The Quality of Opposition Factor in the AL Cy Young Race

The AL Cy Young race has gotten pretty interesting. A month ago, Max Scherzer looked like a near lock, with his glistening 19-1 record and strong peripherals that supported the idea that he’d pitched like an ace this year. However, he’s gone 1-2 with a 4.66 ERA in his last five starts, perhaps opening the door for voters to take a closer look at all the candidates. And when they’re investigating, they just may find that the difference in quality of competition is a pretty significant factor this year.

For example, let’s look at the competition effect for Scherzer and Chris Sale. On first glance, you might think that there wouldn’t be a huge difference here, since both pitch in the AL Central and face a lot of common opponents. However, breaking down their starts by opposition reveals some striking differences. The following table lists their opponents in common first, then the teams that they faced that the other did not.

Opponent Scherzer Sale
Cleveland 4 4
Houston 2 2
Kansas City 4 3
Minnesota 2 2
New York Mets 1 1
New York Yankees 1 2
Oakland 2 2
Tampa Bay 2 2
Texas 1 2
Toronto 1 1
Seattle 1 1
- - -
Opponent Scherzer Sale
Chicago White Sox 5 0
Boston 2 0
Baltimore 2 0
Philadelphia 1 0
- - -
Opponent Scherzer Sale
Detroit 0 5
Anaheim 0 2

Among the teams in common, there aren’t many major differences. Scherzer got one extra start against the Royals, and Sale drew the Rangers one more time, but by and large, they basically matched up pretty similarly in teams that both pitched against. However, when it comes to the teams that one faced and the other did not, the story gets a lot different.

Both Scherzer and Sale have managed to draw five starts during the series in which their teams play each other. For Scherzer, that means five starts against the worst offense in the American League, as the White Sox have scored just 3.7 runs per game this year; for Sale, that means five starts against the Tigers and their 5.0 runs per game offense. This is one of the hidden advantages of pitching on a good team that inflates a pitcher’s W-L record: not only do you get the benefit of run support in helping decisions swing your way, but you never have to face your own offense, lowering your overall quality of competition at the same time.

Nowhere is this don’t-pitch-against-teammates divide larger than with Scherzer and Sale, as Sale essentially was selected out of facing the most feeble AL line-up, while Scherzer wasn’t forced to have a single confrontation with Miguel Cabrera or Prince Fielder this year. Having good hitters on your team doesn’t only get you more runs when you’re pitching, but makes it less likely that you’re going to give up runs, since that’s fewer good hitters you have to face yourself.

For the record, Sale was absolutely brilliant against the Tigers this year, giving up just 10 runs in those five starts, and throwing 39 innings in the process, so it will be tempting for some to wave away this difference as only theoretical, since all those extra starts against Detroit didn’t drag down Sale’s numbers. But by doing that, we’re marginalizing the very best performances Sale had all season, where he shut down an elite line-up and nearly averaged a complete game each time out. We should not just ignore Sale’s dominance against tough competition because he managed to overcome it rather than get beaten down by an offense that regularly beat down lesser pitchers.

For reference, here are the seven pitchers who currently have an RA9-WAR of between +5.5 and +6.5 — using RA9 instead of FIP because this is the pool that voters will likely consider — and the average runs scored by their opponents, per Baseball Reference’s calculations:

Pitcher Opponents RA9
Chris Sale 4.54
Yu Darvish 4.46
Hisashi Iwakuma 4.42
James Shields 4.36
Max Scherzer 4.30
Bartolo Colon 4.29
Anibal Sanchez 4.27

Not surprisingly, we see the two Detroit pitchers near the bottom, with the Oakland A’s — 4th in the majors in runs per game — represented at about the same level, while the pitchers on teams with less effective offenses have each faced teams that have averaged about two-tenths of a run more per game.

This isn’t a large enough factor to be the deciding reason to pick one AL Cy young candidate over another, but the reality is that this race is really about splitting hairs, at least for voters who won’t put much or any weight on win-loss record. These guys have all pitched well this season, and there isn’t a huge separation between the pitchers at the top of the pack. There are a variety of candidates with arguments in their favor, and I wouldn’t say that Sale should win simply because he drew five starts against the Tigers.

