Playing Up or Down to the Competition

Over the course of a season, a baseball team will play a lot of games. Some of those will be against good teams! Some of those will be against bad teams. The games against good teams are supposed to be the tests. The games against bad teams are supposed to be the gimmes. There’s a variety of things people say about those games. Those are the games that “good teams are supposed to win.” A team that stumbles might have “overlooked” the opponent, with a tougher one coming up. Then there’s the line about a team that “plays down to the competition.” People have a lot of things to say about losses to bad teams. People have a lot of things to say about sports.

I want to steer your attention to something. For years, Baseball-Reference has provided a split: stats against teams .500 or better, and stats against teams under .500. I think a lot of us have known about this, but I’m not sure I’ve seen the information cited more than a handful of times. It’s a rough split, but it’s a handy split — teams .500 or better tend to be the good teams, and teams under .500 tend to be the bad teams. Let’s settle for the rough split, for now. So, now I want to steer your attention to the Mariners and the Angels.

This is going to be all about wins and losses. Below, a chart of 2014 data, that ought to be self-explanatory. On the y axis, win percentage against .500+ teams. On the x axis, win percentage against worse teams than that. Naturally, the numbers on the x axis are mostly higher than the paired numbers on the y axis.


So, obviously, there’s a relationship. Better teams will win more games against both sets of opponents. Worse teams will lose more games against both sets of opponents. But, look at that red dot, well above the line. And look at the red dot on the far right, separated from the rest of the pack. Here we see the Mariners, and here we see the Angels, and they’ve so far gone in two different directions.

This table might give you a better sense of things:

Team Win% vs. Good Win% vs. Less Good Difference
Mariners 0.594 0.485 0.109
Reds 0.481 0.475 0.006
Tigers 0.541 0.563 -0.022
White Sox 0.448 0.471 -0.023
Cardinals 0.529 0.559 -0.030
Marlins 0.469 0.507 -0.038
Brewers 0.508 0.547 -0.039
Padres 0.459 0.508 -0.049
Twins 0.418 0.468 -0.050
Braves 0.481 0.540 -0.059
Cubs 0.425 0.485 -0.060
Giants 0.509 0.571 -0.062
Rangers 0.354 0.429 -0.075
Yankees 0.479 0.554 -0.075
Astros 0.402 0.481 -0.079
Phillies 0.425 0.508 -0.083
Blue Jays 0.473 0.556 -0.083
Orioles 0.545 0.633 -0.088
Rays 0.438 0.560 -0.122
Royals 0.500 0.623 -0.123
Pirates 0.451 0.582 -0.131
Indians 0.452 0.587 -0.135
Athletics 0.513 0.655 -0.142
Dodgers 0.475 0.625 -0.150
Rockies 0.314 0.478 -0.164
Red Sox 0.382 0.551 -0.169
Diamondbacks 0.319 0.507 -0.188
Mets 0.382 0.571 -0.189
Nationals 0.448 0.658 -0.210
Angels 0.487 0.754 -0.267

The Angels have won just under half their games against good teams, and they’ve won three-quarters of their games against bad teams. They’ve done what you’d expect, except to an extreme degree. All but two teams have been worse against good opponents than bad opponents. The Reds have been basically even. The Mariners have been freaks.

To this point, the Mariners have gone 33-35 against sub-.500 opponents. Yet, they’ve gone 41-28 against superior opponents, and as I write this very sentence, they’re narrowly beating the A’s. The Mariners have a winning-percentage difference of 109 points in the opposite direction from the norm, and they’re separated from second place by more than 100 points. The Mariners would serve as an example of a team that plays up to opponents and down to other opponents. They’ve handled themselves against the A’s, Braves, Royals, and Angels. Within the past couple weeks they’ve lost series to the Phillies and the Rangers.

Going back to 1914 gives us 2,184 team seasons. In 2,023 of those, the given team had a lower winning percentage against .500+ opponents. Only 11 teams ever have finished with a winning-percentage difference of at least 100 points in the unusual direction, like the Mariners are on pace to do. Right now, the Mariners’ difference would rank sixth-greatest, ever. In 2010, the Cardinals finished with a difference of 137 points, which ranks No. 1. Then you’ve got the 1932 Reds, the 1979 Giants, the 1998 Twins, and the 1918 Pirates. The Mariners have done something not unprecedented, but they’ve done something really weird.

Of course, there are facts, and then there are meaningful facts. It’s tempting to want to try to explain this. We want to try to explain everything. The positive spin would be that Lloyd McClendon gets his team amped up for big games. The negative spin would be that McClendon doesn’t sufficiently get his team amped up for lesser games. You can’t determine whether it’s a good thing or a bad thing, and the key point here is understanding randomness. There’s no rational reason why a team might over-perform against quality opponents, and under-perform against weaker opponents. Not to such an extreme degree.

