Football Data (CSV, JSON) - UPDATED 16/08/17

[Euler]

I agree, we have given plenty of slack and asked kindly. It's not too hard to get on with others. If you have something that works, that's great but constantly drowning out threads endlessly is killing genuine discussion.

[Atho55]

Is the solution to give Jonny his own thread that he confines his posts to. Other forum members then have the choice to read it or not. Any posts he makes in other threads automatically deleted.

[dm1900]

Any chance of an updated data set, but with yellow cards as well? Or is that data not available?

Thanks again!

[welshboy06]

Any chance of an updated data set, but with yellow cards as well? Or is that data not available?

Thanks again!

I'm working on this at the moment. Not been able to find a reliable source for yellow card timings though.
At the moment i'm using the csv's from football-data.co.uk. Which has shots, shots on target, yellows, fowls, reds, goals, result (all for HT and FT) as well as odds for 1x2, o2.5,u2.5 from Pinnacle and an average from a selection of oter bookies. Theres no timings though, just totals for HT/FT. But it'll do for now.

[knome]

I'm working on this at the moment. Not been able to find a reliable source for yellow card timings though.
At the moment i'm using the csv's from football-data.co.uk. Which has shots, shots on target, yellows, fowls, reds, goals, result (all for HT and FT) as well as odds for 1x2, o2.5,u2.5 from Pinnacle and an average from a selection of oter bookies. Theres no timings though, just totals for HT/FT. But it'll do for now.

Have a look at https://www.fctables.com/
Oodles of data... Tooltip pop-ups with reds/yellows and goals.

[welshboy06]

I'm working on this at the moment. Not been able to find a reliable source for yellow card timings though.
At the moment i'm using the csv's from football-data.co.uk. Which has shots, shots on target, yellows, fowls, reds, goals, result (all for HT and FT) as well as odds for 1x2, o2.5,u2.5 from Pinnacle and an average from a selection of oter bookies. Theres no timings though, just totals for HT/FT. But it'll do for now.

Have a look at https://www.fctables.com/
Oodles of data... Tooltip pop-ups with reds/yellows and goals.

Thanks for this. I've taken a quick look and it doesn't seem to be designed in a webscraper friendly way. Seems awkward to drill down in to stats and as yet I'm not able to find an easy to use underlying data source/api. Will check again tomorrow though.

[knome]

I'm working on this at the moment. Not been able to find a reliable source for yellow card timings though.
At the moment i'm using the csv's from football-data.co.uk. Which has shots, shots on target, yellows, fowls, reds, goals, result (all for HT and FT) as well as odds for 1x2, o2.5,u2.5 from Pinnacle and an average from a selection of oter bookies. Theres no timings though, just totals for HT/FT. But it'll do for now.

Have a look at https://www.fctables.com/
Oodles of data... Tooltip pop-ups with reds/yellows and goals.

Thanks for this. I've taken a quick look and it doesn't seem to be designed in a webscraper friendly way. Seems awkward to drill down in to stats and as yet I'm not able to find an easy to use underlying data source/api. Will check again tomorrow though.

Ah! From the awesome work you've done with the HKJC data I just assumed you'd got a good handle on this web scraping malarkey.
I've had a quick play with python/BeautifulSoup and some chrome extensions but I've concluded web scraping is a young man's game.

... or woman's...

[welshboy06]

Have a look at https://www.fctables.com/
Oodles of data... Tooltip pop-ups with reds/yellows and goals.

Thanks for this. I've taken a quick look and it doesn't seem to be designed in a webscraper friendly way. Seems awkward to drill down in to stats and as yet I'm not able to find an easy to use underlying data source/api. Will check again tomorrow though.

Ah! From the awesome work you've done with the HKJC data I just assumed you'd got a good handle on this web scraping malarkey.
I've had a quick play with python/BeautifulSoup and some chrome extensions but I've concluded web scraping is a young man's game.

... or woman's...

Yeah I've scraped a good few websites. But in my past experiences, working with awkward sites involves a lot of spaghetti code, and then they end up getting updated by the developer.
It's always easier to try and find the data source they're using, rather than scrape the full page.

Hkjc was an exception because it was a quick fix and the site was actually pretty easy. Football-data provides up to date CSV files, which can be easily manipulated.

I'll take another look at the above site tomorrow though:)

[dm1900]

Also if no-one has mentioned yet, corner data would be super useful, even if just number of corners at HT and number of corners at FT!

[Euler]

Corners hold a relationship with goals. So if you plot goals you also plot corners.

[dm1900]

Corners hold a relationship with goals. So if you plot goals you also plot corners.

Hmm but surely corners can add more information to game state - there may be two games that are 1-0, except the first game has 10 corners by the home team whereas the second only 2 by the home team. If we think about expected goals in the second half, doesn't the 10 corners indicate aggressive play and so more of an expectation of goals in the first game rather than the second?

[Euler]

Describe how a corner would occur.

[dm1900]

Corner = A ball going over the goal line but last touched by the defending team. So usually via clearance or a save by the goalie - both of which indicate that the attacking team are in dangerous positions. So more corners = more attempts to attack by a team, whereas the number of actual goals may not be so indicative as shown by my example.

[Euler]

So if you look at a corner it's more or less linked to an attacking move and can be related to shots on and off target. So it's effectively a derivate of shots (in general) which is a derivative of goals.

[footysystems]

Corners can be misleading imo the best way to predict goals or expect goals is to divide the number of dangerous attacks by the number of attacks to give you a % > than 70% high expectancy of a goal.