Pitching (almost) always wins championships

A bold couple of statements: The Giants should have been the favorites to win the 2010 World Series—Hint: they weren’t. And if you’re looking for a Cinderella story in 2010, the poor, inadequately-housed Oakland A’s would be a good place to start. I write, you decide…

Over at Bay City Ball, Chris Quick recently did something awesome. In a nutshell, he did a franchise scatter plot for the Giants (New York and San Francisco) using adjusted OPS (OPS+) and adjusted ERA (ERA+):

The idea is that we can split past Giants teams into four quadrants based on hitting and pitching. The vertical axis represents team pitching by adjusted ERA, or ERA+. The horizontal axis represents team offense by adjusted OPS, or OPS+. OPS+ and ERA+ are fairly simple concepts in baseball statistics these days – each is adjusted for league, park, and era (making them good tools to compare the 1905 Giants to the 1985 Giants). The league average is scaled to 100, meaning that if you score 110, you are 10% better than league average.

If you’re a Giants fan, I implore you to check it out—actually, I implore you regardless, as this is just one of many great graphs you’re likely to find on his page. As such, he was a great choice for Neyer’s SweetSpot.

But it was something buried in his comments on the graph that really got my neurons firing: “The old adage that, “Pitching wins championships” rings true for the Giants franchise.” It does.

In true Bill Jamesian fashion, this brings us to a question. Does pitching really win championships? This is something people have looked at quite a lot, and the results have pretty clearly shown this is the case. But using Chris’ method, we can test it once again.

I think you’ll be more shocked by the results than you might imagine.

I’ve pulled both the ERA+ and OPS+ for every modern World Series champion since 1903. There have been 106 winners: The Boston Americans won it all in 1903 and the San Francisco Giants in 2010. It’s also worth noting there wasn’t a series in 1904 because the New York Giants (John McGraw specifically) refused to play what they believed was a far too inferior team, the Boston Americans, who later became the Red Sox, of the junior league. Perhaps, this is one of the origins of the term “Junior Circuit.” The other omission is 1994, which ended in strike rather than a Fall Classic. Brutal.

The Yankees hold the record for World Series championships with 27, while the St. Louis Cardinals trail overall but lead the National League with 10. Other franchises with greater than five are the Oakland A’s—they’ve also played in Philadelphia and Kansas City—with nine, the Boston Red Sox—they were originally called the Americans—and the San Francisco Giants (formerly of New York), and the Los Angeles Dodgers (formerly of Brooklyn) are tied with six each.

I’ve labeled a few teams in the scatter plot that are particularly notable for one reason or another, but mostly because they are outliers.


What you should notice immediately is the plethora of dots above the red line which delineates an average pitching team (ERA+ of 100 or more) and a below average pitching team (ERA+ below 100). There have only been three teams in 106 chances who have won a World Series when their regular season ERA+ was less than 100. They are the 1987 Minnesota Twins, the 2006 St. Louis Cardinals and the 1913 Philadelphia A’s. A team of below-average pitching has only won it all 2.83 percent of the time, which I personally find to be mind-blowing.

More often than not, the team that wins it all is going to land in the top right portion of the graph, which means they’ll have both good pitching and good hitting. But, a team can be below average offensively and still win it all, as 33 of the 106 winners have proven (33.02% percent).

I think this really hammers home the point that pitching really does win championships; that you need at least an adequate stable of arms to have a prayer. Trying to win the Fall Classic without league-average pitching has proven to be about as fruitful as attempting to drive a car without gas. You’re not going to get very far. So, if you had to choose which is more important between offense and pitching, the answer is obvious: run prevention.

I looked at the data a lot of ways, so here are some others:

• Only 22 of 106 winners had better hitting than pitching (20.75 percent)
• Only eight of 40 winners had better hitting than pitching in the divisional era (20 percent)
• Only two of 16 winners had better hitting than pitching in the dead-ball era (12.50 percent)
• Since the offensive-centric Reds of the 1970s, aka The Big Red Machine, only five of 33 have had better hitting than pitching (15.15 percent)
• The average World Series winner had an OPS+ of 103.47 and a median of 104
• The average World Series winner had an ERA+ of 113.84 and a median of 113
• Thus, on average, the winner has an ERA+ of 10.37 more than its OPS+

A Hardball Times Update
Goodbye for now.

