Chicago Fanatics Message Board
https://mail.chicagofanatics.com/

Interesting article on the effectiveness of playoff systems
https://mail.chicagofanatics.com/viewtopic.php?f=92&t=95966		 Page 1 of 1
Author:  leashyourkids [ Sat Sep 12, 2015 11:59 pm ]
Post subject:  Interesting article on the effectiveness of playoff systems
http://harvardsportsanalysis.org/2013/0 ... ostseason/


To this unbiased observer, European soccer has unequivocally the fairest system when it comes to deciding a league champion: by the end of the season, each team has played all the others both at home and away. This means all teams play the exact same schedule, and thus whoever wins the most games by definition deserved to win the league. While there is an inherent element of luck to winning an individual soccer match, a team will never be accused of sneaking a championship because it must have earned its victory (unless it pays referees, but that’s really just Italy).

In the four major US sports leagues though, a regular season is supplemented with a postseason to determine the final champion. As once elucidated by Jim Mora, playoffs are an integral part of American sports culture and regularly attract some of the largest TV audiences in the world. This system is in place for the sake of both practicality—can you imagine a round-robin NFL season?—and the exhilaration fans gain from its “win or go home” aspect.

However, even the most ardent playoff advocate will admit that sometimes this system results in an undeserving winner (most explicitly, the 2007 New York Giants). Because of the variance associated with any individual sports game, the shortened postseason creates opportunities for teams with worse talent to pull off a string of upsets through a mixture of form and often luck (David Tyree) and be crowned champions. The question we wanted to address is, of the four leagues, which most consistently produces the least deserving champions.

To tackle this, we looked at past seasons dating back to 1995, when the MLB first implemented the eight-team playoff system, and thus when all the four leagues were approximately using the system in use today. The NBA has made some small alterations to their seeding rules, and the NFL has expanded to four divisions per conference, but each have kept the same number of playoff teams. The MLB in 2012 added the one-game playoff between the two wild cards, a small change that may have slightly aided the Giants’ World Series run last year but in total has been a minor enough change for us to include 2012 in our analysis.

For each season in each league, we considered the Pythagorean win expectation for each team to give an approximation of their regular season performance and then calculated the rank and z-score of each postseason champion’s regular season performance. So, for example, the 2013 Chicago Blackhawks had a Pythagorean winning percentage of .698, which was the best in the league and gave them a z-score of 2.26, implying that they were over two standard deviations better than the league average during the regular season. However, the Pythagorean metric does not take into account the teams’ regular season strength of schedule; consequently, we also analyzed the champions’ “simple rating.” The Simple Rating System (SRS)—devised by sports-reference.com—aims to provide a figure of how many runs/goals/points a team differs compared to league average for the season, while explicitly taking into account a team’s strength of schedule. It can be considered to be a proxy for the point spread between two teams if they played at a neutral site (and were playing at their season-average level). So when the 2007 Patriots (SRS of +20.1) played the 2007 Giants (SRS of +3.3) in the Super Bowl, had they both played at their average level of regular season performance, the Patriots would have been expected to win by about 16.8 points. Computing the SRS rank and z-score for champions as well, we found the following results:

Pythagorean Expectation

Simple Rating System

League Champions from 1995-2013

Mean Rank

Median Rank

Mean Z-Score

Mean Rank

Median Rank

Mean  Z-Score

NFL

5.3

4.5

1.34

4.7

2.5

1.37

MLB

5.8

4

1.10

6.5

7

1.05

NBA

3.3

2

1.47

3.3

2

1.48

NHL

4.0

2.5

1.34

4.2

3

1.27

 

As we expected, the NBA seems to produce postseason champions most aligned with regular season performance. The long format of the seven-game series and reduced likelihood of upsets on account of the high number of possessions have resulted in the best regular season team winning eight of the nineteen possible championships.

Vividly remembering the LA Kings’ and Baltimore Ravens’ unlikely triumphs, we naively assumed that the NFL and the NHL would battle it out for last spot. However, our analysis revealed that the MLB most consistently produces champions most disparate from their regular season performance. What is remarkable is just how bad the MLB playoffs really are. Owing to the length of its 162 game season, one might think that regular season performance would actually be a fairly good indicator – better, for example, than in the NFL where strength of schedule can have a huge impact – of the overall quality of a team. Given this fairly reasonable assumption, if you chose the eight best regular season teams, or even eight of the top ten because you require four from each league, and then just asked each team to draw straws to determine the World Series, the average winning team would be better determined than by the current system. Only three times has the best team from the regular season ended up winning the World Series.

