The Split Personalities of the Cardinals

One of the most remarkable things about the Cardinals’ clutch hitting — aside from the reality of the Cardinals’ clutch hitting — is the way it refuses to regress the way we’d expect it to. From a Cardinals game recap over the weekend:

At the end of April, the team was batting .327 with runners in scoring position. A month later, the average was up to .333. Three months in, it sat at .335. And after going 2-for-5 with runners in scoring position on Sunday, the Cardinals now boast a mark of .338. No team in baseball history has even come close to maintaining such a pace.

The Cardinals came out of the gate hitting well with runners in scoring position. At least by batting average, every month they’ve only gotten better. That’s not what regression looks like. That’s the polar opposite of what regression looks like. It’s been enough to make some people wonder. Just what are the Cardinals doing, and how are the Cardinals doing it?

What the Cardinals have been doing is coming up with timely hits. That much isn’t a secret. Buster Olney likes to tweet about it, and if you ask any of the Cardinals or the coaches, they’ll tell you it’s about a mature, responsible approach. The players want to be hitting in big spots, and they don’t try to do too much, and that explains why the Cardinals’ numbers are so historically high.

Over nearly a thousand plate appearances, the Cardinals have a team 141 wRC+ with runners in scoring position. They’ve generated a hit in every third at-bat, and they’ve slugged .468. At this writing, Evan Longoria has a 141 wRC+. The Cardinals, with runners in scoring position, have hit as well as Evan Longoria, as a team. That’s going to get noticed, because it’s amazing. We’re almost into August.

But all the talk about the Cardinals’ hitting with runners in scoring position ignores something important: the Cardinals’ hitting with the bases empty. Over more than two thousand plate appearances, the Cardinals have a team 89 wRC+ with nobody on. At this writing, John Buck has an 89 wRC+. The Cardinals, with the bases empty, have hit as well as John Buck, as a team. That hasn’t been noticed so much, because it isn’t sexy, but there’s significance in here.

Believe if you want that the Cardinals are capable of amazing things with the pressure on, but how do you explain the performance under a little less pressure? Those are still mostly important plate appearances, and, where are the results? If the pressure results are a consequence of a solid approach, why has that approach failed the team in other situations? Or, why hasn’t that approach been adopted in other situations? If the Cardinals deserve praise for their clutch performance, they deserve also criticism for their performance with nobody on. I mean, as long as we’re assigning responsibility.

The split is too wide. It will regress. At the very least, any reasonable person would expect it to regress, and if it doesn’t, that’s just because the season stops at 162 games. This is the simple, easy point. The Cardinals aren’t that good, or that bad. They deserve some credit for what they’ve already done, just like Mike Cameron deserves credit for having homered four times in one game, but Cameron never did that again. Cameron didn’t make that a habit.

Here’s easy evidence. A year ago, the Cardinals had a very similar baseball team. Similar team, same manager. With nobody on base, the Cardinals posted a .326 wOBA and a 106 wRC+. With runners in scoring position, they posted a .328 wOBA and a 106 wRC+. Whatever they’re supposedly capable of now, they weren’t capable of last season, with mostly the same roster. Again, it’s really easy to say the Cardinals are going to regress to something more normal.

Might there be some meaningful differences in how plate appearances have been distributed? Well, for one thing, this year almost all of the Cardinals’ hitters have been good. But, David Freese and Jon Jay have gotten 16% of plate appearances with no one on, and 22% of plate appearances with runners in scoring position. Matt Carpenter and Carlos Beltran have gotten 25% of plate appearances with no one on, and 17% of plate appearances with runners in scoring position. No solutions can be found here. This isn’t a distribution thing — this is just a weird thing.

By OPS, there’s a 125-point difference between the Cardinals’ scoring-position performance and their overall performance. No team has posted a bigger gap since 1969, with the 1981 Brewers showing up at 124. Just six teams have come in at greater than 100, with the Cardinals leading the way. With more than two months left to go, the Cardinals are on a historic pace. That much probably isn’t surprising.

Something I’ve more recently become interested in, as far as the Cardinals are concerned, is how regression might look. The Cardinals can be expected to hit worse with runners in scoring position. But they can also be expected to hit better with nobody on, and there are more of those plate appearances. And more success with nobody on leads to more scoring-position opportunities. To what extent might these dual regressions cancel out? What’s the effect on run scoring of more timely hitting? What’s the effect on run scoring of less timely hitting?

The core of the research:

runsgamewoba

That’s runs per game, plotted against wOBA, on the individual-team level since 1969. The relationship is obviously strong, deliberately strong, and it allows us to play around. For example, the Cardinals this year have a .328 wOBA and are averaging 4.97 runs per game. As a group, the 26 teams with .328 wOBAs have averaged 4.58 runs per game. Those teams have also averaged very slightly negative base-running scores, while this year’s Cardinals are a few runs worse.

