Playoff Formats and the Marginal Win
In the weirdest year of baseball history so far, 2020 featured a gigantic playoff field introduced right as the season began, turning a 10-team postseason into a 16-team format. Changing the basic structure of awarding the sport’s championship with no advance notice would have been an odd choice in a normal season. But given the 102-game reduction in the league’s schedule and its resulting small sample size season, it kind of made sense. When the decision was made, it wasn’t a surety that there would even be a season, to the point that people would have been happy if extra-inning games were decided by closers riding ostriches and jousting.
Before the World Series was even completed, commissioner Rob Manfred expressed the league’s desire to keep the new format in a normal season. The players need to agree to changes like this, of course, and that permission wasn’t granted after all MLB offered in exchange was a universal designated hitter. One of the concerns, not officially made public, is that a playoff system that is more of crapshoot will further reduce the already eroded incentives for teams to spend money to improve their rosters. That’s hardly a shock; at the 2020 trade deadline, 10 teams already had projected playoff probabilities above 97%. Combine that with the absence of the normal advantages afforded to higher seeds, and you had a trade deadline that saw only a single team, the San Diego Padres, aggressively improve, and even their moves were almost certainly made with an eye toward 2021 and beyond.
So what is the ideal playoff system? That’s a difficult question, one that’s impossible to answer to everyone’s satisfaction. I can only answer for myself, and for me, there are a few requirements that are particularly important. Basically, I want a system in which regular-season performance matters, thus maintaining one of the core aspects of the game. I also want a playoff system that more heavily awards quality over randomness without making the result a preordained one. The more a championship is decided by randomness, the less incentive there is for teams to innovate and invest.
To keep people interested in baseball’s long regular season, the wins have to actually matter. So what is the value of an additional win under various playoff formats? To answer this, I reconfigured ZiPS to run using four new playoff formats: a 14-team playoff system with byes for the top teams in the first round, a 16-team playoff system like last season, a four-team playoff system with the top two teams advancing, and the no-division format in use before the 1969 season. I then ran 31 simulations for each playoff format, first to get the baseline World Series win probability, then adding a single win’s worth of strength to each of the 30 teams in order. After writing three new models and running a million seasons in 155 separate simulations, I’ve got some hard data I did not previously have. Also, I think I hate numbers and never want to see one again!
Let’s start with the current format:
| Team | Current WS Probability | With One More Win | Change |
|---|---|---|---|
| Los Angeles Dodgers | 14.67% | 15.91% | 1.24% |
| San Diego Padres | 11.44% | 12.58% | 1.14% |
| New York Yankees | 10.87% | 11.99% | 1.12% |
| Minnesota Twins | 8.13% | 9.12% | 0.99% |
| Chicago White Sox | 7.87% | 8.83% | 0.96% |
| New York Mets | 6.64% | 7.52% | 0.88% |
| Oakland A’s | 6.03% | 6.88% | 0.84% |
| Atlanta Braves | 7.88% | 8.71% | 0.83% |
| Tampa Bay Rays | 5.09% | 5.91% | 0.82% |
| Toronto Blue Jays | 3.99% | 4.69% | 0.71% |
| Houston Astros | 4.18% | 4.87% | 0.69% |
| St. Louis Cardinals | 3.84% | 4.45% | 0.60% |
| Washington Nationals | 1.97% | 2.41% | 0.44% |
| Chicago Cubs | 1.92% | 2.34% | 0.42% |
| Milwaukee Brewers | 1.52% | 1.88% | 0.36% |
| Los Angeles Angels | 1.33% | 1.69% | 0.36% |
| Cincinnati Reds | 0.67% | 0.87% | 0.20% |
| Philadelphia Phillies | 0.88% | 1.03% | 0.15% |
| Boston Red Sox | 0.33% | 0.46% | 0.13% |
| Cleveland | 0.27% | 0.37% | 0.10% |
| Arizona Diamondbacks | 0.20% | 0.29% | 0.09% |
| Seattle Mariners | 0.14% | 0.20% | 0.06% |
| Kansas City Royals | 0.07% | 0.10% | 0.03% |
| Miami Marlins | 0.05% | 0.07% | 0.03% |
| San Francisco Giants | 0.02% | 0.05% | 0.02% |
| Pittsburgh Pirates | 0.01% | 0.02% | 0.01% |
| Baltimore Orioles | 0.00% | 0.01% | 0.01% |
| Detroit Tigers | 0.00% | 0.01% | 0.01% |
| Texas Rangers | 0.01% | 0.01% | 0.00% |
| Colorado Rockies | 0.00% | 0.00% | 0.00% |
By adding a single win, the average team gained 0.44 percentage points of World Series championship probability. Considering gains of at least 0.1 percentage points significant — since that’s what we normally round to — 20 teams saw movement by making them one win stronger. Somewhat arbitrarily, I imagine 0.5 percentage points as being major gains; 12 teams meet that threshold. Where we set these lines doesn’t necessarily matter in that we’re simply looking at relative changes due to format. Under the current format, the best teams all see significant benefits of adding a win thanks to division winners not having to play the high-stakes Wild Card game.
