JackassRampant
Joined: Feb 26, 2011
|
  Posted:
Nov 09, 2011 - 01:10 |
|
Hey, all. Is there a way we can use the FC to get some knowledge about the effectiveness of inducements, the head-to-head power of [B] vs [R], and all that? I have some stats that I think we can learn a lot from studying.
I'd be happy with just a table of matches including the following data. Willing to do it by hand, but there's got to be somebody with the code-fu to do it faster.
For each FUMBBL match, we'd learn a lot if we knew:
1) Race
2) Division
3) Highest total inducement (handicap or petty cash) value
4) Score (obv, the most important stat)
5) Cas count in favor
6) Cas count against (to account for inducement cas)
7) Inducements taken, maybe with stars split into categories (Saw, SW/Stab, ST5+, ST4, Other, in descending priority)
OT? If so, shootout?
What do y'all think? is it worthwhile to go down that road? Anyone know how to put that together without doing it by hand? I can do the statistical part, but I meddle not in the affairs of wizards. |
_________________ Lude enixe, obliviscatur timor. |
|
King_Ghidra
Joined: Sep 14, 2009
|
  Posted:
Nov 15, 2011 - 18:25 |
|
I saw Christer do something in chat yesterday that spawned a list of remaining races in the cup, so I assume others may also be easily possible.
I was certainly wondering about the B vs R situation - those left vs those that started.
But yeah, inducement value, that's an interesting one. Be very interesting to see how many teams with say, a 150k+ TV difference down (the 'Wizard bar' to coin a phrase) have won games. |
|
|
Christer
Joined: Aug 02, 2003
|
FUMBBL provides APIs where you can query for a lot of this information in XML format.
There's a bunch of information on this in the help pages:
https://fumbbl.com/help:Fumbbl+API |
|
|
King_Ghidra
Joined: Sep 14, 2009
|
  Posted:
Dec 21, 2011 - 11:31 |
|
Coming back to this, I was trying to see what the racial breakdown was for this fumbbl cup, and then planning to see how that compared with previous tournies. Unfortunately I didn't get very far
Using http://fumbbl.com/xml:group?id=195&op=members doesn't return anything. 195 is the fumbbl cup group, but I'm guessing the fumbbl cup teams aren't actually members of it? In fact I had mixed results getting this to work for past tournies, with some working and some not. I know it gets a bit messy with qualifiers and suchlike, but there aren't any for the fumbbl cup so i'm a bit confused as to what i'm doing wrong (btw adding the particular tourney that is fumbbl cup x (30211) doesn't make any difference).
Looking at a working example, I'm still not sure I understand the results:
http://fumbbl.com/xml:group?id=1275&op=members&t=28709 gives me something resembling the latter stages of the last Ulthuan Invitational. But there should be 16 teams, whereas the xml returns only 10. |
|
|
koadah
Joined: Mar 30, 2005
|
  Posted:
Dec 21, 2011 - 11:54 |
|
|
King_Ghidra
Joined: Sep 14, 2009
|
  Posted:
Dec 21, 2011 - 12:12 |
|
yeah I went in that direction a bit but it doesn't really help me get to races, because the team races are not contained within the match xml, and even if they were it would be a bit of mess to sort out unique teams, etc. and get to where i am trying to get.*
*which btw is: has the representation of participating ag4 teams been markedly different in this vs other fumbbl cups and tournies generally |
|
|
koadah
Joined: Mar 30, 2005
|
  Posted:
Dec 21, 2011 - 12:38 |
|
|
King_Ghidra
Joined: Sep 14, 2009
|
  Posted:
Dec 21, 2011 - 13:11 |
|
I don't see how that makes things any quicker than just manually going through and reading the match reports and clicking on the team names to find out what race they are?
I was looking for an API function that will give me a quick way to pull up the breakdown of races that participated in a group or tourney. Clearly it is possible because as mentioned above I saw Christer running a command in chat that showed remaining races in the cup. Some variation on that is what I am after.
Actually I'm intrigued by how that works anyway, and it is somewhat related to the question I asked above about the numbers returned on the Ulthuan Invitational - how does the tournament know that a team is still in that particular tournment (as opposed to in the group and knocked out)?
I'm not a coder, I was just hoping to use the api to get some easy-to-process raw data and then do the rest in excel or something. Maybe it's not possible. |
|
|
koadah
Joined: Mar 30, 2005
|
  Posted:
Dec 21, 2011 - 13:38 |
|
King_Ghidra wrote: | I don't see how that makes things any quicker than just manually going through and reading the match reports and clicking on the team names to find out what race they are? |
Yeah, I did think that my self. I missed the bit that said the solution had to be quick.
For one tournament just looking at the match reports is probably quickest. But if you want to compare the results against all the tournaments in LRB4...
The API is bit of a coder tool though. Christer has access to the database so he can run powerful queries much more easily. |
_________________
Open [L]eague Tournaments - NOW recruiting!! Old/New style skill progression! - Secret League - All Stars! |
|
koadah
Joined: Mar 30, 2005
|
  Posted:
Dec 21, 2011 - 19:56 |
|
|
dode74
Joined: Aug 14, 2009
|
  Posted:
Dec 21, 2011 - 20:25 |
|
From Koadah's link, for FUMBBL Cup X only:
Code: | Amazon 12
Chaos 25
Chaos Dwarf 23
Chaos Pact 7
Dark Elf 34
Dwarf 23
Elf 6
Goblin 5
Halfling 1
High Elf 13
Human 12
Khemri 8
Lizardman 9
Necromantic 15
Norse 6
Nurgle 14
Ogre 8
Orc 26
Skaven 16
Slann 5
Undead 15
Underworld 2
Vampire 5
Wood Elf 14 |
Total Elves: 67
Total teams: 304
Quote: | has the representation of participating ag4 teams been markedly different in this vs other fumbbl cups and tournies generally |
The percentage of elven (DE, HE, WE, E) teams in each of tournaments 1-10 is:
Code: | 23.44
12.50
21.88
34.38
21.88
19.34
20.31
22.98
18.48
22.04 |
Total participants in the cup from seasons 1-10:
Code: | 64
16
32
32
32
512
256
322
211
304 |
|
|
|
King_Ghidra
Joined: Sep 14, 2009
|
  Posted:
Dec 21, 2011 - 20:41 |
|
|
|