wins. He voted for Felix Hernandez for the Cy Young a few years ago despite the low win totals. He understands that strikeouts and baserunners and stuff matter more than wins. This thinking somehow disappears when he starts talking about Jack Morris and the Hall of Fame but he’s no Hawk Harrellson or Harold Reynolds about these things.

I don’t value wins too much in pitcher analysis, but I don’t think they’re utterly meaningless. For me they’re attention-getters more than anything. When I’m looking through stats from past decades and I see pitchers with big win totals I tend to want to look more deeply at their stat lines to see what kind of season they really had. If I see pitchers who I know (or heard) were good, I notice low win totals and look at their teammates and strikeouts and unearned run totals and things.  I use wins as a signpost, and I’m glad they’re there for those purposes. Growing up in the 70s and 80s wins were much talked about and no matter how much my thinking has advanced, I still key on them some. Having them around is like having the common phrases page in the back of your guidebook while traveling in a foreign country.

Also: wins are fun to talk about outside of analysis. As are bunts and batting average. I like having them around for that kind of fun and I like talking about them as long as people don’t mistake the fun talk for meaningful analytical talk. It’d be a real bummer if those stats disappeared simply because we don’t use them the way we used to.

Ultimately all of this may turn on how you feel about revolutions in general. I’m kind of a cautious guy with small-c conservative tendencies. I’m pro-change and advancement and think science and math and change and the future are wonderful things that we should embrace and not fear.  But I am wary of rhetorical extremes and ideas which posit that the past is crap and must be abandoned if we are to advance.

Kill the win? Nah. Just reduce it’s significance. And keep it around like we keep around record players and manual typewriters and stuff. They’re neat.

  1. Old Gator - Aug 27, 2013 at 9:53 AM

    Great idea! And while we’re at it, let’s have divisional standings based on team ERA too.

    • stevetsra - Aug 27, 2013 at 10:08 AM

      I’m not seeing how easing the importance of pitcher’s wins, which are largely out of their control, has anything to do with basing division standings on team ERA.

      • Old Gator - Aug 27, 2013 at 10:42 AM

        Can’t read the sarcasm font? Time for a retinal scan.

      • baseballici0us - Aug 27, 2013 at 10:55 AM

        wait, was that sarcasm as well? :-)

      • Kevin S. - Aug 27, 2013 at 12:10 PM

        Honestly Gator, the only way to realize that was sarcasm was to see who was posting.

      • Old Gator - Aug 27, 2013 at 3:40 PM

        Well heck, I warn’t hidin noplace.

    • michaelle123 - Aug 27, 2013 at 4:25 PM

      what Marcus implied I didn’t know that a stay at home mom able to get paid $4252 in 4 weeks on the computer. have you read …???? http://miniurl.com/dyRW

  2. cohnjusack - Aug 27, 2013 at 9:58 AM

    Pitcher wins, as is, doesn’t really mean anything. A starter goes five or more innings, leaves the game with the lead that is subsequently not relinquished….and then a whole different set of rules for relievers who for some reason don’t have to go five innings to get a win. Why does that make any sense. STARTER can go 6 innings, give up 0 runs and leave the game up 3 – 0. RELIEVER can then come in, give up three runs in one inning, have the offense score another run and get the win. Huh?!?!

    How about this: Instead of all these seemingly random criteria for awarding it, just give the starter a win if his team wins the game and a loss if they lose it. It’s still not great, but it tells me a hell of a lot more about the pitcher than the current system.

    Under this system, who would you vote for CY Young?

    Pitcher A: 21-12, 2.53 ERA, 227 innings, 1.023 WHIP, 229 K, 63 BB
    Pitcher B: 22-11, 2.73 ERA, 233 innings, 1.053 WHIP, 230 K, 54 BB

    Hell, not much a difference really. So why did pitcher B get 27 1st place votes and pitcher B only got 2?
    Because, despite his team winning almost the same amount when he started, Clayton Kershaw’s “record” was actually 14-9 while RA Dickey’s was 20-6.

    • number42is1 - Aug 27, 2013 at 10:08 AM

      you.. uhh…. care to fix that a bit?

    • waiverclaim - Aug 27, 2013 at 10:11 AM

      Pitcher wins have to do with offense, defense, relief pitching, management, umpiring aka all facets of the GAME, and some genius decided to come up with doling out a stat to a single player when all other players played a part in it. Its stupid, dumb and should not exist.

      If ALL pitchers went 8+ every game, you might have a point, but even still its pretty dumb anyway. Baseball is a team game not an individual one. It was broken from the very beginning and never stopped being a broken, dumb way to value a player. There’s so many other stats that do it better.

