Adrian the Canadian fixes the MLB playoffs

“But he’s Canadian,” you say, “So what does he know about baseball?” Well, he’s from Toronto, the team most screwed by the current system, so let’s give it a shot:

A few days ago, the MLB announced that it was expanding its playoffs to include a second wild-card team. Under the new system, the two wild-cards will play a single game that determines who goes to the divisional series. Response has been, at best, mixed. The strongest criticisms, like this one from ESPN’s Joe Sheenan, have taken a traditionalist perspective. Sheenan worries about what this new system will mean for deep-seated, and still exciting, elements of baseball like the pennant race. He sees the wild-card system as debasing what has historically been one of the most exciting parts of being a baseball fan: following your team through a tense September race to win the division. By Sheenan’s estimation, the old system encouraged top teams to play their best throughout the whole season. If you happen to be one of the two best teams in the league by regular season record but can’t win your own division, tough grapes.

I sympathize with Sheenan and other traditionalists. Baseball’s regular season is long and arduous and does a pretty good job of determining the “best” team (or, at least, a better job than other pro-sports at determining the best team). Meanwhile, baseball playoffs, due to the nature of the game, are pretty close to random. As Billy Beane said, “my shit doesn’t work in the playoffs” – seven and five game series are simply too short to give us a good idea as to which team is best. In the 17 post-seasons since the advent of the wild-card, the wild card team has won five times and made five other World Series. This problem is acute in baseball. Sheenan is right that winning a division is no longer a big deal and that division battles between top teams mean less due to the wild card. He’s also right that it seems somehow unfair when teams like the 2011 Cardinals win the World Series. 162 games should mean something. And epic collapses like that of the 2011 Red Sox are less epic when they land in the second wild-card spot.

But I’m a Torontonian. I have the comfort of knowing that, no matter how good my Jays are, I can begin every season with little hope.  Despite playing in the AL East, in 2008 we won 86 games – the same number as the 2011 Cardinals and enough to win the 2008 NL West – and yet had no hope of making the playoffs! To do so, we’d have had to finish with a better record than two of New York, Boston, and Tampa Bay and beat out another team from another, weaker, division all while playing one of the league’s most crippling schedules. While Toronto may be in a unique position, Toronto is not alone in its general predicament. By late-July, most teams are well served to pack it in. So few teams make the playoffs in the MLB that, for most of the country, the last half of the season is a good excuse for one long late summer nap in most cities.

MLB’s solution has been to push the tension down the table. While the pennant race might be riveting, those cities remain interested in October and expanding the playoffs at least makes a few more cities interested in August. So how can baseball solve this situation? How can baseball keep the regular season interesting and reward regular season accomplishment? While some problems of competitive balance can’t be addressed without baseball undertaking economic reforms, some structural playoff reforms could help balance these two concerns.

In order to make winning your division, and therefore the divisional race, more valuable, we must come up with a playoff system that rewards division champions by giving them a better chance at winning the World Series than wild-cards but keeps fans interested after the divisions have been clinched. Here’s a potential, and potentially radical, solution:

  1. Move back to two leagues of two divisions, the East and the West.
  2. Four teams still make the playoffs like under the current (well, current up until this week) model.
  3. The over-all division champion, as measured by record, gets a bye into the League Championship Series.
  4. The two wild-card teams, one from each division, play each other in a “coin-flip” game to make the Divisional Series against the second division champion.
  5. The second division champion and the wild-card play the divisional series for the right to make the League Championship series.

I think this proposal splits the baby (without killing it!). First, assuming an underlying probability of 50% for each team in each series, the team with the overall best record has a 50% chance of making the World Series – a result that has the benefit of 1) making the best regular season record more valuable than it currently is thereby encouraging the top team to play all out until the last day of the season and 2) comporting with our intuition that the MLB regular season is a good measure of the best team.

Meanwhile, the division winner with the second best record during the regular season has a 25% chance of making the World Series while the wild-card teams have only a 12.5% chance of making it. This preserves the integrity of the pennant race between top teams. While currently, finishing with the second best record in your division and winning the wild-card gives you approximately the same probability of making the World Series as the eventual division champion (approximately 25%), now you risk a 12.5% to 37.5% decrease in your chance of making the World Series! As for potential revenue loss, it’s true that baseball loses between two and four divisional games, but the added interest generated by the more meaningful pennant race in August and September should more than make up for it. Plus, in the rare situation where the wild-card makes a run at the World Series, the “Cinderella Story” is far more compelling than it is now.

Such a solution may not save the Jays, although it will make the baseball regular season a lot more exciting. Will traditionalists object to such a change? Surely. Many of them will complain that my system is too complicated. I don’t think that it is, but nothing short of returning to two divisions, a single LCS, and three teams not named the Mets in NYC will appease them. I think that even George Will knows we’re not going back to that. All I can hope is that, if George reads this blog, he’ll agree with my proposal more than I agree with him every Sunday on This Week.

Advertisements

8 responses to “Adrian the Canadian fixes the MLB playoffs

  1. I think you’re missing part of the appeal of the MLB-proposed system, which generates a large reward for the division title, when you have the two best teams in the same division.

