On Pitching Wins, ERA and WHIP

I don’t think pitching wins should count for anything. I’ve expressed my feelings toward this counting statistic before, but maybe I haven’t gone over exactly why I believe it to be useless.

I think most people are already accepting of the fact that a win and loss record for relievers is meaningless because, as we saw last night in the Blue Jays and Athletics extra inning game, a pitcher like Jason Frasor can supply the worst pitching of a string of relievers and still collect the victory, based solely on what the offense does the next inning. So, let’s focus on wins and losses for starting pitching.

In order to win a baseball game, you need to score more runs than your opponents. This means that the responsibility for any game is broken up between offense and defense. That’s why there are two halves of each inning. Already, we’ve got 50% of the entire game for which a pitcher, at least in the American League, plays absolutely no role in, and in the National League, he normally holds the least amount of responsibility in the entire lineup.

From there we look at defense, where the pitcher does play a role, but even within that 50% of responsibility, the fielders behind him play a role too. How much of one has been estimated before, but I can’t say that I completely understand how those estimates work. However, I would say that fielding plays a significant role in defense, especially considering that modern baseball analysis has come to understand that, for the most part, pitchers don’t control where hits end up going.

And then even from the leftovers of that, once a starting pitcher exits the game, he has to trust in the bullpen to maintain his win. It’s something that he has absolutely no control over whatsoever.

So, how do we judge a pitcher’s worth? For years, we’ve looked at earned run average (which is simply earned runs multiplied by nine and then divided by innings pitched) and walks and hits per innings pitched (which is fairly self explanatory). Unfortunately, these two widely used statistics also account for things that aren’t under a pitcher’s control, like his team’s defense, the park he’s pitching in, and just dumb luck.

That’s why those interested in a clearer way of looking at just the pitcher’s performance prefer fielding independent pitching (FIP) .

Our own Drew Fairservice also tackled the usefulness of FIP and xFIP at the end of last season.

I bring this up because on yesterday’s post game radio show on the FAN 590, one of my favourite baseball guys, Mike Wilner, attempted to justify his claim that Jesse Litsch was one of the best pitchers in the American League in 2008 by suggesting:

You can’t throw away stuff like WHIP and ERA because they actually tell the story of what happened.  They don’t necessarily tell you why it happened or whether it’s likely to happen again, but they tell you what happened, and that’s significant.

To that effect, so do pitching wins. Wins and losses tell you what happened in the game in terms of how they’re attributed to starting pitchers. ERA and WHIP aren’t as limited as pitching wins, but they still give far too much weight to other factors beyond a pitcher’s control. They’re simplistic and somewhat arbitrary ways of judging a pitcher’s worth, and not a very good indicator of a pitcher being good or bad, or at least, not as good as FIP.

If Wilner suggested that Litsch allowed some of the fewest earned runs or walks and hits in the American League, then by all means, use ERA and WHIP to justify that statement. It wouldn’t mean much because it doesn’t take the differences in fielding or official scorers into account, but it wouldn’t matter all that much anyway because he didn’t say that. He said that he was one of the best.

Interestingly enough, you can make an argument that Wilner is correct in his assessment looking only at Litsch’s FIP for 2008. He ranks 21st in the American League in FIP among pitchers who threw 170 innings, but also behind teammates Roy Halladay and A.J. Burnett. It can be somewhat tricky claiming that a pitcher is one of the best in the entire league when it doesn’t even seem as though he’s one of the best on his own team.

In all of the other important peripherals Litsch ranks as average or slightly above, but he certainly didn’t put up the kind of numbers that would indicate that he’s one of the best. Still, based on FIP, you could make an argument.

However, if you really want to disagree with Wilner on his point about Litsch, a better response would have been, “So what?”

You know who else was one of the best pitchers in the American League in 2008? Ervin Santana, Gil Meche, Scott Baker, Andy Sonnanstine, Nick Blackburn and Jeremy Guthrie all rank in the top 25 for FIP in 2008 among American League pitchers with 170 innings. Since then, Litsch has had both Tommy John and hip surgery, and has pitched horribly in limited action across different levels.

I just don’t understand how bringing up Litsch’s success, whether warranted or not, in 2008 means anything as to his ability to contribute to this team as the fifth starter in 2011. I suppose we’ll find out to some degree tonight when he makes his first start of the season against the Oakland A’s.

