In the midst of what will go down as a disappointing season for the Phillies, an interesting detail about the front office’s thinking appeared. This morsel snuck into a Ken Rosenthal article: “…once the Phillies began to slump, their front office’s thinking was, ‘We don’t want to go all-out for the chance to play in the wild-card game and then face the Dodgers in the Division Series.’”
There are separate discussions to be had about whether that’s a defeatist attitude, or even whether the Phillies could have done more at the deadline. That’s for someone else to decide, though. What this statement got me, among others, wondering was: wait, would you actually rather play the Dodgers in a seven-game series than a five-game one? No one would argue that the Phillies are as good as the Dodgers — they’d clearly be underdogs no matter what. But does the extra chance of avoiding the juggernaut make up for the fact that you’re more likely to win in a shorter series?
To investigate this problem, I worked out a simplistic playoff win probability model. For each team, I took their projected rest-of-season runs scored. Then I projected a playoff rotation and how many innings each pitcher would pitch per game. Using those starters’ projected runs allowed per inning and adding in the projected runs allowed per inning by the bullpen (an admittedly inexact science that involves stripping out starters’ projections from the team’s total runs allowed projections), I was able to produce a runs allowed forecast for each starter on each team. Let’s take a look at the Phillies, for example:
With runs scored and runs allowed projections for each game, I used Pythagenpat to create a win percentage expectation against a neutral opponent for each team. Then, I used the odds ratio method to get an expected winning percentage against a specific opposing team and pitcher rather than a neutral one.
For example, per my method, the Phillies have a .540 expected winning percentage with Aaron Nola pitching, while the Dodgers have a .646 expected winning percentage with Kershaw going. Put those together, and the Dodgers have an expected .608 winning percentage when Kershaw faces Nola. That’s on neutral ground — I then added a 4% edge for the home team in each game.
Now that we have winning percentages for each game, we can simulate the outcome of the series some arbitrarily large number of times and look at the results. The results aren’t pretty for the Phillies, whether they play five or seven against the Dodgers. In a seven-game series, my model sees the Phillies winning a measly 22.1% of the time. That’s pretty bad, obviously, and it assumes they get to use their full pitching staff, with Nola getting two starts.
Face the Dodgers in the NLDS, and there are two offsetting undercurrents. Winning three of five as an underdog is easier than winning four of seven — that’s just math. On the other hand, you face home field disadvantage in a higher proportion of the games and your ace can’t start two games due to the timing of the Wild Card game. Again per the model, this works out in Philadelphia’s favor, if only marginally: the Philles have a 24.6% shot at winning a five-game series with the Dodgers.
But that’s just level one. If the Phillies had to play the Dodgers, they’d rather do it in a five game series, even if it means only getting one turn with Nola rather than two. It’s closer than you think, but duh. That doesn’t correctly frame the choice the Phillies face, though. The Phillies won’t necessarily have to face the Dodgers in the NLCS. There’s always a chance that the unstoppable Los Angeles juggernaut will meet an immovable object — two Jack Flaherty shutouts or a few bullpen implosions, say.
To get a truer sense of the tradeoffs, let’s consider a different question. How likely would the Phillies be to reach the World Series if they won the Wild Card game as it’s currently constructed — a game in Washington, followed by facing the Dodgers in the NLDS? How does that compare to a scenario where they face the Braves in the NLDS, and then the Cardinals/Dodgers winner in the NLCS?
Using the method I outlined above, it’s easy enough to run all of these scenarios. First, the bad news for the Philly faithful. Facing Max Scherzer in an elimination game is hazardous to your playoff health. Scherzer and the Nats project to beat Nola and the Phils 64% of the time on neutral ground, 68% of the time at home. That happens in both scenarios, though, so it won’t affect which option is better — it’s just a fact of life.
Next, we have the NLDS against the Braves. The Braves have a pretty decent postseason roster — three pitchers who move the needle, and a fourth starter who isn’t embarrassing. It sets up like this:
The Phillies’ best shot is in Game Three, with Nola going against Max Fried. It’s not a particularly good shot (my odds see them as 48.5% likely to win, even at home), but it’s better than anything the Phillies could expect against the Dodgers. Altogether, it works out to a 31.3% chance of surviving the series. Combine that with the odds of winning against the Nationals, and it works out to a 10.6% chance of reaching the NLCS; or, in layman’s terms, “not all that great.”