The last 18 years of sports have produced some truly fantastic teams, but there have also been some complete duds, and we would like to take this opportunity to individually call out the least-deserving champion of each of the four major sports. In the NBA, the honors unquestionably go to the 1995 Houston Rockets, who were barely half a standard deviation above the league average and were the eleventh-best regular season team. The NHL has a few out of the blue champions, and honorable mention should go to 1995 Devils, 2007 Hurricanes and 2009 Penguins, but all things considered, no one can touch the 2012 Kings. As the eleventh-best team of the regular season and proud holders of an eight seed, they went on a statistically improbable away-game winning streak, only losing one game outside Los Angeles en route to their Stanley Cup triumph. The craziness of that postseason was only amplified by the ninth-best and sixth-seeded New Jersey Devils making the finals as the Eastern Conference champs.

When it comes to the NFL, no one steals Super Bowls like Eli Manning. The 2007 Giants were playing the team with the highest z-score for both Pythagorean expectation and SRS of any champion in our dataset across all four leagues and still managed to win as the eleventh-best team (by SRS, thirteenth by Pythagorean). However, that team is second on the list to the other Eli Manning triumph. The 2011 Giants, who scraped into the playoffs at 9-7, remarkably actually scored fewer points than their opponents in the regular season and were thus a below average regular season team before tearing through the playoffs behind a dominant defensive line, Eli “somehow-more-rings-than-his-brother” Manning and another ridiculous catch.

Last but never least, the MLB can lay claim to the least deserving postseason winner. Noteworthy runners-up include the 2012 Giants (apparently a solid franchise name for sneaking championships) who pitched their way to victory and the 2003 Marlins, who rallied to victory from the wildcard spot behind Dontrelle “ D-Train” Willis and his unorthodox motion, but neither can touch the least meritocratic winner sports might ever see: the 2006 St. Louis Cardinals. The Cardinals were shockingly the twenty-first ranked team by SRS, over half a standard deviation below the league average, roughly equivalent to the 2012 Arizona Cardinals of the NFL. They needed a loss from the Astros in the final game of the season to barely slip into the playoffs with the third-worst record for a postseason team in MLB history at 83-78, coming off a September when they went 12-17. However, for some reason the gods were smiling down on St. Louis as they pulled off three huge upsets—including an epic Game Seven against the Mets in the NLCS—to win it all.

So as we move into October and the time of the year when everyone talks about how clutch various baseball players are, never underestimate the power of the MLB to produce some truly appalling victors. But hey, isn’t the craziness why we watch in the first place?

(For another look at measuring variance across the four major sports, check out this piece from Tangotiger.)
Author:  Douchebag [ Sun Sep 13, 2015 12:26 am ]
Post subject:  Re: Interesting article on the effectiveness of playoff syst
MLB playoffs are a crapshoot.
Author:  Terry's Peeps [ Sun Sep 13, 2015 7:54 am ]
Post subject:  Re: Interesting article on the effectiveness of playoff syst
Defense mechanism.
Author:  leashyourkids [ Sun Sep 13, 2015 9:13 am ]
Post subject:  Re: Interesting article on the effectiveness of playoff syst
Terry's Peeps wrote:
Defense mechanism.


(regarding the NBA) " reduced likelihood of upsets on account of the high number of possessions "

Wanted to make sure you didn't miss this quote. Interesting.
Author:  Terry's Peeps [ Sun Sep 13, 2015 9:51 am ]
Post subject:  Re: Interesting article on the effectiveness of playoff syst
In the playoffs right?

We were talking in general yesterday.

Interesting indeed.
Author:  leashyourkids [ Sun Sep 13, 2015 10:05 am ]
Post subject:  Re: Interesting article on the effectiveness of playoff syst
They aren't really clear, but it seems like they're just making a general statement in that particular section. And the number of possessions is the same in regular season and playoffs, so it would hold true either way.
Author:  sinicalypse [ Sun Sep 13, 2015 10:13 am ]
Post subject:  Re: Interesting article on the effectiveness of playoff syst
i like the european soccer setup.... the regular season is its own competition and given the parity of the schedule it ultimately awards the best team in the league the championship.

then for your playoff/tournament type situations you've usually got 2 domestic cup competitions (one is a lesser one that typically has kids/bench-guys feature in it til the tournament quarterfinals or prolly semis, but nowadays most #1 domestic cup competitions arent considered a huge deal because of...) and the champions league / europa league, which like the world cup starts out with a bunch of 4-team groups who play each other home and away, then the 2 top teams from the 8 groups get put into a 16 team home-and-away knockout format til the final so you get the best of both worlds.