That equation up there on the graph allows us to calculate an “expected runs per game.” Do so and 73% of teams fall within +/- 0.20. I decided first to look at the 20 teams from the sample with the most lopsided performances in terms of hitting with runners in scoring position. These 20 teams averaged an .813 OPS with runners in scoring position, and a .719 OPS overall. They also averaged 4.48 runs per game, against an expected average of 4.37. By base-running, they averaged -1.4 runs.

Then I looked at the opposite 20 teams — those teams with the worst relative performances with runners in scoring position. These 20 teams averaged a .676 OPS with runners in scoring position, and a .731 OPS overall. They also averaged 4.29 runs per game, against an expected average of 4.50. By base-running, they averaged -0.8 runs.

By that evidence, hitting poorly with nobody on doesn’t cancel out hitting well with runners in scoring position, even though there are far more plate appearances in the former sample. And hitting well with nobody on doesn’t cancel out hitting poorly with runners in scoring position. It makes good sense, because those scoring-position plate appearances have a higher leverage to them, as there’s more to be gained from each hit. The Cardinals have hit too well with runners in scoring position and too poorly with the bases empty, but as a consequence of this they’ve scored more often than they should be expected to.

Which is kind of reflected by their projected 4.44 runs per game. The Cardinals have averaged more runs per game than any other team in the National League. They’re projected to be behind the Rockies, and though they should still be good, they should score less often, because this split shouldn’t sustain, no matter how long it’s been sustaining. It doesn’t make sense why this would be reality, so absent any compelling reason, we needn’t accept it as reality. We need to accept what’s already happened, but it isn’t completely reflective of the team’s true ability.

Which is all to say: you could’ve guessed this conclusion. It’s nothing particularly surprising. The Cardinals have been too good at producing runs, and going forward they should be less good. They’re still a very good team, and they presently have the lead in their division so they’re sitting in a comfortable spot. It helps that they ought to start performing better with nobody on base. But when there are people on base, on dangerous bases — expect more stranded runners. The Cardinals can only do so much, from here.




Print This Post



Jeff made Lookout Landing a thing, but he does not still write there about the Mariners. He does write here, sometimes about the Mariners, but usually not.


50 Responses to “The Split Personalities of the Cardinals”

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

    I really admire most everything about the Cardinals organization. Letting Pujols go and extending Craig reinforced this

    Vote -1 Vote +1

    • Jake says:

      I am pretty sure the article concludes that the organization cannot take credit for the success this year.

      Vote -1 Vote +1

      • Steven says:

        Right, and the SF Giants organization definitely wasn’t responsible for their World Series title last year! Sound logic!

        Vote -1 Vote +1

    • Andrew says:

      They didn’t ‘let Pujols go. They still offered him over 200 million and got lucky that the Angels bailed them out.

      +10 Vote -1 Vote +1

      • croggs says:

        I wouldn’t say they got lucky. They gave him an offer that they knew was significantly below what the market was willing to pay to give the appearance that management tried to get Pujols to stay. I think the offer was more to help with management’s perception since King Albert was the most love person in STL over the last decade or so. Maybe they got a bit “lucky” in the sense that there was an unfavorable risk/reward scenario from a pure baseball perspective. However, I think the offer was a calculated risk, and management had a good sense that Pujols was pretty unlikely to stay for what he was offered.

        +10 Vote -1 Vote +1

        • Alexander Nevermind says:

          So the Cardinals offered Pujols a contract knowing he wouldn’t take it just to appease fans? That makes no sense. The Cardinals thought he was worthy of a 9yr/$200mm contract.

          Vote -1 Vote +1

        • oPlaiD says:

          Yeah this is a bunch of malarkey.

          They thought Pujols would sign and he almost did sign. It caught the organization by surprise when the Angels came out of nowhere and offered a huge deal to get Pujols.

          Still, they do deserve some credit for being fiscally responsible in this case. Obviously things would be much much different if they had Pujols still and, say, didn’t have Molina… or more likely didn’t have Wainwright re-signed.

          Vote -1 Vote +1

  2. JT says:

    I was actually comparing the Cardinals to the Reds and Braves just yesterday. These two teams are almost the exact opposite of the Cardinals. They both hit well overall but fall apart with RISP….particularly with 2 out and runners in scoring position. My guess is that there is no way that the Cardinals finish the season hitting .325 with 2 out and RISP and there is likely no way that the Braves finish the season hitting .205 with RISP (or the Reds at .199).

    The NL race is going to be exciting at the end of this season. My guess is that as the regression inevitably happens the Cardinals, Reds, AND Braves will ALL be vying for the best record in the NL.

    Vote -1 Vote +1

  3. Jake says:

    It would be interesting to see if the batted ball and plate discipline data for the cardinals is very different with runners on vs. bases empty.