      • rbj1 - Aug 27, 2013 at 10:50 AM

        Cy Young pretty much did. 815 starts, 749 complete games (according to baseball reference.) So in his day, wins were a function of the starting pitcher. Nowadays, with 7th & 8th inning guys, Ws might have reduced importance, but back in the day they were more important.

        And to have a high win total, an organization has to look around every fifth day and decide you are the best option for the team to get a win that day.

      • grumpyoleman - Aug 27, 2013 at 4:12 PM

        Why do we need any stats if it is a team sport?

  3. scoutsaysweitersisabust - Aug 27, 2013 at 10:06 AM

    Welcome to 1999 Heymen!

    While we’re at it, can we kill RBI’s too? Pretty please! With cherries on top! And sprinkles….

    You know, we may as well kill errors also, since score keepers can’t seem to get that stat right. (It says NOWHERE that a player must make contact with the ball for it to be considered an error. An error is any play that the fielder should reasonably have been able to convert into an out. Stop inflating batting averages by refusing to call an error an error.)

    • cohnjusack - Aug 27, 2013 at 10:14 AM

      I once saw a fly ball hit to Jim Edmonds on a sunny day. Edmonds took a couple of steps in, put his glove up…and the ball bounced out of the glove to the grass.

      It was ruled a double. Because…you know…it was really sunny. Hey, you try catching a fly ball in the sun!

    • Jeremy T - Aug 27, 2013 at 11:02 AM

      Don’t forget saves. Wins, RBIs, and errors are harmful for analysis, but saves are harmful for how pitchers actually manage the game.

    • ctony1216 - Aug 27, 2013 at 12:31 PM

      For a starting pitcher, wins are not easy to come by. You have to out-pitch your opponent, and your bullpen has to hold your lead. Not easy, especially for pitchers on weak-hitting teams with lousy bullpens. That’s why a 20-win season is pretty special, as is 300+-win career.

      To not keep track of those stats would be a shame.

      • scoutsaysweitersisabust - Aug 27, 2013 at 1:26 PM

        As Craig said in the article, go ahead and keep wins, but stop relying on them to be important. As you have yourself said, wins rely on your offense and bullpen, not just your starter’s performance. RBI’s rely on your teammates to hit and get on base before you. Errors rely on the scorekeeper to make a judgement call. (One that’s increasingly failing to be done properly.) In addition, teams have several motivations not to assign errors.

      • grumpyoleman - Aug 27, 2013 at 4:17 PM

        The RBI argument is weak. Miggy gets a third of his RBI’s by driving himself in. If nobody is on ahead of you do something about it.

  4. waiverclaim - Aug 27, 2013 at 10:08 AM

    Name a stat from any sport that had “reduced importance” from previous years/eras. I’ll give you GW-RBI. Now you go:

  5. danaking - Aug 27, 2013 at 10:09 AM

    Pitcher wins don’t mean a lot on a yearly basis, but over a career some things–such as run support–tend to level out. If a pitcher has a lifetime winning percentage of .600 while pitching for teams with a cumulative winning percentage of .500 over a career of fifteen years, it’s a good bet he was a fine pitcher. By the same token, bad pitchers don;t win 200+ games.

    Wins aren’t the be-all and end-all stat some would like them to be, but they can be a worthwhile tool in evaluating pitchers when used in their proper context.

    I’ve been reading Bill James since he was first mass published, and consider myself a relatively early adopter of many advanced metrics. What disturbs me now are how many advocates of those metrics either try to run down the value of almost all traditional stats, or have reached a point where their calculations have themselves, become somewhat arbitrary. (Some of the advanced fielding metrics come to mind.)

    • waiverclaim - Aug 27, 2013 at 10:11 AM

      A pitcher can literally go 0-30 with a 1.00 ERA. That is near impossible to do, but it could be done.

    • cohnjusack - Aug 27, 2013 at 10:16 AM

      It really doesn’t level out that much. Which is why a guy like Jim Bunning can end up with 30 fewer wins than Jack Morris.

      • yahmule - Aug 27, 2013 at 10:22 AM

        Sadly, Bunning never learned to pitch to the score.

      • spursareold - Aug 27, 2013 at 10:26 AM

        Bunning played in the reserve clause era, and was never able to improve his lot by switching teams. After a decent stay in Detroit, Morris moved around.

