When are NBA playoff series actually decided? Is Game 1 more important than Game 2 or Game 3? Such questions are frequently debated by fans, but it's hard to definitively say what essentially finished a series.
In order to make these determinations, I compiled a database, logging the game-by-game results of every single postseason game since the playoff format changed in 2003. That's when the opening round was expanded to a seven-game series. This gives us a sample size of 150 series and 839 games to consider.
There are a few different ways of approaching this, so I used them all. And of course, I used pretty graphs to make it all easier to absorb.
The Regular Season Really Does Matter
One way of determining whether the regular season "matters" is to see if the more successful regular-season team had a more successful postseason. That's easy enough to do just by determining the frequency with which the favorite wins the series.
For the most part, the favorite is the higher seed and the underdog is the lower seed, but there were a few occasions where the lower seed had the home-court advantage. For the purposes of this article, the team which had the home-court advantage was always considered the “favorite,” regardless of what the seeding was or the betting odds.
Here is what the result looks like in a simple pie chart:
I actually found it surprising that there have only been 37 series upsets over the last decade, but considering it’s about 3.7 per year, it sounds about right. But what round do they come in? And what is an upset versus an upset?
Do More Upsets Occur with Less Seed Differential?
In order to quantify that, I looked at seeding differential, where a No. 8 seed upsetting a No. 1 seed would have a value of seven, and a No. 5 seed upending a No. 4 seed had a value of one. Here are the results for that (In case you’re wondering how you can have a difference of fewer than one seed difference, there are times when teams with the same seed in their respective conferences squared off in the finals):
Not surprisingly there is a pretty significant difference when the seed differential is smaller. Nominal upsets happen with greater frequency than major upsets.
This suggests that landing a top-two seed is worth the struggle. Whether you're battling a No. 7 seed or a No. 8 seed doesn't matter much in the grand scheme of things.
However, getting home court in the second round still appears to be helpful. Once you get past that though, it's pretty much a toss-up. This suggests that going all-out to push for the No. 1 seed isn't advantageous.
Do Upsets Occur More Frequently in Later Rounds?
This raises the question of whether upsets happen more frequently the deeper you get into the postseason. So here is a look at the winning percentages by round. There is a clear indication that home court means a lot more in the first two rounds. As teams get deeper, they know how to win on the road, and home court means less:
Do Higher Seeds Win More Games?
Another way of viewing this by looking at how many games on average each seed wins in the postseason. Here is a look at that:
It seems pretty conclusive once again that the regular season matters. The higher the seed, the deeper teams tend to go. It is marginally interesting that there has been more success from No. 8 seeds than No. 7 seeds, but that’s partly an anomaly due to Derrick Rose’s injury last season, which followed the upset of the Spurs by the Grizzlies in 2011.
While the regular season matters, it’s not the only thing that matters. There’s not a vast difference between the top two seeds in each conference. In fact, when the top two seeds have squared off over the last 10 years, the underdog has won seven of 11 times.
Which Seeds Go to the Finals?
More evidence pointing to the relevance of the regular season is that of the 20 teams who have made the finals, all have been at least a No. 4 seed. Here are the trips and records in the finals by seed number:
This supports the finding that a No. 1 or 2 seed is far more important than anything else. In the current format, 65 percent of NBA Finalists have been one of the top two seeds. Of the No. 3 seeds who won the title, only the Pistons were the lower seed in both the Conference finals and the NBA Finals. In only one instance has the No. 2 seed won both series without home court.
All of this indicates that winning successive road series is very difficult, particularly as you face tougher opponents. This explains why even though the No. 2 seeds get to the finals more often than the No. 1 seeds—they tend to lose once they get there.
The bottom line is that getting the top seed is nice, but by no means a guarantee. However, if you don’t finish at least third, you can pretty much write things off.
Looking at seeding tells us a lot about how much the regular season matters, but what happens after the postseason starts?
Series Wins by Game Score
The next thing I wanted to do was look at who wins when the series is at a specific game score. For the chart below, the number on the left is the number of games won by the favorite. The number on the right is the number of games won by the underdog.
So a 2-0 score means the favorite is up two games to none, and a 0-2 score means the underdog is winning. The corresponding numbers represent how often the teams went on to win the series. When the series was tied three games apiece, the favorite won the series 23 of 28 times.
Here are the game-by-game results.
From this it’s difficult to say too much. However, it does seem that if the underdog can steal the home court in one of the first two games, they improve their chances significantly. When they take at least one of the first two games, the series is fairly even, 40-31.
Which leads to the question, what happens if a team wins each game? Here's a look at that breakdown:
What is most intriguing here is that it seems that Game 3 is the most irrelevant game, which was completely contrary to my expectations. Looking at the same data while focusing on winning percentage rather than totals shows this more clearly:
Surprisingly, Games 3 and 5, which would logically be the most pivotal, appear to have the least impact on who wins the series. However, what this does not account for is that not all of those games are the same.
When a team has dropped the first two games of the series, they are desperate in Game 3 and usually come out fighting. When teams are playing with their backs against the wall in Game 3, they are 50-35.
When you look at the series winning by who is winning after the Game 3, we see that Game 3 is actually the most pivotal game of all. Whoever is winning after Game 3 has won 126 of the last 150 series, the most of any game number.
This actually makes sense. After Game 3, both teams have two games left on their home court, regardless of who started with it. All they have to do is hold serve, and they win. If it’s the favorite up 2-1, then they’ve already shown that they can either protect their court or steal one on the road. If the underdog is up, they have already shown they can win on the road.
When you break out just the rubber games, you see how much impact they have in determining the series.
Based on all of this, we can conclude that the rubber matches are critical. Only 23 percent of teams overcome slipping from a tied series to being down one after three, and only 15 percent bounce back from losing Game 5 in a tied series.
While there is no one singular game which guarantees a team will win a series (apart from Game 7 of course), there are some pretty good indications here: Having the home court helps, and winning two of the first three seems to make a big difference too, regardless of who has home court.
Morsels for the Curious
Here are a few intriguing little tidbits that I uncovered that I just found interesting and thought I’d pass along:
*There are 70 different ways that a series can play out in terms of winning and losing series games.The most frequent pattern is the series sweep, which has happened 22 times.
*There have been 20 series which have followed the exact pattern of the favorite winning two, the underdog winning one, then the favorite winning two. This is the most frequent non-sweep pattern.
*There have been three “reverse-sweeps” where the underdog won every game.
*Nine series never “got started,” if the adage about the series doesn’t start until the home team loses a game holds water. In fact, the 2008 Boston Celtics didn’t even have a series start until the second game of the conference finals, as the home team won their first 15 games.