Page 1 of 1

Testing decks

PostPosted: 22 Dec 2011, 10:10
by Marek14
This is a "kit" for decks that should showcase BotArena's full range of cards.

The archive contains three deck archives:

Global -- Each of these decks is made of four copies of each card from a particular set.
Dual -- This set contains "parts" archive with a Python script; you'll need to change the dir1 and dir2 lines on top to directories you use. Basically, this will mix and match the decks to create a set which will, somewhere, contain every single two-card combination that exists. Each deck will be made from two copies of each card from two different sets. I usually conduct testing by combining this with global decks and running random matches.

All decks in Global and Dual sets are set to 40 life to offset the inherent randomness, so a bad start won't screw you up that much.

Commander -- contains "data" archive and a Python script (again, the "path" variable at the beginning has to be changed to correspond to your system) which will make a set of commander decks. The final output is quite large, but it contains every possible commander-card combination.

In all cases, decks that contain the six Wishes we have so far are fitted with appropriate "collection sideboards" so you'd have a lot of options when you cast the Wishes.

All decks are updated for latest BotArena version. I consider this a nice replacement for random deck construction which we don't yet have.

Re: Testing decks

PostPosted: 23 Dec 2011, 04:00
by missalexis
Thank you for posting this. When I try to run it on Win 7 x64 after installing the latest python installer, it seems unable to create the output file. Any idea why this might be?

-----

Python 3.2.2 (default, Sep 4 2011, 09:07:29) [MSC v.1500 64 bit (AMD64)] on win32
Type "copyright", "credits" or "license()" for more information.
>>> ================================ RESTART ================================
>>>
Commander data read: Aboshan, Cephalid Emperor (13441)
ODY
U

Traceback (most recent call last):
File "C:\Users\ahinds\Downloads\archive\commander\data\commander.py", line 242, in <module>
create_all_commanders(masterfile, output)
File "C:\Users\ahinds\Downloads\archive\commander\data\commander.py", line 238, in create_all_commanders
model_commander(q[0], q[1][:-1], q[2][:-1], output)
File "C:\Users\ahinds\Downloads\archive\commander\data\commander.py", line 225, in model_commander
create_commander_deck(commander, set, i, color_data, output)
File "C:\Users\ahinds\Downloads\archive\commander\data\commander.py", line 113, in create_commander_deck
file_out = open(comname,"w")
IOError: [Errno 2] No such file or directory: 'C:\\Users\\ahinds\\Downloads\\archive\\commander\\decks\\Aboshan, Cephalid Emperor - ODY + UTD.txt'
>>>

Re: Testing decks

PostPosted: 23 Dec 2011, 08:08
by Marek14
Have you created the directory "decks" beforehand?

While the dual script dumps the resulting decks one level up, the commander script produced so many files that it slowed down the system moving through the directory in file manager, so I decided to have it dump them into a separate directory.

Re: Testing decks

PostPosted: 23 Dec 2011, 14:47
by missalexis
*Smacks head with hand*. So sorry and thanks again, can't wait to play Commander. :)

Re: Testing decks

PostPosted: 23 Dec 2011, 15:38
by FranAvalon
Also you can do your own Commander decks. Check this post: viewtopic.php?f=22&t=4821

Re: Testing decks

PostPosted: 23 Dec 2011, 18:34
by Marek14
Yeah, these decks would be probably useless in serious play -- they are for testing and sheer randomness.

Two of these against each other still tend to be interesting, as nobody has an advantage, and you often manage to do something cool, but remember that Commander tends to have a lot of undocumented bugs.