It should be a factor, though. This is basically the same thing as a park effect, and it’s pretty well established at this point that a pitcher who posts a 3.00 ERA in Texas was more valuable than a pitcher who put up the same number in San Diego. We could essentially lump this difference in with park factors and refer to them in total as degree of difficulty, since opposing hitter quality affects the value of a raw number in a similar way to a park effect.

In a close race, this isn’t something that should be overlooked. And this year, it is a close race, even with Scherzer being the only pitcher to notch 20 wins. Sale and his 11-13 record are unlikely to get serious consideration, but he’s had a great season, especially once you factor in the schedule he’s had to face.




Print This Post



Dave is a co-founder of USSMariner.com and contributes to the Wall Street Journal.


60 Responses to “The Quality of Opposition Factor in the AL Cy Young Race”

You can follow any responses to this entry through the RSS 2.0 feed.
  1. chuckb says:

    Is your table for RA/9 weighted by number of games against those opponents or does each opponent count equally?

    I’m not that familiar with B-R’s numbers since I spend so much time over here.

    Vote -1 Vote +1

    • Dave Cameron says:

      It’s prorated based on the number of games.

      Vote -1 Vote +1

    • chuckb says:

      I should add that this is really interesting stuff. I would like to see if Fangraphs could come up with some version of its numbers that includes the quality of competition. It would be great (though I know it would be a humongous task) to be able to compare players by the level of competition they faced, especially if it could be done by player rather than just by team.

      In other words, a certain hitter faced pitchers who combined for a FIP of … and a pitcher faced hitters who combined for a wOBA of (whatever). It doesn’t all even out, which I think is Dave’s point here.

      Vote -1 Vote +1

  2. Rex Manning Day says:

    “We could essentially lump this difference in with park factors and refer to them in total as degree of difficulty, since opposing hitter quality affects the value of a raw number in a similar way to a park effect.”

    Confirmed: Degree Of Difficulty coming to Fangraphs custom leaderboards by next season.

    Vote -1 Vote +1

  3. dave k says:

    I know Scherzer gets all the attention, but I think Sanchez deserves it more. If it wasn’t for his midseason injury I think he’d be getting a lot more attention.

    Vote -1 Vote +1

    • Bip says:

      I know (player with a lot of wins) gets all the attention but I think (player with fewer wins but better peripherals) deserves it more.

      There’s going to be quite a few of these every season until Mr Brian Kenney succeeds in his crusade.

      Vote -1 Vote +1

      • dave k says:

        When you put it like that you make it sound like Scherzer’s wins are all luck but that’s not really the case…his peripherals are quite good too and on par with the other contenders.

        To Kevin’s point below, had Sanchez been able to rack up more innings while maintaining the same level of performance I think he’d be in (and perhaps leading) the conversation. Same goes for Felix. Innings matter.

        Vote -1 Vote +1

    • Kevin says:

      That injury really kinda hurt his chances. 27th in Innings pitched. His numbers have been incredible when he’s healthy. But missing almost an entire month is a tough sell especially when other pitchers have done pretty well.

      When trying to win a Cy Young, don’t miss all of june.

      Vote -1 Vote +1

  4. Tim A says:

    I think WaR is unfair to Oakland pitchers. From what I can tell they have an entire team built too beat FIP. Is there going too be some park factor sort of adjustment? All season long this team has beat the xFIP by a mile, and knowing Beane I would surmise that is not an accident, so will pitchers continue getting dinged on WaR totals, simply because they play for a team thats philosophically against traditional pitcher metrics for success?

    Vote -1 Vote +1

    • Dave Cameron says:

      xFIP has no impact on WAR.

      The A’s pitchers have a lower ERA than their FIP because their ballpark has a lot of foul territory and they have good defenders behind them. There is no reason to give the pitchers credit for either of these things, since every pitcher benefits from large foul territory and quality defensive teammates.