    So as you said, under the MLB proposed system with equal teams, the division winners have a 25% chance of winning the pennant, while the wild card winners have a 12.5% chance — a 12.5% premium on winning your division.

    But now suppose the potential playoff teams include two type A teams and a bunch of type B teams. Game outcomes between like types are 50-50, while a type A beats a type B 60% percent of the time, meaning the chance of A beating B in a 1,3,5,or 7 game playoff is 60,65,68, or 71%, respectively.

    Under the current system, assuming home-field is worthless, the A types (in the same division) have a 39% chance of the pennant while the B types have an 11% chance.

    Now let’s instead suppose one A type has to play a single-game wild card playoff, while the other A type wins the same division. Here are the probabilities of the pennant (and LCS probability) :
    A division winner = 43% (68%)
    A wild card team = 23% (41%)
    B division winner, playing wild card game winner = 14% (39%)
    B division winner playing A division winner = 13% (32%)
    B wild card team = 7% (20%)

    So an appeal of the MLB-proposed system is that when you have a great pennant race between two strong teams it does increase the reward of that division winner by a lot more than 12.5%, specifically 20% in this case. This is due to both the benefit to the division winner of the wild card getting upset in the wild card game, and the greater probability of the A-type team losing a 1-game series than a 5-game or 7-game series.

    Now of course you’ll get similar kind of effects in Adrian’s proposed system, but I just think you undersell the benefits of the MLB proposed system. I also worry in Adrian’s system about (1) the effect of an LDS layoff on the team getting the bye to LCS and (2) giving such a huge reward to the No. 1 seed when there’s unbalanced scheduling (or alternatively, having to get rid of unbalanced scheduling).

  2. It seems unlikely that there are both A type and B type teams in the playoffs. For example, last year, only six wins stood between the wild card and the top division champion. Effectively a sneeze in the baseball season. For there to be A-Type and B-type teams in the playoffs, we would expect that A-Type teams would have, on average, 60% winning percentages over B-type teams during the regular season (remember, there can be max two or three A-types in the playoffs, implying, at best, one or two more A-types out side of the playoffs). This means that we should expect A-Types to have a winning percentage well >60% over other non-playoff teams. However, a winning percentage >70% has only been observed twice in the last fifty years and there is rarely a team with an overall winning percentage of >60% in either league (only six times since 05, just under once a year. 59% is pretty common though). Furthermore, numerous times we’ve seen teams that are inferior by any statistical measure make and win World Series in the wild-card era. (it’s also worth noting that individual wins are highly pitching dependent, putting a hypothetical A wild-card at a further disadvantage in the divisional series). In baseball, it seems that teams are sufficiently balanced so that differences between the two teams don’t manifest themselves over a short series.

    Also, it seems odd to worry about both the effect of the bye on the 1st DC AND the large advantage the 1st DC gets. How can the system be both too much and not enough of an advantage for the 1st DC?

    Regardless, appearances are everything in sports. The current system has created the impression of meaningless playoffs tacked on top of meaningless division titles. I think this proposal would restore some perceived meaning to both the regular season and the world series.

  3. Adrian, you’re right that 60-40 weren’t realistic numbers. 60-40 is the difference between the 2009 Yankees and 2009 Twins. But 55-45 is reasonable for the 2004 Red Sox & Yankees and the rest of the AL playoff field, and that is the kind of year I had in mind. That is the kind of year when everyone was saying that the wild card denied us a great AL East pennant race.

    You’re right that there’s a contradiction in my worries, but I’m sure that both worries would come up in the press every year and it’d be a distraction. Like when the Rockies had that huge layoff after the 2007 NLCS, that took something away from the World Series, whether the effect on the Rockies was real or not.

  4. I’m interested in the layoff question, and the value of home field advantage. There’s potentially a way to get at the average effect — use close games as random variation in length of series (and regular season winning percentage, i.e., playoff seeding). However, I wonder if the layoff issue is too team-specific to learn much. Depending on the health of the team, and their degree of reliance on specific starters, a layoff could help or hurt.

  5. It seems to me that a protracted layoff would have less of an effect in baseball than in any other sport. Pitchers can pitch “simulated games” on the days they’d otherwise be starting and hitters can do much the same. I’m deeply skeptical of the idea that players forget what it’s like to play in front of a crowd and “choke” based on ten days rest. I think that the “long layoff” excuse is another tactic that the baseball media uses to dodge the unpleasant fact that the playoffs are nearly random.

  6. Yeah, and in the case of the 2007 Red Sox-Rockies World Series, the Red Sox were just a much better team, at least a 65% chance of winning that series. The AL probably had the top 5 or 6 teams in baseball that season. But according to the media, the Rockies had been so hot and then got swept, so it just had to have been the layoff that was responsible.

  7. Pingback: LOS LIIIIIINNNNNNKKKKSSSSSS!!!!!!!!!!!!!! | Causal Sports Fan

  8. Pingback: Adrian the Canadian: What’s wrong with the BCS and it’s successor? | Causal Sports Fan

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s