Next up, 2020’s format:
| Team | Current WS Probability | With One More Win | Change |
|---|---|---|---|
| Los Angeles Dodgers | 10.25% | 10.72% | 0.47% |
| San Diego Padres | 9.15% | 9.60% | 0.45% |
| New York Yankees | 8.19% | 8.63% | 0.44% |
| Minnesota Twins | 6.95% | 7.37% | 0.42% |
| Atlanta Braves | 6.75% | 7.17% | 0.42% |
| New York Mets | 6.25% | 6.67% | 0.42% |
| Chicago White Sox | 6.84% | 7.25% | 0.41% |
| Tampa Bay Rays | 5.83% | 6.24% | 0.41% |
| Toronto Blue Jays | 5.24% | 5.65% | 0.41% |
| Oakland A’s | 5.52% | 5.93% | 0.41% |
| Houston Astros | 4.63% | 5.03% | 0.40% |
| Washington Nationals | 3.62% | 4.00% | 0.38% |
| St. Louis Cardinals | 3.81% | 4.17% | 0.36% |
| Los Angeles Angels | 2.56% | 2.91% | 0.35% |
| Philadelphia Phillies | 2.43% | 2.77% | 0.34% |
| Chicago Cubs | 2.62% | 2.95% | 0.33% |
| Milwaukee Brewers | 2.26% | 2.58% | 0.32% |
| Boston Red Sox | 1.42% | 1.70% | 0.28% |
| Cincinnati Reds | 1.31% | 1.56% | 0.25% |
| Cleveland | 1.18% | 1.42% | 0.24% |
| Arizona Diamondbacks | 1.13% | 1.36% | 0.23% |
| Seattle Mariners | 0.59% | 0.75% | 0.16% |
| Kansas City Royals | 0.48% | 0.63% | 0.15% |
| Miami Marlins | 0.41% | 0.53% | 0.12% |
| San Francisco Giants | 0.31% | 0.41% | 0.10% |
| Texas Rangers | 0.08% | 0.12% | 0.04% |
| Detroit Tigers | 0.08% | 0.11% | 0.03% |
| Baltimore Orioles | 0.07% | 0.10% | 0.03% |
| Pittsburgh Pirates | 0.07% | 0.10% | 0.03% |
| Colorado Rockies | 0.00% | 0.01% | 0.01% |
Naturally, baseball’s best teams take the biggest overall hit, with the Dodgers losing nearly a third of their simulated championships. That the Dodgers actually did win in 2020 is one of those little practical jokes that history likes to pull, with the most random playoffs in history resulting in the winningest AL and NL teams making the World Series. It would be a mistake to assume predestination here; it only requires changing a few plays last October to get the 29-31 Houston Astros into the World Series.
More teams see gains in playoff probability, with 25 teams now making that 0.1 percentage-point threshold. But now, not a single team meets the 0.5 percentage-point mark. With no bye for division winners, adding a free agent who improves the roster by a win suddenly has less than half the value for quality teams in tight division races like the Dodgers, Padres, Yankees, and Mets. Teams at the bottom, however, improve but remain long shots.
So, let’s restore the divisional advantage with one of MLB’s proposals, a 14-team format that leaves the top team in each league with the ability to skip the Wild Card round:
| Team | Current WS Probability | With One More Win | Change |
|---|---|---|---|
| Los Angeles Dodgers | 12.87% | 13.71% | 0.84% |
| New York Yankees | 9.84% | 10.64% | 0.80% |
| San Diego Padres | 10.67% | 11.46% | 0.79% |
| Minnesota Twins | 7.53% | 8.24% | 0.71% |
| Chicago White Sox | 7.37% | 8.06% | 0.69% |
| Tampa Bay Rays | 5.86% | 6.48% | 0.62% |
| Atlanta Braves | 6.82% | 7.41% | 0.59% |
| Oakland A’s | 5.37% | 5.94% | 0.57% |
| Toronto Blue Jays | 5.05% | 5.62% | 0.57% |
| New York Mets | 6.14% | 6.70% | 0.56% |
| Houston Astros | 4.19% | 4.70% | 0.51% |
| Washington Nationals | 3.08% | 3.48% | 0.40% |
| St. Louis Cardinals | 3.16% | 3.56% | 0.40% |
| Los Angeles Angels | 1.92% | 2.26% | 0.34% |
| Philadelphia Phillies | 1.92% | 2.25% | 0.33% |
| Chicago Cubs | 1.94% | 2.26% | 0.32% |
| Milwaukee Brewers | 1.62% | 1.92% | 0.30% |
| Boston Red Sox | 0.99% | 1.23% | 0.24% |
| Cincinnati Reds | 0.85% | 1.05% | 0.20% |
| Cleveland | 0.77% | 0.97% | 0.20% |
| Arizona Diamondbacks | 0.83% | 1.02% | 0.19% |
| Kansas City Royals | 0.28% | 0.38% | 0.10% |
| Miami Marlins | 0.27% | 0.36% | 0.09% |
| San Francisco Giants | 0.20% | 0.27% | 0.07% |
| Texas Rangers | 0.03% | 0.05% | 0.02% |
| Baltimore Orioles | 0.03% | 0.05% | 0.02% |
| Pittsburgh Pirates | 0.03% | 0.04% | 0.01% |
| Detroit Tigers | 0.03% | 0.04% | 0.01% |
| Seattle Mariners | 0.35% | 0.35% | 0.00% |
| Colorado Rockies | 0.00% | 0.00% | 0.00% |
Under this blueprint, the average team’s World Series probability gains 0.35%, less than the current system’s 0.44%. Still, by giving a significant advantage to each league’s best teams, there’s a real motivation not to coast into the playoffs at 89 wins. Twenty-two teams pass the significant gain line, and the number of teams passing the major gain one goes back to 11 from zero.