Prove me wrong, Ginger Beard Man.

Comments (46)

  1. He… he can’t be worse than Jo-Jo Reyes… can he?

  2. Comparing ERA to pitcher wins in usefulness is like comparing playing in the NFL to slavery.

  3. If you could choose what to do with a pitchers Win-Loss, what would you do? Just eliminate it completely from pitchers stats and only have a Win-Loss attributed to the team?

    If so, how would use Saves/Holds? Or as well eliminate them and not give these stats to pitchers?

  4. 2008 shows that he is capable of putting up above average production in the majors (particularly when you consider the competition he faced, and FIP doesn’t). It’s not insignificant.

    Whether he can get back to that level is anybody’s guess.

  5. If best case for Litch is he regularly pitches 6-7 innings and only gives up only 4-5 runs then thats fine by me for a 5th starter. Somehow I have my doubts he can do that.

  6. “I just don’t understand how bringing up Litsch’s success, whether warranted or not, in 2008 means anything as to his ability to contribute to this team as the fifth starter in 2011.”

    So you are saying, “I don’t understand how past success has any barring on predicting future success”?????

    If you’re point is that, “it’s 2008″ – you have to realize he hasn’t been healthy since 2008…

    That being said I have no idea what your point is..

    I feel people give Litsch shit, because he is an overweight, ginger, who was drafted in the 24th round… I know you’re normally not that superficial, so please explain…

  7. I don’t think there’s any reason to count wins, saves and holds. I like the idea of looking at relievers through the shutdown / meltdown back story that Drew has been posting about that FanGraphs are using. But I think for predictive stats, looking at FIP, command rates, pitch selection / success, are the best things whether it’s a starter or reliever.

  8. My point is that having an above average year in 2008 doesn’t indicate anything for 2011.

  9. I miss McGowan. Wah wah. Could you imagine Romero, Cecil, Morrow, Drabek and McGowan? Dirty.

    I’m not gonna lie, I don’t have a lot of faith in Litsch getting much past the 5th inning on any sort of regular basis, at least for the first few months. Looking forward to him proving me wrong, hopefully.

  10. IMHO, Litsch is the poster boy for the Jays, as kudos for the organization’s uncanny ability to teach any ML average pitcher how to pitch successfully in the tough ALE.

    Just hope that he can pitch this yr like in 2008, such that AA can dangle him in trade during the off season and get a Lawrie-like prospect back. It is just a matter of time that the heavy hitters in the ALE figure out his changeup

  11. “My point is that having an above average year in 2008 doesn’t indicate anything for 2011.”

    It indicated he’s capable..

  12. Four years and two major surgeries ago. Look at the other “capable” pitchers.

  13. You want a list of pitchers who have been successful after Tommy John?

    And as Litsch has said in interviews, the hip surgery wasn’t major, it was just fixing a small tear..

    It’s not like he blew out his shoulder or something..

  14. Since I just posted this in another story about a minute before this story went live, I’m just copying and pasting

    I was thinking about W-L record vs. ERA vs. FIP as statistics go, as well as Avg/RBI vs. OBP vs. OPS, etc. I think I’ve decided they are all (including W-L) important in different ways. They are all really layers. Think as a 5 year old for a second. Hey, is this pitcher good? Well he’s got a winning record, so yeah. Then as you get older you discover the easier to understand stats. Well, the guy is 5-5, but his ERA (simple math of runs divided by 9 = how many runs a picther would give up for a whole game) is 3.12 which is pretty good. Same thing with avg. He gets a hit 3 out of 10 times, not bad.

    Then as you age and discover the deeper layers (and this can take years of baseball understanding… a new fan will not get new statistics) you can start to understand stats like FIP (which I couldn’t explain to you if I tried), BABIP, OPS, WAR, etc. These are all different layers in understanding a player’s worth.

    Now, and this is important, in this understanding, there is no excuse for anybody who has been around the game for a long time, especially a GM, relying on only the first couple level of stats.

  15. Does anyone actually remotely believe he’s “one of the best”? I think the argument is, if you need a #4 or #5 starter on your team, is he one of the best to have? If, based on 2008 numbers, the answer would be yes. But you’re right – using a single year as a reference point isn’t a solid example. However, that second last paragraph really just falls flat.

