Actual vs. Expected Wins: Some Interesting Analysis

[mr beauxjangles]

I did this last night and thought I'd share. A little background. My education is in economics and statistics, and I wanted to start playing around with some stats and start building a model to predict wins and other game statistics. What follows below is just the beginning.

To build a model like this, you have to leverage a technique called regression analysis. I won't bother explaining it, but you can check out Wikipedia if you'd like. In short, regression analysis allows you to estimate the effect of one thing on another thing, all other things being constant. So last night I decided to look at the effect of average margin of victory on win-loss records. Once you estimate the equation that includes the impact of average margin of victory on wins, you can estimate the expected number of wins for each team. Again, what I've done here is very basic, and I'm going to beef it up a lot in the coming months.

In the table below:

The second column is average margin of victory.

The third column is actual wins in 08-09.

The fourth column is what the model predicts based on average margin of victory.

The fifth column is the difference between actual and expected wins.

Basically, this gives us a measure of the degree to which a team underachieved or overachieved based on their average margin of victory.

Those with a positive difference (where expected was greater than actual) are the overachievers. Those with with a negative difference are the underachievers.

Miami and Denver were the biggest overachievers, and there we were at #4, with an expected record of 10-6 based on average margin of victory alone. Green Bay, San Diego, and Phili were the biggest underachievers based on margin of victory.

That's all for now. Hopefully I'll be back with some more interesting stuff in the weeks ahead.

[mr beauxjangles]

damn. no one cares. definitely should have posted in nerdvana.

[Gscottnc5]

You have too much time on your hands. lol. But that is a pretty cool chart

[Darth Biscuit]

I care, but statistics make my head hurt... sorry.

[Lt Pinkerton]

Really cool.

Basically we won more games than we should have just looking at our Margin of Victory, That's probably due to Steve Smith and Jake Delhomme Late-Game heroics. It's weird though, I didn't think Detroits MoV was so low... I guess when you lose every game by about 3-5 points it adds up.

[bLACKpANTHER]

thats cool that it came up as Miami as the biggest over achiever and Philly & San Diego as under achievers.. i dun understand em in the least bit but obviously the numbers dont lie..

[Fireball77]

Let me be the dummy who asks this: how do you arrive at the margin of victory numbers? Take points against and points for and subtract PA from PF then divide by 16? Or do you use the scores of the games, take the point differentials, then average that?

Very interesting topic!

[rayzor]

i like it and i have some things i'd like to ask you about it but the first thing i want to know is what you used as the basis for expected wins?

not really a stats background. my education was in telling people that know how to do stats what to do.

[ncbobcat]

I wouldn't have thought that MoV would be an accurate predictor of record, but that numbers seem to make sense. I've looked to try to find one team's figures that I disagree with and really can't argue with any of them.

Good job!

[jkeough]

1/2 the studies show that 95% of statistics are made up

[PanthaSan]

Sorry I just caught this.

Interesting stuff and it seems like you really put some thought and effort into it.

Can the model predict?

[pstall]

I did this last night and thought I'd share. A little background. My education is in economics and statistics, and I wanted to start playing around with some stats and start building a model to predict wins and other game statistics. What follows below is just the beginning.

To build a model like this, you have to leverage a technique called regression analysis. I won't bother explaining it, but you can check out Wikipedia if you'd like. In short, regression analysis allows you to estimate the effect of one thing on another thing, all other things being constant. So last night I decided to look at the effect of average margin of victory on win-loss records. Once you estimate the equation that includes the impact of average margin of victory on wins, you can estimate the expected number of wins for each team. Again, what I've done here is very basic, and I'm going to beef it up a lot in the coming months.

In the table below:

The second column is average margin of victory.

The third column is actual wins in 08-09.

The fourth column is what the model predicts based on average margin of victory.

The fifth column is the difference between actual and expected wins.

Basically, this gives us a measure of the degree to which a team underachieved or overachieved based on their average margin of victory.

Those with a positive difference (where expected was greater than actual) are the overachievers. Those with with a negative difference are the underachievers.

Miami and Denver were the biggest overachievers, and there we were at #4, with an expected record of 10-6 based on average margin of victory alone. Green Bay, San Diego, and Phili were the biggest underachievers based on margin of victory.

That's all for now. Hopefully I'll be back with some more interesting stuff in the weeks ahead.

*single tear*

[Jangler]

*waits for opposing fans to poo poo*

Even though you did them too, I'm sure somebody will poo poo.

[pstall]

nice chart, did you factor recency bias at all?