I also took the liberty of averaging each teams OPS+ and ERA+ to determine how much better than average they were overall. I then averaged all of those numbers, whereby I discovered the average World Series champion has graded out at 108.7, or about eight percent better than league average.

If you take the most recent champion, the San Francisco Giants, and add their OPS+ (95) and ERA+ (121) from 2010 together, you get a grade of 108. This goes to show their World Series title in 2010 shouldn’t be considered lucky or a fluke, rather, they were a perfectly average champion. What’s more, their excellent ERA+ fulfilled the pertinent requirement of at least adequate pitching. Looking at the data this way would make a lot of pundits feel silly on their pre-postseason picks, routinely expecting them to lose.

It’s also an awful lot of fun to take a look at those teams that stand out… like the 1907 Chicago Cubs. They own the biggest disparity between offense and pitching with a whopping delta of 52 (90 OPS+ and 144 ERA+). Their entire pitching staff probably required back surgery after the season.

That Cubs team was the World Series winner with the best pitching (ERA+ of 144) ever, and it’s really not even close. These guys could really throw the pill. They had four starters with 27 or more starts with ERAs under two. Anchoring the staff was 26-year old Orval Overall (great name, by the way) who threw 268.1 innings with a 1.68 ERA and a 23-7 record. He threw the most innings, anyway; his personal ERA+ was 149.

But he was accompanied by Mordecai Brown (20 wins, 1.39 ERA, 233 IP, 179 ERA+), Cal Lundgren (18 wins, 1.17 ERA, 207 IP, 213 ERA+), Jack Pfiester (14 wins, 1.15 ERA, 195 IP, 215 ERA+ and Ed Reulback (17 wins, 1.69 ERA, 192 IP, ERA+ 148).

The offense did just enough with an OPS+ of 92, and the pitchers were most certainly helped by Harry SteinfeldtJoe Tinker, Johnny Evers and Frank Chance, which, according to Bill James, may have been the very best defensive infield in history, immortalized in “Baseball’s Sad Lexicon,” aka “Tinker to Evers to Chance.”

These are the saddest of possible words:
“Tinker to Evers to Chance.”
Trio of bear cubs, and fleeter than birds,
Tinker and Evers and Chance.
Ruthlessly pricking our gonfalon bubble,
Making a Giant hit into a double –
Words that are heavy with nothing but trouble:
“Tinker to Evers to Chance.”

When you consider what James said (subscription), referring to the famous poem just this past week, there’s a good chance we should be giving that infield an awful lot of the credit:

He was putting in words what was in the air: that this was the greatest defensive infield that anybody had ever seen. And in fact it was.

I know that not everybody agrees with this. I know that there are other people who have looked at this issue and reached a different conclusion. I’ve looked at the issue myself at other times, not knowing as much as I do now, and reached a different conclusion. I’ll have to leave it up to you to weigh this analysis against the others. But it is my opinion that Joe Tinker was very possibly the greatest defensive shortstop in the history of baseball, and that he is a well-deserving Hall of Famer.

Another would be the 1910 Philadelphia A’s. The combination of their 114 OPS+ and 133 ERA+ (grading out at 123.5) makes them one of the best World Series champions of all time.

Everything for the 1910 A’s started with their brilliant, Hall of Fame second baseman Eddie Collins. He led the team that year with a .324 average, .382 on-base percentage (OBP) and 152 OPS+. He was also a spectacular fielder. Home Run Baker (another splendid name), Rube Oldring and Danny Murphy were excellent as well with OPS+’s of 126, 142 and 143.

On the pitching side, Jack Coombs led the league with 31 wins and 13 shutouts. He threw 353 innings with a fantastic ERA+ of 182. Also impressive were Cy Morgan (290.2 innings, 153 ERA+) and Chief Bender (250 innings, 150 ERA+) who helped them to capture the pennant and ultimately win their first World Series championship.

In addition to those 1910 A’s, the 1913 version of the A’s is worth a look for different reasons: They couldn’t really pitch. They hit well enough with an OPS+ of 116 but their 87 ERA+ was pretty dreadful.

Every regular posted an OPS+ of 100 or better with the exception of catcher Jack Lapp. Once again, Eddie Collins led the way, this time with an OPS+ of 163 and 10.4 wins above replacement (WAR). He was closely followed by Home Run Baker who led the league with 12 home runs and 117 RBI—this was the third of four straight years where the aptly named chap would lead the league in round-trippers. He also had a 167 OPS+ and finished with 9.4 WAR. He wasn’t as good defensively as Collins but led the offense.

But that’s not really why I’ve singled them out. I do so because they were the worst-pitching World Series champion in history. Eddie Plank, Bob Shawkey, and Chief Bender all pitched well with ERA+’s of 106, 118 and 125, but the performances of Boardwalk Brown, Bullet Joe Bush and Byron Houck brought them down terribly, with ERA+’s of 94, 72 and 67. Unfortunately, Coombs wasn’t around to help, as he caught typhoid fever in spring training and nearly died. Still, they managed to win it all without him.

Here’s a team you might recall: The 1927 Yankees. At this point, the dead-ball era is over and Babe Ruth is the ballast of the New York lineup. But, as we’ll soon learn, they had excellent pitching as well. To put their pitching into context, realize they had an ERA+ of 122 versus the 121 the 2010 Giants posted. Also, that 122 ERA+ was accompanied by Murderers’ Row (OPS+ 127).

Behind Ruth, they won 110 games, the AL pennant and swept the World Series. Among Ruth were Tony Lazzeri, Bob Meusel and another fine hitter: Lou Gehrig. Gehrig was just 24 but hit .373/.474/.765 for an OPS of 1.240 and OPS+ of 220. He also had a league-leading 175 RBI and 52 doubles to go with 47 home runs. His efforts earned him the MVP.

I probably don’t need to tell you this, but 175 RBI is an absurdly high mark. But it was a golden age for driving in runs, apparently, because every single player in the top 10 of the single-season leaderboard for RBI did it between 1921 and 1938. Gehrig appears in the top 10 three times, and his 175 RBI season is tied for fourth with Jimmie Foxx’s 1938 season. His 1930 season with 174 places sixth, his 1931 season with 184 places second. Hack Wilson’s 1930 takes the cake with 191.

Ruth managed an even better OPS+ of 225 with his 60 home runs and 164 RBI. There really aren’t words to describe a one-two punch like Gehrig and Ruth; it’d be enough to make any right-handed pitcher wet his pants. Even Pedro Martinez, probably.

If their offense wasn’t enough to handle, their staff was also fantastic. They had five starters make more than 20 starts each, and the worst of them, George Pipgras, had an ERA+ of 95. Waite Hoyt, Herb Pennock, Urban Shocker and Dutch Ruether each finished with above-average ERA+’s: 148, 130, 137, and 115. And, coming mostly out of the pen, Wilcy Moore was their best in 50 games with a 2.28 ERA, 19 wins, 213 innings and a 171 ERA+.

It was with both the ferocity that they hit and pitched that made the 1927 Yankees one of history’s greatest teams, if not the greatest. They were incredibly balanced, but it’s possible they’re not quite remembered that way considering their nickname: Murderers’ Row.

The St. Louis Cardinals of 1942 and 1944 were remarkable teams as well, led by their star outfielder Stan “The Man” Musial—he won the Triple Crown in 1943—but I’ll instead skip all the way to the “Amazin’s” and 1969.

The 1969 Mets were known as the “Miracle Mets.” They piqued my interest because, well, they are the classic example of a crummy-hitting team winning it all. They finished with a dreadful OPS+ of 84, worst all-time for World Series winners. But, they had an ERA+ of 122 that saved them. It was just as well, a magical year with Woodstock, landing on the moon and the Jets, Knicks and Mets all winning championships.

Second baseman Ken Boswell was about league average with a 103 OPS+ and center fielder Tommie Agee (122 OPS+) and Cleon Jones (151 OPS+) really saved them. Agee led the team with just 26 home runs and 76 RBI. Only three players hit even 10 home runs.

But, they had the pitching. Tom Seaver led the league with 25 wins and, behind 273.1 stellar innings, would eventually win the Cy Young award with his 165 ERA+. Not far behind was Jerry Koosman who had an ERA+ of 160 in 241 innings. Each of their starters had an ERA+ greater than 100 and Tug McGraw held down the bullpen with an ERA+ of 163 in 100.1 innings. Even 22-year old Nolan Ryan got in on the fun with 89.1 solid innings. They rode those arms all the way to their first World Series title, besting the heavily-favored Baltimore Orioles in four straight wins after dropping the opener.