    On another note, xFIP > FIP, personal opinion there.

  16. Litsch had a 4.29 FIP in 2008 which is far from very good and when you consider what 21st in a 14-team league means; it means he’s an average 2 or 3 starter; not one of the best.

    A couple other things to consider about Litsch was his somewhat low .278 BABIP in 2008 which for a contact pitcher such as him could really balloon his numbers if it were to go up.

    And he also threw his cutter much more often in 2008 at 43.4% of the time and also his slider at 11.4% of the time; compared to 18.8% and 5.6% of the time respectively in 2010. Those two pitches were by far his most effective in the ’08 season and he obviously has not had the feel for them since otherwise he’d be throwing them.

    If he doesn’t get back to using those pitches I don’t believe there’s much hope for him even getting back to his modest, unspectacular 2008-self

  17. Isn’t anyone going to give Parkes props for questionning Wilner’s assertion? How many times have he and Stoeten been accused of parroting Wilner on the DJF blog (the accusation is usually voiced much more crudely than I just did) and now that Parkes takes on Wilner, all he gets is grief?

    Wilner is a smart guy but even smart guys can be called out for saying dumb things. Much like this blog, Wilner does a lot to try to improve the level of baseball discourse in this country (or at the very least he does a lot to discourage idiotic discourse). On the other hand, he will say occasionally something that pretty grossly homerish.

  18. Hey, I was prainsing and providing reasoning for my praise.

  19. Jesse Litsch pitches 2 years in the American League east with an ERA under 4! That’s saying something. You saying “I just don’t understand how bringing up Litsch’s success, whether warranted or not, in 2008 means anything as to his ability to contribute to this team as the fifth starter in 2011″ is non-sensical and idiotic. Really? Past performance has NO effect on how good you are, or future performance? Alex Ovechkin’s past success says nothing about how he will perform in the future? Just a stupid comment, I don’t even know how you could justify it.
    He’s proven that he can pitch. I’m frustrated by the fact that there’s even a debate over who should be the 5th starter between him and Jojo when Morrow comes back. In a horrible way, yesterday was the perfect scenario whereby Reyes had a bad game(increasing the chance that Litsch will be the 5th starter), and the Jays still won.

  20. Haha,
    Funny, but it wasn’t comparing Litsch to Ovechkin, I just don’t understand how you can say that howhe did in 2008, and 2007 for that matter, have no bearing on how he will perform now.
    An athlete’s past performance very often has some bearing on how good he is, and how he will perform in the future.
    Why do you think that his past numbers are irrelevant?

  21. He’s basically a fat, ginger version of Ovechkin…without the hockey skills, of course. You know who’d be a better fifth starter than both Litsch AND Reyes? Rzepczynski, damn it. Discuss.

  22. It was three years and two surgeries ago. And he wasn’t even that good then. I have no problems using a healthy Litsch as a fifth starter but imagining him as anything more than that has no basis in reality.

  23. If he can put up an era around 4 again, then he absolutely should be in the rotation.
    Your right, the surgeries put a question mark on whether or not he can repeat those numbers…but pitching 2 seasons as a starter in this division with an ERA under 4 gives Jesse some leeway in my books, and he deserves at least 5 starts to prove that he can repeat those numbers or at least be a quality starting pitcher. He looked good in spring training. I’m excited to see what he can do. I think if Jesse Litsch is our 5th starter, then we have one hell of a pitching staff.

  24. Peter, you’re going to get trounced for justifying his performance with ERA because it pretty well goes against everything stated on this blog.

    But yes, I think we can all settle this debate by acknowledging Travis’ assertion that Rzep would be a better 5th starter.

  25. I have a couple of small problems with FIP.

    (1) it includes HR’s. Why are they any less to do with luck than any other batted balls? park factors surely have a huge bearing, and whether a ball scrapes over the wall or falls short into a glove/off the wall has an element of luck.

    (2) it says that that balls in play are affected by defence, luck, park factors etc. All true. so it ignores it completely. Why not attempt to adjust for these factors? instead of judging a pitcher on what happens in with what, 25% of the batters he faces? seems odd.