    • spudchukar - Aug 27, 2013 at 11:21 AM

      And the UZR numbers aren’t improving either. Some things are just difficult to quantify, and until defensive metrics get much better they shouldn’t be used, because currently they aren’t as good as a trained eye and often brand players who do not deserve either their criticism nor their applause.

      • Reflex - Aug 27, 2013 at 3:05 PM

        I can’t agree with this, and GG awards are probably the biggest evidence of it. The ‘trained eye’ thinks Jeter has been an amazing defensive SS over his career. The ‘trained eye’ gave Raphael Palmerio a GG when he spent most of the season as a DH. The ‘trained eye’ consistently rewards players who make a lot of diving plays without figuring out if they had to dive only because their range was low or they misread the hit.

        Yes defensive stats need a lot of work. But I’d take the advanced defensive stats out there now over the so-called ‘trained eye’ that gave Michael Young a GG at short.

      • spudchukar - Aug 27, 2013 at 3:30 PM

        No it doesn’t. The trained eye shouldn’t be confused with bad decisions. All the awards you state were made by untrained eye. There is a difference.

      • Reflex - Aug 27, 2013 at 3:41 PM

        The GG is awarded by managers and coaches in each league. That is about as ‘trained’ as an eye can get.

        You are arguing for subjectivity. I am arguing for removing that subjectivity. Advanced metrics may not be perfect, but I never let perfect be the enemy of ‘better’.

      • spudchukar - Aug 27, 2013 at 4:09 PM

        Wrong again. The issue with UZR is how poorly the quantification, which has to use subjectivity. Range factors, how hard a ball is hit, is it a screaming line drive, or merely a soft liner, or somewhere in between, the angle an infielder takes, his positioning and a dozen more factors need to be taken into consideration.

        I had hoped some trained eyes would be given the task to grade/rate defensive metrics, but to date they have gotten worse not better.

        As far as the GG goes, we all know offense is always part of the equation. It shouldn’t be but is.

      • Reflex - Aug 27, 2013 at 4:17 PM

        You keep asserting this but where is your evidence? Is UZR perfect? No. Is it good? Not really. Is it better than subjective ‘trained eyes’ that consistently reward players who by any objective measure are terrible defensively? HELL YES!

        You can make any excuses you want, but at the end of the day the trained eyes you speak of consistently make terribly judgments on the performance of players defensively. I will accept a metric that can continue over time to be improved.

        BTW, on what do you base your assertions that its getting worse? There are a number of advanced defensive metrics, UZR is but one of them, and probably the most basic one still in use. UZR itself is unlikely to improve much given its premise but others will eventually replace it(and are already being used).

        Again though, the idea that a imperfect stat is somehow worse than pure subjectivity especially when those doing the grading are consistently proven to be very bad at it is just ridiculous.

      • grumpyoleman - Aug 27, 2013 at 4:23 PM

        Those GG weren’t awarded by “trained eye”. They were awarded mostly via reputation or a couple of highlights seen on sportscenter.

      • Reflex - Aug 27, 2013 at 8:44 PM

        While I agree that most coaches and managers likely do not watch every play for those in contention, I see that as an argument for why the ‘trained eye’ test is pretty baloney. The real ‘trained eyes’ in an organization are the scouts, and they still only see a tiny fraction of the plays of any given player.

        Its a big argument in favor of objective statistics however, even if they are flawed. At least they can quantify every play rather than just the tiny handful that a few ‘trained eyes’ might see.

      • spudchukar - Aug 28, 2013 at 11:37 AM

        They aren’t objective. That is what you cannot seem to understand.

      • Reflex - Aug 28, 2013 at 2:51 PM

        *sigh*

        Go read the Derek Jeter thread. It goes over a lot of this and the methodology. And yes, I’ll take that over any ‘trained eye’ you want to throw at me. The methodology is sound and a very good start, even if not perfect. Furthermore, they understand why and where it is not perfect and attempt to compensate for that.

        Again, perfect should not be the enemy of better.

      • spudchukar - Aug 28, 2013 at 4:13 PM

        Sigh all you want. The methodology is not sound. If anything UZR metrics are falling apart for their incompetent decisions. They are ridiculed by the knowledgeable, and they, by association are harmful to the meaningful sabermetric contributions.

        One need not have a “trained eye” to recognize the complete failure to date.

        The Jeter reference is risible. In this instance it isn’t perfect that is the enemy, but blind allegiance to a methodology which, to date has proven to be an embarrassment to an otherwise worthwhile endeavor.