The last team I want to address specifically is the 1976 Cincinnati Reds, aka The Big Red Machine. And they didn’t call them that without a reason.

Their pitching was merely adequate (100 ERA+), so their hitters carried the majority of the load with an OPS+ of 120. If you’re looking for balance and stability, this team had it. Every one of their starting position players played in at least 135 games and had over 500 plate appearances. What’s more, every last one of them was an above-average hitter with an OPS+ over 100. The standout among them was Joe Morgan, one of the best second baseman of all time. He doesn’t love sabermetrics, but they love him.

Morgan hit .320/.444/.576 and finished with an OPS+ of 186. He hit 30 doubles, 27 home runs, drove in 111 and stole 60 bases to boot. He led the team by a long shot with 9.9 WAR, which is no small feat amongst Pete Rose (6.5), Ken Griffey (5.5) and George Foster (5.0). Morgan’s season earned him a second straight NL MVP award and fourth straight Gold Glove.

The pitchers did just enough with Pat Zachry tossing 204 innings with an ERA+ of 128 and Rawly Eastwick coming out of the pen and finishing with an ERA+ of 168 in 107.2 fine innings.

The last thing I’ll not are the 1987 Twins and 2006 Cardinals. They appear on the lower left corner, signifying their ability to win a championship despite both below-average pitching and below-average offense. They help to explain the fact that any team making it to the tournament has a shot to win it all. The odds are against a flawed team, sure, but all you really need is a shot.

The fact that each of these below-average winners appeared relatively recently shouldn’t surprise us. Once divisional play began, a really great team could be knocked out in the League Championship Series (seven games), versus heading straight to the World Series. A smaller sample of games will always result in more upsets, which is a simple concept. What’s more, when the Wild Card team was added shortly after the Giants won 103 games in 1993, only to find themselves outside the tournament, such an occurrence became even more likely with the five-game Divisional Series.


Returning to our question: Does great pitching win championships? The answer to that question is a resounding yes. History has shown that it’s not only preferred, but very nearly an absolute necessity to have, at the very minimum, a league-average pitching staff. Beyond that, a quality offense helps, but pitching is king.

With this in mind, I think the 2011 version of the Yankees is once again in trouble, the Phillies have one heck of a shot at winning it all, and the Giants absolutely have to be considered a threat to return to the Fall Classic. Those teams that can’t throw a team ERA+ of 100 out there are probably better off packing it in then heading to October.

What’s more, the A’s, who had and OPS+ of 92 and ERA+ of 116 in 2010, are probably a whole lot closer to getting back to October than anyone thinks. They’ve returned a solid bullpen and enforced it to the point that it’s petrified, and added several offensive pieces that should push them closer to a league-average offense. With that young starting pitching, much like the Giants’, they are on their way.

You might have to squint a little to see it, but you might even say they’re better positioned than the powerful Yankees. Still, if the Red Sox can stay healthy I think they are probably 2011’s best bet. But that’s why they play the games, because the best bet doesn’t always win.

Stats throughout were pulled from baseball-reference

Newest Most Voted
Inline Feedbacks
View all comments
Tom B
13 years ago

I think the 2011 version of the Yankees is once again in trouble

You only mention the Phillies and Giants as WS favorites…  What AL teams are going to be facing them? 

My money is still on the Yankees, pitching or not (pitching that wont be nearly as bad as the naysayers want to believe).

It would be nice to see breakdowns by slot. Top3 starters, relievers… lumping the entire staff with a 4/5 guy that doesn’t or hardly contributes to a playoff run seems unfair.

Rory / Paapfly
13 years ago


Maybe I wasn’t clear, but I was trying to pinpoint teams that had elite pitching to win it all. The A’s could sneak in, I mentioned that as well. And I even slipped in the Red Sox as 2011s best bet.

You’re probably right about the Yankees’ pitching, it should be fine. With that offense, then can win. But I think they’d be better off moving some of that offensive talent for some value on the mound.

