NBA Playoffs Upsets: How Many are There and Why do They Happen?

After years of hoping and sort of promising I would eventually have hard proof, in 2010, thanks to the Quest for the Ring (QFTR) NBA Playoffs Games, Series, and Coaches Database, I now have and am delivering hard proof that results of NBA playoff games and series are sometimes decided by which team has the better coach.

Almost 9% (not far off from one in ten) of playoff games are where the difference in the coaching causes the worse team (the underdog) to defeat a better team (the favorite). This means that almost 9% of all series (or 9 games out of 100) are won by the team with the inferior players because it had a better coach or coaches than did the team with the better players. To put it another way, the coaching difference upset the player difference and the team with the worse players won those games.

Stepping back and looking at the even bigger picture, about 24.6% (not far from one in four) of all playoff games were and will continue to be in the future won by the wrong team. Among all these upsets, there are three causes in general: the coaching, the injuries (with the very unusual suspensions included in this category so technically the category should be called “players unavailable”) the players (separately from coaching). In numbers, there were 537 playoff game upsets from 1980 through 2010 out of a total of 2,186 playoff games. 1,649 playoff games were won by the favored team.

One of the many outputs of the Playoff database is that QFTR has produced a truly and very good estimate of how many of those upsets are due to coaching. This was reported out in the previous paragraph as almost 9% of all playoff games. This turns out to be slightly more than 35% of the upsets. In numbers, there were 189 playoff game upsets due to coaching and nothing else from 1980 through 2010 out of a total of 537 upsets and out of a total of 2,186 playoff games.

Now if I make a rough but reasonable and educated estimate that among the upsets not caused by coaching three fourths of them are due to injuries (hurting the favored team more than the underdog) and the other fourth of the upsets are due to the players of the underdog outplaying the players of the favorite team (unexpectedly and independently from coaching) than the following overall result is achieved:

NBA PLAYOFF GAME UPSETS BROKEN DOWN
Total of all Upsets: 24.6% of all playoff games are upsets
Upsets Due to Injuries: 48.7% of all upsets which is 12.0% of all playoff games
Upsets Due to Coaching: 35.1% of all upsets which is 8.6% of all playoff games
Upsets Due to Players: 16.2% of all upsets which is 4.0% of all playoff games

Note: as explained shortly, in percentage terms you have almost the exact same breakdown for playoff series as you do for playoff games. The percentages you see above are almost exactly the same for playoff series.

Since as already reported above 189 out of the 2,186 playoff games played between 1980 and 2010 were wins that should have been losses (and at the same time losses that should have been wins) caused by coaching and nothing else, the other 1,997 playoff games were played out and determined with the coaching differential not affecting who won and who lost. These games were determined according to how the players played and according to the injury situation. (As in all sports, the home court advantage plays a role as well). For those 1,997 games, the coaching differences affected the score to one extent or another but did not affect who won the game.

PLAYOFF SERIES, UPSETS IN GENERAL, AND UPSETS CAUSED BY COACHING
The QFTR Playoff Teams and Coaches Database contains all 433 playoff series from 1980 to and including 2010 (except for a few round one best of three series that were played from 1980 through 1983; those best of three series would introduce too much statistical error into the key calculations we make so they had to be excluded). 104 of these 433 series (24% or almost one in four) were upsets where the team that was supposed to win (aka as the favorite) lost.

As already reported, for the playoff games we can directly calculate the number of upsets and also the number of upsets caused by coaching. With respect to series, we know the exact number of them which were upsets. Although technically we have not and can not directly calculate exactly how many of the series upsets were won by the underdog team as a result of coaching (and then how many were due to injuries and how many were due to the players) we can indirectly calculate (make a good approximation for) all of these using the same percentage breakdown we have for games, since series are simply groupings of games and since the law of averages applies.

Doing this, we can make a good estimate that 37 or 38 playoff series out of 433 playoff series (8.4% of them) between 1980 and 2010 were won by the wrong team due to coaching. Coaching was responsible for about 35% of the 104 series upsets. Note that these are almost exactly the same percentages that were involved with the games.

QFTR estimates that 50 or 51 playoff series out of the 104 upsets and out of the 433 total playoff series were won by the wrong team due to injuries (and on rare occasions due to some other reason for one or more players not being available). 48.7% of the upsets and 11.7% of all the playoff series were won by the wrong team due to players not available, mostly because of injuries.

Finally, QFTR estimates that 17 playoff series out of the 104 upsets and out of the 433 total playoff series were won by the wrong team due to the players (on their own, separately from coaching). 16.2% of the upsets and 3.9% of all the playoff series were won by the wrong team due to the players on their own, separately from coaching.

The vast majority of the general public and a fairly large majority of people who think they know basketball will claim that the percentage of playoff games and series won by the underdog due to players without regard to coaching is much higher than it really is. I have a strong hunch that the general public believes that rock bottom minimum 10% of all playoff games and series are upsets due to players, whereas the real number is about 4% and it is 5% at the most. In fact, many people probably think that the percentage is around 20%.

