THE POWER RANK

Championship series win probabilities for the 2014 MLB playoffs

By Leave a Comment

boy_hitting_foreheadIt’s an embarrassment, but I’m not going to hide from it. All the favored teams by my numbers (and the markets) lost in the division series.

And it’s not like the favored teams lost in a game 7 coin flip. The favored teams won two games (Dodgers over Cardinals, Nationals over Giants).

As Billy Beane said about his analytics in Moneyball, “my shit doesn’t work in the playoffs.”

Someone tweeted that quote at me before the division series started. I blew it off. If you’re betting on heads, you want the coin to come up heads 52% instead of 50% of the time. Always.

But Billy’s words have new meaning after the division series. It’s not that analytics don’t work in the playoffs. It’s that we should appreciate the randomness of a short series.

Let’s also remember that the team that wins the World Series might play more playoff games than a college football team does all season.

Here are numbers for the championship series.

  • San Francisco has a 52.6% chance to beat St. Louis. No, there is Cardinal Devil Magic in this prediction.
  • Baltimore has a 56.6% chance to beat Kansas City.

These win probabilities start with my MLB team rankings, which take run differential and adjust for strength of schedule. Also, for the first time, I adjust for cluster luck based on the regular season.

In addition, the projections consider starting pitching through xFIP, an ERA type statistics that captures the skill of a pitcher through strike outs, walks and fly ball rate.

Daily predictions for each game appear on the predictions page.

As of noon Eastern on October 10th, the markets give both San Francisco and Baltimore an implied odds of 55.1%.

Is St. Louis safe from an upset against Pittsburgh? A 2013 MLB playoff preview

By Leave a Comment

mlb2013_baserunsThe St. Louis won more games than any other NL team this season, winning the Central by 3 games over Pittsburgh. Moreover, they’re ranked 2nd in The Power Rank while Pittsburgh is 9th. It should be easy to call a series win for St. Louis.

Not so fast.

To dig deeper into these two teams, consider the idea of cluster luck. Some teams tend to cluster their hits together and score more runs. Other teams spread their hits out over the innings and score fewer runs.

One can quantify this luck using run creation formulas. These equations take box score statistics such as at bats and hits to estimate the number of runs a team should have scored. Deviations from this expectation are random.

You can read more about this in my cluster luck article on bettingexpert.

Pittsburgh at St. Louis

Cluster luck has some dramatic consequences on this series.

This season, the Cardinals have scored 60 more runs than expected, almost 3 standard deviations away from the mean. They scored more runs than any teams besides Boston and Detroit despite having average power numbers.

While the Cardinals have gotten lucky, the Pirates have not been so fortunate. They have scored 34 fewer runs than expected. Moreover, their league leading 577 runs allowed has not been the result of luck. This remarkable total is only one less run than expectation.

The visual above accounts for cluster luck by ranking teams by expected runs scored minus allowed. St. Louis only leads Pittsburgh by 10 runs over the course of the season.

To see the rankings of all 30 teams, click here.

The gambler’s perspective

While I originally looked into run creation a few years ago, Joe Peta, author of Trading Bases, inspired me to get back into the analysis. He has his own way of quantifying cluster luck that seems consistent with my work.

Joe thinks the Cardinals still have an edge. His preview shows that the Pirates pitching and defense was spectacular during the first half of the season but only average the second. In a 5 game series, he’s predicting Cardinals in 4.

Why should you trust his prediction? Joe took his methods to Las Vegas and turned a 41% profit. To read more, buy his book Trading Bases.

Outlook

This series has a special meaning to me since two of my best friends root for the opposing teams.

I texted the Pirates fan a few weeks back about this cluster luck analysis. When the Pirates started to lose the division in the season’s last week, I got a death threat.

Cluster luck had an even more drastic effect on the Cardinals fan. Despite his Ph.D. in Mechanical Engineering from Stanford, he texted this back.

You have just made me doubt all numbers ever.

Nothing like analytics to make baseball fans crazy.