And of the pitching isn’t what it needs to be, that’s exactly what theyll do. Go get it

Tom B
13 years ago

Rory/Paapfly – Ahh i see that line about the Sox now, my brain must have glazed over towards the end there smile

Rory / Paapfly
13 years ago

Understandable, as that was 3,000 words on.

13 years ago

Really cool article.  On the surface it would seem that you could craft a team with a great offense and top heavy, but overall not good pitching and if you make the playoffs, hide the 4th and 5th starters and 2 or three relievers.  In reality, it would appear that either those teams don’t make the playoffs, they fail once they get their, or the “top heavy” part of the staff is good enough to compenstae and give the staff an ERA+ of 100 or better.

It would be really cool to see this expanded to all playoff teams.  It could be divided into the 4 team and 8 team playoff eras to see if good pitching is necessary to get you into the postseason too.

It would also be cool, but more difficult, to look at it the other way.  How many teams with good pitching didn’t win the championship.  Not sure how to present that data.

Rory / Paapfly
13 years ago

Mike, thanks.

I did exactly what you mentioned, but only did 2010 to 1997. As you might imagine, it’s a lot of work.

I was trying to see if there was a decent way to predict playoff winners. One way I tried was to weight pitching 2/3 and hitting 1/3. What I found is that this method was hit and miss. It’s generally a good thing to have good pitching, obviously more so than hitting. But, once you’re in a five or seven game series, the poorer teams often wins. Often enough, anyway.

Based on this data, it’s even more surprising to me than before that the Braves didn’t win at least two championships during their run. They had a great team in 1998, and in 2002 had unbelievable pitching (133 ERA+)

13 years ago

For the low OPS teams how were their runs scored compared to average?  Generally RS and RA corresponds to W/L and ERA lines up with RA.  But if OPS+ doesn’t correlate with RS+ for a team (e.g. they have a super sinkhole catcher or SS that exaggerates the badness of the rest of the lineup, and in the case of the ‘68 Tigers, gets replaced in the WS; or they steal a lot of bases successfully) that might skew a couple of these teams to the upper left that should be upper middle.

Eric R
13 years ago

Looking at that graph, there are about the same number of teams at:

<=95 ERA+ and <=85 OPS+
<=100 ERA+ and <=90 OPS+
<=105 ERA+ and <=95 OPS+
<=110 ERA+ and <=101 OPS+
<=115 ERA+ and <=105 OPS+

So, a 10 point gain in ERA+ seems to be worth about an 11 point gain in OPS+.

Dan Hennessey
13 years ago

Hey Rory,

Did you pull out pitchers for all of the NL teams?  Because 100 is the average OPS+ for all non-pitchers.

Also, I’d be interested in seeing the same graph for WS losers and all playoff teams.  I wonder if this is predictive just for winning the WS, or also for making the World Series/playoffs.

13 years ago

Thanks for the link, Rory! I’ve enjoyed reading you so far at THT.

Something I didn’t mention when I wrote up the post was ERA+ is based on ERA. And, as we know, ERA tends to encompass not only some parts of pitching, but a large part of team defense. So, I think it goes to say that excellent pitching and defense are a huge part of what defines winning teams.

Also, that 1907 Cubs staff is nuts.

13 years ago

Wow.  Kudos for not stringing this along into a mini-series of articles.  There’s certainly enough data to support multiple pieces; the fact that you braved through most of it on one post is amazing.

I gotta admit, I’m dying to see the remaining plot points identified.  I’d be curious to see if any other trends reveal themselves in concert w/ contemporary trends and/or certain designated eras of baseball…or even just the impact of the DH on the OPS+/ERA+ figures. 

A whole lotta possibilities to explore w/ this data.  Good stuff.

Rory / Paapfly
13 years ago

Dan, I did not. That presents problems and makes the graph misleading… Which would have been usefully YESTERDAY.

It’s something I can fix to see what conclusions can be drawn, but it’ll take time.

Tangotiger posted it and the masters of sabermetrics are doing a good job of showing me where I went wrong. It’s been a good learning experience, which is undoubtedly a good thing.

Chris, you’re right for sure. The defense is an integral part of the pitching, and there’s little doubt Tinker, Evers and Chance were a big reason for that 144 ERA+.