And, yeah, it turns out to be mostly random. Let’s take the 15 teams who finished with the greatest positive differences. They averaged differences of 107 points, in favor of playing stronger opponents. The year before, they were 193 points worse than that. The year after, they were 215 points worse than that.

Now let’s take the 15 teams who finished with the greatest negative differences. They averaged differences of 360 points, in favor of playing weaker opponents. The year before, they were 222 points better than that. The year after, they were 196 points better than that. Those 2010 Cardinals? That year, they were better against good teams by 137 points. In 2009, they were worse by 151 points. In 2011, they were worse by 88 points. The problem with breaking things down by year is that teams aren’t identical in consecutive seasons, but you’d figure that an ability like this wouldn’t completely erode, if it were real.

Instead, it seems to be in the same family as clutch, or hitting with runners in scoring position. There are clutch performances, but there generally are not consistently clutch performances. There are a lot of numbers we can use to describe the past, but only some of those numbers can be used to try to see into the future, and it’s important to understand which numbers go where. Maybe it’s not actually important, for us on the outside, but if you’re going to know something, you might as well also try to know what it means. The Mariners, so far, have been a weird-ass team, and the Angels have been a differently weird-ass team, but you’d still prefer the Angels going forward, because you mostly care about overall performance and certainly no one’s going to go into October unmotivated.

The benefit of numbers you don’t expect is that you have to think about them, and in so doing you might develop some new connections. But not every number means what it seems. Numbers aren’t liars, but sometimes they’re more than happy to just string you along for the hell of it.

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.

25 Responses to “Playing Up or Down to the Competition”

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

    Why not?

    Vote -1 Vote +1

  2. Ken Griffey Jr. says:

    I’m available for parties (cough) dh’ing (cough) and end of season motivational speaking.

    Vote -1 Vote +1

  3. db says:

    This is such a great article. It would be easy to build a narrative around these facts (on either extreme), but the narrative would largely be bullshit as Jeff shows. I assume the comments will be filled with potential narratives and explanations, but the short answer is that stuff happens.

    Vote -1 Vote +1

    • SimonSays says:

      The slightly longer answer is that it’s possible Felix and ‘Kuma faced off against the better teams more often because of the Mariner’s scheduling. Not that a situation like that makes a narrative. “Better pitchers give your team a better chance to win!”

      Vote -1 Vote +1

      • Walter says:

        That was my first thought as I was reading it – while I’m sure luck is a huge factor – I’d be curious to see how teams who are positive outliers like that, look in terms of pitching staffs…are they a “Spahn and Sain and two days of rain” staff (like the Mariners this year), and were they able to arrange their starters such that their good ones pitched against the best teams while the lesser starters pitched against the lesser teams?

        Vote -1 Vote +1

        • Mike Pozar says:

          Yes, would be great to look at what starters faced sub-.500 vs. over-.500 teams. The Mariners in fact have altered Felix’s starts a couple times – once they pushed his start out to face the A’s (instead of the Astros, I believe); most recently they pushed it out 2 days to have him face the Nationals instead of the Rangers.

          Vote -1 Vote +1

        • snack man says:

          Yeah, and Felix was thanking the manager for getting to start vs the Nationals.

          Vote -1 Vote +1

      • Bryan Cole says:

        This is a pretty good idea. From

        Felix: 15 GS vs. teams over .500, 2.10 ERA; 13 GS vs. teams under .500, 2.38 ERA
        Iwakuma: 12 GS vs. teams over .500, 2.92 ERA; 11 GS vs. teams under .500, 2.88 ERA
        Seattle: 41-28 in 69 games vs. teams over .500; 33-35 in 68 games vs. teams under .500

        So three more games. That’s not nothing. If you (naively) turn three of those under-.500 losses into wins, and three of those above-.500 wins into losses, the difference shrinks to .021.

        Vote -1 Vote +1

  4. Stevie O says:

    Curious to know is if this is a product of the Mariners’ great pitching and suspect offense. With a below average offense, they generally hit all pitchers the same(below average). However, their pitching plays any night, keeping them in most games because they don’t have a great offense. The teams in contention in the AL (sans Royals) all have above average offenses while some are great (O’s, A’s, Angels). Does keeping these elite offenses in check, swing the odds in the Mariner’s favor because they are “winning” the matchup by playing a low scoring game against a high scoring team? While the Orioles might not have a significantly above average pitcher, The A’s and Angels do/did (RIP Garret Richards ;( ) but those teams are used to playing with 4-5 runs a night. Does it get in the heads of the Mariners players that keeping the sore low is better for them and gives them a little more confidence because the pitchers are doing their job against great offenses? I guess we can’t know for sure. My guess is that it is random (like you said) that they win the close games against the good teams, but lose them against the bad. But I would like to know if the teams you listed above with swings in winning % over .100 all had great pitching staffs and below average offenses or another combination.

