My games are always about the Mafia, more specifically how they’re able to strategize with there being few distractions to confuse the town. Naturally, this isn’t the case with everybody. Seamus’s Capo di Tutti Capi series, for example, is about the individual choices people make. A lot of other hosts take complicated mechanics and run the games to see if it works out. A lot of others use it as more of a roleplaying opportunity than anything else. However, I’ve believed for a long time that high drama comes from elsewhere. Like the typical mad scientist, I like to take a mafia team, put them under ever-increasing pressure, and see what comes out after a certain period of time.
I suppose I had always thought this on some level, but I didn’t start truly thinking about it until a bit before the game started. There had been a string of mafia victories on the .Org. Dating back, Chicago Soirée, the Scourge of Ephesus, The Prometheus, The Full Monty, and Midgard II have all ended in town losses with only Fillet Royale breaking the streak in Large games.
My curiosity on this matter only increased when White_eyes brought up the supposed “Curse of GH” in the sign-up thread, where the town usually always wins. I knew this was true, having, you know,
hosted the games, but I never really thought about why this was the case. Then, I started looking at the differences my games have from the others. Obviously, I’m well known for hosting vanilla games. But is that the only reason for the town’s disproportionate success in the Mafia series and now the Godfather series as well?
As it turns out, pretty much. The way the games on the .Org have evolved, we’ve always placed more emphasis on the mafia than the town, and the result is numerous complex, sublime, and most importantly
effective strategies used by the mafia against a town that is always a couple of steps behind. It had gotten to the point where TinCow and I were having a discussion about possibly putting in substitute objectives for the town as opposed to killing the mafia, and whether those goals could even consider the game a mafia game anymore. Anyway, at the risk of going off on a tangent, the point is that, other things equal, the mafia beats the town on the .Org a significant majority of times.
I have tried to compensate for this, both consciously and unconsciously, by making my Mafia series experience for the mafiosi something akin to the garbage pit sequence in the original
Star Wars – things look good at the start, but eventually the walls close in on you from all sides. You and your partner have to first put all the barricades up to buy time, and second think of a proper escape strategy (i.e. endgame plan) since those barricades won’t hold forever. This is doubly true in the Godfather series, speaking from experience. I remember Silver’s Godfather 2 keeping me up at nights, just constantly thinking about strategy and tactics. When do we switch to one kill per night? Do we at all? If we ever go back to two kills, how many holes will that blow in our alibis? Is the twist potentially helpful or harmful? Will nightkilling this person come back to bite us in the future? Will pushing for this person’s lynch? What happens if the Godfather comes under suspicion while I’m still alive? I think it’s this kind of evening of the playing field that always made so attracted to the series in the first place, and why it wasn’t too much of a jump for me to host it.
An alternative explanation for the amount of mafia victories on the .Org as opposed to town victories is game setup itself as opposed to superior strategy. Many of these games have lots of roles, which I believe skews things in the mafia’s favor for the following reasons:
-
More ambiguity, and thus room for deceit. It’s easy to infiltrate the town and fake a role PM in a game with a ton of roles, especially secret ones.
-
Town exhaustion, especially when there’s multiple mafia factions to be dealt with. The fact that it’s so hard to lynch a mafioso anyway makes a game with multiple “families” that much harder. The best example that comes to mind is the recently completed game, “The Scourge of Ephesus.” Even in Ephesus, when the town was on its game and lynched four villains (Sigurd, CA, Chaotix, and boudica, in that order), an anti-town faction still won.
-
Extra firepower. A sudden spike in kills per night can be devastating to the town, especially in the mid- or even endgame. Since the potential for crossfire is low (for the same reason as why it’s so hard to lynch a mafioso) it all adds up to the town in trouble.
Now, this isn’t to say that these games aren’t enjoyable, because they most definitely are. This isn’t to say that they’re unbalanced in favor of the mafia, either, because I know how much time these hosts put into making their games work. As stated in my first paragraph, perhaps the overall theme of their games is simply different than mine.
Personally? I believe that it’s a combination of factors. .Org culture combined with multi-role games combined with hosts looking for something else rather than what I usually strive for. However, the fact remains that the win ratio is not 50-50.
A lot of the veterans in the Gameroom have made names for themselves by being mafiosi: Sigurd, khaan, TinCow, Kagemusha, myself to an extent. The number of people who have achieved reputations as excellent townies is much lower. As a matter of fact, I can only think of two truly great townies: Kommodus, who doesn’t play anymore, and Sasaki Kojiro, who plays intermittently. Likewise, a lot of the excellent mafiosi listed above do not have the same skills as a townie. khaan freely admits that he’s an abysmal townie. TinCow doesn’t go that far but I recall him acknowledging that he’s a better mafioso than he is townie. Personally, I know townie play is and always has been the weakest part of my game.
