Is this a valid measure?
I know this comes up in all sports almost every year, but currently with the NBA play-offs in full bloom I'm hearing the fighting increase between the Kobe fans and the LeBron fans.
You might notice I'm from Ohio, but I'm not trying to make this about Cleveland teams or LeBron in particular, but this seems to highlight the discussion well.
It seems that a player can't reach that final tier of greatness until they win a championship. While I understand why this occurs (namely the entire season is set up with the goal being a championship, and there are certainly psychological reasons why we remember championship teams more easily), is it a valid measure of an athlete or a team?
Besides the obvious criticisms (how can you judge a great player on a bad team versus a great player on a great team where the latter SHOULD be winning more often), I wanted to take at a look at the raw probabilities (in an admittedly somewhat amateurish way for expediency).
So, to generalize this even more and restate the point, the team that has the best record in a league (I'll stick with the NBA for this exercise) isn't given much credit. If they don't win the championship, what's the point?
I'm going to make some assumptions and take a few shortcuts here, but I'll note them as I come to them.
In the last 10 years of the NBA, the #1 team in the regular season has had an average win percentage in the mid-.700s.
Assuming this is a valid indicator of their odds of winning ANY given game, let's transfer that into the play-offs with a few adjustments. Seeing as your competition is better than the league average in the play-offs, it seems like your probability of winning would likely drop. My rough estimates (which you can take issue with) would be that in the early rounds it would be in the lower-mid .700s and in the later rounds it would be in the .500s to lower .600s. On average, I'll just guesstimate .667 (2 in 3 chance). I'll hold this number steady for ease of calculations (which isn't a horrible assumption as it will approximately even out in the end).
So, given that average, what are the odds that the best team in the NBA wins the championship?
They would be expected to have a roughly 45% chance of winning. More than half the time, a team that is NOT the best team would win (based on my above oversimplifications).
Statistically speaking, the best way to measure something is to take LOTS of samples (i.e., the regular season). Taking a handful of samples (the play-offs) is more likely to result in errors.
I realize there are other factors influenced by the regular season having less emphasis (i.e., the ;possibility of teams not playing as hard in the regular season), but most teams seem to play hard the whole way through fighting for home court advantage and pride. And, once you hit the play-offs, you would expect all (or most) teams to play harder, so it should even out to some extent. So, even if a team is coasting a little, would it change their odds that much when they hit the play-offs?
I just find it interesting that in the above scenario, even if you had the theoretically "best team" in the league for 2 straight years, there is still a 30% chance that you would not win a championship.
Now, stepping out from under my statistician hat, as a fan, I understand that by placing so much emphasis on the play-offs, you encourage the athletes to play harder and create a more exciting game for the fans. Does that translate into a better measure of who the best team is, though?
Also, I just checked to see how many of the teams with the best regular season record won the NBA finals in the last decade, and it happened 4 out of 10 times (not that this proves anything, but it fits nicely into my analysis). If those 10 teams with the best record WERE the best team that year (i.e., the had the greatest chance of winning any given game against any given opponent at any given time), only 4 or 5 of them would be expected to survive all the way through the play-offs (which is what happened).
So, what does that say about the team that "wins it all"? Just due to chance, around half the time, the best team isn't going to win the championship.
Technically, it is harder (or less likely for the team) to make it to the finals and lose twice than it is to win a championship once (according to my numbers). We don't reward teams (or players) for that, though.
Thoughts?
I know this comes up in all sports almost every year, but currently with the NBA play-offs in full bloom I'm hearing the fighting increase between the Kobe fans and the LeBron fans.
You might notice I'm from Ohio, but I'm not trying to make this about Cleveland teams or LeBron in particular, but this seems to highlight the discussion well.
It seems that a player can't reach that final tier of greatness until they win a championship. While I understand why this occurs (namely the entire season is set up with the goal being a championship, and there are certainly psychological reasons why we remember championship teams more easily), is it a valid measure of an athlete or a team?
Besides the obvious criticisms (how can you judge a great player on a bad team versus a great player on a great team where the latter SHOULD be winning more often), I wanted to take at a look at the raw probabilities (in an admittedly somewhat amateurish way for expediency).
So, to generalize this even more and restate the point, the team that has the best record in a league (I'll stick with the NBA for this exercise) isn't given much credit. If they don't win the championship, what's the point?
I'm going to make some assumptions and take a few shortcuts here, but I'll note them as I come to them.
In the last 10 years of the NBA, the #1 team in the regular season has had an average win percentage in the mid-.700s.
Assuming this is a valid indicator of their odds of winning ANY given game, let's transfer that into the play-offs with a few adjustments. Seeing as your competition is better than the league average in the play-offs, it seems like your probability of winning would likely drop. My rough estimates (which you can take issue with) would be that in the early rounds it would be in the lower-mid .700s and in the later rounds it would be in the .500s to lower .600s. On average, I'll just guesstimate .667 (2 in 3 chance). I'll hold this number steady for ease of calculations (which isn't a horrible assumption as it will approximately even out in the end).
So, given that average, what are the odds that the best team in the NBA wins the championship?
They would be expected to have a roughly 45% chance of winning. More than half the time, a team that is NOT the best team would win (based on my above oversimplifications).
Statistically speaking, the best way to measure something is to take LOTS of samples (i.e., the regular season). Taking a handful of samples (the play-offs) is more likely to result in errors.
I realize there are other factors influenced by the regular season having less emphasis (i.e., the ;possibility of teams not playing as hard in the regular season), but most teams seem to play hard the whole way through fighting for home court advantage and pride. And, once you hit the play-offs, you would expect all (or most) teams to play harder, so it should even out to some extent. So, even if a team is coasting a little, would it change their odds that much when they hit the play-offs?
I just find it interesting that in the above scenario, even if you had the theoretically "best team" in the league for 2 straight years, there is still a 30% chance that you would not win a championship.
Now, stepping out from under my statistician hat, as a fan, I understand that by placing so much emphasis on the play-offs, you encourage the athletes to play harder and create a more exciting game for the fans. Does that translate into a better measure of who the best team is, though?
Also, I just checked to see how many of the teams with the best regular season record won the NBA finals in the last decade, and it happened 4 out of 10 times (not that this proves anything, but it fits nicely into my analysis). If those 10 teams with the best record WERE the best team that year (i.e., the had the greatest chance of winning any given game against any given opponent at any given time), only 4 or 5 of them would be expected to survive all the way through the play-offs (which is what happened).
So, what does that say about the team that "wins it all"? Just due to chance, around half the time, the best team isn't going to win the championship.
Technically, it is harder (or less likely for the team) to make it to the finals and lose twice than it is to win a championship once (according to my numbers). We don't reward teams (or players) for that, though.
Thoughts?




