How The Cubs Went 124-33 Last Year (sort of)

In Uncategorized by myles48 Comments

If you want to be technically correct (the best kind of correct), the Cubs went 103-58-1 last year in the regular season. That said, I was curious what their record was in each 7 game tranche (as that's the number of games in the World Series. The Cubs won the World Series in 7 games last year.), and since it's easy to build that architecture to expand to 5 game series, I did that as well.

The Cubs crossed the 3 wins (or losses) threshold on game 3 of the season, and after that, I went through the entire season and calculated each time the Cubs had won at least 3 of the last 5 games played (and 4 of the last 7 games played). The game the Cubs tied, I just disregarded altogether. 

Rolling 7-game series record: 124-33
Rolling 5-game series record: 125-34

Something interesting to note: the Cubs didn't lose a 7-game series until game #37, where they lost both games of a double header (#32 and #33) and then lost two in a row to Pittsburgh and Milwaukee. Game #36 was where the Cubs lost their first 5-game series (same series of games: #32-#36, obviously).

There is a purpose to this analysis, and that's an attempt to calculate how evenly distributed the wins are. Imagine a team that plays 100 games, and wins 70 of them. That seems like an extremely dominant team; on the whole, that's a great winning percentage (113 win pace). However, if that team won the first 70 games, and then lost the last 30, you can easily imagine catching this team in a "cold spell" and sweeping them. My eternal crusade in sports analysis is emphasizing just how important variance is, and how maximizing it/minimizing it is so incredibly important and underpins nearly everything a team does. Good teams want as little variance as possible, and the Cubs were a very even team last year. To put this into contrast, the Indians won 94 games last year, but in 5-game series they only improve to 99-58, and 102-53 in a 7 game series. 

We can actually measure how evenly distributed a team's wins are by calculating how many n-game series victories you'd expect given a win percentage. For instance, the Cubs won 64% of their games last year (.639752, to be slightly more precise). That means that in a 5 game series, you'd expect them to exactly 3 games (.64)^3 * (.36)^2 * [(5!)/(3!)*(2!)] = 34% of the time. You can expand that to x out of n in a formula I'll spare you, and you can expect a 64% win rate team to win approximately 75% of it's 5 game trials. That's 119 wins in 159 trials, and the Cubs won 125. That means the Cubs distribution of wins was better than average, exactly what you want from a high performing team. 

  5 game delta 7 game delta
Cubs 6 1
Indians -5 -5

In a 5 game series, the Cubs were distributed slightly more evenly than expected, and the Indians were slightly less evenly distributed. Oddly enough, this was EXACTLY what both teams would have been trying to do in this instance. The Cubs were a better team (strictly based on winning percentages in the regular season), so they want those wins spread out as evenly as possible because they are favored in any given game. The Indians want to bunch their wins together, because their path to victory is to be on a hotter streak than the Cubs are, and the more spikes you have, the more likely you are to hit on one.

The playoffs are a collection of 1, 5, and 7 game playoffs. It's sort of unfair that 162 games can be superseded by these series. However, that's what happens, and the Cubs were well prepared for it.  

 

Share this Post

Comments

  1. Rice Cube

    This was an analysis I recently enjoyed. And part of why we remained perhaps irrationally optimistic despite being down 3-1, due to Cubs ability to rattle off 3 wins in a row throughout the season.

      Quote  Reply

  2. cerulean

    Hendricks struck out Freese on the third pitch—nine pitches and many foul balls later, he double to give the P’s the lead.

    Robot umps please.

      Quote  Reply

  3. cerulean

    Perkins,

    I noticed that too. I know he has said that he purposefully takes some off to get better movement (or perception of movement), but 85mph is a lot slower than last year’s 88mph that was a couple ticks slower than the 90mph he could hit.

      Quote  Reply

  4. Perkins

    cerulean,

    I’d feel more comfortable if he started dialing it up a bit in later innings as Verlander used to do. Especially since guns read a bit hotter than last year with the change in data sourcing.

      Quote  Reply

  5. cerulean

    Cole’s velocity is also down a couple ticks. I wonder if the radar system has a lot more variance than the previous one by dint of being new.

      Quote  Reply

  6. Rizzo the Rat

    Wait, Duensing was tasked with retiring three straight righties? The maximum number of consecutive right-handed hitters he should face is zero.

      Quote  Reply

  7. Perkins

    I wonder if Arrieta and Hendricks are facing a dead-arm period or something. In Arrieta’s case, the results have been there, and Hendricks looks to have decent movement and command, even if the results aren’t what you’d hope for.

    I’m just spitballing, but it seems reasonable, especially in light of last year’s World Series run.

      Quote  Reply

  8. cerulean

    By the W, Myles, this analysis is pretty cool. A season standings updated every week (if you could automate it mostly) would be interesting to see on this site. Also, confusing for newcomers when they see obstructed standings that seem way off. I feel botching an apparent credit to the site would discredit it enough to be on this Web Log.

      Quote  Reply

  9. cerulean

    Jeimer Candelario has only one single in 33 plate appearances. He’s certainly no Tony Campana.

    Now, one would guess that he is in a slump, but his OBP is over .400 and his SLG is more than double it. Finally, the Cubs have some promise at 1B/3B.

      Quote  Reply

Leave a Comment