It’s been a pretty exciting pennant race for a handful of lucky baseball teams. The Red Sox got to play spoiler. The Astros got to play spoiler. The Marlins got to play spoiler. The Reds got to play spoiler. The Rays got to play spoiler. The Diamondbacks got to play spoiler. The Phillies got to play spoiler. The Rangers got to play spoiler. The Rockies got to play spoiler. The Cubs got to play spoiler. The Braves got to play spoiler. The White Sox got to play spoiler. The Padres got to play spoiler. The Twins got to play spoiler. The Mets got to play spoiler. Last night, the Blue Jays got to play spoiler.

But for most teams, the baseball season is a cruel march toward irrelevance, and as we anticipate the final weekend of the season, there is no longer any real suspense. All the playoff teams are basically set, due in no small part to the heroes of the American and National leagues: the spoilers. Which raises the question: Is this all a bunch of bullshit?

We crave meaning in life, and we perhaps crave meaning most lustfully when life appears meaningless. So we tell the story of the spoilers, the teams who are just playing out the string but have one final opportunity to affect history. Now, one might note that there is more than one way to screw over a pennant contender; one way is to beat them, while the other, equally effective way is to lose to their opponents. But one might not note that, too. One might prefer the story of the spoiler.

Let’s believe that story for a moment. The story of the spoiler suggests first one very basic way of looking at the sport: that incentives, not just abilities, decide who will win. Sabermetrics has a close relationship to economics, and economics loves the incentives, and yet we generally refuse to acknowledge a lot of these types of incentives. That’s not because they don’t exist or they couldn’t possibly move the needle on performance so much as they seem dwarfed by the very real incentives that always exist for big leaguers: many millions of dollars that are dependent on how well a player produces personal statistics. A beer leaguer might see his performance suffer from lack of interest when he’s trailing by 14 runs, but a big leaguer knows that every RBI counts the same toward his final statistics. If those overwhelming market-based incentives exist (generally) for all players, then we might not feel the need to look for other incentives.

But the story of the spoiler argues just the opposite. In this narrative, as my podcast co-host put it on Hang Up and Listen recently,

You get this strange belief in the spoiler team at this time of year that’s been bad all year but it’s suddenly dangerous. How spiteful are we saying these players are? That they haven’t been motivated by their own chances of winning all year but suddenly…

In the spoiler story that posits extra danger playing a team that’s extra out of it, the assigned incentives seem almost certainly a leap too far. Playing spoiler probably can’t make a bad team good, for the reason that Ben pointed out: There were unlimited incentives already for them.

But what about the spoiler story that posits merely that teams that are extra out of it are at least as dangerous as they were when they were in it? In other words, that this story keeps teams that suck from being teams that roll over and play dead? An old baseball truism is that you can’t take any big-league team lightly, that any big-league team is dangerous in any given day, and what the spoiler story tries to do is justify that belief all the way to the end of the season.

Now, back to the incentives: If we believe in incentives, we believe that a team that is in a pennant race would have more incentives to win, that they might be playing extra hard and doing whatever it takes to get the win. Surely, they’ll be trying hard in April and June, too, but the season is long and the job is a drain and you can certainly accept a story that says being in a pennant race might lead teams to get a better night’s sleep, to take their batting practice a little more seriously, to be especially focused for every pitch on defense, to bust extra hard out of the batter’s box. If one team is giving 100 percent and the other is only 90 percent, then the team that’s trying ought to prevail more often. So: Maybe being a spoiler just means matching the 100 percent?

These are the terms here. Now, onto the work.

Warning! Simple Mathematical Details Befitting A Non-Mathematician Like Myself Ahead!

We pulled every game played in the past month, going back to August 23rd. To beef up our sample, we pulled every game played after August 23rd last year, too. For each game, we found a likelihood of each team winning, using each team’s expected winning percentage (which, to refresh your memory, is schedule- and league-neutral; it is essentially the site’s best estimate of true talent level, using observed performance, PECOTA projections, and updated depth charts), then adjusting for home field advantage (based on the league wide 53.7-46.3 home-field edge over the past five seasons) to get a true true talent level for each team in each game, then running log5 to find the chances that the home team would win the game.

We broke teams into three categories:

  • Spoilers were teams that entered play with less than a 1-in-200 chance of making the playoffs.
  • Concerned teams had adjusted playoff odds between 15 percent and 97 percent.
  • Unconcerned teams had adjusted playoff odds of 99.5 percent or higher—in essence, not only guaranteed to make the playoffs but guaranteed to win the division.

(Some teams, you’ll note, would be excluded. We wanted every team to essentially know which category they were in. A team with 8 percent playoff odds might consider themselves contenders or spoilers, depending on the day and the organization.)

So what happens when a spoiler plays a concerned team?

  • In games at home, spoiler “should” have won 68.4 games. They won 64.
  • In games on the road, spoiler “should” have won 44.2 games. They won 48.

I broke them up by home/road because I anticipated the possibility that, late in the season, teams otherwise out of it might be more focused while playing in front of their home crowd. But that was a non-issue. Totals: 112.6 expected wins, and 112 wins. Spoilers played exactly as our expected winning percentage + log5 predicted they would.