    Batted Ball:

    Men on Base:
    BABIP: .364
    LD%: 22.6%
    GB%: 45.4%
    FB%: 32.0%
    HR/FB: 10.7%
    IFFB%: 5.9%

    Bases Empty:
    BABIP: .282
    LD%: 22.5%
    GB%: 43.9%
    FB%: 33.6%
    HR/FB: 8.4%
    IFFB%: 10.6%

    Vote -1 Vote +1

    • Ben says:

      So the IFFB% drops in half, the line drives stay about the same and the ground balls increase while the fly balls decrease. The HR/FB also increases 25%. That should lead to a wildly better BABIP, which it does.

      Vote -1 Vote +1

    • Jake says:

      Runners in Scoring Position:
      BABIP: .386
      LD%: 24.5%
      GB%: 45.5%
      FB%: 30.0%
      HR/FB%: 8.4%
      IFFB%: 5.0%

      Vote -1 Vote +1

  4. Matty Brown says:

    I am quite fascinated by the 2013 Cardinals. I went to their team stats page and was amazed by the gaudy and consistently strong offensive numbers on the team. I then noticed something of equal interest and relevant to this article.

    They have surrendered a lot of runs on defense and baserunning. As a team they are at -21.3 Fld and -4.2 Bsr. Their team stolen base leader has 4 steals, and as a team they only have 23 steals.

    While their overall play is impressive, and measuring Fld and BSr is far from perfect, we see a team almost strictly focused on hitting and succeeding in this day and age of sabermetrics.

    Vote -1 Vote +1

  5. Eddie says:

    St. Louis leads the league in Faustian Bargains Above Replacement (FBAR).

    +10 Vote -1 Vote +1

  6. John says:

    Jeff, what is the purpose of regressing (and I use that term lightly) Runs/G on wOBA? What does that tell us? wOBA is a metric derived from regressing offensive events on runs.

    Vote -1 Vote +1

  7. Peter says:

    Is it possible they’re stealing signs? A man in scoring position means a man on second, most of the time, and a man on second has a pretty clear view of the catcher. Maybe baserunners are relaying signals to the hitters so they know what’s coming

    Vote -1 Vote +1

    • frivoflava29 says:

      I hate to support what sounds like a conspiracy theory, but seeing as how this was brought up with the Cards & Matt Cain a couple of months ago… at the same time, I have nothing against the Cardinals and so I’d say that I merely see this as an interesting speculation.

      Vote -1 Vote +1

      • FromSomeFrontOffice says:

        If the Cards are stealing signs on the basepaths, then it’s the opposition’s fault for not hiding their signs better. That’s just part of the game.

        +6 Vote -1 Vote +1

  8. leeroy says:

    I find it pretty interesting that on the other end of the spectrum you have the Pirates hitting .220 with RISP (.242 bases empty) and only 1.5 games back.

    Vote -1 Vote +1

  9. Super BAZUNGA TIME says:

    ALL THINGS COME TO REGRESS

    Vote -1 Vote +1

  10. OldDogScout says:

    Firstly, what a giant leap forward for anyone on this site to actually now acknowledge clutch hitting exists.

    As to the puzzlement over the delayed regression to the mean by the Cardinals BA-RISP, are you also tacitly now acknowledging that it might be a sustainable skill, even if limited to the short term?

    What if the relationship is non-linear? Or, there’s an interaction effect not yet fully explained in your modeling attempts. Or, even still that you’re not getting the data at all which would explain the phenomena you’re grasping for an answer to?

    Food for thought.

    -6 Vote -1 Vote +1

    • “Sustainable in the short term” is another way of saying “unsustainable”.

      +20 Vote -1 Vote +1

      • Spencer D says:

        He makes an interesting point regarding linear scoring. Should someone get on base, Babip increases, because the defense is more likely to try and throw out the guy in scoring position rather than the hitter. Once started, it could become a self reinforcing process, especially as pitchers that put men on base are more likely to put more men on base. Add in the potential play by play psychological aspects too.

        Vote -1 Vote +1

        • OldDogScout says:

          You’re on the right track, pursuing this further using the Retrosheet data you will find the answer.

          Hint: cumulative sequences & interaction effects

          Despite Mr. Sullivan’s snarky reply which didn’t offer any attempt to solve the challenge, your assumption holds merit.

          Think outside the typical linear regression to wOBA & regression routine that’s slavishly devoted to continuous practice on this site.

          You will find an interesting new plateau awaits.

          Vote -1 Vote +1

      • OldDogScout says:

        Particle interactions while short lived, relative to other longer term phenomena, still can reach a critical mass responsible for release of a large amount of energy.

        Think on that for a minute & how it might just relate to game theory that’s not easily identifiable within aggregate stats, but can be isolated within event level data such as Retrosheet.

