A Series Of Possibilities
The Cubs will likely be favored in every series they play on the way to the World Series. But their toughest opponent this October might just be probability.
This has been a season for the ages for the Chicago Cubs.
The team has its largest division lead in more than 100 years, the possible National League Most Valuable Player in Kris Bryant, and a trio of possible Cy Young candidates (the award for best pitcher) in its rotation.
They cruised to a postseason berth, and will likely be favored in every playoff series they play (depending what you think about the Boston Red Sox). By many estimates, they’re favorites to win the World Series.
And by favorites, that’s around a 1-in-4 chance of winning. Or if you’re a Cubs fan, a 3-in-4 chance of losing.
Even taking into account all of the Cubs advantages — the superstar lineup, stacked pitching staff, and homefield advantage through the National League playoffs — it’s more likely than not that Cubs fans will be disappointed again this year. Not because of goats or black cats, but probability.
Baseball Prospectus writer Russell Carleton laid out the math in a 2014 article. According to his figures, to even have even odds to win, a team would need to have the talent to win 113 games (the Cubs finished with 103).
All else being equal, with eight teams in the playoffs after the two Wild Card games, the Cubs have a 1-in-8 chance. But even taking into account just how good the team is, that only jumps to around 1-in-4 according to the projections at Baseball Prospectus and 538, while fellow stats site Fangraphs has them closer to 1-in-5.
The sites runs thousands of simulations to come up with their figures, taking team strength and other factors into account. Even with that, the Cubs only get up to around 25 percent.
“You have to assume the Cubs are huge favorites in every series to get them to 25 percent,” said Sam Miller, a former editor-in-chief at Baseball Prospectus and current writer at ESPN. “The problem with baseball, the problem for the Cubs, is that baseball’s not like football, you never have a team that’s 95 percent likely to win a game. Even if you put the Cubs up against a last place team in a five-game series, which is a short series, you’d have that last place team winning some of the time. And they’re not facing any last place teams once they make it to October.”
To test that idea, we built our own simulation (see how we built it here). Here’s what the playoffs would look like if the projected favorite won each series, and their chances of doing so in each round.
The Cubs would be favored at each step, but only win the World Series in 25 to 30 percent of the trials.
The problem with the real world is that teams don’t get thousands of trials, just one. While the Cubs have given themselves as good a chance as they can, all anyone will remember is how this single one plays out.
What does it mean that the Cubs only have a 1-in-4 chance? To test that out, we built a simulation you can run yourself. At the click of a button you can play your own version of the MLB playoffs as many times as you want.
What makes baseball unlike a sport like basketball, where 538 gave the Golden State Warriors a 42 percent chance to win the NBA Championship?
“Baseball is at its core, a game that is heavily driven by luck,” Miller said. “By baseballs landing where nobody is standing for no particular reason and almost by accident. This sort of random fluctuation is baked into it more than any other sport, and seven games aren’t enough to wipe that away.”
That means there’s very little a team, even one as good as the Cubs, can do to improve their playoff chances. Random occurances — say a unheralded second baseman hitting home runs in every game — can sink even the best of clubs.
Still, Miller said he’s optimistic about the Cubs’ chances of winning eventually. The team has a good young core of players, a strong front office and plenty of cash and depth to acquire any missing pieces.
“If you ask me, ‘Are the Cubs likely to win a World Series this year?’ I’ll say they’re still an underdog to the field,” Miller said. “But are they likely to win in the next five years? I’d say overwhelmingly they are. They are definitely over 50 percent to win a World Series in the next five years, and that’s a good position to be in.”
Miller’s main piece of advice for Cubs fans is not to live and die by a World Series title alone, which he says from experience as a San Francisco Giants fan. Appreciate how special this team has been and savor every moment.
“This is a team that has been remarkable in so many ways,” Miller said. “You’re watching future Hall of Famers right now playing at age 22 and you get to see them from day one knowing they are going to be stars, and that’s a privilege not a lot of teams fans get to experience.
“The moment of peak joy a championship provides is irreplaceable, and I don’t begrudge any fan wanting that desperately; it’s what we work for. But the day-in-day-out experience of being a Cubs fan has improved a lot of people’s serotonin levels this year whether they know it or not. I don’t know that there are three teams in my lifetime that I would rather be a fan of than this team, and that’s true whether or not two unlucky bloops go against them and knock them out of the NLDS this year.”
Update Nov. 2: Here's how the 2016 MLB Postseason played out
How did we make this?
Even before the release of Moneyball, baseball fans have loved data and statistics. Because of that, public researchers have been incredibly generous in sharing their findings and methods.
In building this simulator, we started by virtually pitting teams against each other using the log5 method created by the legendary Bill James, a father of Sabermetrics. The formula estimates how often one team will beat another based on their season winning percentage.
Initially we only used team winning percentage, but eventually switched to BaseRun winning percentage at Fangraphs. The measure estimates a team’s expected runs scored and allowed based on its component statistics (singles, doubles, triples, etc …), and then uses that to estimate an expected winning percentage. This was used because it provides a better estimation of a team’s true talent level, and not simply because it makes the Cubs look even better than their actual won/loss record.
For each game, a random number was generated to determine the winner. Games were played until one team won each series, advancing to play the next until a World Series winner was crowned.
Eventually we switched to use an odds ratio method, as described here by The Hardball Times. This allowed us to incorporate homefield advantage, defined as a 54 percent winning percentage. We applied a series-wide homefield advantage to each game after computing the expected BaseRuns winning percentage.
Series probabilities shown in the first bracket were generated using a cumulative binomial distribution, using the single-game probability computed as described above.
Baseball icon by Grant Fisher from the Noun Project