How to Make the NBA All-Star Game Competitive
Players don’t compete because they don’t care. To fix this, the All-Star Game should impact players' chances of winning a championship. Let’s see what the math says.
The NBA All-Star Game remains broken. It doesn’t make sense for players to risk their bodies for a game whose outcome doesn’t matter to them. The result is lackadaisical defense and an abundance of uncontested three-pointers and dunks.
The question then becomes: How do we make the players care? Here’s a rundown of recent failed attempts:
For many years, the All-Star Game pitted the Eastern conference against the Western conference. Players didn’t really care.
From 2018 to 2023, captains (leading vote-getters) drafted their own teams. The captain’s honor was at stake! Players didn’t really care.
This past weekend, there were four teams that played in a two-round tournament. Players didn’t really care.
Looking to the future, there’s talk of a U.S. vs. international friendly game, modeled on the NHL’s “4 Nations Face-Off.” Maybe players will care?
With all this failed meddling, I can think of only one real way to make players care about the All-Star Game. Make it impact their teams’ actual chances of winning a championship. These days, NBA championships are what players care about most—or at least that’s what they say. Improving the winning players’ odds of a championship is the sort of incentive the NBA should seek to provide via the All-Star Game.
One way to quantify the importance of the All-Star Game is to look at how teams’ chances of winning a championship change if their representatives win the All-Star Game. If an All-Star Game is important, this impact will be great; if an All-Star Game is unimportant, this difference will be close to zero. For the NBA, this difference is currently zero.
Baseball’s All-Star Game Provides Only Modest Incentive to Play
The NBA isn’t the only major sporting league with an All-Star dilemma. In particular, Major League Baseball experimented with allowing their All-Star Game to impact championship odds. From 2003 to 2016, whichever team won—the American League team or the National League team—that league’s representative in the World Series was granted home field advantage, meaning they potentially got to host more games than their opponent.
That sounds like a big deal, right? Let’s see what the numbers say.
The best-of-seven World Series has a “2-3-2” format, meaning one team hosts the first two games, the other team hosts the next three, and the original team hosts the last two. By winning the All-Star Game, the corresponding league got to host up to four home games in the World Series (games 1, 2, 6, and 7), while the other league only hosted up to three (games 3, 4, and 5).
Home field advantage in baseball is relatively modest compared to basketball. In the 2024 regular season, the home team won 1267 of 2429 total games, or about 52.16 percent. Assuming the home team has a 52.16 percent chance of winning any given game, what is the probability the team with home field advantage wins the World Series against an evenly matched opponent?
This is a classic sort of problem here at “The Fiddler.” (Indeed, it resembles a 2021 puzzle from FiveThirtyEight’s “The Riddler” about how many games the World Series lasted, on average.) Let’s work it out briefly, supposing more generally that the home team wins each game with probability p. In particular, we want to find the probability that the team that hosts games 1, 2, 6, and 7 wins the World Series as a function of p.
The team wins in four games with probability p2·(1−p)2.
The team wins in five games with probability 2p3·(1−p)2 + 2p·(1−p)4.
The team wins in six games with probability 3p5·(1−p) + 6p3·(1−p)3 + p·(1−p)5.
The team wins in seven games with probability p7 + 9p5·(1−p)2 + 9p3·(1−p)4 + p·(1−p)6.
Since a team must win the World Series in precisely four, five, six, or seven games, the probability that the team hosting games 1, 2, 6, and 7 ultimately wins is the sum of the four expressions above. Here’s a graph of how this probability varies with p:
If the home team always wins (i.e., p = 1), then the team that hosts four games is guaranteed a victory. And if there’s no home field advantage (i.e., p = 0.5), then both teams have an equal chance of winning the World Series. But thanks to the curvature of the graph, for all other values of p greater than 0.5, the team that hosts games 1, 2, 6, and 7 wins the World Series with a probability less than p.
So if the value of p is indeed 0.5216 (based on the 2024 regular season), the team that hosts games 1, 2, 6, and 7 has a 50.68 percent chance of defeating an evenly matched opponent. Meanwhile, the team that hosts games 3, 4, and 5 has a 49.32 percent chance of winning.
In other words, winning the All-Star Game and gaining home field advantage increases your chances of winning the World Series by about 2.7 percent compared to losing the All-Star Game.
It’s not zero, but it’s not much.
Applying Baseball’s Rules to Basketball Wouldn’t Do Much
Perhaps the NBA All-Star Game could be made more competitive by following baseball’s example. You could have the Eastern Conference face off against the Western Conference, with the winning conference hosting up to four games in the NBA Finals.
While the outcome of the All-Star Game has only a marginal impact in baseball, as we just showed, there are two reasons awarding home field advantage might work better in basketball. First, the NBA Finals operate in a “2-2-1-1-1” format rather than baseball’s 2-3-2 format, meaning one team hosts games 1, 2, 5, and 7, while the other team hosts games 3, 4, and 6. Second, home field advantage is greater in basketball than in baseball.
It turns out that the order of the games doesn’t matter—at least in theory. As long as one team hosts four games and the other team hosts three, their respective chances of winning the series are independent of the order in which those games are played. Of course, the order of the games could affect a team’s psychology, jet lag, etc. But according to the straightforward mathematical model, the 2-3-2 format and the 2-2-1-1-1 format produce equivalent results regarding the teams’ respective probabilities of victory.
