If anyone's interested, the topic I said I would create for discussing behaviour in a mafia game is here.

Sorry Rubik, this is the last time I'll derail your thread.

day/night length was fine

majority not ending the day probably puts scum into a somewhat worse position. It is probably a good mechanic.

Choosing roles was fun.

Roles having multiple one-shot abilities does not strike me as a mechanic that inherently influences the game, in this execution it was not remarkably different from every player having a standard role. There were too many power roles in this game, I believe.

The number priority system did not seem to play into any sort of decision making or into any sort of puzzling out what happened during prior night actions. I mostly ignored it during the game.

Daytime abilities went smoothly.

I thought the days were a tiny bit long. Night length was okay.
Days not ending upon a majority is okay in my books.
Choosing roles was fun.
I would have preferred roles with static abilities, but that's just me.
I didn't pay any attention to the number priority system during the game, but thought it was an effective idea before and after the game was played.
I thought there was too much public mod confirmed day abilities.
Oh, a lot of them. Some were very innovative. I won't point out my favorites because they've inspired me to include some things like them in my upcoming games.
How long did it take you to come up with all of those roles?

Thoughts on the game, as a design:

I think the game was pretty well rounded, overall. Naturally it's hard to create a "choose your role" game with variety that is also fairly balanced. The only way to truly balance such a game is to give the player choices that add up to basically the same thing, which takes the fun out of this style of game. I don't feel like any of the abilities were over-powered, as every ability presented was easy enough to argue against (I will caveat that I have not had time to peruse all of the unused characters, I'm merely speaking on that which was chosen). At no point did I feel like either side didn't have a chance of winning, which speaks to care given to balance to me.
I enjoyed the numerical system in it's theory (since it didn't seem to actually come into play in this game) and would like to see more of that. The only issue I had with ability resolution would probably be the one doc to rule them all methodology, but that's probably personal bias. I look at doc'ing like a regeneration ability in Magic. You get doc'd, that prevents one kill. You take multiple hits, you'll need a doc'ing for each one. But like I said, that's probably just a personal bias issue.
On the subject of days, I felt they were too long. A week is, in my opinion, a good measure of time to accomplish everything you need to get done in a day (and by good measure I mean max length). The nights I felt were probably appropriate. With all abilities being one shot, a little more care was called for in making night decisions and strategizing for the coming day (even if you were strategizing with yourself). I think the game also benefitted from days not ending early because of majority. With the plethora of day abilities available having a guaranteed window helped plan and execute those better as well as not placing any pressure on having to be chained to a timetable.

My thoughts on the gameplay itself:

Town made a couple of great plays this game. First being Rag's night 1 ploy. Granted the argument presented against it Day 1 and the subsequent almost non-lynching of your truly is a valid counter point, but the ploy was well timed, which makes it the great play it was. Even if town was unfortunate enough to get caught in the cross hairs, it's early enough for town to recover from and still had more chance of hitting scum than not hitting scum (not to say it had more chance of hitting scum than town, but to say that it was very likely that at least one of the people caught up was likely to be scum).
Second being Neo's self protect shot at the end of the game. This really put scum in a bind. Even if he shot town, he removed himself from the possible suspect pool. Protecting himself guaranteed scum couldn't interfere with his actions either without detriment to themselves. Adding kills on top of this only narrowed the suspect pool even more. While it's true that scum could have pulled out a mislynch following this, I don't foresee us taking it all the way to a 1-1 finish.
Zinger's flip ploy probably would have been better had he not employed it on a day I was vote blocked. Since I couldn't complete my side of the action that day, there was literally no incentive for me to flip on the rest of the scum team, as giving up the scum team would likely result in a retribution killing the following night resulting in still losing. Granted I had no intention of doing so anyway, but that's beside the point.
Scar's play was decent at clearing KoD, but ultimately required him to become confirmed town, whether by ability or death, essential to that information being trustworthy anyway.
Conversely I felt scum played a pretty sub par game. Not to say the whole game, just when it mattered. <_< Mostly I'm talking about day 3. Because the role I chose required me to be the angel of death, I kinda put on the persona of unofficial face of the mafia in the early game. I was the vocal one, I was the one executing the kills. This allowed Garren and Katsura to pretty much fall into the shadows. But when I died neither stepped up to sow the discord in the town Day 3. To me, this was the real turning point of the game for scum. Had we put more pressure on key areas on that day, the rest of the game probably would have been drastically different. Those of you who dove into the mafia thread can probably see that I was still trying to direct traffic there, often calling peoples moves before they made them but to no avail. Let this be a lesson people, lurking wins nobody games.

One argument that I'd like to address now, I voted for True on a whim. Rag at the time was simply an easy scapegoat, but offered nothing more. Killing True presented a mislynch, but for much better reasoning. Simply put, it put me in a better perspective light to kill True than Rag. I wasn't afraid of him because I could kill him in the night via my protection ability, but that wouldn't garner me a more favorable light.

Fun fact: I used no abilities this game. Also I chose Umibouze because I thought it'd be better, for the team, but I really wanted to play Kagura. I think I've addressed everything I wanted to.

It was quite obvious to me that Numbers and KOD were together, even the time of their posts matched. My other suspects were Scarlet and Fruit at the same level, and Neosilk a level below. These were my impressions at day 1.

My plan was to threaten them with a kill, but not use it at night, making them waste a protective ability. Instead using my first ability to kill one of them at the start of day 2 with just my vote, as I had a killing protective ability too and I was thinking they could have one and use it if I named my kill.

And of course use my own killing protective ability at night 1 if I proved right. Of course If Kod was lynched that would reveal KOD was a mason with Numbers, maybe clearing Numbers in my eyes.

So you see. The best play for mafia was to throw KOD into the fire (and I was counting on that too if I was wrong). Numbers could not change his stance without looking suspicious, but he just needed to make one of his scum buddies to vote KOD, lynching him and proving he was a mason. Effectively painting me bad and clearing at the same time my suspicions of him. The plus is he could hide his killing with my own, making Ragnarokio plot not work.

I am a winner.