      • Reflex - Aug 28, 2013 at 5:23 PM

        http://hardballtalk.nbcsports.com/2013/08/27/an-illustrated-analysis-of-derek-jeters-defense/

        That article and what it links to. Including the methodology being used. You keep beating up on UZR. Fine, don’t use it. There are lots of others out there, this article pointed out a couple and used one that did a damn good job too. Pretending that defensive analysis is all bunk and the mythical ‘trained eye’ is better is just closing your eyes and putting your fingers in your ears.

        I always figured you were smarter than that. Go read please.

      • spudchukar - Aug 29, 2013 at 11:19 AM

        Please tell me how you know that one methodology is better than the other. Impossible I know.

      • Reflex - Aug 29, 2013 at 4:22 PM

        Maybe its because I can read their published methodology, I can see the data they are working with, if I think it is flawed, I can adjust for that flaw, and I can produce consistent, repeatable results.

        Furthermore, the alternative you seem to be suggesting is a mythical ‘trained eye’, which has no ability to see every play, no ability to see every player, and no statement of who this mythical observer actually is given that you throw away coaches and managers(an presumably understand scouts see only a small sample as well). Such a person would be subject to personal biases, uncorrectable since their methodology is internal, and has a very very long track record of getting it wrong as a result.

        Am I misrepresenting your ‘trained eye’ comment? If so, who is this individual you speak of? Tell me how to properly evaluate defense in a open, examinable, debatable and unbiased fashion?

        This may be because I am a software engineer, but I have built a career on quantifying things that observers generally feel are unquantifiable, and then developing metrics that permit one to measure both overall improvement and relative improvement(and regression). I do not see anything in baseball, and I’ve watched it actively since I was 7 years old, that makes me think that defense is some incredibly mystical thing that makes such metrics impossible, and while I do NOT believe any of the existing metrics is the end-all/be-all *yet*, I think they are headed in the right direction and already provide usable information that is better than the subjectivity that came before.

        I would like to hear your specific criticisms as to why that is not the case, if you have any. Such criticisms are how you improve a metric, and when I design a process in my job I invite such analysis and criticism in my work.

      • spudchukar - Aug 29, 2013 at 4:53 PM

        I am not a software engineer. I am a baseball fan. I played for 3 years in the Cards minors, many semi-pro seasons after that. I have 3 degrees, much of it in Neurology and Cognitive Science, with minors in Biology and Physiology.

        Plus, I watch baseball intensely. I care little for periphery baseball obfuscations, all I care about is the game itself, and how it is played.

        I have seen no evidence whatsoever that UZR and other defensive metrics have improved. A very knowledgeable eye, like mine, destroys the current defensive advanced statistics.

        You create a straw man, when you use GG awards, as examples of “trained eye” failures. They have been criticized for decades as a combination of hype and biased glorifications.

        What you fail to recognize is there will never be a way to compare advanced defensive metrics, without using the same subjective measurements you hope to avoid. Who would be the arbiter?

      • Reflex - Aug 29, 2013 at 5:18 PM

        I am not a software engineer. I am a baseball fan. I played for 3 years in the Cards minors, many semi-pro seasons after that.
        I find this very impressive and it certainly enhances your credentials from the perspective of playing the game. It does not necessarily make you a great talent evaluator, however. I have hundreds of millions of users of my software(not an exaggeration), many of them using it on a daily basis, professionally and even in administrative functions. However despite that fact the vast majority of them are not qualified to make technical statements about how we design, develop and QA the software we deliver. Their observations, while useful in planning features and user interface, are not useful in quantifying many, if not most other metrics in the development process.

        I have 3 degrees, much of it in Neurology and Cognitive Science, with minors in Biology and Physiology.
        Again, I find this impressive. I also know to get those degrees you had to go through some statistics classes and understand more than a bit about case design and quantifying the subjective. Those are not new concepts. My significant other has several science degrees as well, including in Zoology, Agriculture & Natural Resources and Natural Resources, and is working on a PhD in Human Dimensions. Most of her statistical work is in collecting human biased observations and finding ways to adjust perceptions to reality. It all looks very fuzzy and grey area, but her team has several papers on the topic and can consistently predict human reactions to policy questions with merely mathematical models. In other words, they can quantify the unquantifiable.

        Plus, I watch baseball intensely. I care little for periphery baseball obfuscations, all I care about is the game itself, and how it is played.
        I agree with you here, which for me is part of why I like to understand its component parts in detail, and how to evaluate them. I understand that most fans do not or that they admire other aspects, and I’m fine with that.