Let’s go really old-school and look at the playoff format that dominated before the divisional era: two leagues, with the best team in each league going straight to the World Series:
| Team | Current WS Probability | With One More Win | Change |
|---|---|---|---|
| Los Angeles Dodgers | 26.52% | 29.90% | 3.38% |
| San Diego Padres | 17.51% | 20.46% | 2.95% |
| New York Yankees | 15.13% | 17.72% | 2.59% |
| Minnesota Twins | 8.27% | 10.14% | 1.87% |
| Chicago White Sox | 7.92% | 9.71% | 1.79% |
| Atlanta Braves | 5.60% | 7.04% | 1.44% |
| Tampa Bay Rays | 4.51% | 5.81% | 1.30% |
| New York Mets | 4.20% | 5.35% | 1.15% |
| Oakland A’s | 3.54% | 4.61% | 1.07% |
| Toronto Blue Jays | 3.11% | 4.06% | 0.95% |
| Houston Astros | 1.88% | 2.56% | 0.68% |
| St. Louis Cardinals | 0.56% | 0.82% | 0.26% |
| Washington Nationals | 0.51% | 0.77% | 0.26% |
| Los Angeles Angels | 0.30% | 0.46% | 0.16% |
| Philadelphia Phillies | 0.13% | 0.21% | 0.08% |
| Chicago Cubs | 0.11% | 0.19% | 0.08% |
| Milwaukee Brewers | 0.07% | 0.13% | 0.06% |
| Boston Red Sox | 0.06% | 0.09% | 0.03% |
| Cleveland | 0.03% | 0.05% | 0.02% |
| Cincinnati Reds | 0.01% | 0.03% | 0.02% |
| Arizona Diamondbacks | 0.02% | 0.03% | 0.01% |
| Kansas City Royals | 0.00% | 0.01% | 0.01% |
| Baltimore Orioles | 0.00% | 0.00% | 0.00% |
| Colorado Rockies | 0.00% | 0.00% | 0.00% |
| Detroit Tigers | 0.00% | 0.00% | 0.00% |
| Miami Marlins | 0.00% | 0.00% | 0.00% |
| Pittsburgh Pirates | 0.00% | 0.00% | 0.00% |
| San Francisco Giants | 0.00% | 0.00% | 0.00% |
| Texas Rangers | 0.00% | 0.00% | 0.00% |
| Seattle Mariners | 0.00% | 0.00% | 0.00% |
The average gain is the highest so far, with 0.67 percentage points per team. The gains, however, are limited to only a very few teams, with the break point somewhere around 85 wins. One can argue that this isn’t necessarily healthy for baseball, either. Teams like the Dodgers see larger gains by adding free agents, but those individual free agents would have a lot fewer suitors despite the larger average improvement. Auctions require bidders! This is likely the best playoff format for crowning the best team but may not be a healthy one for the game for various other reasons.
As the last playoff format to look at today, I ran a modified version of the divisional era format. Rather than making up new divisions — there are teams in different leagues and two teams that did not exist during the two-division era — I simply looked at the two best teams in each league.
| Team | Current WS Probability | With One More Win | Change |
|---|---|---|---|
| Los Angeles Dodgers | 22.16% | 24.25% | 2.09% |
| San Diego Padres | 17.06% | 19.05% | 1.99% |
| New York Yankees | 13.09% | 14.86% | 1.77% |
| Minnesota Twins | 8.22% | 9.70% | 1.48% |
| Chicago White Sox | 7.91% | 9.33% | 1.42% |
| Atlanta Braves | 7.06% | 8.40% | 1.34% |
| New York Mets | 5.47% | 6.67% | 1.20% |
| Tampa Bay Rays | 5.05% | 6.17% | 1.12% |
| Oakland A’s | 4.15% | 5.14% | 0.99% |
| Toronto Blue Jays | 3.67% | 4.58% | 0.91% |
| Houston Astros | 2.44% | 3.14% | 0.70% |
| St. Louis Cardinals | 1.03% | 1.41% | 0.38% |
| Washington Nationals | 1.04% | 1.41% | 0.37% |
| Los Angeles Angels | 0.50% | 0.72% | 0.22% |
| Chicago Cubs | 0.30% | 0.45% | 0.15% |
| Philadelphia Phillies | 0.31% | 0.46% | 0.15% |
| Milwaukee Brewers | 0.20% | 0.32% | 0.12% |
| Boston Red Sox | 0.13% | 0.21% | 0.08% |
| Cleveland | 0.08% | 0.13% | 0.05% |
| Cincinnati Reds | 0.05% | 0.09% | 0.04% |
| Arizona Diamondbacks | 0.05% | 0.09% | 0.04% |
| Seattle Mariners | 0.01% | 0.02% | 0.01% |
| San Francisco Giants | 0.00% | 0.01% | 0.01% |
| Kansas City Royals | 0.01% | 0.01% | 0.00% |
| Baltimore Orioles | 0.00% | 0.00% | 0.00% |
| Colorado Rockies | 0.00% | 0.00% | 0.00% |
| Detroit Tigers | 0.00% | 0.00% | 0.00% |
| Pittsburgh Pirates | 0.00% | 0.00% | 0.00% |
| Texas Rangers | 0.00% | 0.00% | 0.00% |
| Miami Marlins | 0.00% | 0.00% | 0.00% |
Not surprisingly, this one comes smack-dab between the 10-team system and the old-school one. With 17 teams at worse than 100-to-1 to win the championship compared to 13 today, it has some of the same problems as the pre-divisional system, though they aren’t quite as severe.
