<<<Soccer Correct Score Odds Historical Data>>>

[Datapunter]

Re: <<<Soccer Correct Score Odds Historical Data>>>

Anyhow, i found a Java program that filters out the correct score odds from the historic data but does nothing more than that. It basically reduces a 150Mb file down to 15Mb. You could filter it down some more by excluding all in-running data as well as records with very low volume. End result should fit nicely into excel.

See here: http://www.punterslounge.com/forum/f21/betfair-historic-data-java-filter-program-s-86012/

[Datapunter]

Re: <<<Soccer Correct Score Odds Historical Data>>> beuher2000, I've posted the java program with the database version in the other thread.

---

After you've run that program you can review the events from the import_event table: select * from import_event;

---

then you can see all the correct score data like this: select * from import_event as ev join import_market as mr where mr.marketID = ev.cscoreID; (warning: this will give a lot of records ! )

---

then you can pick 1 event and note cscoreID, let say that is 23567 and just the single market like this: select * from import_event as ev join import_market as mr on mr.marketID = ev.cscoreID where mr.marketID = 23567;

---

If you want to examine the actual selection names you can add them like this: select * from import_event as ev join import_market as mr on mr.marketID = ev.cscoreID join import_selection as sl on mr.selectionID = sl.selectionID where mr.marketID = 23567;

---

if you'd like to know the total volume matched then you can get that like this: select *,sum(volume_matched) from import_event as ev join import_market as mr on mr.marketID = ev.cscoreID join import_selection as sl on mr.selectionID = sl.selectionID where mr.marketID = 23567 group by mr.selectionID;

---

Next step for you is to get it working then you'll need to determine what odds data you want to extract. Could be average, could be early odds, maybe odds with most matched volume. From that you can make a specific extract and match that using the date/teamnames with results data from other sources. I never said it wouldn't take some work nor be easy :lol