Is Ray Allen really the Celtics' most consistent scorer?

[Big Ticket]

That seems to be a very universally believed fact among Celtics' fans and others around the league.

I was curious... so I ran a quick little test of the standard deviations for Allen, Pierce, and Garnett's scoring outputs this season.  For those that don't know, standard deviation, in this case, is found by taking the average, or mean, for their points (so just PPG), squaring (^2) the difference between each individual game scoring and that PPG average, find the mean of all these squared deviations, then take the square root.

The lower the standard deviation, the more "clustered" the individual points are around their PPG... which is the easiest definition of consistency I can think of.

Anyways...

Ray Allen
Mean: 18.1
St Dev: 6.97

Paul Pierce
Mean: 19.4
St Dev: 8.25

Kevin Garnett
Mean: 16.3
St Dev: 4.24

So... as it turns out... Kevin Garnett is actually the most consistent offensively (scoring wise) on the team this year.

Now, the reason behind this is because Ray Allen (and Pierce) has far more single game performances well above his average (29, 31, 35, 36 for Allen) than KG (no 30 point games) does.  Ray Allen is also more likely to score in single digits (5 games vs. 2 for KG), but I think most of his deviation from the mean is driven by those above.   My point is... being inconsistent in a positive fashion is obviously not a bad thing, but it does technically make you more "inconsistent".

There's additional analysis that can be done by finding the skewness of their distributions and what not, but it is above me and not too interested in doing it.  Just thought I'd put this out there though.

[jluce]

Well, as you stated, the problem lies in this. Inconsistency to everyone are games that are well below the player's average. Or an overall bad shooting night. Any game that meets the average or more is the definition of consistency in sports. You can't can't factor inconsistency in if they go way over their average. Nobody looks at it like that in sports.

[Big Ticket]

Well, as you stated, the problem lies in this. Inconsistency to everyone are games that are well below the player's average. Or an overall bad shooting night. Any game that meets the average or more is the definition of consistency in sports. You can't can't factor inconsistency in if they go way over their average. Nobody looks at it like that in sports.

That's true, but there is something to be said for staying consistently close to your average.  Meaning, would you rather have a player go 10, 40, 12, 38, 8, 41, 9, etc... or someone putting in 24, 26, 25, 22, 30, 21, 25.... their averages are pretty similar, the 2nd has a much lower standard deviation.  I would prefer the 2nd... knowing what you'll get every nigh is big. 

In finance (my career and why I work with stats like this) standard deviation of mutual funds is a very useful stat... and a lower S.D. is considered less risky and ultimately better if average returns are equal.

[jluce]

Well, as you stated, the problem lies in this. Inconsistency to everyone are games that are well below the player's average. Or an overall bad shooting night. Any game that meets the average or more is the definition of consistency in sports. You can't can't factor inconsistency in if they go way over their average. Nobody looks at it like that in sports.

That's true, but there is something to be said for staying consistently close to your average.  Meaning, would you rather have a player go 10, 40, 12, 38, 8, 41, 9, etc... or someone putting in 24, 26, 25, 22, 30, 21, 25.... their averages are pretty similar, the 2nd has a much lower standard deviation.  I would prefer the 2nd... knowing what you'll get every nigh is big. 

In finance (my career and why I work with stats like this) standard deviation of mutual funds is a very useful stat... and a lower S.D. is considered less risky and ultimately better if average returns are equal.

Yeah, it works in finance. My point (and you brought it up as well) is that in sports, if you go way over your average it isn't "inconsistency". People would call Lebron pretty consistent b/c he doesn't usually fall waaaaaay under his average. In games he goes way over it would be a great night, not an inconsistent night. But I get your point.

[Big Ticket]

Well, as you stated, the problem lies in this. Inconsistency to everyone are games that are well below the player's average. Or an overall bad shooting night. Any game that meets the average or more is the definition of consistency in sports. You can't can't factor inconsistency in if they go way over their average. Nobody looks at it like that in sports.

That's true, but there is something to be said for staying consistently close to your average.  Meaning, would you rather have a player go 10, 40, 12, 38, 8, 41, 9, etc... or someone putting in 24, 26, 25, 22, 30, 21, 25.... their averages are pretty similar, the 2nd has a much lower standard deviation.  I would prefer the 2nd... knowing what you'll get every nigh is big. 

In finance (my career and why I work with stats like this) standard deviation of mutual funds is a very useful stat... and a lower S.D. is considered less risky and ultimately better if average returns are equal.

Yeah, it works in finance. My point (and you brought it up as well) is that in sports, if you go way over your average it isn't "inconsistency". People would call Lebron pretty consistent b/c he doesn't usually fall waaaaaay under his average. In games he goes way over it would be a great night, not an inconsistent night. But I get your point.

Yeah, we are in complete agreement.  I think it still qualifies as "inconsistent"... but that's why I said in the OP that it was far from a bad thing in the upward inconsistent case.