        I have seen no evidence whatsoever that UZR and other defensive metrics have improved. A very knowledgeable eye, like mine, destroys the current defensive advanced statistics.
        I’d really like specific examples of where it fails, since the methodology is open you should be able to provide this, especially given your background. Furthermore, YOU could contribute, the people working on these things are using people like you to better understand what they are seeing, you can help make it better, or you can take potshots from the sidelines.

        You create a straw man, when you use GG awards, as examples of “trained eye” failures. They have been criticized for decades as a combination of hype and biased glorifications.
        The problem with this reasoning is that if those who vote on GG awards are not the ‘trained eye’ you speak of, being as its coaches and managers, the very people who train the people being evaluated defensively, then who the hell is the trained eye? You seem to be speaking out of both sides of your mouth when you say “trained eye”, you want the authorities making the evaluations, except for all those authorities who make evaluations that you don’t agree with. Again, *who* is the trained eye? Is it just you personally? Are you a defensive evaluator genius?

        What you fail to recognize is there will never be a way to compare advanced defensive metrics, without using the same subjective measurements you hope to avoid. Who would be the arbiter?
        All metrics are on some level subjective. That is true of literally every single thing measured in the history of ever. It is not a disqualification of the approach. In baseball, who decided that going over the wall is a home run? Given the differences in fence heights and distances isn’t a home run an arbitrary stat? Why do we count it the same in Cincy as PetCo? At some point decisions are made on what is important, and then statistics are gathered around those decisions.

        There are a number of things that can be measured repeatedly and reliably in baseball defensively. Range to the right, range to the left, range back, range forward, ball handling, transitioning to the throw, speed and accuracy of throws, etc etc etc. In fact the Jeter article I mentioned above specifically contrasts Jeter with Brendon Ryan and the results are striking, obvious, and most importantly, quantified. At the *very least* it permits measuring players against each other, and at best as data collection improves it permits measuring players against players of what will eventually become the past(defensive data using this methodology will likely never be older than the 90’s).

        Its a huge improvement. And it was not simply UZR, which measures only a few things, some well, some questionably. And yes, I’d like to see you go read that article and tell me why its wrong, and if you want bonus points, explain how to improve it. A trained eye that cannot explain what they are seeing may as well not exist beyond the owner of such eye. If you cannot communicate your knowledge, well, its kind of pointless to the rest of us.

  6. baseballici0us - Aug 27, 2013 at 10:16 AM

    I agree with Crag’s view of the win.

    However, on a slightly different note, Quality Starts is by no means a perfect stat either. I have serious problem with a pitcher going 6 innings while allowing 8 hits, 3 earned runs, and having 3 k’s resulting in a QS; yet someone goes 9 innings, allows 4 runs on 6 hits and strikes out 9 and does not receive a QS.

    If you want to go by numbers, that’s a 4.5 ERA for 6 innings vs a 3.9 era for 9 innings. Additionally, and for some teams this is more valuable than a QS can ever be, they give the bullpen the night off.

    • Jeremy T - Aug 27, 2013 at 11:06 AM

      Quality starts were a step in the right direction, because they ignore what’s happening when the pitcher isn’t on the field. That doesn’t mean the stat doesn’t have flaws, and that’s why people continued to look for more comprehensive stats. Still, as a quick, easy-to-calculate, generally accurate stat, the QS does a pretty good job.

      • danaking - Aug 27, 2013 at 12:35 PM

        The Quality Start was intended to measure how often the pitcher gave his team a reasonable chance to win through 6 innings. So long as it is accepted as that, it’s fine.

        Even using the worse example of QS (6 IP/3 ER), which would you rather have: a pitcher with 10 quality starts and a 4.50 ERA, or a pitcher with five complete game shutouts and 5 outings where he goes 3 innings and allows 6 runs in each game? Both pitchers have ERA of 4.50.

      • baseballici0us - Aug 27, 2013 at 1:19 PM

        They’re indeed a step in the right direction but it needs to be fixed/altered.

        Danaking….I would take the second guy…5 complete game SO are a guaranteed 5 wins for the team. The first guy may have 10 QS but it doesn’t mean the team won any of those 10 games.

    • philliesblow - Aug 27, 2013 at 12:16 PM

      I’d like the quality start stat a lot more if it was 7 innings and 3 runs or less. A team can win a lot of games if the starter can eliminate the need for a middle reliever. Also a 3.86 ERA is a whole lot better than 4.50.

      • Jeremy T - Aug 27, 2013 at 12:57 PM

        I don’t think the 4.50 thing is quite right, though. 4.50 is bad because nobody in real life pitches the exact same game every time out, so you’re probably going to have a few terrible games and a few pretty good ones. After a couple minutes research, it looks like the 4.50 thing was apparently debunked all the way back in 1992: http://207.56.97.150/articles/qstart.htm

        Again, I’m not saying it’s the best way to analyze a pitcher. I’m saying it’s OK, and much better than the Pitcher Win.