      Vote -1 Vote +1

      • jsolid says:

        Here’s what they do benefit from.

        The Oakland park factor for HR (from 2012) is .94, one of the lowest in the league. (the 2B PF is on the low side also.) So from this i will infer that fly balls benefit pitchers more in Oakland than most other stadiums. So what did Billy Beane do? he went and collected fly ball pitchers. The team pitching fly ball rate is 40.9% Not only does that lead the league, its not particularly close; the rest of the league is 30.0-36.8% (plus the Pirates, lowest at 27.1%).
        So i think we can assign some WAR, not to the pitchers, but to that clever GM over there.

        Vote -1 Vote +1

        • Tim A says:

          If the team knows it works, and does it on purpose, then the pitchers get no credit for fine tuning there style too work within the teams wheel house? Also they would have an absurdly low, even with park factors, HR rate if you remove AJ Griffon, and his league leading HR allowed from the mix( most of which were the solo variety. It is far easier too be the team longball king, when you have the lowest WHIP on the team as well).

          Vote -1 Vote +1

        • David says:

          a’s hitters are leading the league in hr’s since the break, and are third for the season.

          Vote -1 Vote +1

      • Tim A says:

        I’m referring more too the teams philosophy progressing over the last few years too go more Bartolo with it. I have noticed that, since they have a good park and defense they ignore strikeouts. The rotation as a whole seems too have adopted the Bartolo Colon style, whereby they don’t focus on strikeouts. They throw strikes, challenge the hitter, walk almost no one, pitch deep in games, and generally seem to thrive without striking out batters at the same rate as other teams that perform around the same level in run production.
        All season long it has seemed too me that they get no love for WaR with the rotation, and despite top tier run prevention, they have middle of the pack WaR. I also will note that team wise they rank badly in D compared too other teams at -13 D runs as a team per fangraphs.
        The projected RoS team, has for months been projecting them too regress as far as run prevention goes, and was projecting thru the whole summer for a team that hasn’t had a losing record in any of the last 10 months straight too turn into a sub .500 pumpkin. I guess my point, is that it seems too me, as an A’s fan that watches the team closely, that both your projection and WaR systems hate the type of baseball the A’s play.
        I think the sample is starting to get large enough that this is a sustainable team skill, and I am getting a little sick of correctly miss judging the projections. For a site I love, dedicated too the types of baseball discussions that are interesting, Oakland gets about the same Love as Seattle from Fangraphs/National BB press, and I personally think this is just criminal.
        Best team in Baseball for the last 2 years combined, with unreal consistency that no one has ever heard of.

        Vote -1 Vote +1

        • Tim A says:

          I guess this is a long way of saying, Bartolo has been more important too his team then any other AL starter this year. By the same logic that CY comes from only a playoff team. Most valuable pitcher should be in relation too what the team would have looked like without them. Bartolo carried this team thru periods of struggle, and was winning consistently with very little run support when the team needed sterling 8IP 0-1 run allowed starts he did it day in and out.

          Icing on the cake he’s Fucking Forty!!!! BC 2013 Cy Young in my opinion, nobody else had a tougher road too walk(or not walk as the case may be), in order too achieve top level success then him, and it’s not even close.

          Vote -1 Vote +1

        • vivalajeter says:

          If Colon was actually the best pitcher in the AL this year, then it would make perfect sense for them to give the CY Young to Johan Santana.

          Vote -1 Vote +1

      • Rick says:

        As Tim A mentioned further down: you state the A’s have a good defense, but they don’t according to the stats. Could you further revise your comments to include the facts being different than you assumed?
        I am really curious as to what the A’s pitchers are doing to perform so much better than their WAR, FIP, etc. Is the only explanation, “oh, must be all b/c of the foul territory.”

        Vote -1 Vote +1

    • Spit Ball says:

      Lots of foul ground generally helps pitchers more than hitters.

      Vote -1 Vote +1

      • Spit Ball says:

        Sorry Dave, didn’t mean to double up on your comment, It was not there when I entered it but the clock says you beat me to it.