Here’s the quick-and-dirty summary for the five formats we looked at.
| Format | Average Improvement | Significant Gains | Major Gains |
|---|---|---|---|
| Old-School World Series | 0.67% | 14 | 11 |
| Current System | 0.44% | 20 | 12 |
| 14 Teams, Two Byes | 0.35% | 22 | 11 |
| 2020’s 16-Team System | 0.28% | 25 | 0 |
| Four-Team Playoffs | 0.55% | 17 | 11 |
So, what does this all mean? One can make the argument that whether by design or dumb luck, MLB has stumbled into a reasonable playoff configuration. The regular season still matters and while a third of the league makes the playoffs, there are real benefits to winning the division.
We’re far more likely to get more teams in the playoffs than fewer going forward, but I think these numbers demonstrate some of the risks involved of simply shoving more teams into October. Having more teams make the playoffs doesn’t have to destroy the regular season or eviscerate the value of adding a Mookie Betts or a Yu Darvish. But it will have to involve the league designing a postseason that not only gives more teams hope but also makes teams actually care whether they win 90 games instead of 88, or 98 instead of 96. This is absolutely necessary because of how even teams are, relatively speaking. The NBA postseason works with 16 teams because low seeds mainly get destroyed by the top seeds. No MLB team can be the Jordan-era Bulls or peak Warriors.
Three professors, Michael J. Lopez, Gregory J. Matthews, and Benjamin S. Baumer, looked at this specific issue a few years ago and estimated that to match the NBA’s playoff history of the team with the better record 80% of the time, baseball would need to have best-of-75 series. I don’t think I need to go into why that is impractical.
I think that for a larger playoff field to be a good idea requires MLB to continue to re-think what kind of advantages can be given to the better teams. The bye was a clever idea, but there are limitations of how many byes you can give top teams before you’re almost penalizing them by making them cool their jets for weeks. That likely makes a straight-up application of the KBO’s gauntlet also a non-starter. What it might take is for MLB to explore lopsided playoff series. If you’re going to give a chance for a 78-84 team to knock off a 105-win team, you need to think of a bigger hurdle than winning two of three games. Why not require a four-game sweep?
If expanding the playoffs is going to mean more to the owners than simply a way to fatten their wallets and thin those of the players, some creativity will be needed. We’ll see.
Dan Szymborski is a senior writer for FanGraphs and the developer of the ZiPS projection system. He was a writer for ESPN.com from 2010-2018, a regular guest on a number of radio shows and podcasts, and a voting BBWAA member. He also maintains a terrible Twitter account at @DSzymborski.
Dan, can you see how this would differ depending on the number of teams in MLB? I’ve been banging the drum on expansion for a while and while it seems likely now than it was a couple years ago, I’ve thought that the idea of 16 teams making the playoffs was a lot more palatable when there were 36 teams than 30. But maybe because the expansion is so huge it wipes out those gains.
The thing is that the expansion teams are pretty hard to project given Dan is using real rosters – you’d have to project an expansion draft and figure out who comes off of existing squads, etc.
And then -where- you put those expansion teams matters a lot, because there’s a big wins advantage to having one in your divison with the unbalanced schedule.
I, for one, would like to see the advanced stats on closers riding ostriches while jousting.
Apparently ostriches can carry only about 150 lbs. With MLB looking to increase kids’ interest in baseball, perhaps the jousters could be children…
Per this article (http://wearefanatics.com/baseball-players-shapes-sizes/), Ronald Torreyes’s (lightest player at 151 lb) value would sky-rocket in this system.
Free advice: DO NOT f*ck with ostriches. They are Bruce Lee level at kicking people right in the face but Bruce Lee lacked claws.
Cool, thanks!
Well, the results were kind of expected.
Great study, and it explains perfectly why I like the 8-team playoffs. In my opinion, the best teams should get in, and then it should be fun to watch them play each other. I really despise the footballization of the baseball playoffs where teams get byes, there’s a random crapshoot play-in game, etc. Give me an elimination game any day, but make me earn it. I think if you could pair that structure with a disincentive for organizations to undertake lengthy rebuilds and a rejiggering of the economic structure that makes the rebuild such an attractive strategy, both the players and the fans would be happier. That leaves the owners, and, well, that’s the problem, isn’t it?