      • baseballici0us - Aug 27, 2013 at 1:14 PM

        That’s the entire opposite of what I said, Phills. I’m saying 9 innings and 4 er can still be considered a QS. Hell, 5 innings with no er could also be a QS.

      • 18thstreet - Aug 27, 2013 at 1:26 PM

        I understand the complaint, but the stat is quality start, not HIGH quality start. Thus I’m okay with six innings, three earned runs. I’m not about to do the research, but if the QS stat covers roughly one-third of all starts, then I think they’ve drawn the line correctly. A high-quality start would cover the best 10 percent of starts.

        Okay nerds: prove me wrong!

  7. pappageorgio - Aug 27, 2013 at 10:17 AM

    Agreed, what we need to do is stop counting runs at the end of the game.

    Winning and losing should be based on some complex statistic that measures all the sabermetric numbers that the stat-heads think are the most important and calculated at the end of the game. This way, 90% of the fans (or the players for that matter) at the stadium will have no idea what’s going on until the results are announced at the end of the game and everyone will have to stay until the end to see who wins. They will also make additional profit from the sale of scientific calculators at vending areas.

    Or….maybe….we could add some sort of sabermetric wild-card game to the playoffs (yes…..the Angels are in the playoffs until Trout retires!!!!!)

    No where is my Captain Crunch decoder wheel?

    • youknowwhatsgoodforshoulderpain - Aug 27, 2013 at 10:19 AM

      Don’t be foolish. Wins are a team stat, not an individual stat. Nobody is saying wins aren’t important to the TEAM, just not terribly valuable in evaluating a pitcher.

      • pappageorgio - Aug 27, 2013 at 10:37 AM

        Should we not count QB completions/TDs/Ints in football? Because other factors are included…talent of the WR (who could just be awesome or suck), or the CB covering (maybe a QB is in a division of great CBs), or maybe there is an awesome RB on the team.

        Should we not include goals in soccer or hockey because of differences in the quality of the goaltenders they face? Quality shots are stopped all of the time. How do you account for the quality of a guys teammates on that line (which help him get goals). How do you account for that, so many variables?

        I’d make a basketball analogy….but that game sucks, so who cares.

        The point I’m trying to make is that there are hundreds of variables that go into almost any action or statistic on a sports field. The statheads trying to use that argument to get a particular stat thrown away because it doesn’t fit in to their particular sabermetric argument….it’s just dumb. Every statistic will be influenced by another player(s).

      • Jeremy T - Aug 27, 2013 at 11:10 AM

        That’s true, context affects everything, but there are certain stats it affects more than others. One of the nice things about baseball, from an analytical point of view, is that it’s broken down into a series of individual events between individual players. The pitcher win completely ignores that little gift.

      • Liam - Aug 27, 2013 at 11:28 AM

        Every stat has some context to it, but that doesn’t mean we should throw up our hands and just use the broadest, most context dependent stat possible. Even if context free stats are impossible, we should still be striving to reduce it and focus more on what the individual himself did. “Killing the Win” is a step in that direction.

      • pappageorgio - Aug 27, 2013 at 1:09 PM

        I just threw up in my mouth a little.

        You just made the sabermetric argument I argued against. All stats are viewed in context…..but we don’t like that stat, so we’re going to say it relies TOO much on context.

        Yeah…OK.

        If you don’t like wins, fine….assign whatever meaning to them you like. I really do tire of all the saber people telling everyone they don’t understand BB since bill james wrote his book (and really intensified since Moneyball came out).

        It’s a game, enjoy it. I look at wins and say “hey… M Sherzer(sp?) is pitching pretty well this year” not “this guy is the best in the game”…..because obviously his team must be playing somewhat well behind him.

        “kill” the win stat? S**U

      • Jeremy T - Aug 27, 2013 at 1:29 PM

        If you really do look at the win that way, then you’re probably not the problem that Kenny is trying to fix. The problem isn’t so much that the pitcher win is counted, but the significance given to it by writers and fans. “Kill the win” is hyperbole, but sometimes hyperbole is a good way to make a smaller point more clear.

    • grumpyoleman - Aug 27, 2013 at 4:26 PM

      Lets just play the games on some game system and then we can talk over our headsets about how cool it was to see Nync back together again while it plays out.

