News Shop
Events Chat

Balance Team Battle Tournament Idea

I have been following the Fantasy Strike discord thread. I think the team battle idea with random character select from roster construction is a good idea. Here is my idea.

  • Each player picks 3 character stable
  • Best of 5
  • TO randomly generates all matches
  • First three matches must include complete stable
  • Final two matches will be randomly generated to use 2/3 of character list


CKR - :zane: :setsuki: :onimaru:
mysticjuicer - :troq: :onimaru: :argagarg:

CKR RNG - 1 3 2 1 2
mysticjuicer RNG - 2 3 1 1 3

MUs would be:

:zane: vs :onimaru:
:onimaru: vs :argagarg:
:setsuki: vs :troq:
:zane: vs :troq:
:setsuki: vs :argagarg:

As you can see, I am still at a slight disadvantage, but most of my characters are hard CPed by his list. My focus in this match would be to win game 1 and to try and get one of the next three to finish out game 5 with my advantage matchup. This of course is a hypothetical, but I did use the RNG to create the matchups. I think that this method would alleviate the hard CP nature of the game. It would be a good test of skill. I think Sirlin’s idea for fantasy strike can be used here.


Here is a five character stable example.

CKR characters :setsuki: :geiger: :onimaru: :zane: :menelker:
mysticjuicer characters :rook: :troq: :argagarg: :onimaru: :setsuki:

CKR RNG - 5 2 1 3 4
mysticjuicer RNG - 4 5 3 2 1


:menelker: vs :onimaru:
:geiger: vs :setsuki:
:setsuki: vs :argagarg:
:onimaru: vs :troq:
:zane: vs :rook:

I think I like the 5 character stable better. It increases character diversity and decreases the risk of a hard CP.


I will happily add it to the list of tournaments that I’m willing to run. This formulation seems a bit less fiddly than some of the other ways I’ve seen proposed to the the Fantasy Strike tournament flavor into Yomi. I like to 5-stable version too, because it allows for the possibility that you won’t even have the chance to play with two of you chars of you don’t play well.


Would you be able to automate the RNG into the conversation?

It’s definitely possible. The code is a bit crufty, but I think adding that randomization wouldn’t be too hard.


I just checked back at the tournament planning poll, and Fantasy Strike Style is winning. I want to run some numbers on the hypothetical metas for a 3-char all random, a 5-char all random, and a 3-char only win once per character, and see if any are particularly broken.

my only concern about this format is that if everyone picks the top tiers the matches would be still heavily plagued by the carousel.
EG: if I had to PtW I’d pick zane-troq-geiger-degrey-grave, leaving only the 5th spot as an actual variable

Some will pick top tiers and some will pick low tiers. With a five character selection this is less of an issue than with a three character stable. The main point is that it will eliminate the CP carousel. It is not meant to eliminate character selection entirely. Most players will probably mix in top/mid tier characters with a few out there adding in a low tier character or two.

btw, I play to win. I would pick Zane/Setsuki/Onimaru/Argagarg/Geiger. That is only a 40 percent overlap. It looks like we have a different valuation on the path to winning. I think others do, as well.


Just for fun, I will do the @legion top tier vs CKR list.

CKR RNG - 2 4 3 5 1
Legion RNG - 2 5 1 4 3

CKR vs Legion

:setsuki: vs :troq:
:argagarg: vs :grave:
:onimaru: vs :zane:
:geiger: vs :degrey:
:zane: vs :geiger:

I think it would would be fun, and we would both have a good chance of winning. Compare that to the carousel.

:zane: vs :troq:
:geiger: vs :troq:
:geiger: vs :zane:
:troq: vs :zane:
:troq: vs :geiger:

I stand by my argument that it reduces hard counterpicks and adds character diversity.

1 Like

When I crunched the number for the strictly-fantasy-strike style (3 characters, win once with each, selected at random), Zane/Troq/Geiger definitely came out on top in the overall rankings, but the spread wasn’t as lopsided (IIRC). I think with 5 chars, no repeats, the spread will likely be flatter.

As a funny aside, I actually voted for the Fantasy Strike format in the poll. I thought it was similar to the Triple Threat format that we ran in the past. I had no idea about the random character select element. I just happened to be reading Sirlin’s discord comments about it in the Fantasy Strike discord. If you want to hear his arguments for why it is a better format, then do a search in the discord. His logic is sound. That man does have a knack for his craft.

Yeah, I’ve read the stuff he wrote about it. It definitely eliminates counterpicking at the individual game level, but it doesn’t completely mitigate the effect of unbalanced characters. As I mentioned before, when I crunched the numbers, a team composed of dominant characters remained dominant, even when the counterpicking aspect was removed.

I’m actually about to sit down and try and crunch the numbers a different way to see if I can evaluate what meta would evolve over time by a bunch of agents playing tournaments together.

1 Like

I do agree that it does not eliminate power levels. There is no mitigation for unbalanced characters without restricting character choice to an extreme degree. That is the nature of assymetrical games. Some characters are better than other characters.

I try to beat the MU sheet. If I have a 45% chance to win the match with my roster, then I will try to increase that to 55%. I do my best to put myself in the best position possible, but sometimes that is just not what happens. I just have to be better.

With all that said, it would be difficult to believe that this doesn’t go some way towards balancing the cast. If Zane is the strongest character, surely only having access to him in 20 percent of your games does even the playing field by some degree. This format does have some character select restriction built in without getting rid of character select entirely.

1 Like

Oh, yeah, it definitely flattens the power curve. My lens on it is from the perspective of a tournament runner: what can I do to minimize the effectiveness of the dominant strategy, so that more strategies are viable. In the 20XX meta, for instance, the dominant strategy seems to mostly be: counterpick with one of Zane, Troq, or Geiger (I don’t yet have a good handle on what the dominant blind-picking strategy is, in the abstract, although I think that would be influenced a lot by your particular opponent).


The format isn’t going to solve tiers, but it does prevent counter-picking. So if you’re Zane you’re probably going to get a 6-4 MU because most of Zane’s MUs are 6-4. But if you’re Rook, you’re probably not going to get Geiger or BBB, unless your opponents have picked both and you get unlucky in the MU RNG. So it’s more that it gives characters with one or two hard CPs more room for play compared to a standard CP format.

Zane, Troq, Geiger are still great picks if you can play them well, but Rook, Quince, Gloria, Oni aren’t terrible picks.


Imo with a 3 char rooster, any PtW player would pick Zane, troq and geiger/degrey.
Simply because they are very very strong and have no loopsided mu.
Imo it would be interesting if the players were forced to pick 1 high tier char, 1 mid tier and one low tier. Would spice things up :chibilum:

5 character stable with first to 3 sounds pretty hot to me.


yeah, 5 chars would make things more interesting, but the OP proposed a 3 chars stable. Hence my objection :stuck_out_tongue:

1 Like

Why did you pick five characters, then? Seems disingenuous.

because i misread XD I saw the 5 random MU and thought it was a 5 char stable. Then I read the OP again and realized my mistake, but did not edit cuz the point imo is still valid.
(The point being that even with a 4 char stable the first 4 spots are locked in a ptw mentality.)
Just my 2 cents, man.
Sorry if it came out wrong, hope you are not offended :frowning: