BetBurger | Live and Pre-game surebets
RebelBetting - Turn betting into investing

3.5 years odds data analysis

Make bookmakers cash cow machines
head-on
Gaining experience
Gaining experience
Karma: 3
Posts: 4
Joined: Fri Nov 21, 2014 7:34 am

3.5 years odds data analysis

Mon Sep 04, 2017 8:19 pm

Hello everyone,
First of all, sorry for my english.
2 weeks ago I did a great research of around 25000 football predictions and results from 2014 till 08.2017. Then I collect all pinnacle odds from 5000 of those games (I selected most obvious picks, where almost every one predicted same team win). Pinnacle odds are next: close odds, 30 mins before kick off (b.k.o.), 3 hours b.k.o., 12 hurs b.k.o., 24 hours b.k.o. 4 hours b.k.o. and open odds.
Of course not all of those games has pinnacle odds (some small markets) but around 4800 has odds. and around 1% of those odds data is wrong because of parsing mismatches.
I spend few days with analysing data I gathered, and I found some nice trading opportunities. Also there is some profitable streight betting results with around 350% of profit from 3 and half years (I can calculate ROI but Im tired of this data already).
I never tried to use betfair and trading software so im not going to trade.
Second thing - I was thinking  to start betting acoording to system I discovered in this data, but for me its easier to go on upwork and find some software development job instead of gambling.

In past I did some arbing and then live arbing. Then I started to dive into data science and python programming.

So I wrote all this because I didnt find enough valuable info (for example ROI 30% or something like this).
It looks like this data contains value for traders and sharp punters. So I just thinking about selling this data.

I am wondering if some one interested in such data?

Feel free to ask whatever you want about it.

First questions was "how I collected that data?"

One of websites I have scraped is oddsportal.com

Best regards
dealer wins
Totally Pro
Totally ProTotally ProTotally ProTotally ProTotally Pro
Karma: 43
Posts: 966
Joined: Mon Sep 15, 2014 12:16 pm

Re: 3.5 years odds data analysis

Mon Sep 04, 2017 9:32 pm

As an arber I dont care about data, past results, bias, beating Pinnacle odds.  I just care that my back bet has higher odds than my lay bet lol
Never trust a goose!!!
Huba
Gaining experience
Gaining experience
Karma: 1
Posts: 11
Joined: Mon Nov 04, 2013 10:23 pm

Re: 3.5 years odds data analysis

Tue Sep 05, 2017 8:59 pm

Where did you get odds: "30 mins before kick off (b.k.o.), 3 hours b.k.o., 12 hurs b.k.o., 24 hours b.k.o. 4 hours b.k.o."?
head-on
Gaining experience
Gaining experience
Karma: 3
Posts: 4
Joined: Fri Nov 21, 2014 7:34 am

Re: 3.5 years odds data analysis

Wed Sep 06, 2017 5:53 am

Huba wrote: Where did you get odds: "30 mins before kick off (b.k.o.), 3 hours b.k.o., 12 hurs b.k.o., 24 hours b.k.o. 4 hours b.k.o."?
mellbet.com
bm1125
Gaining experience
Gaining experience
Karma: 0
Posts: 7
Joined: Wed Jul 12, 2017 3:26 pm

Re: 3.5 years odds data analysis

Thu Sep 07, 2017 9:42 am

Would you mind sharing the data you've scraped from oddsportal ? not your strategy of course just the raw data. I'm using R and I don't know how to scrape data from web and so far I have used football-data.co.uk for analyzing. Would be nice to have another source :)
blackjim
Gaining experience
Gaining experience
Karma: 1
Posts: 45
Joined: Thu Nov 22, 2012 9:18 pm

Re: 3.5 years odds data analysis

Thu Sep 07, 2017 10:42 am

I think he's not into sharing. He's selling.

Probably it would cost you less to just learn some simple JS and scrape it yourself. It's not that hard to make a simple script to do that.

On the other hand, I'd wish I knew some R to analyse the data I get :)
head-on
Gaining experience
Gaining experience
Karma: 3
Posts: 4
Joined: Fri Nov 21, 2014 7:34 am

Re: 3.5 years odds data analysis

Thu Sep 07, 2017 11:31 am

bm1125 wrote: Would you mind sharing the data you've scraped from oddsportal ? not your strategy of course just the raw data. I'm using R and I don't know how to scrape data from web and so far I have used football-data.co.uk for analyzing. Would be nice to have another source :)
If you are using R then I think JS or python scraping is like a piece of cake with R skills :)

about football-data.co.uk:

Im not sure what kind of odds they show. Close odds or open odds or what? So its like those odds can be wrong. Imo thats why its better to analyse close odds or odds from certain time, and of course only sharp bookies.

Later I will send you PM with 20 rows of data just as sample. And of course I dont want to share this data because I spend some time on this research. So its better to keep it for myself if no one interested.
head-on
Gaining experience
Gaining experience
Karma: 3
Posts: 4
Joined: Fri Nov 21, 2014 7:34 am

Re: 3.5 years odds data analysis

Thu Sep 07, 2017 11:39 am

blackjim wrote: I think he's not into sharing. He's selling.

Probably it would cost you less to just learn some simple JS and scrape it yourself. It's not that hard to make a simple script to do that.

On the other hand, I'd wish I knew some R to analyse the data I get :)
You can try Python libraries like pandas. There is a lot of tutorials on youtube. Its easier then R for sure:)
bm1125
Gaining experience
Gaining experience
Karma: 0
Posts: 7
Joined: Wed Jul 12, 2017 3:26 pm

Re: 3.5 years odds data analysis

Sat Sep 09, 2017 7:33 am

blackjim wrote: I think he's not into sharing. He's selling.

Probably it would cost you less to just learn some simple JS and scrape it yourself. It's not that hard to make a simple script to do that.

On the other hand, I'd wish I knew some R to analyse the data I get :)
Yes I guess that what I'll do. Take it as a new challenge :)

head-on wrote:
bm1125 wrote: Would you mind sharing the data you've scraped from oddsportal ? not your strategy of course just the raw data. I'm using R and I don't know how to scrape data from web and so far I have used football-data.co.uk for analyzing. Would be nice to have another source :)
If you are using R then I think JS or python scraping is like a piece of cake with R skills :)

about football-data.co.uk:

Im not sure what kind of odds they show. Close odds or open odds or what? So its like those odds can be wrong. Imo thats why its better to analyse close odds or odds from certain time, and of course only sharp bookies.

Later I will send you PM with 20 rows of data just as sample. And of course I dont want to share this data because I spend some time on this research. So its better to keep it for myself if no one interested.
Sure man.

About football-data.co.uk , they record odds twice a week, Tuesday and Friday noon for midweek and weekend games (respectively). They also have closing odds of pinnacle.

Return to “Arbing, matched betting and trading talk”