    Vote -1 Vote +1

  5. One Mississippi says:

    Interesting observation is that the Angels just super-swept the A’s, and lost to the Lastros. So five games ago, they would have been even more extreme. Also, running a Monte-Carlo simulation would be a snap given team winning pct, and opponent winning pct, to find out what number of those 2200 team-seasons would finish with better records against worse teams among a statistical distribution.

    Vote -1 Vote +1

  6. joser says:

    As some people have suggested, there is some oddness in the results when we look at starts by Felix Hernandez in particular. The first game started by Felix against a over-.500 team that the Mariners lost was his 22nd start, against Baltimore on July 25. By that point the Mariners had already lost 6 games started by Felix, but they were all against sub-.500 teams (two each against Texas and Houston in a disastrous April, and then another one against Texas and one vs the Padres in June). But by that same point the M’s had won 9 games started by Felix against better-than-.500 teams (four vs Oakland, three vs the Angels, and one against each of the Yankees and Indians). Since then, with Felix on the mound, things have been more “normal”: they’ve won against Atlanta and Toronto but lost to Cleveland, Detroit, and Washington. The lone Felix start since July 25 against a bad team was vs Boston, and they won that.

    Essentially the Mariners win games for Felix when he’s pitching against the best of the AL West (as they did again today), and can’t win for him when he pitches against the worst.

    Somebody with a little more time might want to look at Iwakuma, and also overall run differential vs “good” and “bad” teams.

    Vote -1 Vote +1

  7. Tom says:

    Actually, I’m beginning to think McClendon is a good manager!

    Vote -1 Vote +1

  8. bookbook says:

    It may be random, but it’s frustrating as hell to watch. I’d bet, on a more granular level, that the M’s pitching does what it’s supposed to do–pitch better against worst teams overall. Yet, the offense seems only able to perform against better pitchers. The M’s have been shut down by a fine collection of AAA arms this year. It may be a function of a team-wide lack of patience, which gets in the way of letting bad pitchers defeat themselves. Or…

    Darn it. Random. My human mind isn’t willing to accept it, despite the evidence. Next, you’ll tell me there’s global warming despite the snow I distinctly remember from last winter.

    Vote -1 Vote +1

  9. Evan says:

    The good news is the Mariners have a tough schedule!

    Vote -1 Vote +1

  10. pft says:

    Baseball is such a long season that cumulative W-L records does not tell the story of how good a team is at any point in time.

    Such studies IMO are useful only if they include an adjustment for how well the opponent was playing at a given time. After all, its not always who you play but when. Going up against a 500 team on a hot streak can be tougher than plating a 600 team on a losing streak.

    Then there is the pitching matchups. Who is the better team among 2 true talent 500 teams, the one with the front end of their rotation going for them or the one with the back end. With the unbalanced schedule and 3 game series, such mismatches occur far too often.

    I guess thats part of the Luck that has a 500 team playing +/- 6 games (1 standard error)above or below their true talent level

    Vote -1 Vote +1

  11. PackBob says:

    When even the worst teams have some of the best players that play baseball, this is a great example of randomness.

    Vote -1 Vote +1

  12. Cleveland Spiders says:

    Great, another ‘objective’ Jeff Sullivan article that uses basically one stat and tries to make a conclusion from it. What about pythagorean record, base runs record, using better than a rough split, etc? This article is just another piece of Sullivan propaganda with no data to back up his opinion. Worthless article.

    Vote -1 Vote +1

    • Word says:

      You’re overstating things. This was almost quite a good article. I agree that some of Jeff’s recent posts have been pointless. But this is an interesting idea, and I would have loved to see it explored more fully, to see how much is explained by pythagorean record, base runs record, etc.

      I suspect your criticism would be more effective if you expressed it with a bit more civility.

      Vote -1 Vote +1

  13. DD says:

    I thought this was going to be an article about the greatness of Jack Morris.

    Vote -1 Vote +1

  14. bjoak says:

    I think this is a nice start. It would be interesting to see correlations of what the good against .500+ have (defense, hitting, etc.) vs. what the bad have. Yes, the outliers have a lot of random noise, but there are 50% above and below the line every year with thousands of team seasons. There should be some conclusions there.

    The good pitching beats good hitting theories others mentioned are valid theories that should be explored. I would think there are rational reasons why this could happen. Here’s another: since neither pitchers nor hitters have a terrible amount of control in where a ball is hit into the field, one would think a really good defensive team would do an equitable job of getting to balls against both good and bad teams, therefore having a better than normal record against good teams.

    There’s just too much here to chalk it up to “no rational reason.” It could really be a much better sample to tell us what types of teams would do better in the post-season.

    Vote -1 Vote +1