        Vote -1 Vote +1

    • Pale Hose Kyle says:

      This is offtopic of course.

      Having the lowest HR/FB rate in the AL could have more to do with how you build your stadium than how you build your team. Having the lowest BABIP in the AL could have something to do with defense. I think those are reasonable justifications for “dinging” A’s pitchers. They didn’t do that.

      Vote -1 Vote +1

      • Tim A says:

        -13D as a team this year why does everybody keep saying the D did it? Either they have better D then the system says, or they are being hurt by the defense not helped, get it straight people.

        Vote -1 Vote +1

        • Nathan says:

          I’m not sure what “-13D” refers to. I’m not seeing any team fielding numbers like this for the A’s.

          In any case, this isn’t a zero-sum situation. The defense’s UZR (or whatever number you’re looking at) does not have to be an exact compliment to the pitching staff’s RAR/WAR. Your view that it’s “one or the other” is a false dilemma.

          My feeling is that the mistake you’re making in your thinking that leads you to this false dilemma is that you’re not consciously considering that the statistics you are citing are context-neutral. The purpose of statistics that measure the value of a player is to give us a context-neutral assessment of what a player did, because with very few exceptions, no player can control the context of a situation that he’s put in on the baseball field.

          When you add in the context (e.g. runners on when this catch was made, empty bases when that strikeout happened), then you get the stuff that causes the value measurements to appear as if they are “wrong” because they don’t match the “results.” But really, nothing is wrong except the lack of understanding about which numbers contain context, and which ones do not.

          Vote -1 Vote +1

  5. Scott J Marcus says:

    In looking at the teams that Sale and Scherzer faced, one can see that Scherzer faced Boston twice, and Baltimore twice — one great offense, and one fairly good one. On the flip side, Sale did face the Angels twice, and they are a decent offense.

    But given the extra starts for Scherzer against the Red Sox and Orioles, I see the difference between their difficulty level as fairly small.

    Vote -1 Vote +1

    • what...? says:

      Cameron already listed the ra9’s of their opponents. There’s a 0.24 RA9 difference, which I’d say is not insignificant.

      Vote -1 Vote +1

    • Brandon says:

      The Angels are pretty darn close to the Red Sox in wRC+ (109 vs. 113) so I’d say they’re closer to a “great” offense than a “decent” offense.

      Vote -1 Vote +1

    • Free Bryan LaHair says:

      the pressure of having to pitch at an elite level knowing the run support will most likely not come is a major factor for me. Sale would’ve cruised through lineups on a team with run support similar to Detroit’s. what intrinsic pressure did scherzer or sanchez ever have to deal with when they woke up every 5th day with a peace-of-mind knowing that their lineup was better than everybody else’s?

      Vote -1 Vote +1

  6. isavage30 says:

    Not sure how meaningful a quality of competition analysis is. Sale’s an interesting guy to choose for this, because what destroyed his numbers was not the 5 starts against Detroit, it was the 4 starts against Cleveland (6th in the AL in wOBA). He gave up almost 1/3 of his earned runs on the year in those 4 starts, and would have a good Cy Young case if you could eliminate those 4 starts.

    How would you break this down in a meaningful way? Pitching a shutout against Detroit may be more impressive than doing the same against Chicago, but what about giving up 22 runs in 23 innings to Cleveland’s average offense? (which would look even more average if not for the 4 games where they got to beat up Chris Sale). How does that work to penalize his overall numbers, especially considering the Indians’ team wOBA is higher than it would be because of the starts against Sale? I would think this tends to even out enough, that breaking it down to who faced who just confuses things.

    Vote -1 Vote +1

    • Jamie says:

      Also, Sale is a LHP, and the Tigers have been very bad this year against LHP, which would lead one to wonder whether his success against the Tigers should be downgraded given how little success the Tigers have had against any LHP.

      Vote -1 Vote +1

    • Pale Hose Kyle says:

      Sure, it’s a thorny situation, but we treat park and league adjustments the same way, don’t we?