As for home field advantage, the home team won 668 of 1230 total games (about 54.31 percent) in the 2023-24 regular NBA season. Home field advantage appears to have roughly twice as great an effect in basketball as it does in baseball.
Again, for simplicity, let’s assume all 30 basketball teams are equally skilled. If your team’s conference happens to win the All-Star Game and you now enjoy home field advantage in the NBA Finals as a result, your chances of an NBA title are about 5.6 percent better than if your conference had lost the All-Star Game.
It’s greater than the impact in baseball, but I’d still call it fairly modest.
Incentives for a Competitive All-Star Game
If awarding home field advantage in the NBA Finals is insufficient motivation for players to demonstrate effort, there are other options.
It’s worth noting that home field advantage in the NBA Finals would only benefit a team once they actually made the Finals. An alternative incentive could help teams make the playoffs, which would additionally enhance their chances of winning a championship.
I’m old enough to remember a time when the Western Conference was stacked, and the Eastern Conference was known as the “Leastern Conference,” dominated by a LeBron James team year in and year out. In the 2009-10 regular season, when the top eight teams from each conference advanced to the playoffs (rather than the current play-in format), the eighth seed in the East (the Bulls) had a record of 41-41, while the eighth seeds in the West (the Thunder) had a records 50-32. At the time, there was some chatter that the top 16 teams by record—regardless of conference—should advance to the playoffs.
But what would happen if the All-Star Game helped make this determination? Suppose the game returned to its East versus West roots. If the Eastern Conference wins, then the East advances nine teams to the playoffs, versus seven from the West. If the Western Conference wins, then the West advances nine teams to the playoffs, versus seven from the East.
First off, how would this impact a team’s chances of winning a championship?
For simplicity, let’s again assume all teams are of equal ability. To win the championship, a team must first qualify for the playoffs, and then emerge victorious from the playoffs. There are 15 teams in each conference, and let’s suppose eight from each conference advance to the playoffs. Then each team’s chances of winning the championship is 8/15 (the probability of making the playoffs) times 1/16 (the probability of winning among the 16 teams that made the playoffs). This product is 1/30, which makes sense—each of the 30 teams has an equal chance of winning the championship.
Now, let’s suppose that the All-Star Game determines which conference gets to send nine (as opposed to seven) teams to the playoffs. If your conference wins the All-Star Game, then your team’s chances of winning the championship are now (9/15)·(1/16), which simplifies to 3/80, or 3.75 percent. If your conference loses the All-Star Game, then your team’s chances of winning the championship are now (7/15)·(1/16), which simplifies to 7/240, or about 2.92 percent.
While these probabilities remain fairly small, the relative difference between them is significant. Winning the All-Star Game means your team has a 28 percent greater chance of winning the championship than if your team had lost the All-Star Game.
These stakes are about 10 times greater than those the baseball All-Star Game experimented with. Would players be properly incentivized? I think so!
Of course, there are some drawbacks to this proposal. First, I don’t know if any of this is even possible, given the current collective bargaining agreement between the NBA and its players. Fine.
Another potential issue would be perverse incentives. Suppose a team currently leads the standings in their conference. Any of their players who are selected to the All-Star Game might not care so much about preserving the eighth and ninth playoff spots for their conference. Worse, they may prefer that the eighth- and ninth-seeded teams in the other conference make it to the playoffs instead of the corresponding seeds in their own conference. That could incentivize them to not try (or even throw?) the All-Star Game.
A remedy for this would be to hold the All-Star Game earlier in the season, perhaps a third of the way through. This way, teams would not be as confident about their ultimate placement in the standings. They would strictly prefer that more teams from their own conference make the playoffs, thereby improving their own chances in case they have a losing streak down the stretch or one of their star players is injured.
And, as I already suggested, an added bonus would be that the playoffs could reflect any lack of parity in the conferences. If the West really is so much better than the East, then they can prove it at the All-Star Game, and more West teams will make the playoffs.
I can picture it now: A return to East versus West in the All-Star Game. The winning side gets more teams in the playoffs. The best of the best battle it out, knowing their championship odds are on the line in a meaningful way.
Come on, Adam Silver—make it happen!
I was just thinking about this the other day! I want to point out that out of the four major sports leagues in North America, football and basketball have a deeper inherit all-star indifference problem than baseball and hockey. This is because it's really hard to "not try" to play your best in baseball or in hockey. For baseball, pitchers don't deliberately toss the ball to the batter in the all-star game. They have to pitch as normal, and batters have to bat as normal. Plus, each player might get to play just one (or even half) inning, making the experience more precious for them and motivating them to make that inning count. For hockey, the sport is played on a low-friction surface already so even if you want to go slow, you can't. Also, you are on the ice for at most 30 seconds and it might be the only 30 seconds you will get for the entire game, so you are motivated to make it count.
Basketball and football, on the other hand, are not played on low-friction surfaces. It's okay for one to not play defense to reduce the risk of injury. This makes for boring games. The problem with NBA is that, for NFL, they figured out the situation and made the pro-bowl a casual flag-football event serving as an appetizer one week before the Superbowl. For NBA, it's still marketed as a marquee mid-season matchup. So, NBA has gotten the worst deal out of the four sports simply due to the nature of the game.
One risk with this proposal is that the a good performance in the All-Star game by the players from the weaker league would lead to even more skewed playoffs with 9 teams from the weaker league and only 7 teams from the stronger league.
To use your example of 2009-10, the East did win the All-Star Game. That would have removed the 50-32 Thunder from the playoffs, and included the 40-42 Raptors instead.