When baseball executives meet physicists.

By Leave a Comment

What happens when baseball executives show up at a physics and computer science conference? I had the opportunity to find out this past Saturday at the Sportvision Pitchf/x conference in San Francisco. Sportvision might be the most recognizable small company in Silicon Valley, as they make the ubiquitous yellow first down lines on American football broadcasts. As a company with expertise in sensors and software, Sportvision has embarked on a new mission to making a digital record of every major league baseball game. This effort started with Pitchf/x, a technology that tracks the trajectory of pitches. The data for every pitch thrown in the majors in the last three season is available online. The baseball blog community has made a huge contribution in analyzing this data, and Sportvision highlights this work at their conferences. Many of the talks on Saturday focused on Fieldf/x, Sportvision’s new technology with the more ambitious goal of tracking every player and event on the baseball field.

It was a fascinating cast of people who attended the meeting. Greg Moore, an almost impossible combination of Southern California baseball jock and computer programmer, gratiously invited me. As part of the business team at Sportvision, he had the unenviable task of giving the last talk of the day. However, he managed to wake up the crowd after eight hours by promising a Mindf/x technology by 2015. One of the first people I met was Doc Rosenthals, who practiced neurosurgery for 29 years before quitting to pursue his passion for baseball. He and his collegue Fred Vint have started Scientific Baseball, a company that uses modern technology in youth baseball development. The man who organized the talks was Alan Nathan, the physics baseball guru. Alan retired from the Physics department at the University of Illinois Urbana-Champaign to spend more time on his physics of baseball research. In an excellent talk, carefully tuned to the non-physics audience, he explained his most recent results in ball bat collisions. Then there was Allison Binns, the young Harvard PhD in Sociology who left academic life to work in the Boston Red Sox’s front office. We met later that night when the entire conference attended the Giants game. In a quintessential San Francisco moment, she and her colleagues used the WiFi to agonize over every pitch of the Red Sox game playing on an iPad.

The conference also provided a great chance to tell people about the Power Rank. Let’s check in and see how the second half predictions have worked out. The Rays and Yankees are tied atop the AL East, with Boston 7 games behind. It would be nearly impossible for the second place team in this division to not get the Wild Card into the playoffs. In the AL Central, Minnesota has a comfortable 4 game lead over the Chicago White Sox. While the Power Rank forecast did well in these two divisions, it is failing miserably in the NL Central. Here, 15th ranked Cincinnati has a 6 game lead over 8th ranked St. Louis. Since sweeping the Reds three weeks ago, the Cardinals have lost 7 games to bottom four teams in the division, none of which are higher than 24th in the Power Rank. Six games is a big deficit to make up, even for a significantly better team.

1. Tampa Bay, 81-50, 1.23
2. New York Yankees, 81-50, 1.21
3. Atlanta, 76-55, 0.81
4. Minnesota, 75-56, 0.68
5. San Diego, 76-54, 0.66
6. Boston, 74-57, 0.62
7. Philadelphia, 73-58, 0.56
8. St. Louis, 69-60, 0.47
9. Texas, 74-57, 0.42
10. Chicago White Sox, 71-60, 0.41
11. Toronto, 68-63, 0.36
12. Colorado, 69-61, 0.32
13. San Francisco, 72-60, 0.31
14. Florida, 65-65, 0.28
15. Cincinnati, 76-55, 0.20
16. New York Mets, 65-66, 0.20
17. Oakland, 65-65, 0.13
18. Los Angeles Dodgers, 68-64, 0.09
19. Detroit, 65-66, -0.08
20. Los Angeles Angels, 64-68, -0.08
21. Washington, 57-75, -0.42
22. Cleveland, 53-78, -0.60
23. Arizona, 53-79, -0.62
24. Chicago Cubs, 56-76, -0.75
25. Milwaukee, 62-69, -0.78
26. Houston, 60-71, -0.79
27. Seattle, 51-80, -0.80
28. Kansas City, 55-76, -1.00
29. Baltimore, 48-83, -1.01
30. Pittsburgh, 43-88, -2.03

Click here for a visual display of these rankings.