i've only been into euro soccer for a few years but i can tell you that it works out swell. the cup tournaments get worked in at various points throughout the season. the main FA cup matchups start in january for top-league teams.... and typically the major teams get to play a little crappy side or two early on, altho having been along for the ride during arsenal's last 2 successful FA cup runs i can tell you that basically you only have 1 or 2 games against legit opponents where going into the match it's like "yeah we can/prolly-will lose this" so basically it's more like you don't lose the cup as opposed to go out and win the cup. this doesnt have the chutzpah/OOMPH of the american playoffs where even if you beat a 7-9 seahawks team it's still a legit playoff WIN (save the usual detractors) --- in europe, tho, winning the champions league is the #1 thing you can do ("winning europe!") followed by your domestic league, then your domestic cup, then your domestic #2 cup (winning any 2 of these is a "double" and any 3 a "treble" and prolly winning 4 is "FC Barcelona") and then all the goofy meaningless FIFA cups come in like the club world championship. the euro teams are expected to win that easily and if they do they get to wear a little golden patch with a trophy on it on their uniforms for the rest of the season! isnt that special?

so yeah, especially in baseball i think it'd be nice if there was added emphasis on the pennant.... you know maybe at the end of the regular season the team with the best record in each league gets a trophy and parade and all that... but with MLB marketing saying "IT'S ALL ABOUT OCTOBER!" you end up with the typical american sports thing where basically the regular season is just a giant circlejerk that's all about seeding for the tournament, aka THE ONLY THING THAT MATTERS TO WINNERS LIKE US AMERICANS!

and given all the ratings/revenue/$$$ generated by the uptick in interest in the playoffs, my pennant-esque situation will never happen. i guess i'll just have to stick to euro soccer for that!
Author:  Jaw Breaker [ Mon Sep 14, 2015 2:09 pm ]
Post subject:  Re: Interesting article on the effectiveness of playoff syst
Clearly the biggest factor for MLB is the starting pitcher. In the other sports, pretty much the same team plays every game of the playoffs (with minor lineup changes), while in baseball there is a huge variation in the odds for each game depending on SP.
Author:  Rod [ Mon Sep 14, 2015 2:12 pm ]
Post subject:  Re: Interesting article on the effectiveness of playoff syst
Jaw Breaker wrote:
Clearly the biggest factor for MLB is the starting pitcher. In the other sports, pretty much the same team plays every game of the playoffs (with minor lineup changes), while in baseball there is a huge variation in the odds for each game depending on SP.


Which is exactly why a starting pitcher's W/L record is pertinent.
Author:  rogers park bryan [ Mon Sep 14, 2015 2:20 pm ]
Post subject:  Re: Interesting article on the effectiveness of playoff syst
Didnt read whole article but the ratio of playoff games to reg season games would seem to hold true

NBA and NHL are like 25-40% 16-28 games.
NFL playoffs 25% of reg season
Baseball is 7-11% (11-19 games) of the regular season


Not surprising that NBA and NHL would see less parity
Author:  Hatchetman [ Mon Sep 14, 2015 2:22 pm ]
Post subject:  Re: Interesting article on the effectiveness of playoff syst
in other news, water is wet.
Author:  good dolphin [ Mon Sep 14, 2015 4:20 pm ]
Post subject:  Re: Interesting article on the effectiveness of playoff syst
The desired effectiveness of the playoffs is to put money in the owners' pockets
Author:  leashyourkids [ Mon Sep 14, 2015 4:33 pm ]
Post subject:  Re: Interesting article on the effectiveness of playoff syst
good dolphin wrote:
The desired effectiveness of the playoffs is to put money in the owners' pockets


Of course.

Titles are hard.
Author:  veganfan21 [ Mon Sep 14, 2015 5:40 pm ]
Post subject:  Re: Interesting article on the effectiveness of playoff syst
Interesting article indeed.

Quote:
the shortened postseason [NFL] creates opportunities for teams with worse talent to pull off a string of upsets through a mixture of form and often luck


I also think you can extend this analysis to the regular season, which is very short as well. That's why I think even a team like the bears with all the inadequacies can contend for a spot simply due to the rampant parity the game of football creates, mainly through injuries and other randomness that levels the playing field. The one constant is most likely good coaching, and if you have that you've separated yourself from teams that have squandered their chances because the coaches have no idea what they're doing.
Author:  Darkside [ Mon Sep 14, 2015 5:50 pm ]
Post subject:  Re: Interesting article on the effectiveness of playoff syst
Joe Orr Road Rod wrote:
Jaw Breaker wrote:
Clearly the biggest factor for MLB is the starting pitcher. In the other sports, pretty much the same team plays every game of the playoffs (with minor lineup changes), while in baseball there is a huge variation in the odds for each game depending on SP.


Which is exactly why a starting pitcher's W/L record is pertinent.

Sweet Jesus.
Author:  IMU [ Mon Sep 14, 2015 6:16 pm ]
Post subject:  Re: Interesting article on the effectiveness of playoff syst
I just cannot read this on a Monday. :(
Page 1 of 1 All times are UTC - 6 hours [ DST ]
Powered by phpBB® Forum Software © phpBB Group
https://www.phpbb.com/