But they might get to dodge the Dodgers, right? Sure, but not as often as the Phillies would hope. The Cardinals are a pretty good team, and they have a star pitcher, but they can’t really hold a candle to the Dodgers:
Maybe you think Dakota Hudson’s projection is a little pessimistic, and that’s fine, but the Dodgers are still a big favorite in the series. The odds ratio method gives the Dodgers a 67.7% chance of advancing. One of the only benefits for the Phillies is that the Dodgers project to need five games in 25.6% of their series wins, which means less Kershaw in the NLCS. As the Phillies only get one appearance from Nola against the Braves, he’ll be rested to start the first game for them no matter what.
With these percentages in hand, I ran the Phillies’ chances against four potential opponents: the full-strength Dodgers, the down-a-Kershaw-start Dodgers, the full-strength Cardinals, and the down-a-Flaherty-start Cardinals. Both teams have home field against the Phillies and are favored, but there are at least advantages to be had. It shakes out like so:
|Opponent||Likelihood||PHI Series Win %|
|LAD, Kershaw Game 3||18.3%||22.2%|
|STL, Flaherty Game 3||11.8%||40.9%|
Bad news, Phillies fans. Shifting the starters around doesn’t make much of a difference. Enough series go seven, and the difference between two good pitchers is so low, that it only adds a tiny amount to their win probability. Add up their odds across each of the scenarios, and the Phillies have a 27.4% chance of winning the NLCS assuming they reach it. That works out to a 2.9% chance of making the World Series if they got to play the Braves in the NLDS.
I’ll save you the walkthrough, but I re-ran the math assuming Philadelphia plays the Dodgers in the NLDS, the way the postseason is currently structured. In that scenario, they reach the NLCS only 8.4% of the time. The Braves beat the Cardinals 60% of the time, and the Phils advance past the winner of that series 33.5% of the time. That works out to a 2.8% chance of reaching the World Series. In the end, it hardly matters whether they play the Dodgers in the NLDS or not — both options are awful.
What if they could win the division, though? Playing the Nationals is a real buzzsaw, one that makes every later round’s vagaries inconsequential. This time, I’ll swap the Braves into the Wild Card and the Phillies into their spot, unlikely as that may seem. In this wild world, the Phillies advance to the World Series 9.2% of the time, more than triple their odds in the scenario where they make the Wild Card game, but still a lot less than you might think.
In the end, the biggest problem facing the Phillies at the trade deadline wasn’t that they had to face the Dodgers in a hypothetical NLDS matchup. Their problem was that they don’t project to be very good, while the Cardinals, Nationals, Braves, and especially Dodgers all looked solid. It’s not particularly sensitive to which teams make the playoffs, either: substitute the Cubs in for the Cardinals and the odds get worse for the Phillies, not better.
After doing all the math, considering the tradeoff between series length, starter schedules, and quality of opponent, the order in which you play teams isn’t very important. The best way to advance in the playoffs is simply to be good. Take the Nationals, for example. They’re in the same bind as the Phillies; win the play-in game, and they’ll face a rested Dodgers team.
But talent matters. Their odds of beating the Phillies and Dodgers back-to-back are higher than 30%. The Cardinals, likely to be the worst true-talent team to make the NL playoffs, project as 13.4% likely to reach the World Series if they win the NL Central, significantly higher than Philadelphia’s odds even if they could magically win the NL East.
The current system penalizes mediocre teams who don’t win their division, no doubt. But that effect is overblown. The real penalty is for being mediocre, not for failing to win the division. You have to play all the best teams in baseball in the playoffs, and if you aren’t very good, it’s likely to show.
Now, these calculations aren’t perfect. I’m hacking together projections and projected win rates; perhaps you could quibble with the exact details. That doesn’t invalidate the general point, however — reaching the playoffs is nice, but being good in the playoffs is the real key to success.
It would be really interesting to hear teams talk in these terms — “We don’t think we’re good enough right now to try to get better.” It probably won’t happen, but next time you hear someone bemoan the structure of the playoffs, imagine them saying that instead. It will probably be closer to the real reason they decided it wasn’t worth it to improve slightly.
Ben is a contributor to Fangraphs. A lifelong Cardinals fan, he got his start writing for Viva El Birdos. He can be found on Twitter @_Ben_Clemens.