The AL winners 1973 to 2010 should be good, but NL teams after that and each team prior to 1972 would need some tinkering.

Steve Treder
13 years ago

“The AL winners 1973 to 2010 should be good, but NL teams after that and each team prior to 1972 would need some tinkering.”

Yeah, I was going to point that out.  Dang!

I understand the logic of changing the manner in which OPS+ is calculated on bb-ref.com, but this exercise provides a vivid illustration of how frustrating it is when a league-normalized stat doesn’t center on 100 at the league level.  They solved one problem but replaced it with another.

Dang! again.

13 years ago

I’d be curious to see the numbers with RS+ and RA+ instead of OPS+ and ERA+, to keep a consistent scale.  Just looking at the 2010 season, for example, the best hitting team was the Yankees, with 5.30 RS/G and a 109 OPS+.  Compare that to the Giants, with a 121 ERA+ and 3.60 RA/G.  The league average was 4.38 R/G.  Take any year at random, and in most cases, the upper range for ERA+ will be greater than the upper range for OPS+, even though the range of RS to RA tends to be quite similar.

It’s certainly an interesting approach to the question, and the graph is presented in a way that emphasizes the point nicely.  The question is whether the results hold true with an equal scale, though, or whether the findings are in large part a product of different baselines.

Steve West
13 years ago

The chart is a little misleading because of the spacing and axes.  Trying to compare 10 points of OPS+ to 10 points of ERA+, the distance should be the same in both directions.  As it is, the x axis is about twice as wide as the y axis is tall.  It makes it look like the 27 Yankees are as far away from 100 as the 07 Cubs.  In reality the Yankees are 127 and the Cubs 144.  If you resize to make the boxes square and extend the axes on both sides to 150, I think it would give a much clearer emphasis to the pitching side, at least visually.

Russ S
13 years ago


If you’re going back and re-getting the data, it would be interesting to see what the losers look like as well.  And if you’re up for it, I was wondering if I could “borrow” your new data (with winners and losers) to run a few paired analyses to see if that would give any further insight. I have a feeling that you could get a lot cleaner picture of the relationships by controlling for strength of opposition, which would naturally be done through pairing.

Walt Davis
13 years ago

Sorry, but there’s a fundamental flaw.

“League average” OPS is based only on position players.  Team OPS of course includes pitcher hitting.  This means that, for the entire history of the NL and the pre-DH history of the AL, league average TEAM OPS+ is not 100.  The 2006 Cards had a 97 OPS+ in a league with an average team OPS+ of 94 so they were above-average.  The 1907 Cubs had a 92 OPS+ in a league with an average team OPS+ of 93 so they were barely below-average.

Shift the Y-axis so it crosses the X-axis at around 94-95 and you have a more accurate chart (post-DH AL winners screw things up though—the Twins were a legit below-average offense).  In that chart, almost all the winners are, no surprise, above-average in both offense and defense.

In short, a median OPS+ of 104 is, for most of baseball history, about 10 OPS+ points above average and a lot closer to a median ERA+ of 114 than you think.  As a crude fix, divide team OPS+ by league-average OPS+ then re-do your plot.

There are also potentially issues about how ERA+ is figured (see Tango’s work on that) and that OPS+ undervalues OBP some.

Rory / Paapfly
13 years ago

Thanks, everyone. I’m aware of the mistake in the league-average OPS+ and will be posting an addendum on THT Live.

13 years ago

“1907 Chicago Cubs….the very last Chicago Cubs team to win it all”

Oh dear.

Jeroen Blok
13 years ago

Nice analysis I have one important question though, does this also hold true for the regular season? Or is there something about the post-season format that just makes pitching more important?


Dave Studeman
13 years ago

I’m having problems with this line of inquiry.  Regarding Bill Petti’s work, all he showed was that teams with higher run differentials in the past ten years tended to reach that run differential through better pitching than better hitting.  That’s not the same thing as saying that that pitching is better than hitting.  Plus, OPS+ and ERA+ are not the same scale—you can’t draw quadrants at 100 for each of them.

To me, it’s simply a matter of looking at run differential and asking two questions:

– Do better pitching/fielding teams show a tendency to outperform their Pythagorean projection?  I haven’t found that to be true in my studies, but I could be wrong.

