Twitter | Search | |
Sam Larner
Which backrowers at RWC19 leaned towards defence and which towards attack? I didn't include players who didn't play over 140 mins apart from some like Yato and Parisse who are big names. Better back row than O'Mahony, Yato, and Polledri??
Reply Retweet Like More
Sam Larner Jan 9
Replying to @SamLStandsUp
In answer to my last question - yes - Ritchie, Yato, Polledri....
Reply Retweet Like
thedeadballarea Jan 10
Replying to @SamLStandsUp
Not sure how hard it'd be but what would be really good (visually) is if the players could be colour coded by squad - so South African players in Green, English in White - you'd be able to visually see if they group together or all play differently. (does that make sense?)
Reply Retweet Like
Sam Larner Jan 10
Replying to @thedeadballarea
It does. I can probably do that, that's at the limit of my knowledge but a good thing to practice
Reply Retweet Like
Arron Ludlam Jan 10
Love this visual - any insight into the data used for scoring?
Reply Retweet Like
Sam Larner Jan 10
Hey Arron, yes so this is all normalised to minutes played. Attack takes tries, passes, carries, metres made, defenders beaten etc but weights some more highly. Defence takes tackles, minuses missed tackles, turnovers won, lineout steals etc again weighting them. Hope that helps?
Reply Retweet Like
Paul O Connor Jan 10
Replying to @SamLStandsUp
This is really cool. Where did you get the raw data from?
Reply Retweet Like
Sam Larner Jan 10
Thank you Paul, I was out at the World Cup covering that and my colleagues @CameronJDPetrie and collected this data from the information on the website. I'd love to do the same for the Pro14 and Prem but unfortunately cannot get the data yet.
Reply Retweet Like
Sean Mc Jan 10
Replying to @SamLStandsUp
Fantastic stuff, wondering how you source your data out of ESPN? Self built? Just finished my own project scraping ESPN stats and would love to compare builds
Reply Retweet Like
Sam Larner Jan 10
Replying to @Sean_Mc50
Thanks Sean! I was actually in Tokyo covering the World Cup and we collected the data from the match reports. There wasn't a simple way unfortunately. I would love to learn to scrape automatically.
Reply Retweet Like
Tom Gorman Jan 10
I guess Polledri's stats are also skewed as Italy played against two of the 'worst' teams in the competition?
Reply Retweet Like
Sam Larner Jan 10
It depends on whether you think that skews them or not. I'm always wary of mentally removing players from consideration because they played against weaker opposition.
Reply Retweet Like