    (3) it’s a model. some very smart people figured that these things are the right things to judge a pitcher on. but they aren’t necessarily right. they could well be, but i don’t know that and you don’t know that. so stating that “FIP is the better way to judge a pitcher and ERA is flawed” is perhaps being more sure of oneself than one out to be.

  26. “Litsch had a 4.29 FIP in 2008 which is far from very good and when you consider what 21st in a 14-team league means; it means he’s an average 2 or 3 starter; not one of the best.”

    Actually, that is very good when you consider what division he pitched in.

    I really don’t understand why everyone is so quick to criticize Litsch, he’s only been a quality pitcher for this team.

  27. @btc – nobody says FIP is better just because. The ideas of that pitchers could only control so much took hold and then, after much mathematical aerobics, they discovered the coeffieicents and components that best predicted future success. Year-over-year results for hundreds if not thousands of pitchers.

    FIP (or xFIP) is better because it has proven itself over a long trial. It isn’t 100% airtight but it isn’t pulled out of thin air witchcraft either.

  28. Dave Cameron at fangraphs generally states that FIP is better for analyzing how a pitcher performed in a year and xFIP to predict future success.

    What bothers me with this whole situation is that Wilner is constantly smug with people who make claims as to why a player is good or bad, but when someone calls him out he goes berserk and loses his cool. If you want to call people out, fine, I personally think it’s hilarious, but when someone reasonably argues against you for making an erroneous comment be a little more professional.

  29. I’m no SABRmatician, but by ignoring balls in play doesn’t FIP also ignore a guy who can consistently generate weakly hit balls?

    I can remember games when Halladay would cruise along, striking out 3 guys all game, but having 15 guys leave the field shaking their heads after grounding out weakly. I’m not saying Litsch is anywhere near Halladay, but when he was at his best a huge part of his game was throwing that cutter and getting guys to ground out.

  30. The reason I think xFIP is better is because it takes that which was criticized about the FIP formula and improves upon it, e.g. solving the home run problem because xFIP becomes universal due to park factors, etc.

  31. I think wins/losses wont be scrapped but it should be altered… If a starting pitcher has a quality start and his team wins the game. He should get the win, regardless of the score he left in. Either that or just call wins, quality starts. The way it is now is just so primitive.

  32. I agreee that Wilner is wrong that GBM is a top pitcher in baseball, but with the fastball back at 90, the movement on the pitches, and his ability to throw strikes, this guy is pretty good. At least as pretty good (if you throw out last year’s numbers) as Brett Cecil. It looks like whatever good numbers the little chubber dude throws up, someone has an unverifiable reason to say he’s crap. I thought this blog was all about the numbers — what we know as opposed to what we might believe.

  33. Keep in mind the fielding independent isn’t totally independent.

    A good fielding defense, keeps a pitcher in the game longer, and gives him more overall innings which drives down his FIP…

    It’s hard to say how much of an affect this has, but I’m sure Halladay’s FIP would be better with the Phillies this year, than if he we’re playing with the Brewers for example..

  34. Halladay’s FIP didn’t change much after moving to the Phillies from the Jays, but his ERA did decrease. FIP really is fielder independent. You can look it up yourself on Fangraphs. That’s the awesome thing about modern stats. You don’t always have to trust your sports gut to get into the game.

  35. Haha, it’s “get way into the game”

    But anyways, I think FIP is a more accurate way to show what a pitcher is worth but I agree that it sorta ignores the ability to induce weak contact. What is the response to that? It seems relevant to Listch’s skill set doesn’t it? Sure theres luck involved in BABIP but arent some pitchers are just more likely to have a lower BABIP by the way they pitch?

    I’m not willing to throw out ERA and WHIP because as Wilner said they are the story of what happened regardless of who got lucky… and if someone gets lucky over and over (not to say that litsch has..) then shouldn’t we question our assessment of them?

    Anyways I mostly agree… Listch is a 4th or 5th starter… if he gets back to 2008 form that would be wonderful.

    • one more thing for @RRE – very few pitchers are able to sustain BABIPs below the working average of .300. Balls find holes, whether you’re Roy Halladay or Casey Janssen. Unless a pitcher’s line drive rate gets out of hand, they’ll give up nearly the same amount of hits.

