Panic! At The 12-Game Mark
Here’s a question that you, the fan, might have in a few weeks: My team is 4-8. Is it time to get worried? In a regular 162-game season, that’s not exactly a puzzle: nope! Things are fine, my friend. Sip a frozen drink, or crack open a bag of chips, or whatever it is that you do to relax. Maybe watch a baseball game!
In a 60-game season, things will feel more consequential. That’s convenient, because they’ll also be more consequential. There’s a saying that you can’t make the playoffs in your first month, but you can miss them. This year, that’s out the window. Start out 9-3, and you’re feeling pretty good. Start 3-9, and it’s time to start booking remote cabins for your October socially-distanced vacation.
I wanted to be a little more quantitative than “it’s important to start well.” My first thought was: let’s fire up the Playoff Odds Calculator (an arcane table in David Appelman’s basement that spits out the odds of each team making the playoffs after he mutters some incantations) and give a team a hot start to see what it does to their odds. After that, we can fire up the ZiPS-ulator (a sentient robot living in Dan Szymborski’s attic) and compare.
There’s a problem with that plan, though: it depends on the messy vagaries of our estimation of a team’s strength. A 4-8 start is less painful for the Dodgers than the Diamondbacks because we think the Dodgers are better. The same goes, to a varying extent, for every team in baseball. The schedule matters, but so do our idiosyncratic views on team talent.
Instead, I wanted to create a true neutral scenario, one that you can apply in a general way without specifically digging into what we think of your team. I built a bubble baseball universe that consists of only a single five-team division. Each team has the exact same talent level, and there’s no home field advantage. That means that a 162-game season would, on average, look like this:
Team | Record | Games Remaining | Playoff Odds |
---|---|---|---|
1 | 0-0 | 162 | 20% |
2 | 0-0 | 162 | 20% |
3 | 0-0 | 162 | 20% |
4 | 0-0 | 162 | 20% |
5 | 0-0 | 162 | 20% |
Want another boring and uninformative chart? Here’s a 60-game season in the same format:
Team | Record | Games Remaining | Playoff Odds |
---|---|---|---|
1 | 0-0 | 60 | 20% |
2 | 0-0 | 60 | 20% |
3 | 0-0 | 60 | 20% |
4 | 0-0 | 60 | 20% |
5 | 0-0 | 60 | 20% |
A few quick notes: I used actual scheduling rules. In the 162-game season, each team plays each other team in the division 19 times, and the balance come against out-of-division opponents (our teams are also .500 teams against them). In a 60-game season, they play each division opponent 10 times and then 20 games against out-of-division teams.
With this generic model in place, we can totally ignore any specific pronouncements of team strength. Imagine a 162-game schedule after 12 games split evenly among division opponents. Give one team a two-game lead, say 9-3 against a second place team that goes 7-5. Then, let randomness reign:
Team | Record | Games Remaining | Playoff Odds |
---|---|---|---|
1 | 9-3 | 150 | 34.3% |
2 | 7-5 | 150 | 23.1% |
3 | 6-6 | 150 | 18.6% |
4 | 5-7 | 150 | 14.9% |
5 | 3-9 | 150 | 9.2% |
What I mean by “letting randomness reign” is that I simulated a massive series of coin-flips. Every division game is a coin-flip that hands one team or the other a win. Every non-division game is also a flip; the only difference is that if the non-division team wins, that win just disappears into the ether. We’re not modeling the other divisions, after all. No Wild Cards, nothing else: this is just a simple “Can you win the division?” simulation. I simulated an entire season that way, then repeated the simulation a million times to produce the odds seen above.
In that pure randomness world, a hot start really does increase your chances of making the playoffs. It’s not a lot, but it matters. Getting ahead is good, and one team falling meaningfully behind is good as well. Even a one-game lead helps out a decent amount:
Team | Record | Games Remaining | Playoff Odds |
---|---|---|---|
1 | 8-4 | 150 | 28.7% |
2 | 7-5 | 150 | 23.5% |
3 | 7-5 | 150 | 23.4% |
4 | 5-7 | 150 | 15.1% |
5 | 3-9 | 150 | 9.3% |
Actual divisions won’t be pure tossups, naturally, but it’s still a useful mental model. The fewer teams you have to battle for supremacy, the better. A few games in hand matters too — a two game lead in April isn’t the same as a two game lead in August, but it’s not meaningless.
