Compared to the one-and-done format of the NCAA tournament and the NFL playoffs, the NBA's best-of-seven series are designed to reward the better team, not the more fortunate one on a given day. Even the NHL, which also uses the best-of-seven format (along with MLB's League Championship Series and World Series) sees the better team win less frequently because the role of randomness is larger in a single game in those sports than the NBA.
But there's a danger of the notion that the better team wins an NBA playoff series descending to circular logic. Who wins a series? The better team. How do we know they're the better team? Because they won the series.
We know that isn't always the case. True upsets happen, even in a best-of-seven. But how often?
Simulating the Best-of-Seven
The simplest way to answer that question is to look at how often the lower-seeded team wins a playoff series: 74.4 percent of best-of-seven series, per WhoWins.com. That's much higher than the corresponding figures in the NHL (64.9 percent) or MLB (55.9 percent).
However, we also know that having a better regular-season record doesn't always translate to being a better team, both because of randomness and factors like injuries. Sometimes an upset is in fact a case of a better team demonstrating its ability.
If regular-season results aren't sufficient to determine true talent (and they aren't), then we need to control it ourselves. Using the same simulation I run to estimate playoff odds, and randomized point differentials based on the average distribution during the past decade, I played out 1,000 seasons through the playoffs. For each series, I tracked whether the team with the better true point differential won the series. That happened almost an identical amount to total upsets -- 75.2 percent of the time.
Odds better team wins
As the chart at right shows, the percentages decrease throughout the playoffs, for fairly obvious reasons. When teams are closely matched, an upset is more likely. It's also easier for the team with the lesser true ability to win a few more games during the regular season and claim home-court advantage in a 2-3 or 1-2 matchup. But even in the first round, the lesser team advances 21.5 percent of the time -- typically two of the eight series.
Making decisions based on playoff results
All of this might be nothing more than an academic exercise were it not for the decisions that hinge on the outcome of playoff series. Tonight, the Golden State Warriors, Indiana Pacers and Oklahoma City Thunder could be eliminated, while the Houston Rockets will try to extend their series Friday in Portland. If these favorites lose, speculation will mount about the fates of their coaches -- Mark Jackson, Frank Vogel, Scott Brooks and Kevin McHale -- respectively. Other teams will react to early departures by overhauling their rosters.
None of those moves are necessarily mistakes. Brooks' shortcomings as a tactician late in games are well documented, Jackson's rotations remain an issue into the postseason, Vogel has presided over one of the great second-half collapses in league history, and McHale has been outcoached by Blazers counterpart Terry Stotts.
But if these decisions are being made on the basis of playoff outcomes -- and there is widespread speculation Jackson needs to make the conference finals to stay, while sources told ESPN.com's Marc Stein that Vogel is "coaching for his job" in this series -- that's clearly a problem.
The close margins of this year's playoffs, as documented by Tom Haberstroh in Per Diem earlier this week, have ensured that randomness has played a larger role than usual in the outcome of series. If your decision on whether to change coaches hinges on whether Serge Ibaka's shot was before or after the buzzer, or whether Stephen Curry was fouled on a 3-pointer, or whether James Harden makes the final shot, it's probably not a very good one.
Odds better team win in '14
Because of the disparity between the conferences, this year's matchups were bound to be more random than usual. Running the same simulation with 2013-14 true talent ratings, the weaker team wins the playoff series 27.7 percent of the time, including 26.9 percent of first-round series. (The NBA Finals is actually likely to be far less random than usual because of the West's strength.)
For this season, none of this matters. Superior teams that lose a series to a lesser foe who matches up better, gets hot at the right time or catches important breaks won't get a consolation trophy. In terms of maximizing a team's potential going forward, it could be crucial. While it's tough to be patient in the wake of a disappointing playoff loss, overreacting to randomness could be costly for years to come.