Cardinals Romp Over Dodgers

By 2 Comments

St. Louis completed a sweep over the Los Angeles Dodgers this weekend, taking four games with a +13 run differential.  The Dodgers have been an outlier all season with a great record but poor ranking, usually in the bottom half of the league.  They continued their awful streak tonight with a 5-2 loss to the San Francisco Giants.

Major League Baseball, Second Half Forecast

By 4 Comments

With the conclusion of the All Star break, the Power Rank offers a weather forecast for the second half of the season. Let’s start in the AL East, the division in which Major League Baseball still begins and ends. The New York Yankees and Tampa Bay Rays have traded the top position in the Power Rank back and forth all season while the Boston Red Sox have consistently been a distant third (see rankings at the bottom of the post). Boston will have a tough time catching either New York or Tampa Bay. Since the Power Rank rates the Yankees and Rays as equal, one has to look at the remaining schedule to see who has the advantage. Fortunately for Tampa Bay, they have 12 games remaining with the lowly Baltimore Orioles while the Yankees only have 6. Advantage Rays. Clearly, the second place team in the division gets the AL Wild Card.

The most interesting forecast comes from the AL Central, a division in which the Chicago White Sox hold a slim lead over the Detroit Tigers and the Minnesota Twins. Detroit is screwed. Not only are they the lowest team in the Power Rank but they have a significant number of games against AL East teams while Minnesota and Chicago play Oakland and Seattle. The Twins and White Sox seems to have similar schedules; the White Sox may have a slight edge with 7 games against the Baltimore Orioles while the Twins have 4. The Power Rank puts the Twins ahead of the Sox, but will this be enough to make up the 3.5 game lead in the standings?

As for the remaining divisions, the NL Central has the Cincinnati Reds and the St. Louis Cardinals racing neck and neck for the division. The Cardinals are a significantly better team, and they hold the ultimate trump card: 12 games against the Pittsburgh Pirates to 6 for the Reds. It must be wonderful playing in a division with so many games against the Chicago Cubs, Houston Astros, Milwaukee Brewers and the Pirates. Bring up your prospects, throw them out their against division rivals. But the Power Rank has both the Cardinals and Reds quite low, as the algorithm expects more wins (and big wins) from the leaders of weak divisions. As for the other three divisions, expect San Diego, Atlanta and Texas to comfortably win their respective divisions, with the NL Wild Card going to the Colorado Rockies.

1. NY Yankees, 56-32, 1.26
2. Tampa Bay, 54-34, 1.19
3. Boston, 51-37, 0.89
4. San Diego, 51-37, 0.77
5. NY Mets, 48-40, 0.64
6. Atlanta, 52-36, 0.60
7. Colorado, 49-39, 0.58
8. Philadelphia, 47-40, 0.51
9. Minnesota, 46-42, 0.50
10. Florida, 42-46, 0.46
11. Texas, 50-38, 0.41
12. St. Louis, 47-41, 0.41
13. San Francisco, 47-41, 0.38
14. Chicago White Sox, 49-38, 0.22
15. Toronto, 44-45, 0.19
16. LA Dodgers, 49-39, 0.19
17. Cincinnati, 49-41, 0.04
18. Detroit, 48-38, 0.04
19. LA Angels, 47-44, -0.12
20. Oakland, 43-46, -0.14
21. Kansas City, 39-49, -0.41
22. Washington, 39-50, -0.46
23. Cleveland, 34-54, -0.58
24. Seattle, 35-53, -0.65
25. Arizona, 34-55, -0.79
26. Chicago Cubs, 39-50, -0.82
27. Milwaukee, 40-49, -0.82
28. Baltimore, 29-59, -1.10
29. Houston, 36-53, -1.23
30. Pittsburgh, 30-58, -2.17

Check out the site for a full explanation of these rankings and the numbers next to each team.