One thing that is stunting this growth and creating balance problems for mods, is the awkwardly restrictive cap of 12 players in Mini Normals. To the casual Mini Normal player, it is forgivable to gloss over the concept of a possibly unbalanced game for your faction. Likewise for Mini Normal mods, how balanced your set-up truly is, is a difficult question to answer and most won't understand or forsee the subtle intricacies of certain role interactions and the amount of power needed to offset scum in a 12-player game. And for the most part, it is not their fault, as their knowledge of set-up design (likewise for reviewers), is based on a small sample of similar games they have played in or read. They soak up a small pocket of (usually recent) games to solidify their understanding of balance, and will produce games based on this ideology.
Obviously, the more experience playing and modding games, the more chance you have of producing a balanced set-up (coupled with the more experience your reviewers have). But because of our current queue system, the overwhelming majority of Mini Normal mods are first-time mods, which makes the cycle of unbalanced games hard to snap, if these newer players are largely emulating what they see in games run by first-time mods. A mandatory review committee (something hopefully happening soon) will go some way to improving the quality of Mini Normal set-ups we produce, but I think it will only be useful to a certain point; correcting wildly unbalanced games, providing support for confused newer mods and a gradual shift towards greater standardization of roles. My possibly controversial stance is the majority of Mini Normal games will still be unbalanced under the umbrella of 12 player set-ups, even with a slew of competant reviewers behind each game.
The problem with balancing games, is you can only account for what you think is wrong. If you see no problem overall, you cannot fix it. And I think the main problem is the lack of players acknowledging or even realising there is a problem, because the problem itself can only be spotted when you view the overall picture, rather than a handful of games. Here is the up-to-date win/loss statistics for closed 3:9 Mini Normal games;
This is considerably disturbing when you take into account, this is the most common 12 player set-up (by some way), and is the most biased toward scum. We are obviously doing something wrong here, and I will talk about my proposed solution in a minute, but I have some more numbers to show you first. Check out the last 50 completed 3:9 games;
Something equally is alarming is how comprehensively mafia is winning some of these games. Looking at those last 38 mafia wins, have a look how many scum the town actually manages to eliminate:
Finally, of the last 50 completed Mini Normals of ANY TYPE, 3:9's have been run in 60% of those, which runs contrary to the overall make-up of 3:9's in Mini Normals, of about 45%. This shows the trend is slowly moving toward 3:9's as the default Mini Normal set-up, which is very scary when you see how easily and how often scum has been demolishing towns.
The main questions we need to ask ourselves is why is this happening, and what can we do to change this? Both these questions are very important as the change can't be made without accurately theorising and understand the why. Here are my positions;
The main awkwardness stems from the initial question of how many scum to put into a 12 player game. Two is too few for a 12 player game, unless you are running something close to mountainous, and three is too many without pumping the town with swingy powerroles to offset this extra scum. I think the biggest misconception is how difficult it actually is to lynch three scum, when the town only has three mislynches up it's sleeve. In a sense, it is almost a race of
This is the main problem, as I see it - the low rate of scum lynches Day 1, which puts town behind early in a contest where teams are striving for the same goal (number of lynches/mislynches-wise). Whilst Day 1's are largely believed to be about generating information to improve lynch odds on future days, you now also have a higher ratio of scum to townies in the game, which in a sense, offsets some of the advantage town gains over scum - more information. And I have serious doubts that many towns can scumhunt at a 60% success rate without serious help from PR's.
Of course, with an even number of players starting the game, the amount of mislynches needed for scum
An extra number from the beginning now changes the dynamics to four mislynches needed versus three scum lynches. It also serves to eliminate some swing, as a blocked scum kill at night won't give an extra mislynch to town, but will merely require it to no-lynch at some point. 13 player set-ups also have benefits beyond the realm of 3:9's too, as 3:1:8's need an extra townie for balance, as town wins rely on cross-kills. It also enables 2:2:9 set-ups, which are a far more attractive option than 2:2:8, as worst case scenarios can see towns lose after N2. An extra townie still keeps towns in it.
The only thing 12 player games have in their favour is familiarity, and the only reason 12 became the magic number, is because this is how many players the very first game used. It is has served it's use long enough, but it is completely arbitrary, and if we seriously want to improve the quality of games (namely balance), we need to rethink this structure, because it is naturally impeding balance of our most common set-ups (outside newbie games). Of course, thoughtful and creative mods can overcome the obstacles of the number 12, but it is very difficult to do right. Rather than individually nitpicking and focusing on each 12-player set-up, knowing it is a tricky number to work with, ticking up to 13 players means that the set-ups that don't come under as much scrutiny as this will still have a good chance of being balanced.
With the soon-to-be introduction of mandatory reviews for normal games, this to me feels like the perfect opportunity to review the base number of players mods build their set-ups around and introduce something new, that will go a long way to producing fairer games. If you've read this far, thank you! I hope I've been convincing, because this is something that should be talked about again.