Ever since the start of the year, I’ve been trying to change that to become a more complete player. I think it’s gotten off to a somewhat decent start, but judging by losses in Ephesus and Chicago Soirée, there’s still a lot of work to be done. I remember reading Sigurd’s commentary for Ephesus, pointing out my “predictability”. That little revelation left me pretty shaken, and I knew that my generally by-the-book, conservative playstyle had to have a makeover, whether it was to be simply more thorough in my examinations of the thread or something more radical.
At some point during the game, The Godfather, Part 3 ceased to be just a standard game for me, but more of an arena in which all of the various townies’ philosophies did battle. To the victor (even if the mafia won, there would still be a victor, for it would be whoever was most on-target) would go my respect, and even possibly my adaptation of that particular style in future games.
At one end of the spectrum, we had those who believed in pure analysis; the true left-brainers of the Gameroom. This town style’s champions were most notably
Seamus Fermanagh, who focused,
Holmes-like, on discrepancies in behavior as compared to other games, and
TinCow, who always looked to write-ups for clues. At the other end was
Reenk Roink, who preferred voting by retaliation, and of course, intuition (or, as he preferred to call it, “direct esoteric knowledge that surpasses systems of logic and reasoning, language, and sense perception”
). Other styles and their champions that were present in the game were
seireikhaan’s usage of vague PMs to try to flush out mafiosi,
Sasaki Kojiro’s doing multiple rereads of both the thread and individual posts to try and make sense of them based off of present knowledge as opposed to the past’s obscurity, and, of course,
Askthepizzaguy’s raw, promising, but ultimately flawed analysis program Skynet. I’m going to examine each strategy a little bit in-depth and then rule on their ultimate effectiveness, at least in this game:
seireikhaan’s “flush out” strategy: A simple, time-honored strategy: Probe for a reaction. Heck, it’s what everybody does in the thread a lot; the only difference is that khaan did it in private. It can and possibly has worked in the past – it could freeze an inexperienced player and cause him or her to stumble. However, as I mentioned above, I think a lot of players are beyond that by now, especially as mafiosi. In addition, khaan based his initial PM targets off of incomplete reasoning (he failed to determine the true reason for his kill) and thus his efforts were doomed from the start. All in all, I would not use this as a primary strategy in games, although it does have some effectiveness as an augmentation.
TinCow’s “go to the source” strategy: TC has always looked at subtleties in write-ups and analyzed them (and to a lesser extent, behavior) to determine the current state of affairs for the town. In my past two games, Mafia VIII and Godfather 3, he was always at the forefront of trying to decipher my recurring clues: the letters and songs, respectively. He also did much of the same work for Ephesus and probably other games as well that I can’t recall at the moment. This strategy does have its uses; in games where hosts do that sort of thing it can do an amazing job of painting the overall picture and helping the town get on the right long-term course. However, that’s the catch:
in games where hosts do that sort of thing. Sometimes the host could be extremely sadistic and throw in a bunch of red herrings. Sometimes the host could give control of a lot of write-ups to a mafioso. Sometimes what the host puts in there is just so complicated and difficult to make anything out of that ultimately it saps the town’s energy. I do have to give TC credit for being the first to try to make sense out of the songs (and, I think, determine that I was the boombox guy), which ultimately (and somewhat unintentionally) helped the town in the end. However, ultimately, this is another good augmentation strategy, but I think overall there’s too much at risk to solely rely on the write-ups.
Seamus Fermanagh’s deviation analysis:
Holmes without
Holmes, basically. Seamus looks at behavioral patterns and post count, compares them to other games, and bases conclusions off that. It has traditionally been effective and Seamus gets credit for ID’ing Beefy early on for his high post count (beyond one standard deviation). However, there are a couple problems with this, one of them more about the person than the strategy. Seamus, for whatever reason, just simply isn’t as listened to as are some of the other big players like Sasaki and ATPG. I’m not sure why this is; we all know that he’ll usually hit one mafioso before the end of the game. Maybe he lacks the killer instinct the other players have, the switch that some of the other players turn on and doggedly pursue a certain target until he either answers them to satisfaction or is lynched. Maybe, in an effort to stay purely unbiased and scientific he refuses to make conclusions unless he’s absolutely certain, which in turn holds the rest of the people back from following from his lead. I’m not quite sure. The second problem is that this strategy may not hold water in the future. Again, as .Org mafia players get consistently better and better, the patterns and deviations that Seamus looks for will be less and less.
