Forum Brawl #4 - Suggestions and Feedback

[Vangelios]

 May be less class to vote?  artist, decks, cards, "forum in general" and bonus are enough to me,
 Beacuse like this is now I'm afraid to imagine 12 teams.

[Treldon]

May be less class to vote?  artist, decks, cards, "forum in general" and bonus are enough to me,
 Beacuse like this is now I'm afraid to imagine 12 teams.

(Un)Fortunately we don't have to worry about handling 12 teams. We should worry about whether or not we have a forum this time next year...

[Naesala]

May be less class to vote?  artist, decks, cards, "forum in general" and bonus are enough to me,
 Beacuse like this is now I'm afraid to imagine 12 teams.

I assume you mean fewer roles? We do not intend to cut any roles at this time.

[Vangelios]

May be less class to vote?  artist, decks, cards, "forum in general" and bonus are enough to me,
 Beacuse like this is now I'm afraid to imagine 12 teams.

(Un)Fortunately we don't have to worry about handling 12 teams. We should worry about whether or not we have a forum this time next year...

I Agree gold donor, you is all right

[andretimpa]

May be less class to vote?  artist, decks, cards, "forum in general" and bonus are enough to me,
 Beacuse like this is now I'm afraid to imagine 12 teams.

I assume you mean fewer roles? We do not intend to cut any roles at this time.

I'm thinking of changes that would reduce it to 4 submissions per round (plus challenges and bonuses). Basically cutting Off-Topic (that has felt redundant for the last 2 brawls and could be merged with Writer/Forum Expert), the bosses assuming one of the roles instead of having separate tasks (would require changes in the way auction is managed) and running one combo task every round (and not making combos overwhelming to vote). I'd first ask around the reasons why there was such little voting from non brawlers.

[russianspy1234]

Brawl 2 had lots of combo tasks.  Like 4 roles to one task.  More like that would reduce the tedium of voting, which I agree is an issue.  The biggest way to deal with that is unfortunately out of our control.  It seems that not many people other than brawlers are voting.  If we had significantly more voters than brawlers, we could just rely only on them.  Another possibility that works with the relatively small scale of the event is this:  Bring back the expert panel, with a twist:  No voting.  After each round, there is a shortish (few days) period of discussion in the "voting" thread between brawlers and spectators alike, where we all compliment or criticize each other's work.  After that, the brawlmasters and expert panel choose the winners of the round.  This would decrease the impact of a brawler not participating in the vote as well as completely eliminate the issue of brawlers just voting for their team and one other.

[andretimpa]

Putting more than two roles in a combo makes the voting more difficult to be objective, as there are too many variables at play. 2 roles seems lik the best way to make combos. Of course, we could just make more combos.

[artimies7]

I thought the number of combos was alright, as a competitor. I know for the ones we did have, doing the art for a card or whatnot, were appropriate and made sense. It's a bit harder to do something when you're waiting on your teammate to finish their half and I think that would be magnified with 3-person combos, though. More 2-person combos is for sure viable.

An expert panel would be an interesting change, though. Could we try something of the sort out for another competition, just to get the logistics down? It'd be a big pain to sort out all the little stuff in a brawl all itself.

[ddevans96]

An expert panel would be an interesting change, though. Could we try something of the sort out for another competition, just to get the logistics down? It'd be a big pain to sort out all the little stuff in a brawl all itself.

I'll discuss this with the other COs, see if we can find a good implementation for it.

[Naesala]

Brawl 2 had lots of combo tasks.  Like 4 roles to one task.  More like that would reduce the tedium of voting, which I agree is an issue.  The biggest way to deal with that is unfortunately out of our control.  It seems that not many people other than brawlers are voting.  If we had significantly more voters than brawlers, we could just rely only on them.  Another possibility that works with the relatively small scale of the event is this:  Bring back the expert panel, with a twist:  No voting.  After each round, there is a shortish (few days) period of discussion in the "voting" thread between brawlers and spectators alike, where we all compliment or criticize each other's work.  After that, the brawlmasters and expert panel choose the winners of the round.  This would decrease the impact of a brawler not participating in the vote as well as completely eliminate the issue of brawlers just voting for their team and one other.

We decreased the number of members in any given combo task because people complained it didn't provide enough time for people to all work together, and it's harder to vote on them fairly.

The expanded expert panel has problems. Firstly, it removes one person per role who might have competed from the competition. While vrt and OldTrees, who don't compete, might fill part of these roles, it'd be difficult to fill all spots and have enough people to run the event. If we had a larger community, I'd be more in favor of this. Next, We get to the problem of people disputing with the "experts", as it becomes largely on person's opinion what the scores come out to. Even with discussion, and a panel, ultimately the "card" person would be ranking everyone. We have discussed before, and it seemed most people would prefer democracy in this case, as do I. I don't think I would be for expanding the expert panel sans a bigger community regardless, but if CO's want to "test" this that would be fine.

[OldTrees]

The value of consistently placing +1 place is: 7/9/11/11/11 points.
The maximum difference in auction prices is: 10 (1 pt bid by a favored boss vs 10pt bid by another boss)
This difference means Brawl assumes that the average place for all brawlers in a role will be within 1 place.

I remain convinced that this math is suboptimal. Let's see what data we get from this Brawl to support or dash my doubt.

Brawl 2 stats
Brawl 3 stats

*My preliminary prediction is that the average point gap(from all 6 rounds but ignoring auction costs) in each role will be 21+(aka greater than twice the maximum auction gap)

Artist: 67-20=47
Card Designer: 56-27=29
Deckbuilder: 63-27=33
Forum Expert: 60-12=48
Off Topic: 62-33=29
Writer: 67-30-37
So the point gap ranged from 29 to 48 over the auctioned roles.

1) Is it ok for the point gap to dwarf the auction by at least a 3:1 ratio?
2) Would it be better if the point gap was closer to the auction spread?
3) How would you make these spreads closer? Would increasing the auction cap work?

Given that replacements are people who were either less desirable than the others or like me posted a topic in the final hours of the auction, charging average price seems harsh.  Charging at all seems harsh given how little control the boss has on if IRL takes one of their members

4) How should replacements be handled?

[JonathanCrazyJ]

I have seen the run-down of the DB role somewhere, are the rundowns of other roles around somewhere?