      Vote -1 Vote +1

      • isavage30 says:

        Well, something like park should be an independent variable, but personally, I think the Fangraphs team pitching WAR, for example, is a pretty crazy thing. How do the Mariners, with a 4.33 ERA and 3.93 FIP, have a higher WAR than the Rays (3.76/3.74)? Do you really think the Mariners have better pitchers than the Rays? That stuff is useful at the extremes, looking at individual pitchers in Colorado, for example, but in the aggregate I think it gets very murky.

        When you’re looking at something like this, it’s even more complicated because there is no independent variable. The White Sox offense is affected by their performance against Scherzer in those 5 starts, etc. You’d have to look splits, and at the actual players in the lineup on the days they faced the particular teams, and look at those players’ performance when they’re not facing the pitcher you’re evaluating … you face the Twins now sans Mauer and Morneau, it’s different then facing them earlier in the year. Or the White Sox, with Rios. So even looking at team wOBA could be hugely misleading when you’re talking about a handful of starts. I just don’t see how you’re going to get a result that’s more meaningful than their actual performance against the teams they faced, and it would be absurdly complicated to be done right. Looking at a flat ra/9 for opponents and making a judgment off of that I would say is next to meaningless, as in the case of Sale, where what hurt his numbers against Cleveland … what does the fact that he shut down Detroit and got be up by Cleveland and didn’t get to face his own team do for his Cy Young case? I have no idea

        Vote -1 Vote +1

  7. Wes says:

    So Scherzer held opponents to a 2.74 FIP, compared to their RA9 of 4.30 (a difference of 1.56).

    Sale held opponents to a 3.20 FIP, compared to their RA9 of 4.54 (a difference of 1.34)

    Even if we account for the quality of opposition, then, it seems that Scherzer has the edge over Sale.

    Vote -1 Vote +1

    • pft says:

      Yet its Darvish who is really the man who should be the front runner and be compared, Both Darvish and Sale pitched in much more of a hitters park than Scherzer.

      Vote -1 Vote +1

      • Ian says:

        Except that Comerica was a better hitters park than the other two this season…other than that, for sure you’re right.

        Vote -1 Vote +1

      • pft says:

        Parks themselves don’t change much from Y2Y, and minor changes in temp and wind won’t affect them that much. Extreme Y2Y variance is likely a SSS artifact. Like using 1/2 year of UZR, you have to use multi-year park factors and consider the handedness of the pitcher and hitter (eg Safeco is not as friendly to RHP’ers as LHP’ers).

        Vote -1 Vote +1

  8. Jacks says:

    Cleveland is definitely Sale’s kryptonite. In 23 innings vs CLE this year, Sale has given up 22 ER. That inflates his ERA vs winning teams by quite a bit: 3.35 vs teams .500 or better, 2.16 vs losing teams. Turns out that mirrors Scherzers: 3.35 vs winning teams, 2.5 vs losing.

    Vote -1 Vote +1

  9. CFIC says:

    last time I checked, Felix Hernandez was having him quite the season

    Vote -1 Vote +1

  10. Bob Welch says:

    “…perhaps opening the door for voters to take a closer look at all the candidates.”

    You may be giving said voters a little too much credit.

    Vote -1 Vote +1

  11. Feeding the Abscess says:

    More than just RA/9 per team, Sale has faced teams that have done very well left-handed pitching:

    Cleveland, 1st in MLB in wOBA versus lefties: 4 starts
    Oakland, 2nd: 2 starts
    Detroit, 3rd: 5 starts
    Texas, 4th: 2 starts
    LAA, 5th: 2 starts
    Tampa, 7th: 2 starts

    17 of Sale’s 29 starts have come against teams in the top 7 in baseball against LHP, and 11 against the top 3. Whoever said the Indians offense is average is simply wrong. Same goes for the person who said the Tigers offense against LHP has been ‘very bad’.