Now that we have a baseline for 162-game seasons, let’s go to 60. We’ll take the same inputs as before and flip the same coins, just fewer times. First, a two-game lead:
Team | Record | Games Remaining | Playoff Odds |
---|---|---|---|
1 | 9-3 | 48 | 44.7% |
2 | 7-5 | 48 | 23.6% |
3 | 6-6 | 48 | 16.4% |
4 | 5-7 | 48 | 11.0% |
5 | 3-9 | 48 | 4.4% |
Yeah, uh, it’s good to be ahead early in a short season. No surprises there, I suppose. I had hoped for some kind of countervailing effect where the increased number of divisional games dulled the advantage of being out front, some counterintuitive gem. But the impact of a short season is simply too great. If you’re up early, you make the playoffs a lot. Even a one-game lead over two teams is valuable:
Team | Record | Games Remaining | Playoff Odds |
---|---|---|---|
1 | 8-4 | 48 | 34.4% |
2 | 7-5 | 48 | 24.7% |
3 | 7-5 | 48 | 24.7% |
4 | 5-7 | 48 | 11.6% |
5 | 3-9 | 48 | 4.7% |
A 10 percentage point edge might not sound like much, but for a single win, that’s valuable. Let’s extend things slightly to the 20 game mark; for our purposes, that’s one more game against each divisional rival and four games against out-of-division teams. Three games is pretty juicy:
Team | Record | Games Remaining | Playoff Odds |
---|---|---|---|
1 | 14-6 | 40 | 54.3% |
2 | 11-9 | 40 | 21.1% |
3 | 10-10 | 40 | 14.5% |
4 | 8-12 | 40 | 6.3% |
5 | 7-13 | 40 | 3.8% |
To be clear, this is with evenly matched teams. It’s just hard to hunt down someone ahead of you. After a third of the season played in this way, each team still has to play each other division rival six times; you’d think that the second- and third-place teams would win their fair share of races just by doing well in their head-to-head matchup with the division leader. But a three game lead with 40 to play is simply tough to match.
This isn’t exactly how things will work. Teams aren’t evenly matched, for one. If the Giants start out 9-3 and the Dodgers 7-5, the Dodgers will still be favorites to win the division. But it’s still instructive, and in a general sense, we can do a little equating. A two game lead 12 games into the season gives a team a 44.7% chance of winning their division in a 60-game season. That’s the same as a two-game lead after 114 games in a regular season, but that’s not a helpful comparison; after 114 games, you probably have a strong sense of how your team feels this year. After 12 games, you’re still in the dark.
Consider this possible start in a 162-game season instead:
Team | Record | Games Remaining | Playoff Odds |
---|---|---|---|
1 | 11-1 | 150 | 46.1% |
2 | 7-5 | 150 | 21.3% |
3 | 5-7 | 150 | 13.6% |
4 | 4-8 | 150 | 10.8% |
5 | 3-9 | 150 | 8.3% |
That makes the math easier. A two game edge early in this abbreviated season should feel roughly like a white-hot 11-1 start in a regular season. It doesn’t mean you’re bound to make the playoffs, or anything of that nature, but it’s more than a small boost to your odds. No one is holding a coronation for an 11-1 team, but it would certainly be a storyline. That’s the same as a 9-3 team this year, or an 8-6 team in a division where no one else is above .500. And don’t even get me started on what happens if you start out 11-1 this year.
To sum up: starting out hot is always better than you think it is. It’s especially important this year in the abbreviated season; more so than you probably think it is. Don’t let anyone tell you that a few games doesn’t matter. It might not, of course. These are all probabilities, and none are locks; leads can crumble. They can also not crumble, however. In a short season, every game counts, even those pesky early ones.
Ben is a writer at FanGraphs. He can be found on Twitter @_Ben_Clemens.
Great piece!
I’ll admit I assumed, as I’m sure most readers will, that the “12 game Mark” the title referred to was the 12 games Australian RHP Mark Hutton appeared in for the ‘96 Yankees, but I was pleasantly surprised by the totally unexpected direction the article took.
This is probably the greatest comment I have ever read.