My thought on the tanking concept:
Teams draft in groups based on win% and those groups orders are based on % increase in payroll. Example:
All teams under a .400 win %
Tigers
Rangers
Pirates
Ranked 1-3 on spending
All teams .400-.449
Mets
Rockies
D’Backs
Orioles
Red Sox
Royals
Angels
Ranked 1-7 by spending (so I’m guessing the Mets without checking) pick next. So the Mets would potentially have the #4 overall pick.
Maybe Payroll as a % of Revenue would be better here ?
“The NBA postseason works with 16 teams because low seeds mainly get destroyed by the top seeds. No MLB team can be the Jordan-era Bulls or peak Warriors. ”
This is exactly why, even as a hockey slappy, I dont like the NHL playoff system. The NHL game is far closer to a 50/50 proposition like MLB than NBA, and low seeds beat high seeds regularly. Similarly to MLB last year, that the best team in the league, Tampa Bay, won the most random playoff year ever shouldnt be taken as predestination. The NHL playoffs are a random crapshoot, albeit an exciting random crapshoot.
My preference is always toward “you have to win something in the regular season to have a chance to win the league.” I find it hard to swallow, especially in MLB with a 162 game season, that a team that couldnt win their division gets a chance to win the entire league.
And just a year before Tampa Bay won it all they were swept in the first round by the 8 seed Blue Jackets after having one of the best regular season records of all time.
Great point. It’s always said that the Stanley Cup is hardest trophy to win because of what you describe.
Would be cool to see a similar analysis like Dan’s for various NHL playoff formats.
Depends on your vantage point I guess – hardest to win for a top-seed, but probably the easiest (along with the World Series) for a “barely eked into the playoffs” team. We’ve seen the occasional Wild Card world champs in MLB and lower-seeded champs in NHL. I don’t expect to see a 6-8 seed NBA champion in my lifetime.
I think it’s described as hard to win more due to the grueling nature of playoff hockey, and the sheer quantity of playoff games you need to win.
I think the lack of off days proved, or at least gave me an idea, that better teams have a much better than 50/50 if you make the playoffs more similar to the regular season.
As a Braves fan, even up 3-1 I wasn’t confident because of the unreal Dodgers depth. A wild card team with 2 starters on fire can carry a series if they pitch games 1,2, 4, 5, and 7 (short test).
Imo the format for the div series should be 4-1 (first 4 at top seed), LCS and WS at 4-2-1.
” The more a championship is decided by randomness, the less incentive there is for teams to innovate and invest.”
I disagree with this statement. Having such a heavy weighting towards best team wins and limiting playoff participants you essentially push any team that isn’t expected to win 90+ games to tear down and rebuild. This leads to the boom and bust cycles we’re seeing in baseball now with a large chunk of teams not even trying to win. You chose basketball as an example of the deterministic nature of play off series but what about football and hockey where randomness and upsets are the norm? I think a lot of the assumptions laid out in this article arent backed up with what we’ve seen in baseball and other sports. If there’s no reward for being just an above average team then few teams in the league have anything to strive for in the short term, limiting free agency spending and attempts to win immediately.
If things are random, then there is no “trying to win” on anyone’s part since trying doesn’t get you much. You don’t end the boom-bust cycle, you just lower the bar for how good a team needs to be on the boom side. Why would any team care about winning 100 games instead of 80 games if it doesn’t really matter all that much?
If games were decided by literal coin flips, then every team would pay every player the league minimum, regardless of the playoff format.
Hmm yes, this argument taken to an unrealistic extreme seems to fall apart, pat on the back, nice job.
In all seriousness though, I think the issues facing the league right now aren’t helped by giving the best teams further incentives to get better and middling teams continuing dwindling reasons to improve. In a perfect world where owners want to win because they’re invested in the fortunes of the team beyond the operating revenue and sky rocketing valuations this wouldnt be an issue. But in a world where cutting pay roll and sitting on profits regardless of team performance is a viable strategy I think there needs to be more of an incentive to try to win. Again using another sport as an example the best hockey teams still add in the off season and try to win. I don’t think baseball teams will just see a roster capable of winning 85 games to squeak into the playoffs and think “Nice, done.”
Dan, I agree with Yaro on this. Obviously, complete randomness is absurd. But it is a “for profit” entertainment business. Owners want profit and prestige. Their employees want to be successful (this mostly means to win… and, of course, make money). Their customers are the fans. The fans enjoy the sport as a whole, but also want their teams to win and have individual players they appreciate and cheer for (e.g. Mike Trout on a not so successful team). While the avid (rabid?) baseball fan may want fewer post-season teams, the majority of fans (customers) want to see their teams have a chance at the post-season. And, owners of those teams would be more inclined to invest if they felt they had a legitimate shot to compete. So, there should be a middle ground compromise that can accommodate all parties. A system that minimizes the number of post-season teams and is strongly rigged against lower seeds will not incentivize more owners to invest, but a system that has more teams in the post-season (with a logical limit, of course) will create more incentive for more fans to be engaged (and that also means more money spent and more money the owners have to invest).
I also agree with Sadtrombone about expanding the league, except only add two more teams for a total of 32 (at least for now). Then restructure the league with 16 teams each in the AL and NL and four divisions in each league. Shorten the regular season to allow for more post-season playing time. Model the post-season like the NFL in so far as who makes the post-season (with 12 post-season teams out of 32 total teams). The two best division winners in each league get byes in the first round. The others play best of five series, no breaks. From that point forward it becomes a normal best of seven series. This method would provide a middle ground compromise for all parties. It provides more incentive for more fans to stay engaged, and more money to incentivize owners to invest.