  36. @drew… when you say “best predicts future success”, that depends on how you measure success. i assume you measure it by FIP. this was my point about it being a mathematical model – they created a stat that remains stable year on year, by regarding only certain elements of their game. so your previous FIP is a good predictor of future FIP. this doesn’t necessarily mean that FIP (or FIPx) is the best way to judge a pitcher. It could well be, and i’m definitely not saying it’s a bad stat (i haven’t looked into it deeply enough), but there is an element of circular reasoning here that makes me wary

    • @btc – best predicts future success by runs allowed. Isn’t that what we’re all trying to predict, the number of runs we expect a pitcher to surrender?

      @RRE – two things: you can’t tell me “earned run average” is the story of what happened regardless of who got lucky. As soon as you include that proviso, you’re attempting to qualify things that can’t be qualified. If Rajai Davis doesn’t run down that ball in the gap last night, does it go as an error? Are those runs unearned? If a ball falls in front of a hard-charging Juan Rivera, is that an error?

      You can’t make allowances for some bad defense but not all bad defense. If a player get to a ball and boots it, is it really that much worse than not getting to a ball in the first place? If you want to use straight up runs allowed, go nuts. Earned runs? No thanks, I don’t care what Dave Perkins thinks constitutes a “reasonable effort.”

  37. @drew… but if (as according to the video) runs allowed fluctuate wildly, and FIP stays relatively constant, surely that’s true of the future performance as well as past performance?

    maybe the point is that one season FIP gives a better predictor of career runs allowed than one season of runs allowed does. i don’t know, i’m just playing devil’s advocate until i’m thoroughly convinced of FIP’s merits! today i received my copy of “The Book…”, maybe that will assuage my doubts!

  38. @IkeKrizzule
    Halladay’s career FIP is 3.42 and ERA 3.32. Even the definition of a weak ground balls has a FIP that seems to predict his ERA quite well. Also his BABIP is .292, very close to the average of .300.

    If you like ERA as a measure of what happened that is okay. But, (x)FIP is a better predictor of ERA in the future than ERA itself! So using past ERA to predict future ERA is foolish. Litsch had a “one of the best” ERAs in 2008. Fine, but his 2008 FIP indicates that he is unlikely to repeat that ERA in the future (let alone being 3 years and 2 surgeries away).

  39. Let’s compare apples to apples shall we?

    Wilner’s comments about Litsch were based on his assertion that as a FIFTH starter, Litsch is one of the best FIFTH starters in major league baseball.

    Not defending Wilner’s assertion. I mean, I hope Wilner is right on this. But Litsch’s value in the rotation is not becoming a top-3 starter but rather a back-end starter.

    • That’s not true AP. He’s said twice now that Litsch was one of the best pitchers in the American League in 2008. That’s what people are taking issue with.

  40. @MarkV… note though that FIP is (somewhat artificially) designed to be in the same range as ERA. the average FIP will always (or at least for the forseeable future) match the average ERA, because the factors involved were chosen to fit the data. as far as i know, FIP= (13*HR+3*BB-2*K)/IP + 3.10. why these numbers? because they give the answer we’re looking for.

    The fact though that an individual’s FIP apparently reflects his career ERA does lend weight to it though.

    i wonder if there can be significant differences between pitchers’ slugging on balls in play? while you’d certainly expect the same ratio of grounders to get through for different pitchers, and the same ratio of line drives etc, a pitcher who gets a large percentage of grounders would surely be more successful than one who gets a large percentage of liners, even if their FIPs are identical?

  41. @btc Good point. I agree that there is something lacking in a model that has so many parameters that are not determined empirically but chosen only to have the model fit the data. That is why I personally prefer tERA. Although the differences between FIP and tERA are not that great.

    I think that it would be difficult to give up a lot of liners (out of ground balls, flies, and liners the line drive has the highest run value) and be successful (the range of LD% is quite tight (2009-2010) 18+- 4 versus GB% at 47 +- 19).

    I think a high line drive rate would show up in FIP because if the hitters are making great contact on pitches they are probably less likely to strike out and more likely to give up homers. High line drive pitchers probably don’t make it to the majors in the first place. Just guessing here though.

Leave a Reply

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