Although they are regular events, there are many fewer straight player upsets in basketball than in the other major sports. Let's consider what happens on the average in two years of NBA playoffs. In those two years, the total number of playoff series will be thirty. Out of these thirty series, on average there is only a little more than one series upset caused by players and not caused by coaching. In that same two years of playoffs, about 2 1/2 series out of 30 will be upsets caused by coaching. And in that same two years of playoffs, about 3 1/2 out of 30 series will be upsets caused by players unavailable, mostly due to injuries. In total and on average, a little more than seven series out the the thirty will be upsets and almost 23 of the 30 series will be won by the favored team.

The two main reasons why the general public grossly overestimates how many upsets are achieved by players are (a) The general public grossly underestimates the number of playoff games and series determined by coaching and (b) The general public does not understand that the exact nature of basketball keeps the number of upsets caused by players by themselves very low and much lower than the number of upsets in many other sports.

As a specific example, consider that basketball is at one extreme and American football is the opposite extreme. In football there are many upsets caused by players who make interceptions, recover fumbles, and make big plays from scrimmage. Basketball does not have these kinds of single, quick, but game changing plays. The disadvantage from a basketball turnover is trivial compared to the disadvantage from a football turnover.

BAD AND GOOD PLAYOFF COACHES AND WHY THEY ARE HIDDEN FROM THE PUBLIC
The playoff series that get determined by which coach (and which coaching) is better are mostly ones where the players (and teams) are not all that far apart, meaning these are generally close series where six or seven games are played in the best of seven series. In almost all (but not every last one) of the series where the favorite is a heavy favorite, the coaching differential is not enough to change the result. Since the series determined by coaching are often the ones where the favorite is only a slight favorite, the fact that the coaching determined who won remains secret from the public, because for one thing, generally with these series when the lesser team wins it is not considered a significant upset and so no one looks around for answers as to why it happened.

Moreover, the public is handicapped because most of the public makes the inaccurate assumption that the vast majority of playoff upsets are caused by players whereas, ironically, of the three reasons for playoff upsets, the players are by far the least common reason. Players not available due to injuries (or once in a while for another reason) and coaching are reasons number one and two.

So great coaches steal series quietly, with their accomplishments generally not known by the general public. And bad playoff coaches lose playoff series quietly, with the public in ignorance of this as well.

In fact, this Report is almost certainly the first time in history that anyone has proved in any way shape or form that coaches determine who wins or loses some NBA playoff games and series.

So if you think that NBA coaches don't matter or that they matter very little I now have all the proof I need to know for sure that you are dead wrong. It’s true overall of course that players are much more important than coaches. For one thing, the players have to be pretty good or the team won't be in the playoffs in the first place even if it has a great coach. But coaches allow lesser players to win over better players almost 9% of the time in the playoffs. Also, coaches are responsible for more than twice as many upsets in the playoffs than are players on their own.

Since players get paid between about one million and twenty million dollars a year, with the middle paid player at roughly five million, good coaches are worth three or four million dollars a year, which is what many of them actually get. This is interesting: even though the man on the street typically thinks that NBA coaches have very little impact on who ultimately wins playoff games and series, the pay structure seems to agree with QFTR because the NBA pay structure pays coaches roughly in accordance with how valuable they actually are in winning playoff games (relative to the players).

Of course, you may think that all of them, players and coaches alike, are paid too much, which is another subject entirely.

Coaches don't score and they don't defend, so when coaches cause the lesser team to win they do so by managing the game and or by motivating players to be better than they are normally.

OVERLAP AMONG THE UPSET FACTORS
In some series there is some overlap between the three upset factors: injuries, coaching, and players on their own. In most upsets one of those three factors will be the overwhelming reasons and the other factors will be trivial or not in effect at all. But technically the lesser team can win the series from any one of the factors or from any combination of them. The percentages and numbers we reported out for upsets in this Report in some cases refer to the dominant reason instead of the only reason. However, the coaching numbers and percentages are derived and extracted from the underlying data in such a way that upsets identified as caused by coaching are entirely caused by coaching and there is no other reason than the coaching for those upsets. For the other upsets, coaching may play a small role, but the bigger two reasons will be players not available, usually due to injury, and players by themselves.

SOME OF THE MAIN FACTORS THAT DETERMINE WHICH COACHES ARE GOOD AND WHICH ONES ARE BAD PLAYOFF COACHES
Many other QFTR Reports go into this critical subject in great detail so this is just an extremely brief stopover so to speak. Broadly speaking, I think the biggest overall factor that determines whether coaches win or lose in the playoffs is which coaches understand exactly what basketball is and which don't. Which coaches understand exactly how important offense and defense is in playoff games? Which coaches think you can largely or even just partly ignore offense or defense and still win a lot of playoffs games? Which coaches understand how players and all their various characteristics interact with the game and which don't?

At the game management level, which coaches make sure all players who can help win playoff games play and make sure that the players who can not at all help win playoff games don’t play? In other words, which coaches really understand who their best players really are and which ones don’t? Which coaches know when to take a player out of the game and which ones don't?

This was just scratching the surface. These are just some of many questions the answers to which determine whether someone is a good playoffs coach or not. Read other QFTR Reports for much more on this subject.

We hope you enjoyed this "voyage to the bottom of the sea" type of Report where we revealed previously never reported (and probably never known) information about upsets in professional basketball playoff games and series. This Report was certainly very necessary for and important in the QFTR mission to explain exactly how playoff games are won and lost. For one thing, this Report lays the groundwork for even more astounding discoveries in the months and years ahead.