    I would think that any degree of difficulty stat should factor in offenses against pitcher-handedness rather than overall offense, as some teams show a wide disparity in performance against differing pitcher-handedness. For example the Astros: one of the three worst offenses in baseball by wOBA against RHP, the Astros come in at a respectable 12th against LHP.

    Vote -1 Vote +1

    • isavage30 says:

      But one of the reasons Cleveland’s numbers against lefties are so good is they beat up Sale himself. 110 of their 1890 plate appearances, almost 6% of their LH plate appearances, were against Sale, and they hit .340/.391/.570 in those plate appearances. Where would they rank against non-Sale lefties?

      Vote -1 Vote +1

      • Feeding the Abscess says:

        Sale did well against Detroit and destroyed the Angels in both his starts against LAA. Both teams would have would ranked better against lefties had they not faced Sale.

        Vote -1 Vote +1

        • Feeding the Abscess says:

          Besides, Cleveland had a .335 wOBA. Give them league average production in those 6% of PA that were against Sale and it’d be, what, .333 or .334? That would place them second behind Oakland. If you adjust Detroit and LAA upward a point or two each, Cleveland’s drop is negated, if not overly compensated for.

          Vote -1 Vote +1

        • isavage30 says:

          My point being, the idea of quality of opposition factoring into Cy Young race is way more complicated than just looking at opponent RA/9. Even when looking at an opponent by their split, you are looking at that split influenced by the performance against the pitchers you are analyzing. I think you can dissect stats too deeply, where instead of making things more clear and getting more to the reality, you’re just confusing it even more and the new information makes it more difficult to conclude anything. Breaking down pitcher performance by their opponents being one of those cases.

          Vote -1 Vote +1

        • Feeding the Abscess says:

          Quality of opposition is factored into analysis of AL versus NL pitchers. Why can’t that be distilled into a formula that can apply to pitchers from the same league? The same team?

          Vote -1 Vote +1

        • isavage30 says:

          There are problems with the park and league adjustments too. Personally I take them with a grain of salt, one of the reasons WAR isn’t meant to be viewed as accurate to the decimal point. I don’t honestly know much about how league adjustments are calculated. With the NL, the drop-off of facing the pitcher is pretty much a league-wide thing though, with some outliers like Greinke who are actually decent hitters. But the adjustment would be made league-wide with one common fact of light-hitting pitchers being the primary driver of the variation between the leagues, while the adjustment Dave’s talking about here is on a team-by-team basis, and there’s a hell of a lot more variation on individual games against individual teams than there is when you’re looking at the league as a whole Say Darvish faces Detroit and shuts them down. Looks good, but what if Miggy’s hurt, Peralta’s suspended, Victor Martinez is getting a day off? Do you account for that? If you don’t break it down to actual lineups faced, then it’s not definitely not accurate.

          If your goal is to see who pitched the best in relation to the toughness of their outs, I think there’s a lot more variation in just day-to-day streaks, injuries, and lineups faced, than breaking it down as who faced what team and thinking that has some greater meaning than their overall results. Just because Scherzer doesn’t have to go against Detroit and Sale didn’t get to face the White Sox doesn’t say that much, and I don’t think you can break that down into a formula with any degree of accuracy/reality. You might be able to do something on a division-by-division basis that’s meaningful, but getting more granular than that and I’d think it’s pretty much nonsense.

          Vote -1 Vote +1

  12. I Agree Guy says:

    “Chris Sale will be nothing more than a reliever.” – me, circa 2010

    Thankfully I didn’t quit my day job.

    Vote -1 Vote +1

  13. JS7 says:

    Felix Hernandez should be in the conversation as well.

    Vote -1 Vote +1

  14. Michael Allen says:

    “Having good hitters on your team doesn’t only get you more runs when you’re pitching, but makes it less likely that you’re going to give up runs, since that’s fewer good hitters you have to face yourself.”

    You don’t say.

    Vote -1 Vote +1

  15. FeslenR says:

    good information Dave,

    the issue is “traditional writers who votes” usually just look at traditional stats.

    Vote -1 Vote +1

Leave a Reply

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

You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <strike> <strong>