Senior Research Project, Help Appreciated

Peyton2MarvinN06 · 02-08-08, 04:13 PM

Dan, you have a whole bunch of statistics right at your fingertips here on SBR... Here is the link... http://www.sbrforum.com/Scores/NCAA-.../20080208.aspx

Go there and click on matchup... I believe you can go back to games such as the Duke/UNC matchup by click on the particular date on the calendar in the top right hand corner of the page...

As far as injuries go, I can't really help ya there... There are a lot of sharp fellas on here that may have an idea, but I've never come across anything that says if a particular star is out for a team how much it will affect the point spread of a game... I don't think that Carolina would have won that game even if Ty Lawson would have been playing... Duke is a team on a mission right now and they looked very solid the other night!

You might do something along the lines of best teams Against The Spread and worst teams ATS and compare some of their statistics to see how much their defensive and offensive statistics are affecting the outcomes of their games... Typically in NCAA B-Ball, homecourt advantage rules out! Hand down! If teams matchup even on paper, then I'd say atleast 3 out of every 4 times the home team is going to win...

I have seen a lot of games this year where the home team is anywhere from PK to +2.5... It seems as though the home team that is a small dog has been winning a HIGH % of these games outright... I'm sure that someone on here can show you the correlation between being a small home dog and a small road favorite... Small road favorites in the NCAA typically are the better team on paper, but like I said before, the home team in these close matchups almost always covers or wins the game outright...

Hope some of that helps... I'm sure there are a few guys on here that would be more than glad to provide you with some statistics ATS and some other stuff!

Good Luck on the project! Let me know if I can be of any more assistance...

P2M

Crayzee · 02-08-08, 05:00 PM

my favorite strategy of all is betting against big favorites when their next game is vs a big opponent

or they just came from a big comeback or wire to wire game
best of all is if you get both of these scenarios happening and its what you call a sandwich
example:
say nc just beats clemson at clemson in a hard fought game that goes almost to the last shot or ot

next game: nc is a 21-1/2 pt favorite over say nc wilmington or st johns

then the next game is nc at duke or md
my theory is bett big on the 21-1/2 point dog in the middle game
because nc might not be 100% "up" for the game

i havent ever tracked the results but it seems to be somewhat successful

BigOrangeTitans · 02-08-08, 06:02 PM

Espn.com has good box score info

regularguy · 02-08-08, 07:41 PM

You might want to check out thespread.com. Click on NCAAB and go to the Betting Charts. Browse around there. Interesting info.

joeykrowe · 02-10-08, 01:23 PM

Originally posted by regularguy

You might want to check out thespread.com. Click on NCAAB and go to the Betting Charts. Browse around there. Interesting info.

Yea thespread.com was great. I hadn't stumbled across it just yet.
Thank you guys for your suggestions and info. It will all help me in the upcoming weeks as I get my model together.

Any more case studies that you want done, I have twelve seasons worth of data to study.

curious · 02-10-08, 01:54 PM

Originally posted by joeykrowe

Hello, my name is Dan Wright and I'm a senior econ student trying to write and defend a research thesis. I have chosen to study the college basketball point spread market.
As a longtime fan, I'm just getting into handicapping. For my project I need a large set of data from which to mine my research questions. I came to get some question's answered from the horse's mouth.

First, are there any box score databases publicly available? Like, say, the Duke-UNC game on Wednesday the stats would describe FGA-FGM, FTA-FTM, Reb, Blocks, etc.

Second, is there any historical injury data out there? One of my questions focuses on the impact of an injured starter. Going back to the Duke-UNC game, with Ty Lawson not playing I took Duke at +4 and won some cash.

Any other information or comments that you guys could supply would be terrific. Anyone have any theories that I could test, I'll report them all back to you and you'll even get mentioned in an academic publication.

Thanks,
DW

P.S. sorry if I sound like a snot, I just got done writing an academic paper.

There are many papers on this topic. Every economics, statistics, trading, and modeling journal out there has very advanced work on the topic of the sportsbetting spread market behaving like an inefficient financial market. All the information you ask for is available from any number of paid data services.

joeykrowe · 02-10-08, 05:18 PM

Originally posted by curious

There are many papers on this topic. Every economics, statistics, trading, and modeling journal out there has very advanced work on the topic of the sportsbetting spread market behaving like an inefficient financial market. All the information you ask for is available from any number of paid data services.

I've read roughly 20 academic papers on the topic (Sauer, Woodland and Woodland; Paul et al.; etc.). They all concluded, and it would be surprising to many people on this forum because they DO make money, that in the long run the PS betting market is efficient. That means that in the long run you may only break even or, in most cases, lose money. However, they were testing very simple betting strategies: straight underdogs, big underdogs, etc. I'd love to read one that concluded that in their entire set of data that the markets were inefficient (not Camerer 1989, he had bad assumptions)

Also, where would one find those fee-based data services? And how much do they cost?