  8. youknowwhatsgoodforshoulderpain - Aug 27, 2013 at 10:17 AM

    They may not be “utterly meaningless”, but wins are pretty darn close to it. It’s not a low-level stat. It’s a compilation of performance of offense, defense, and the pitcher’s performance. It’s difficult to ascertain exactly what role the pitcher had without lower-level #s.

    I place more value on opp BA, ERA, K/9, and WHIP than silly wins. I’m a big fan of strikeout pitchers though too, and I know a lot of guys that prefer long-distance guys that pitch to contact and rely on their defense to get out of innings quickly. But that can always go south on you. I’d prefer the hitter to not make contact and remove all doubt.

    • spursareold - Aug 27, 2013 at 10:29 AM

      Crash Davis: Relax, all right? Don’t try to strike everybody out. Strikeouts are boring! Besides that, they’re fascist. Throw some ground balls – it’s more democratic.

  9. hk62 - Aug 27, 2013 at 10:53 AM

    I don’t really have an argument on the way wins are allocated – losses is another story though. I think that if a starter leaves the game with his team down say 3-1 and they eventually lose 5-4 having never taken the lead or tied the game, the starter shouldn’t get the loss. He left after surrendering only three runs and his team scored more than he gave up – the reliever that surrendered the run #5 should get the L. Would “grade” starter and reliever performance better that way.

    Back to the importance of wins – they are kinda the whole reason to play right – so they are an important stat. Just not the defining stat for a starting pitcher (Nolan Ryan being the poster child for this line of thinking folks). The references above to awarding wins based on whether a team eventually won the starter’s start (words are wierd sometimes) is something every club has done with their internal stats for ever – but they do not “grade” the pitcher’s performance – that takes combination of multiple inputs – some traditional and some “advanced” in nature.

    Is the idea to try to have everyone (teams, players, reporters (including bloggers), and fans) on the same page regarding who is the “best” and how individual pitchers “rank” within MLB? Because that isn’t going to happen – ever. If its for selecting the CY or justifying CY votes, which I believe (I know it is) is still people casting ballots by rank – why is there so much angst about the debate? Seems to me that some folks are trying too hard to get more people to agree with them. If you don’t want opinion to effect things like the CY, build a model – get everyone to agree to it – and eliminate the human factor. (I am not for that by the way).
    I love the “silly” wins comment youknowwhat-

  10. ryanrockzzz - Aug 27, 2013 at 10:56 AM

    As a young person who has some old sentiment, I would not want eliminate the win as a way of looking at a pitcher. I actually agree with Craig on all points today. To me, a win has developed more into something fun, a broad way to look at a player for the casual fan/up and coming baseball fan. Baseball unlike any other sport, is filled with stat heads who want to bang the gavel and take a 10 layered approach to everything, which is fine. But not everyone wants to do that. Not everyone wants to break down a crazy game into formulas, sometimes they just want someone to say “Hey- look at this, this and this. They are easy to understand, and can start to show you how well someone plays.”
    Usually most pitchers at the top of their game get 10-15 wins in a year, so it does have some validity to the casual fan. Since 90% of baseball fans aren’t going to look at ERA, K/9, WHIP, the W probably won’t go away anytime, but like Craig said the value you of how every baseball fan looks at it has to change since we do have better metrics. I think it has it’s place in the overarching sense. Also, I just have a problem teaching my 5 year old about WHIP or WAR- he is much more into win’s and Spiderman.

    • km9000 - Aug 27, 2013 at 12:19 PM

      I think any kid who roots for a team with a lousy offense has realized how inherently unfair pitcher wins and losses are.

      I just wish Kenny would do a better job of conveying sabermetrics to the lay audience. Reynolds can bark back with a simple “No, it’s not!” or “I know who’s good by using my eyes!” and Kenny will just grin or something. Or MLBN could just bring in more like-minded personalities.

  11. jdanknich - Aug 27, 2013 at 11:00 AM

    My thoughts on this can be summed up in one word: Whatever.

  12. dowhatifeellike - Aug 27, 2013 at 11:40 AM

    The simplest way to reduce the importance of Pitcher Wins is to stop listing their W-L record first when showing their stats. Almost anything else would be better: ERA, K/BB ratio, WHIP, etc.

    Personally, I’m much more interested in the team’s record on the days that a pitcher starts. If a guy is on a crap team but they’re 17-10 when he starts, he must be doing something right.

  13. km9000 - Aug 27, 2013 at 12:14 PM

    If they had never assigned wins and losses to pitchers before, and they started now, there’d be a riot among the traditional baseball people.

    “So you can give up 10 runs in 5 innings, but if your team scores 11, you get a ‘win’? What a joke!”

