Does Playing Well Going Into the Playoffs Matter?
Posted by Neil Paine on April 21, 2010
One aphorism about the playoffs in any sport is that you can't just "flip the switch", or shift some imaginary team "gear" from neutral into overdrive the moment the postseason begins. Instead, you need momentum -- you need to be healthy and operating like a well-oiled machine going into the playoffs, so that you can peak in the postseason and hopefully win a championship. Even one of the NBA's most notorious examples of "switch-flipping", the 2001 Lakers, won 9 of their final 10 games going into the postseason before unleashing the most dominant playoff performance ever on their opponents. So it seems like a no-brainer: you can't succeed in the playoffs unless you played well at the end of the regular season.
But let's test it. Here are the regular-season winning percentages for all playoff teams from 1984 (when the playoffs expanded to 16 teams and 4 rounds for all teams) through 2009, broken down by time period and playoff rounds won:
Rounds Won | Jan-Apr | Feb-Apr | Mar-Apr | Apr | Last41 | Last25 | Last10 |
---|---|---|---|---|---|---|---|
0 | 0.556 | 0.558 | 0.561 | 0.565 | 0.556 | 0.561 | 0.560 |
1 | 0.621 | 0.624 | 0.633 | 0.631 | 0.623 | 0.630 | 0.629 |
2 | 0.674 | 0.677 | 0.675 | 0.671 | 0.678 | 0.675 | 0.666 |
3 | 0.696 | 0.687 | 0.684 | 0.663 | 0.695 | 0.682 | 0.656 |
4 | 0.745 | 0.753 | 0.755 | 0.708 | 0.750 | 0.749 | 0.724 |
As you can see, there's a definite pattern between playing well at the end of the season and advancing deeper into the playoffs. But there's a glaring problem with this chart: teams that advance further into the playoffs are better teams anyway, and therefore are more likely to have played well at the end of the season simply because of this fact, not because of any momentum effects. We need to control for the team's overall performance and isolate teams that played better in the final stages of the regular season than they did at the beginning of the year if we really want to measure the importance of momentum. This means comparing the percentages above to the teams' winning percentages during the rest of the season (i.e., taking WPct in January through April and subtracting WPct in October through December, etc.):
Rounds Won | Jan-Apr | Feb-Apr | Mar-Apr | Apr | Last41 | Last25 | Last10 |
---|---|---|---|---|---|---|---|
0 | 0.015 | 0.012 | 0.013 | 0.013 | 0.010 | 0.014 | 0.010 |
1 | 0.008 | 0.012 | 0.020 | 0.013 | 0.011 | 0.018 | 0.013 |
2 | -0.005 | 0.004 | -0.003 | -0.007 | 0.003 | -0.002 | -0.012 |
3 | -0.001 | -0.017 | -0.018 | -0.042 | -0.003 | -0.021 | -0.046 |
4 | -0.005 | 0.014 | 0.013 | -0.043 | 0.010 | 0.005 | -0.024 |
Well, that's a very different story... As you can see, there isn't a great deal of a relationship between performance down the stretch of the regular season and playoff rounds won. Although eventual champions did tend to play at their best from February on, the teams that performed even better down the stretch were the ones who lost the earliest in the playoffs, and conference champions who lost the Finals tended to play worse throughout the 2nd half of the season! Perhaps this is because strong teams who have nothing to play for after locking up their seed (the kinds of teams who tend to make, and win in, the Finals) ease off the gas pedal down the stretch, while teams fighting to get into the playoffs scrap and claw for every win in the final stages of the regular season. But whatever the reason, it doesn't appear that you have to necessarily be at your best at the end of the regular season to be successful in the playoffs, just as long as you bring your "A" game once the postseason begins.
April 21st, 2010 at 7:22 pm
As far as last 10... you get a lot of teams resting their starters, like the Cavs dropping their last four with LeBron on the bench.
April 21st, 2010 at 9:15 pm
Methinks team 'hot streaks' do not typically last nearly as long as the entirety of the playoffs.
So, if you have had a 10- or 20-game hot streak to end the season, there's actually little chance that streak will last another several weeks.
I'd say it's essentially random. History shows that deep playoff teams have tended to cool off toward the end of the season. Or maybe 'last 25 except for last 5' would be the best indicator.
Optimally, a team gets better as it advances thru the playoffs. Rather like the '95 Rockets.
April 22nd, 2010 at 7:31 am
I very much applaud this research.
But I have been wondering; is there a way to take the "playoff over-achievers" (i.e. the teams that had the most playoff wins per season wins: think of teams such as 95 Rockets and 01 Lakers), and see how well they played at the end of the season? In other words, compare the overachievement in playoffs to overachievement during the end of the season?
I believe that the Magic shall succeed in these playoffs like a 60+ win team, because their second half of the season was stellar.