I'm sorry if this belongs in a different section, but are the historical lines from sportsdatabase accurate? I am guessing not because my initial attempt at a model beats the line way too often (by beats the line I mean lower MSE, and a win% of close to 60% over a large sample, and this is on the backtesting half of my database). I thought I was calculating something incorrectly but I haven't found any errors. sportsdatabase is the only thing i've found that i can get lines from without having to scrape (i don't know any languages except some r and some vba). now that i think of it, i'm not even sure if their lines are opening or closing. it would be really nice if their lines were accurate, but i'm pretty sure they aren't.
-
-
I know they were not using closing lines and there is definitely mis-information. If you run SEC searches in ncaabb LSU is not included for some weird reason.
I asked one of the guys who works there (maybe created the site) why the didn't use closing lines and he said something like, "Nobody bets into the closing line so why would we use that?"
Overall, I enjoy the site. Wish I could program in python and mysql and make my own database in the same fashion. Hopefully within a year....Comment -
They say that lines they use are being typed in manually by some person from Las Vegas and these are not opening or closing lines, something in the middle. Which bookie is that they don't tell.
From what i've tested SDB lines are easier to beat than opening or closing lines from this site or covers.com.Comment -
This is correct, anyway from what I have witnessed. I have used SDQL for roughly three years and with great success the last two years. I feel that I am somewhat decent with SDQL now especially with MLB where I am currently up roughly 30 units on the season. Last year I finished with a little over 40u. I currently have over 600 all team situations within their database that come viewable to me for capping, only when they are active.
Are there any SDQL users here? Does anyone have interest using SDQL to help with capping. I would love to start a thread if there is any interest. Or perhaps, if there are any inquires where someone wants to investigate any historical situation and how it has performed over the years. For example, How the Yankees do as a dog when trying to avoid a three game sweep on the road? Or how teams do after losing in extra innings on the road and are at home in their next game? Or how teams do in their next game after hitting 4 or more HRs and still lose? I can easily translate tons of situations using SDQL and give great historical numerical data.
Post away situations you want investigated in MLB!!!Last edited by JMon; 06-13-13, 05:25 PM.Comment -
and the record for a team for first five ??Comment -
over at pregame, Dr Ed MeyerComment -
data validity is a tough nut to crackComment -
I have wondered the same thing with Sportsdatabase and have found some rather sizable differences between their lines and other sources. I am not sure if that reflects poorly on them or the other sources but their stated method of hand entering lines which are not the closing line does not inspire 100% confidence.
Having said that I had some limited success with some MLB systems using their data.Last edited by Cyyyyk; 09-01-13, 05:28 PM.Comment -
I know this is a necro but still good to know.
Ed gets really pissed when you question the lines he uses. Go check pregame out someone said something about his lines and he went into a diatribe.
Basically he uses a 'widely available' or tries too. I also think he skews some 'key' numbers if the vig is crazy. So if a game is -3 -130 or something like that he calls it a -3.5. Or -3 +130 he calls it a -2.5.
Depending on how many games you get back you can verify it. Covers goes back to 2006 (they used to go back a lot further) with their lines and moves. But it takes a bit more effort. But if youre looking for spreads around 'key' numbers it is best to double check.Comment -
-
Hi JMON !
i have just arrived to subscribe to this site after some tries ! it's a few month i want to do it... always problem of email they says ... no answer from owners...
well, that's not a problem.
i noticed your post few months ago, so i can't answer...
and i'm a SDQL's User since about 2 years and 4 with handicapping. but not a pro of that (i can't do all situation !) so i'm interested with your project or simply to exchange some point of view as for example on the fact that this database don't take all matchs of history and the odds are not closing line and not opening line... so some problem with that i found...
see you soon. i hope so !!Comment