    “So if you’re a reliever in the top of 9th, and you give up 5 runs, and your team makes a comeback in the bottom of the 9th, you get a ‘win’?”

    “So if you pitch 8 shutout innings, and your team scores 1 run in the 1st, you get a ‘win’, but not if they don’t score until the 9th?”

  14. ctony1216 - Aug 27, 2013 at 1:31 PM

    In the old days, starters finished games too. So it made sense to credit the pitcher with a win or loss. He pitched better than his opponent.

    Now, starters rarely finish, and so it’s harder for a starter to get a W. So when a starter does get 15-20 wins in a season nowadays, it means he’s consistently out-pitching his opponent and putting his team in a position to win. And that means something.

  15. sjhaack - Aug 27, 2013 at 1:42 PM

    Just so we’re on the level here Craig, you just made an argument for keeping a number due to nostalgia of having it around when you grew up. I don’t entirely disagree with you, but that’s where we are with this stance.

  16. coloradogolfcoupons - Aug 27, 2013 at 2:29 PM

    Bad start=less than 5 innings, more than 3 runs
    Fair start=5-6 innings, more than 3 runs
    Good start=6 innings, 3 runs or less
    Very good start=7 innings or more, 3 runs or less
    Excellent start=7 or more innings, 2 runs or less
    Great start=8 or more innings, 2 runs or less
    Fantastic start=9 innings, 2 runs or less
    Awesome start=9 innings, 1 run
    Unbelievable start=9 innings, 0 runs

    Add the modifier “Fu*king” before the name of any start that allows fewer runs than the max for that category: i.e. 7 innings of 1 run ball would be a “fu*king excellent start”: 8 shutout innings would be a “fu*king great start”

    I realize this needs a little work, so feel free to chime in with recommendations. Maybe we could come up with a pitchers “Blue Book”, the ultimate referral guide to a pitcher’s stats on a game by game basis

  17. grumpyoleman - Aug 27, 2013 at 4:09 PM

    Let’s just give them all trophies and get thuis over with.

    • km9000 - Aug 27, 2013 at 6:56 PM

      Giving a pitcher a win despite having a crappy game (but strong run support) seems like the closest MLB comes to participatory trophies.

      • grumpyoleman - Aug 28, 2013 at 10:24 AM

        I love love those trophies you get for getting a win.

  18. bender4700 - Aug 27, 2013 at 6:31 PM

    I’m more in favor of ignoring media members who try to alter sports to fit their wishes. The NY media member from the other day who sited himself as a source for claiming Rex Ryan was going to be fired.

    The “media” is ridiculous anymore. Blogs listed as sources. “Reporters” listing their own articles as sources. Out of hand.

  19. bender4700 - Aug 27, 2013 at 8:36 PM

    A lot of people keep talking about Scherzer’s win loss, saying look at the run support. Despite his ERA, K’s, and WHIP being so good.

    I don’t think it’s nearly the same as RBI’s but it’s not meaningless.

    I’d have a better time with this debate if it wasn’t just an example of the media being the media.

  20. grumpyoleman - Aug 27, 2013 at 10:13 PM

    They subjectively quantify every play because every play is different no matter how similar they look

  21. ndrick731 - Aug 28, 2013 at 1:31 AM

    Gee and I thought baseball was about winning games. I guess we should start using that sabermetrics bs to decide who is in first place. Well I take the 20 game winner over the 12 game winner with the better made up stats every year.

    • Jeremy T - Aug 28, 2013 at 10:21 AM

      This argument keeps getting made, and I can’t figure out if people just don’t read the responses, or if different people keep on coming up with the same argument.

      Baseball is about winning games. That is undeniable. It takes an entire team to win a game. I’m pretty sure most people would agree with that one, too. So why do we call something a “win” and give it to one person?

      I don’t understand what you mean by “made up stats”, either. I mean, if you’re saying hits or strikeouts or something actually happened, and everything else is made up, then sure. But the stat that we call a “win” is just as made up as quite a few of the newer stats.

Leave Comment

You must be logged in to leave a comment. Not a member? Register now!

Featured video

A managerial overanalysis epidemic
Top 10 MLB Player Searches
  1. M. Bumgarner (3089)
  2. J. Shields (2902)
  3. T. Lincecum (2691)
  4. T. Ishikawa (2381)
  5. M. Morse (2362)
  1. Y. Cespedes (2003)
  2. L. Cain (1644)
  3. B. Roberts (1637)
  4. B. Posey (1490)
  5. U. Jimenez (1473)