Askthepizzaguy and Skynet: ATPG’s much-maligned program was intended to be a more-than-suitable replacement for
Holmes, analyzing players on a variety of factors. However, it had a ton of kinks and questionable parameters. Reading ATPG’s Skynet post-by-post analyses for certain suspicious players, I raised my eyebrows at some of the “suspicious” criteria, such as overall agreement with his ideas or not, and whether a person’s being defensive or not. ATPG does get credit for his effort and putting pressure on Tevash to reveal (in which a lack of follow-up led to his downfall), but Skynet’s greatest flaws were that it was, I think, subjective, and more importantly without context. Yes, Reenk disagrees with you, but if he always does that, wouldn’t that make him
less suspicious? Yes, a player might have a certain, mafia-level amount of posts, but does that person normally have that amount? A person may say something suspicious here, but knowing what we know now, does it make sense given that person’s situation at that time? Add ATPG’s effort into application of context and Skynet could have been a major weapon.
Reenk Roink’s operating on a higher level: Frankly, I’m not sure how to describe this one, so I’ll just go with the time-honored maxim: “Reenk is Reenk.” I remember during the pivotal penultimate round, Reenk changed his vote to White_eyes on the grounds that he “woke up and determined White_eyes was guilty”. I think for purposes of discussion we’ll call this gut instinct. Reenk will laugh at me for trying to scientifically prove that going with your gut is more or less accurate than any of the other methods, and I know I can’t. All I’ll say in this case is that two noted gut advocates, Reenk and AggonyDuck, have been more right than wrong, and that at least one more (
Silver Rusher, the creator of this series), for all his wonderful qualities, couldn’t name a mafioso to save his life. So either you have it or you don’t. I suppose that your gut instinct could, over a long period of time, improve with experience, but you wonder if it’s your gut or just something small in a person’s post that alerts you. Anyway, with his instrumental role in getting first White_eyes and then Beefy lynched, Reenk and his style is definitely one of the victors of the overall struggle, so much so that I decided to adapt his strategy in Andres’s “King Wahaha” mini-game and just say “screw it” to all of my previously-held principles. In the second round, I said that I had pretty conclusive evidence that taka was guilty, even though I had no such thing. I refused to explain myself, and while the town lost that game, it turns out that taka *was* guilty. This was simply my application of Reenk’s usual style. I think I’ll continue doing so in future games, although to a moderate extent, since doing so too much would probably fry my brain.
Yes, Reenk’s style is good, especially for the people that “have it”, but in this game one other style trumped his:
Sasaki Kojiro and the search for context: There’s something in the Godfather series that brings out the best in Sasaki. First in GF1 he, as the Mafia Mastermind manipulated everyone, including me as the Godfather, to a Total Mafia Victory. Then in GF2, despite pretty much everyone believing in his mafiosery, he almost singlehandedly turned the tables on the bad guys and directed the town to victory. Now in GF3, despite being rusty, he still correctly identified the entire trio of villains and significantly aided in the town’s comeback. How did he do this? Simple – by following the standard Sasaki method of operation. Make early and mid-game accusations, get said people lynched, analyze the direction the game goes in, and then around the endgame reread the thread and revise your conclusions based on how much sense the living people made back then knowing what we know now. Sasaki’s long post that probably doomed White_eyes and Beefy linked then and Tevash together by looking at their overall patterns and the implications of their actions. In other words, he did what Skynet did not by putting things into context. Sasaki’s town philosophy takes a lot of work, particularly in longer games with a lot of information (like mine), but on the other hand it is precisely these types of games where he is most accurate. Results is definitely proportional to the amount of effort you put in. This is true as host, mafioso, and townie, and Sasaki displays it here with another great game as townie.
So yes, Sasaki’s philosophy was the most effective, at least in this game. Am I going to use it? Possibly, although I’m not going to totally copy his style. Every strategy I’ve listed above, as well as the less obvious ones that I didn’t, has its merits, including my own personal, unrefined style that I’ve tried to work on over the years. An amalgamation may be best, copying Seamus’s numerical devotion with Reenk’s hunches, combined with Sasaki’s application of context.
In the end though, town style is your own personal decision and an extremely important one at that. Perhaps, if people spend enough time improving and perfecting theirs, the .Org Mafia players could resolve the discrepancy in Mafia:Town victories in the best way possible for all sides: the town closing the gap by improving their play. After all, improved play on both sides can only lead to better and more exciting games for all involved.
Re: The Godfather, Part 3 (Concluded)
Well done town!
Reply With Quote
It never once occured to me that someone would random.org kill me; I'm such an easy lynch target that I assumed nobody would kill me without spotting my fish for CoP. Plus, I myself am so careful with my kill targets that I almost never pick a random kill, which clouded my mind to the possibility of random killing.
and for reference....
.
.
.
So a big sorry to Sasaki. I just found it hilarious though when ATPG accused us of being mafia partners when i was the one that killed him! 
Originally Posted by TosaInu
Add that to Sasaki defending Reenk up and down and Reenk's odd behavior, and Beefy's very cool behavior...
Bookmarks