        If one were to consider players as particles, both which repelling & attracting properties, wouldn’t it seem reasonable to assume their data also holds evidence of those interactions involving their mass?

        Food for thought for the inquiring mind.

        Vote -1 Vote +1

      • Jamie says:

        Despite Mr. Sullivan’s snarky reply…he’s right. Challenge. Solved.

        Vote -1 Vote +1

        • Rob says:

          Nope. Circumstances are different in 2013 — the Cardinals have a new hitting coach who interacts with his team in different ways. HUMAN factors which cannot be reduced to sabermetic numbers-babble.

          Vote -1 Vote +1

    • Simon says:

      Your post has a lot of long words in it. Doesn’t make more sense for it.

      Vote -1 Vote +1

  11. EDogg1438 says:

    It shouldn’t be surprising that Freese and Jay bat more often with runners on….they hit behind the Cards best hitters, who are usually on base more.

    Carpenter and Beltran bat behind the 8 and 9 spots in the order, whom you would expect to get on base less in front of them.

    Vote -1 Vote +1

  12. nota bene says:

    Could it be possible that it’s the opposing pitchers who are changing their approach somehow bases-empty vs men-on? Or that defenses are over-shifting or something?

    And then what happens if there are follow-on effects to this trend? In other words, are managers going to be more likely to IBB Craig (for example) with a man on third, because they get scared of the clutchiness?

    Vote -1 Vote +1

    • olethros says:

      In Craig’s specific case, he’s reaching a point where there might be some sort of true talent with men on/RISP. What the hell that would be, I have no idea, but he’s been consistently better with men on for years now.

      Vote -1 Vote +1

      • Lanidrac says:

        Sabermetrics and statistics may try to argue otherwise, but in the end these players are human and psychological effects do matter. There have been too many cases of statistical outliers for me to doubt that clutch hitting is indeed a real skill.

        Vote -1 Vote +1

  13. ErnestoSalvaderi says:

    Maybe the Cardinals scout pitchers based mostly or exclusively on what their tendencies with runners on or pitching from the stretch?

    This article just proved that hitting with runners on has more impact on run output despite less pa. Obviously it wouldn’t be a bad strategy, especially for a team of veteran hitters like the Cardinals have.

    Vote -1 Vote +1

      • jj says:

        maybe they just got unlucky in 2012 and 2013 is the real skill level. haha.

        I would like to see some spray charts with runners on vs when bases empty. that may say something about their approach if nothing else. Molina seems to go the other way a lot more (and I have noticed RF playing in on him quite a bit – he could get thrown of at first from RF if the fielder is positioned perfectly) and I have read others saying going the other way and up the middle is what they are trying to do with RISP.

        Vote -1 Vote +1

      • Rob says:

        Easy — different hitting coach.

        Players are human beings, not dice. Their performance varies according to human factors, not randomly.

        Question: If the Cardinals DON’T “regress” by the end of the season, will you admit you might be wrong? Or will you just continue to insist the reality in front of you is one you “needn’t accept”?

        Vote -1 Vote +1

  14. Jose Bautista says:

    They have perfected the art of stealing signs while on the basepaths.

    Vote -1 Vote +1

  15. Guest says:

    Doesn’t this article just reinforce the theory that clutch hitting wins games?

    Vote -1 Vote +1

  16. OldDogScout says:

    Tendencies, pitcher/defense adjustments, & hitting lanes as counter strategies all have relevance.

    Thank you, very much for those who contributed thoughtful commentary points vs merely snark or down votes.

    It’s disappointed, however, to see the quality & depth has significantly changed on this board since it’s analytical high point several years ago..

    Would it have mattered for the analytical curiousity & resulting thread commentary if you all knew the request was coming from a mlb club’s front office?

    I’m sure Dave Cameron might be interested in knowing that & how it was received.

    Vote -1 Vote +1

  17. Jeshua says:

    The article mentions that the team is basically the same as last year and can’t explain why they’re hitting better this year with RISP. I would think at least mentioning the fact that they have a NEW HITTING COACH this year could somehow play into that.

    Vote -1 Vote +1

    • Lanidrac says:

      It’s possible and may lead credence to my theory. I think the Cardinal hitters have changed their approach to hit more line drives. Notice how they had five hitters with 20+ HRs last year, and with all five still there and healthy (outside of Holliday’s current minor injury), only Beltran is on pace to do it again. On the other hand, they have three of the top four in the NL in batting average.

      Another way to look at it is that they have a team 24% line drive rate and 6.3% HR/FB ratio this year compared to 21% and 7.7% last year.

      Vote -1 Vote +1

  18. jj says:

    The cards less their SP are batting a combined .289 as of today. If we just look at that and say they are a pretty good hitting team is it really that surprising that they are hitting so well with RISP? Can it just fall under random variation of when they are getting their hits?

    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>