NAF Data for Number-Crunchers

[plasmoid]

Hi Creeping Darkness (Kyrre?),
that's snazzy!

Would it be possible to add a Year-filter? Or possibly just the option to look at 2011 and later?
For me the data for CRP and beyond is vastly more interesting than that of previous rule sets.
Cheers
Martin

[plasmoid]

Hey Sann,
do you do requests?
I've sent you a PM just in case.
Cheers
Martin

[Creeping Darkness]

Hi Creeping Darkness (Kyrre?),
that's snazzy!

Would it be possible to add a Year-filter? Or possibly just the option to look at 2011 and later?
For me the data for CRP and beyond is vastly more interesting than that of previous rule sets.
Cheers
Martin

Thanks!

You should be able to filter for any year(s) by clicking on the parts of the line in the win % graph that correspond. So if you wanted 2011 on, hover over until you see the 2011 caption, then click and drag to select everything to the right. The main data should then update to only show the selected years.

You can do the same thing for country, too, if say you want to explore how the matchup differs between Canada and the US.

(And no, not kyrre, just citing the data he kindly posted to this thread! I'll update the attribution to make it clearer!)

Cheers
-Morgan

[sann0638]

Ah, confusing! Will update my attribution too!

[kyrre]

I think I found a bug, Creeping Darkness.

Lizardmen - Human is not the same as Human - Lizardmen for Norway.

I found out due to Humans - Lizardmen for Norway does not include my very annoying draw I ended up handing my opponent while I was being clever stalling the game. I have since vowed to never stall another match.

I was considering making a version of the CSV containing all result match duplicated so that both teams would be considered the home team. I think it might simplify spreadsheet formulas and possibly when using with Tableau. Would that change be useful to you? The tradeoff is that the number of lines will double for obvious reasons.



https://i.imgur.com/bGnuBio.png

Click on the link for clarity.

[kyrre]

About casualties: The column that somewhat stupidly is called "casualties?" contain the total number of casualties on the tournament. The rationale was that if this number is greater than 1 the tournament could be counsidered to record casualties. The opposite should certainly be true, if "casualties?" is 0, casualties was not recorded for the duration of the tournament. home_cas, away_cas include the actual casualties for the match.

If someone uses the data to calculate average casualties, do not count games with "casualties?" 0.

[sann0638]

Hi Kyrre,
yes, that's what I did when manipulating the data, made everything a lot easier! Then just had to remember to filter out half the games when counting the games

Mike

[kyrre]

I have added columns home_nationality (24(X)) and away_nationality(25(Y)) that contain the listed nationalities for coaches. In addition to the usual nations, some coaches listed as natives of either Old world or undisclosed. Old world is kyrre-speak for coach did not provide a nation or I have failed to figure out where they are from. Undisclosed is the land of mystery and privacy.

I have also created a CSV file where all matches are repeated with home_coach and away_coach reversed.

[kyrre]

Data for all NAF matches played until recently for use with spreadsheets, tableau or whatever you want really.

Moved here: viewtopic.php?f=29&t=45583

[Creeping Darkness]

Thanks kyrre, I'll check that matchup out and see what went wrong.

I did use a pile of formulae to try and balance the home/away, but must have missed something somewhere. Not sure if we need to duplicate the excel - in tableau we can probably replicate the effect with a union back to the original, with home and away reversed?

Thanks for clarifying the 'Casualties?' column. I did filter the games by whether the tournament tracked casualties or not - but added the number of cas per tournament to get that effect. This will be much easier

Also, not sure if it's been rectified in the latest instance of the data, but I noticed a game with 72 casualties recorded... filtered that out, obviously!

(sorry for accidentally impersonating you too )

[kyrre]

(sorry for accidentally impersonating you too )

Yeah. I am not sure what that was all about. While I am quite the dark creeper, I have never considered myself a creeping darkness.

cheers.

[sann0638]

Updated data for today, added a tab for Plasmoid, and added "most common matchups".

[plasmoid]

I commend you, sir, for that smashing tab!
Thanks

[plasmoid]

How often are the stats updated?

[sann0638]

Well, interesting (not interesting) story - I found a way of automating it using Alteryx, but only on my home computer, then couldn't get a free license (pesky money) so couldn't do that, then tried to get Tableau to talk to the database, and that had its own problems. This was all to avoid the manual downloading of tables.

In short, whenever I get round to it...

Which will be soon, next week at the latest. In a show this week (other major hobby!), but then will have some time.

Mike