I literally said that we might have a middle ground now.
Making the playoffs requires the playoffs actually be meaningful. If a World Series is cheaply earned, then making the playoffs is less itneresting.
I think another thing to consider is that baseball fandom is very localized. People are fans of their teams and their interest is directly tied to the competitiveness of their local team. This is why we need the majority of teams to have a whiff of competitiveness. This is in contrast to the NBA where casual fans watch teams from nowhere near them play games because of the star power of the leagues elite. People watch LeBron games because he’s LeBron regardless of the team he plays for. Baseball doesnt have this ability.
Tanking in baseball is nearly always about dollars, and is it worth it to spend them chasing results. Not the results themselves. Has playoff expansion caused marginal teams to spend more of them? I think that’s debatable. It looks to me like teams recognize the crap-shoot nature of the wild card, for example, and aren’t willing to spend to achieve it. Where some of them ARE willing to spend, is when they are already playoff-worthy and want to lock down a division.
The luxury tax actually creates a bit of a perverse incentive that a hard cap would not. With the tax teams can and do spend to exceed it, with the plan that they will subsequently jettison payroll a year or three later when they judge their window to have closed. What the BoSox and Cubs are doing is an entirely expected outcome of such a structure.
Too late to edit:
My unstated secondary point is that in other sports tanking is usually about chasing a high draft pick. Salary floors prevent too much in the way of a payroll tank job, and I think that most fans find tanking for high draft picks to be more palatable than tanking just to save ownership some cash.
There is an optimal range for the balance between luck and skill. If there was no luck, no one would watch a game with such a talent mismatch among teams. If every game was a coin toss, no one would watch either.
Baseball in 2021 has much more of a concern with having too litlle luck than an excess of luck.
Another issue is that the two best teams in the league can be in the same division. I still think the NL West runner up should have some advantage over whoever winds up with the NL Central.
Maybe the top wild card should be eligible to cross over into another division, like the CFL does, and the rest of the wild cards (including the 84 win Cards) can duke it out somehow to see who gets that fourth spot.
Or have four ‘division’ slots, and however many wild card slots, with the fourth division being the ace wild card team.
Relatedly, I think the NHL reseeds the playoffs each round…I believe the NBA does not…and I think the NFL just moved to it last year with the introduction of a 7th playoff team in each conference.
NBA doesn’t really need to give any extra incentive to the top seeds; the #1 seed will generally steamroll the 8th seed and will usually be a heavy favorite over the 4/5 winner. The NHL does reseed to try and create a little order out of the chaos of the playoffs.
The NHL moved to a divisional playoff format in the mid-2010’s, so it’s 1v4 and 2v3 from each of the 4 divisions in the first round; the winners play each other in the second round; and the two remaining teams in each conference play. this year, it’s the same for the first two rounds, but they’ll re-seed 1-4 for the Cup semi-final and the Cup Final.
Thanks for dropping the knowledge! I am admittedly not as up-to-speed on the NHL as the other major sports. It’s interesting watching this year unfold with teams playing a divisional slate only.
The main thing I dislike about the MLB playoffs is the randomness. It’s also the main reason I hate college basketball March Madness. To reduce the randomness, the stronger teams need the playoff format to give them an added advantage. Skipping an entire series is good, but the getting-out-of-rhythm concern seems legit to me. But there are several other kinds of advantages that can be considered. For example, if one team in a best-of-7 playoff series has 10 more regular season wins, then that team effectively starts the series with 1 win so they only have to win 3 instead of 4 while the weaker team has to win 4. If they have more than 20 regular season wins, then they start the series with 2 wins. And there’s a rule that the better team has to win at least one game. This creates real underdogs and some drama if the weaker team wins the first couple games. An alternative would be to start each game of the series with a 1-run lead at the start of each game in the series. Or give them a 2-run start if their win advantage over the opponent is larger, etc. This type of formant gives the stronger team an edge without giving an immediate win. There are many ways this could be done. Note that it can also incentivize stronger teams to keep playing hard at the end of the season even they’ve locked up the division because they can be playing to improve their playoff advantage.
I looooooooooooooooove the chaos of March Madness opening weekend. Seeing Duke lose to a Mercer or a Lehigh? (Or lose to anyone, really?) Gimme gimme gimme.
Add two teams and create four 4-team divisions in each league, for an 8-team playoff field. No second or third place team should ever make the playoffs.
That would be awful. You’re going to get a 75 win division leader in the playoffs, and a 98 win runner up going home.
With 32 teams, you go two leagues, two 8-team divisions, four WCs per league, division winners get a bye on a 3-game WC round.
i think i’d rather see the system the NFL had until this season in a 32-team MLB, i.e. 2 wild cards per league and the top two division winners get a bye, since this gives an incentive for a team not to merely be the best in a crappy division (like whoever will win the NL central this year)
Appreciate the time you took to write (and research) this topic. As much as I hate one game playoffs from a FO perspective, they are exciting. I don’t necessarily think this system is “bad”… but I do wonder if pushing the LDS to 7-game series rather than a 5 would eliminate some of the randomness with ultimately crowning a winner.