So does that mean the spoilers are keeping pace with the contenders’ elevated interested in the game? In other words, elevating their own interest? Next we compare their expectations vs. performance against non-concerned teams, those teams that have already locked up a postseason spot.

What happens when a spoiler plays a non-concerned team?

  • In games at home, spoiler “should” have won 28.7 games. They won 26.
  • In games on road, spoiler “should” have won 20.6 games. they won 25.
  • Totals: 49.3 expected wins, 51 wins. A seemingly insignificant difference.

But wait! Maybe these spoilers are now merely lowering their effort level to match the lowered effort level of the already clinched. So the last thing we need to do to finish this circle is see whether the unconcerned teams’ lack of incentives leads them to underperform against concerned teams, who have elevated incentives. After all, it’s much easier to see that teams that have clinched don’t really care about playing spoiler: See, for instance, the Angels and the Mariners last week, when Mike Scioscia not only filled out a starting lineup with no regular starters, but sent Tony Campana up as the tying run in the ninth—with Mike Trout and Albert Pujols, among others, available on the bench. If these teams are playing with de-elevated levels of interest, and contenders are playing with elevated levels of interest, and level of interest matters in a sustainable way, then we’ll see it here. So: what happens when an unconcerned team plays a concerned team?

[Results not found]

Problem is, there aren’t nearly enough of these to create even a moderately small sample. There’s like 17 games over the past two years. It’s pointless to try.

So we know this: Out-of-it teams play about as well as they should be expected to even when facing teams that are trying like hell to make the playoffs. And out-of-it teams play about as well as they should be expected to even when facing teams that have absolutely nothing to play for other than keeping fresh.

What’s this tell us? It could tell us incentives are clearly at work: The teams that need to win elevate their game, and those honorable spoilers fulfill their role admirable. More likely, to my eye—though not clearly enough to say there’s proof—it tells us that the incentives just don’t matter, or don’t matter enough to matter. Does it mean they’re not real? Not necessarily. But if Mike Trout, maybe the best player we’ve ever seen, can only swing one win every 20 or so, how much do we think that better night’s sleep, that extra serious batting practice, that little bit of added focus on defense, that hard sprint out of the batter’s box can swing? Your answer before you read this article was probably “not as much as Mike Trout,” and nothing we’ve found today will change your mind.

Sorry, spoilers. You're not special.

Thanks to Rob McQuown for research assistance.

Thank you for reading

This is a free article. If you enjoyed it, consider subscribing to Baseball Prospectus. Subscriptions support ongoing public baseball research and analysis in an increasingly proprietary environment.

Subscribe now
You need to be logged in to comment. Login or Subscribe
San Diego starter Ian Kennedy surprised reporters today when he announced the Padres plan to play spoiler in this season ending weekend series with the Giants. "We had a team meeting and we decided we really hate the Dodgers a lot more than the Giants. So we're going to roll over," the starter for the series opener announced. "I'll just be grooving fastballs down the middle of the plate all night." When asked about the team's motivation during the 2-1 series loss to the Dodgers earlier this month, Kennedy remarked, "Yeah, we really wanted to beat those dicks, but if you think about it there's only so much you can do to win. Losing, on the other hand, you have a lot of control over." When Padres manager Bud Black was asked about benching some of his players to teach them the right way to play, Black said, "We've really only got one decent hitter, so once I sit Smith down, I don't really have to worry about our offense accidentally scoring too many runs. I just hope we can make Puig cry. That would be a great finish to the season."
"Warning! Simple Mathematical Details Befitting A Non-Mathematician Like Myself Ahead!"

Can't. Stop. Laughing.
What kind of pompous fool would use a phrase like that?
Spoils and the Spoiled Spoilers who Spoil them!
A Good Spoil Spoiled
I realize that this article and research were mostly tongue and cheek, but when you are dealing with a sample size of 100 some odd games and looking at win/loss records, it is impossible to even come close to ruling out anything other than the most unlikeliest of hypotheses.

Let's say that I hypothesize that the effect I am looking for is 5%. So, in 100 games, rather than winning, say 50 games, a group of teams is supposed to win 55 games, according to my hypothesis. Unless I find something like a winning percentage of 65%, there is nothing I can really conclude from a statistical perspective. If I get a 50% result, that could easily be a Type II error. Easily. If I get a 55% result - easily a Type I error.

Basically, when looking for small effects in small samples, don't even bother. Seriously. You are more likely to mislead yourself and your readers than you are to find anything reliably interesting.

For example, your results in 100 and 200 games tells us nothing of interest. Before you even get your results, you know that a result of "expected = actual" is not going to tell us anything interesting other than, "if there is an effect, it is likely not really large." Well, did we really think it was large going in? Again, looking for small effects in small samples. Exercise in futility.

I learned that from doing hundreds of these types of analyses. I've done so many of these that I get results all over the place (as expected) and I long ago realized that I could never make heads or tails of them for a lack of statistical significance one way or another (either accepting or ruling out the null hypothesis).
Creating a nice pattern of hyperlink vs regular text in that first paragraph. Great job!