Comments on Advanced NFL Stats: Epic Bulls-Celtics Series

Ty-Thanks. The HCA is a league-wide average.

2009-05-16T11:13:00.000-04:00

Ty-Thanks. The HCA is a league-wide average.

Terrific work, as always, Brian. Do you set the h...

2009-05-16T03:26:00.000-04:00

Terrific work, as always, Brian. Do you set the home court advantage as a constant, or do you adjust it according to each team's home court efficiency differentials? In other words, if one team tends to play better at home, do you take that into account?

Keep up the great work.

Ty Willihnganz

Here is a link to a post on the methodology. The model was originally applied to NCAA games, but this NBA model is based all new data from every game of the last 2 full NBA seasons.

It's an empirical logit regression model based on home court advantage, score difference, and time remaining. Foul bonuses, the possession arrow, and time outs are not factored in.

Cheers. I guess as the model is built from previou...

2009-05-05T11:01:00.000-04:00

Cheers. I guess as the model is built from previous data that it will probably be correct. It's always worth checking though (I'm sure nearly every mathematician has attempted to build a model to use for gambling purposes that's works when using past data but proves useless for future games).

I don't have enough games to make that meaningful ...

2009-05-05T10:46:00.000-04:00

I don't have enough games to make that meaningful or worthwhile yet.

Out of interest, have you reactively tested the da...

2009-05-05T10:08:00.000-04:00

Out of interest, have you reactively tested the data produced by the graphs to the real-life result to confirm accuracy (i.e. for all games when home team WP is 0.5 at half time, have the home teams won half of those games)?