What’s wrong with a best of 75? We would get to have discussions like “yeah, that series was over in 58 games, but it was closer than that portrays! The series run differential was only 34!”
Sigh.
I wonder if the focus on World Series % is misguided, though. You’re effectively looking at how many already strong teams will be in a position to gear up for a push deep into the playoffs, right? Having your eye on the final prize makes more and more sense, as you reduce the number of teams involved.
But I wonder if, as the number of teams increases, it instead makes more sense to look at the effects of adding a marginal win on simply *making* the playoffs. Given how unpredictable and random the playoffs would be, I would think teams would shift their attention to simply qualifying? (And if the number of teams incentivized to improve is greater, in this instance, than in the more exclusive playoff systems.)
But do they? Did the wildcard expansion spur on a payroll surge in teams sitting on the margins of that worthy contest? I don’t believe that it did. Not much.
Now then, a lot of that disinterest is due to the coinflip nature of the game. If more playoff spots were implemented as multi-game series, you might see more of an impact.
You look at win probability as it pertains to winning the World Series. Winning the World Series is certainly one incentive for teams to invest. Although if you assume most teams are profit maximizers (which is fairly debatable for certain owners), then you probably need to consider other incentive points as well.
For example, we are fairly confident that there is a discontinuous revenue bump (for the current, yes, but also for the subsequent season) for making the playoffs. The bump is even bigger for winning the World Series. Your different proposals show how they impact world series probabilities, but you should also show how they impact the probability of making the playoffs. What you could potentially show, in this theoretical exercise, is that the incentive to invest in talent may be dependent on the relative returns to making the playoffs versus the relative returns of winning the world series.
You touch on this when you mention the issue with just having the top two teams play in the World Series. My point is that the incentives to invest within a 16-team format versus the 14-team format versus the current 10-team format cannot be analyzed just by looking at World Series probabilities.
But there’s a bump for the playoffs *because* it gives you a chance at the World Series. Playoffs only matter because they come with a chance to win the World Series. Nobody trades more playoff appearances for *fewer* World Series. And the more we cheapen the World Series, the less it matters. You could call the team that wins the most games in spring training the High Lord Emperors of Springtime Fancies and give them a massive trophy, but people won’t care unless there’s a perception that it’s meaningful.
The purpose of playoffs *is* winning the World Series. Playoffs mean nothing when divorced from the context of winning a World Series or a similar accolade that is perceived as crowning a champion.
It’s like saying an ice cream company shouldn’t just focus on ice cream but also on ice cream scoops. But ice cream scooper only exist because of the demand for ice cream. Without the ice cream, any improvement on the ice cream scoop is irrelevant.
I think that’s certainly part of it, but I still question whether making the playoffs itself lacks revenue consequences outside of its implications for potentially winning the World Series.
A good example from a consumer demand perspective is March Madness. In that case, the expectation of winning a championship for the majority of the teams in the tournament is basically zero. And yet the bracketology industry seems to perform fairly well, especially for teams on the bubble (arguably teams on the bubble garner more interest than teams with better odds not on the bubble). Based on this, there is clearly some sort of interest (leading to revenues) in simply MAKING the event outside of its impact on the probability of winning a championship. What that is for baseball is in question (you are arguing it is zero), but I wouldn’t claim definitively it is zero without further investigation.
I make this argument to highlight the consumer preference argument. Incentives to compete are obviously different for college basketball, but presumably, consumer preferences to making the playoffs would parallel, no?
But making it *is* an event when the spots are scarce. Fairfield getting into the NCAA tourney isn’t anywhere near as big a deal if half the MAAC is getting into the tourney every year. There are 351 Division I basketball teams. That’s like an MLB playoff with 5.8 teams (realistically six, of course)
Imagine that the NCAA announce that they’re expanding March Madness to 187 teams (16/30th of the pool). “Just making it” is a lot less exciting now.
Agreed. But this contradicts your earlier point that consumer interest for making the playoffs only feeds through the probability of winning the World Series. If that were the case, this “making it” event would not exist.
I agree the value of this event is diluted for NCAA if they expand to 187 teams. But can you say definitively that it is still zero? And going back to baseball, this effect should be different depending on how many teams DO make it in. So under the 16-team playoff, there is both a diminishing incentive to invest due to decreasing WS odds AND less value in actually making the playoffs (although that is counteracted by more teams making it, so it once again depends on the magnitude of these effects to calculate total incentive changes).
I don’t think it contradicts my point. The World Series is the lure of a playoff appearance in baseball *because* there are already far too many teams to make it an actual novelty factor. A third of teams make the playoffs and more than half did last year. Maybe if were talking two or four team playoff systems.
“Do fans care about making the playoffs besides for the opportunity to win a championship?” Seems like a good empirical question to test. Can look at how playoff exclusivity (field size) impacts it as well.
You’re right, though I do think there’s still marginal value increases in the wide playoff format of increasing your odds of being in the field to benefit from the RNGfest.
You don’t need to look at stuff outside of WS odds to measure that, but look at the scenarios where a team’s odds increase by an enormous relative magnitude: from 0.5% to 1%, etc.
If the value of the marginal win is doubling your odds (even if they’re bad), it can be worth it.