– Is it inherently easier to reach a larger run differential through defense than offense?  That’s a complicated question, but it can be studied.

I don’t see that any of the work posted thus far answers those questions.

13 years ago

I’ll suggest an even more fundamental flaw.  The Y axis doesn’t measure pitching.  It measures defense.  Until fielding is separated from pitching, the title and entire point of the article is very misleading.

Marc Schneider
13 years ago

I’m not a statistician, but I don’t see why the fact that ERA includes fielding invalidates it as a way to measure pitching overall.  Good fielding will help bad pitching but it won’t make a bad pitching staff good. If the pitchers are giving up home runs and walking people, the defense is not going to help much.  On an individual pitcher basis, you could argue that ERA (or ERA+) is a misleading measure, but would that be true for a staff as a whole?  The Big Red Machine has always been known for having great defense (is that a myth?) yet the pitching was still mediocre.

13 years ago

I left a comment over at The Book blog as well, but I wld like to see these type of exercises treat the season and postseason as two separate events. I’m not convinced the data doesn’t get skewed by all the factors that change from in season play vs. playoffs.

13 years ago

Out of curiosity – where do the 1988 Dodgers fall on this axis?  They were clearly pitched to that championship from Opening Day through to the World Series.

Rory Paap
13 years ago

‘88 Dodgers had a 90 OPS+ and 114 ERA+. That’s basically just what you suspected.

13 years ago

Re: Studes “- Is it inherently easier to reach a larger run differential through defense than offense?  That’s a complicated question, but it can be studied.”

More and more I’m wondering if we are correct in using linear functions for non-linear processes.  We all know Pythagorean is non-linear.  Yet we value basically everything using linear functions.

The question raised above is regarding larger run differential when it is possible to win more with a LOWER run differential.

I covered this in my blog once – http://obsessivegiantscompulsive.blogspot.com/2008/06/hey-neukom-my-giants-business-plan_27.html – and that was based on a THT article’s finding that each run given up results in an exponential rise in the number of runs scored to maintain the same winning percentage.

To give an example, the Giants RA in 2011 was 3.60 runs/game.  To win 90 games, they needed to score 4.10 runs per game, or a 0.50 run differential.  The NL average RA was 4.35 runs/game.  To win 90 games, a team with that average would need to score 4.96 runs/game, or a 0.61 run differential.  With a run differential of 0.61, the Giants would have won around 92 games.

There is roughly a 10% loss of win efficiency as RA rises, for every 0.1 runs given up, the team needs to score 0.11 runs to keep their win total the same.  Or vice-versa if the team is getting better and better at not allowing runs, the team does not need to score as much as other teams with higher RA to achieve the same winning percentage.

I understand the need to measure players as if there was a neutral team with a neutral park and so forth, but when looking at team-statistics and success in the playoffs, that is not neutral.  Teams don’t win with neutral-modified team stats.

For example, based on the above loss of efficiency of winning, as ERA+ rises, the OPS+ necessary to win with it drops.  Yet this skewing is not captured in the chart above (and nothing against the chart or article, which I both loved, just wanted to bring up this point that has been bothering me for a while and see if anyone agree with me or not) or any other charts I’ve seen before like this. 

Generally, most people don’t realize that this efficiency exists when you have a great pitching staff, and just assumes that the offense has to be as good as usual to compete. 

And that is because all of the stats in baseball are linear in nature and yet wins and losses are a non-linear function.  I don’t think anyone has tackled that dichotomy yet, to much extent, other than what that THT author, David Gassko, did:  http://www.hardballtimes.com/main/article/pitching-runs-created/.  As he noted, “A run saved is not equal to a run scored.”

He created a metric called Pitching Runs Created, which does try to capture the value to the team when a pitcher is better than average.  Not sure where to get that stat now, though.  And I’m not sure how it could be used to create a chart similar to the above.

Dave Studeman
13 years ago

Well, you can get Pitching Runs Created in the THT Annual.  Just sayin’

Yes, according to the Pythagorean Theorem, a run saved is worth more than a run scored when added to a .500 team, but is that necessarily the best way to judge such a thing?  For teams with significant negative run differentials, a run scored is worth more than a run allowed.

However, there may be a bias in the Pythagorean Theorem too.  Here’s something to read: