home court advantage(ncaabb)

[zofo1234]

Whats up guys. Ive been lurking the site for 2 weeks now and finally decided to post.

My question is this:

How much of an advantage is playing at home worth in college basketball expressed as a point value? Im currently using a model that factors in this number but Im not sure if the value that Im using is valid.

Thoughts anyone?

zofo

[Justin7]

I have heard about 4 or 4.5... But it's dangerous to use a uniform rating for all teams. It changes based on the team, how far the visitor travels, and the difference in caliber. I think home field becomes worth more the bigger the skill difference.

[frostno98]

Home Teams with bad records favored -1 to -3.5 seems to cover a lot, as unimaginable it is. It usually in those low profile conferences. Div II teams football conference.

[curious]

The "average" is given by providing an increase to the home team's offensive and defensive efficiency by 1.04% and a decrease to the visitor team's offensive and defensive efficiency by 1.04%. This kind of works out to 4 points a game (more or less).

I think this "average" is pretty much useless for several reasons. Some teams are ferocious at home no matter who they are playing. Western Kentucky springs to mind. Other teams are ferocious at home when playing regional rivals. Some teams suck at home no matter who they are playing. Other teams suck at home if the visiting team means nothing to them. Other teams suck at home if the visiting team is a top ranked team or a strong rival.

There are trends across all NCAAB games that show these kind of biases, for example teams that totally suck and are usually big dogs, but are now a small fav at home tend to cover more often than they should.

I think you have to really study the home court in question to have meaningful statistics on this.

Then you have the problem of "is this actually a home court" because so many games are played in "neutral" courts. For example, if Western Kentucky plays a game at Cairo Ill, is that a "neutral" court or a home court?

[Peep]

I have 28,800 college games in a database.

The average HT is favored by 5.25 pts, the home team wins by an average of 5.40 pts.

This probably slightly skewed in that Podunk College is more apt to play Duke at Duke, than the other way around.

[curious]

I have 28,800 college games in a database.

The average HT is favored by 5.25 pts, the home team wins by an average of 5.40 pts.

This probably slightly skewed in that Podunk College is more apt to play Duke at Duke, than the other way around.

Do you have betting odds in this database? Moneylines? Spreads?

[Peep]

Do you have betting odds in this database? Moneylines? Spreads?

I have point spreads. I could convert them to ML's if I wanted, but no reason for me to. I can for instance find out that "five point favorites have won Straight up 1124 out 1607 games in college hoops since 1996".

Or 69.9%.

These ML odds the database puts out are similiar to what the books offer, they are not stupid about it.

[Wheell]

One of the things I did when attempting to solve that problem was examine both the lines and the results for all the divisional games played between teams that have a home and home during the regular season. That actually got me 90% of the way there to where I currently am. I know this misses a lot of issues (extreme travel, unfamiliarity with venue, conference based ref bias), but it does a pretty good job despite those limitations.