I was working this morning and I kept asking myself (from an analytic perspective): What are the keys to the success for a basketball team that at the beginning of the season, had the equivalent odds to win a NBA title as the New York Knickerbockers?
I understand this is a short-term projection, and this information should be taken at that, but if you dig into this a little, it's very simple to see why this isn't a fluke, (in my opinion).
I made a few adjustments to the numbers in order to get consistent results. For example, I eliminated games where all the starters rested (like the 76'ers) and used a points per game average (minus 3 points) for values in games where individual starters were rested or injured.
Here are the results:
The conclusions were astounding:
Please disregard bench players at this time, I am still working here...
You always hear someone say: "they are playing great team basketball" but I actually have the data to substantiate these comments. To all of those who don't understand averages and calculating deviations, it's really not that complicated and a useful tool for analysis. 1 Standard Deviation from the mean (average) in simple terms, means that there is a 68% probability that future data points will fall within the specified range. So the first player's data for example, Milsap, averages 17.6 PPG in 33 MPG. There is a 68% probability that he will play +/- 3.3 minutes off his average, and score +/- 5 points off his PPG average. Also, FYI, if you calculate for 2 standard deviations, your probability that the data will fall within the given range is 95%. So 2 SD's for Milsap in this example would be approx 6.6 minutes, and 10 points (you just multiply).
Anyways, the point is look across the board at every starter's deviation result. Minutes played are almost identical, and PPG deviations are also nearly identical!! The minutes logged on a nightly basis are going to be extremely consistent, within 2-3 minutes of game play and the points scored are going to be withing 3-5 points! Furthermore, just to put that in perspective as far as consistency, James Harden has a standard deviation in all games this season of (+/-) 9 points, meaning 2 SD's (95%) is upwards of 18 points per data sample! That is 2-3 times the deviation of Hawks starters!
So as a sports bettor, can we fade this kind of data? Are we going to use the general rule of thumb to fade the public, or find a contrarian point of view...or sell this team high? Well, that's up to you to decide. In my opinion, this is substantial data that helps justify this team's success and I do not see an end in sight.
This is the Hawks formula for success and as you can see, rarely has there been less than 4 starters in double digits, or any inconsistency in their playing time. They have a system and they stick to it. Right now they're as good or better than any team in professional basketball...and it is the reason why we see sharp action on this team, tonight, and just about any night they take the floor.
Hope this helps! Thanks!
-Mike