Getting Data

JoeyBoots2008 · 02-10-2006, 07:51 PM

Hi,
I am an economics PhD student, with extensive background, trying to get into sports gambling. I'm trying to get a database for a sports (probably NBA, but I'd be interested in more as well) with spread for every game, home team, away team, perhaps some statistics of the team up to that point, etc. I am hoping to use my math skills to find a successful strategy. Does anyone know the best way to get such a database? Is there a website I could go to? Would somebody have databases that they'd be willing to give away, in exchange for sharing what I find?

Thanks much.
My e-mail is thsgambler26@gmail.com

ChuckyTheGoat · 02-11-2006, 03:39 AM

Best I know of are:

1) Davler Sports
2) covers.com, if you're willing to do some cut and paste.

May be some similar sites, maybe jimfeist.com. You can create your own databases from some of these sites, if you're willing to do a little legwork on the cut/paste.

kahlmyishmael · 02-12-2006, 04:18 PM

If one had microsoft office could not one fashion a macro from the desired databases to do the cutting and pasting inot either an excelXP OR accessXP table? Maybe some VBA programming could help?

JackDaniels · 02-16-2006, 11:01 PM

Anyone interested in gathering data on lines, scores, statistics, etc... and/or want to share, trade, sell etc.... on NBA, NCAA Football and NCAA Basketball send me a private message.

Thanks

KingOfTheSquares · 02-28-2006, 11:56 PM

Hi Chucky & company,

Yeah, I know haven't been around for a while...real estate market has just been too good, at least until Katrina hit. I've got some old sports dbs that I'd be willing to swap if anyone is interested.

I might even throw out some picks this NFL season if a new model we're working on matures in time.

KOTS

ChuckyTheGoat · 03-01-2006, 05:08 AM

Hey, KOTS:

What up, old timer? Keep turning that profit on the real estate, man. Good to hear you're winning.

See you down the road, man. Best of luck.

KingOfTheSquares · 03-02-2006, 06:15 PM

Chucky,

Yeah, real estate, when it is hot its 1000 times easier than turning bux in sports with damn litttle risk. It got so that the time I was putting into working data bases and model building seriously impacted r.e. research. And with 2 bad seasons out of 3 in the NFL, I just said screw it. BTW, any NFL season under 55% is just not worth the time in my book.

But, you know, I just miss working with the numbers...so I'm back working with a team of IT guys who seem to be interested in working a lot of the stuff that was sucking down so much of my time. One of the guys is a biostatistician who has much more of a stat background than I and has a different perspective on the nums than I have.

So, I guess I'll pop in from time to time.

BTW, if anyone wants to talk about or share any data, let me know via the MW messaging. I'm a little reluctant to post my e-mail since there is so much spamming and wierd stuff going on. I have data on almost every major sport, some going back deep into the 80s.

KOTS

ChuckyTheGoat · 03-02-2006, 08:03 PM

KOTS:

I may contact you down the road.

I hear you on the NFL. There are a few interesting trends going on. Imho, the ratio of database to non-database guys used to be 50:1. Now maybe it's 5:1. A lot of the winning subsets have been repeated ad infinitum.

The theory is that the linesmaker can learn from his mistakes over time, and I believe that to be true here. Hard to find a lot of generic value on NFL lines. The line pressure from the fav and the dog are often about even.

Good luck w/ the real estate, bro. Very interesting discussion.

KingOfTheSquares · 03-04-2006, 10:32 PM

Chucky,

So you think as many as 20% are using dbs built by other folks, or are building themselves, or using stuff like what Micro Brothers publish? Or maybe just using them for trend data?

Wow, that really seems like a lot. I always seemed to be in a very, very small minority. I'd say I was at least 1 out a 1000 using my own predictor.

j · 03-09-2006, 12:40 AM

Isn't the question percentage of $ bet influenced by DB discovered info, not sheer % of handicappers using it.

I agree with Chucky that it's a lot higher now that it used to be, although you can find DB guys on both sides of a game backing differently defined subsets.

KingOfTheSquares · 03-09-2006, 01:37 AM

J,

I agree. I once thought that it was a damn small amount of $ bet derived from numerics. Maybe we need to distinguish the guys using dbs for bets like this: "since '94, Pitt covered as a homedog 75% of the time against dome teams during October and November, after a new moon, when the away team had been smacked by 28+ points in it's last game." So, bet on Pitt even though you uncover that the occurance happened 3 out of 4 times in the period you have data. This type of db bet has been proven folly many times over even when the z-stat shows an occurance that is 3+ Standard Deviations off norm.

Usually the numeric folks tend to line up on one side a pretty sizable amount of the time. I always had to see at least 3 pts worh of difference from the line. When others applied the same criteria, either we were on the same side, or the calced dif wasn't strong enough to get down on a bet.

DrunkenGoon · 03-20-2006, 02:23 PM

I have about 8 seasons worth of NHL up to the lockout season, didn't bother updating after. It's probably useless as it's a different game nowadays.