That’s probably the one piece of analysis that is worth adding to what you did here.
If the point of adding more playoff teams is to get more playoff games, why not make each round of playoffs longer? There were best of 9 game World Series once upon a time right? Larger sample should give better, deeper teams an advantage. No off days in all rounds of playoffs (like 2020 except WS) ought to also benefit deeper teams.
9 game WS
7 game LCS
7 game Division series
1&2 seeds in each league get by, 3 plays 6 & 4 plays 5 in best of three Wild card. All games hosted by better seed.
12 team playoff. 63 max total playoff games. Incentives for top 2 in each league with the by round. And for 3/4 seeds to host wild card series.
This year’s Christmas Day world series game, sponsored by Macy’s!
A win is not a win, especially in the 5 team per league format. If a team get a win with a better 5th starter, this has almost no impact on their playoff success. If a team gets a win with a better ace starter – like Yu Darvish for Davies – the playoff success odds can change radically.
I think the argument is that if a team improves their total underlying talent by “1 win”. Where it comes from is relevant to the extent you mentioned, but for this theoretical exercise, it probably isn’t too big of a deal to assume that the same “1 win increase” would equally reflect itself for any team in the league.
One thing I really liked about last year’s format where the off days were cut out before the World Series was that it tested the roster in a way much closer to regular season construction.
Part of the reason the Dodgers and Rays advanced to the World Series despite the wider playoff field is that the playoff format better rewarded the -reasons- they had great regular season records (ie significant team depth) in ways the regular playoff format didn’t.
The year before, a nearly equally-deep Dodger team lost in 5 games to a Nationals team that only used like 6 individual pitchers in the entire series because their bullpen was so unreliable. That Nationals team, as good as it was, would have had absolutely no chance in the 2020 playoff format as they would not have been able to concentrate all of their innings onto the few reliable pitchers on their staff the way they did in 2019.
Yeah, I think the limited off days was a lot of fun, and could be a good way to incentivize team depth. It would be cool to see numbers around what that does to the marginal value of a win.
An interesting extension of this discussion would be to fold in auction theory. The MLBPA should be looking to maximize the total returns from free agency over time, which is an asymmetrical auction problem. (note that viewing this over time allows us to generally ignore the issues of different players and different positional needs in any given year, though it does amplify issues like pitcher injuries, reliever volatility, and varying market sizes, all of which can be adjusted for systematically).
But the observation is that there’s a game theoretic sweet spot between the number of teams which see some marginal win value vs the magnitude of that value on a per-team basis. You want to create a market in which free agents have more viable bidders, but you want those bidders to actually be incentivized to spend something meaningful. Auction theory provides a framework in which to answer these questions.
Combining these two works together would actually give a pretty nice way for MBLPA to consider the expected value over time of various proposed playoff formats, which they can then weigh against the expected direct monetary return from the profit sharing, and finally fold all of this in with the negotiation calculus of using an expanded postseason as a bargaining chip to get other things with quantifiable value.
At least, if I worked for the MLBPA, that’s what I’d do.
A suggestion for re-balancing high-seed playoff advantage in a “wild card” playoff that I kind of like is:
Best-of-3 series
over (up to) two days, with Day 2 being a double header
All at the higher seeded team’s park
Higher seed only has to win one game, lower seed has to win 2
Alternately (in order to guarantee play on two days), higher seed has to win 2, lower seed has to sweep.
Aside from anything else, I doubt MLB would agree to anything that has a playoff game starting as early as a double-header would have to, but it would provide a pretty exciting format.
You could have an expanded playoff and greater regular season impact via a pool approach imo.
Top 8 in each league get in.
3 div winners play double round robin to seed 1-3 hosted at the top seed. You’d have double headers and schedule it to favor the top team. So like:
The 5 wild cards play their own round robin at the top WC park to see who advances. These are all played consecutively so it still favors deeper teams.
Day 1: EvF, GvH, D off
Day 2: FvG, HvD, E off
Day 3: GvH, DvE, F off (hurr hurr)
Day 4: HvD, EvF, G off
Day 5: DvE FvG, H off
8th seed (H) basically doesn’t get an off day. Top WC seed (D) starts day 2 after everyone else used their ace on day 1
Then after that it’s business as usual with div/lcs/ws
Maybe we can approach the “ideal postseason format” from another angle.
Let’s say the goal is to make sure that the best team in the league (as determined by seasonal Fangraphs WAR or some other measures with less luck interference), after a 162-game season, has >99% chance of getting into the postseason. I use 99% as the MLB is slightly more than a century old.
We should be able to simulate the chance of the current best team (and all single-season best team in the past) getting into postseason, and compare.
I think in any and all playoff formats with wild cards, the “true talent” best team will be in the playoffs >99% of the time. 162 games is such a large sample, and the addition of Wild Cards mean we just don’t see the “hard luck” teams with the second best records left out of the playoffs anymore like we did in 1993 (SFO with 103 wins!) or 1987 (Toronto with 96). Definitely could have happened pre-wild card though.
I think the easiest/most obvious penalty/reward in the early rounds is to give the team with the better record home field for the entire round. That also helps the system push through to the CS and WS where we can re-add travel and shorten playoff rotations to the classic 3.5 SP.