Arsonist

[not happy]

Arsonist should be able to douse 2 ppl at night, since when he burns them around half of ppl he doused is already dead anyway. Furthermore it's easier to find arsonist than SK (since sk can always say hes a doctor if only investigated) so it still will be harder to win for arso than sk.

[Dark.Revenant]

I'm not changing the arsonist.

If you feel he's UP, pair him with another arsonist.

[not happy]

If arsonist is covered in gasoline by another arsonist and decides to burn gasoline covered ppl does he also die?

And I think arsonist is UP, cause he has the same goal as SK, is easier to identify, and can only kill ppl at around half the speed of sk.

Lets say that arsonist douses 2 ppl per night. then after 3 nights he will have 6 ppl doused, but during that time lets say that 7ppl were already killed (around 2 per night due to mafia/sk/vigilante and sometimes 1 per day). So statistically 7/13*6 = 3.23 of those ppl he doused are already dead, so on 4th night statistically he will be able to kill only 2.77 ppl, so it's not too much - and it's ALREADY if we allow him to douse 2 targets. Now on 4th night arsonist statistically is able to kill only half of it (1.39) which makes him imba underpowered compared to SK that on 4th night has already around 3-4ppl killed (depending on doc/ double kills).

[Dark.Revenant]

The strength of the Arsonist is the ability to drop the game from a lot of players to very few players in an instant. Instead of having to go through trials where the chance of being lynched is high, he can let the town's size remain large until he suddenly kills off everyone.

Once the bus driver is up, the arsonist will be all the more powerful.

[not happy]

Let's see how many death can cause statistically SK and arsonist after each night (assuming 2 ppl die per night-day, if more ppl would die then arsonist is even weaker)

arsonist sk town size at the beginning of night
0 nearly 1 14
11/13 nearly 2 12
1.51 nearly 3 10 (1.51 = 9/13 + 9/11)
1.95 nearly 4 8 (1.95 = 7/13 + 7/11 + 7/9)
2.10 nearly 5 6 (2.10 = 5/13+5/11+5/9 + 5/7)
1.87 nearly 6 4 (1.87= 3/13+3/11+3/9 +3/7+3/5)
0.95 nearly 7 2 (0.95= 1/13+1/11+1/9 +1/7+1/5+1/3 - there's just 1 person left (beside him) and there's no guarantee he has doused her b4)

So statistically at most arsonist one night may kill 2.10 (4th night) ppl so I don't see how it relates to "The strength of the Arsonist is the ability to drop the game from a lot of players to very few players in an instant." You may be a good programmer but u are a very bad mathematician. And I made a small mistake in my previous post when counting probabilities, but those ones here are now accurate.

[not happy]

hmm, after thinking a little they r not fully accurate, but they are close enough to give the clear idea that arsonist is strongly UP

[Supersun]

Arson is very strong off a random though. I mean generally you know when you hit an SK from a random, but an arson you don't know he's there until either an investigator/sheriff finds him or until it's too late.

But as a single role, yeah the Arson is definitely weaker then the SK, but that's not necessarily a bad thing. I mean there isn't a reason to buff a citizen is there? Sometimes you just need a somewhat weaker role to not throw off the balance, and like DR said when the Bus Driver comes out he'll have the ability to bluff an investigator. Sure it won't be as powerful as SKs bluff, but especially off a random where people won't know you are there that's a very powerful position to be in.

(Also I always thought the Driver was a mafia role)

[Dark.Revenant]

You're forgetting to model the fact that the Arsonist may simply wait.

For simplicity's sake, let's compare 1 Arsonist vs 9 Citizens and 1 Serial Killer vs 9 Citizens. We'll start at night. () is chance of survival after the phase.

Serial Killer Strategy: Kill every night.
Night 1 - 1 vs 8 (100%)
Day 2 - 1 vs 7 (88.9%)
Night 2 - 1 vs 6 (88.9%)
Day 3 - 1 vs 5 (76.2%)
Night 3 - 1 vs 4 (76.2%)
Day 4 - 1 vs 3 (61.0%)
Night 4 - 1 vs 2 (61.0%)
Day 5 - 1 vs 1 VICTORY (40.6%)

Serial Killer Strategy: Wait one night, then kill every following night.
Night 1 - 1 vs 9 (100%)
Day 2 - 1 vs 8 (90.0%)
Night 2 - 1 vs 7 (90.0%)
Day 3 - 1 vs 6 (78.8%)
Night 3 - 1 vs 5 (78.8%)
Day 4 - 1 vs 4 (65.6%)
Night 4 - 1 vs 3 (65.6%)
Day 5 - 1 vs 2 (49.2%)
Night 5 - 1 vs 1 VICTORY (49.2%)

As you can see, an unintuitive shift in strategy allowed the Serial Killer to gain a significant improvement on his chances (40.6% to 49.2%). The game is not about how many people you can kill, it's about how well you can WIN. Let's apply it to the Arsonist. [] is the amount of doused people after the phase, if the Arsonist were to survive. {} is the chance for a particular scenario to happen.

Arsonist Strategy: Douse and then kill every night, as long as the doused target is alive. (Oh holy steaming shit this is going to suck to write.)
Night 1 - 1 vs 9 <1> (100%) {100%}
Day 2 - 1 vs 8 <1> (90.0%) {88.9%}[*] 1 vs 8 <0> (90.0%) {11.1%}
Night 2 - 1 vs 7 <0> (90.0%) {88.9%}[*] 1 vs 8 <1> (90.0%) {11.1%}
Day 3 - 1 vs 6 <0> (78.8%) {88.9%}[*] 1 vs 7 <1> (80.0%) {9.72%}[*] 1 vs 7 <0> (80.0%) {1.39%}
Night 3 - 1 vs 6 <1> (78.8%) {88.9%}[*] 1 vs 6 <0> (80.0%) {9.72%}[*] 1 vs 7 <1> (80.0%) {1.39%}
Day 4 - 1 vs 5 <1> (67.5%) {74.1%}[*] 1 vs 5 <0> (67.5%) {14.1%}[*] 1 vs 5 <0> (68.6%) {9.72%}[*] 1 vs 6 <1> (70.0%) {1.19%}[*] I delete scenarios with chances below 1%
Night 4 - 1 vs 4 <0> (67.5%) {74.1%}[*] 1 vs 5 <1> (67.5%) {14.1%}[*] 1 vs 5 <1> (68.6%) {9.72%}[*] 1 vs 5 <0> (70.0%) {1.19%}
Day 5 - 1 vs 3 <0> (54.0%) {74.1%}[*] 1 vs 4 <1> (56.3%) {11.3%}[*] 1 vs 4 <0> (56.3%) {2.82%}[*] 1 vs 4 <1> (57.1%) {7.78%}[*] 1 vs 4 <0> (57.1%) {1.94%}[*] 1 vs 4 <0> (58.3%) {1.19%}
Night 5 - 1 vs 3 <1> (54.0%) {74.1%}[*] 1 vs 3 <0> (56.3%) {11.3%}[*] 1 vs 4 <1> (56.3%) {2.82%}[*] 1 vs 3 <0> (57.1%) {7.78%}[*] 1 vs 4 <1> (57.1%) {1.94%}[*] 1 vs 4 <1> (58.3%) {1.19%}
Day 6 - 1 vs 2 <1> (40.5%) {49.4%}[*] 1 vs 2 <0> (40.5%) {24.7%}[*] 1 vs 2 <0> (42.2%) {11.3%}[*] 1 vs 3 <1> (45.0%) {2.11%}[*] 1 vs 2 <0> (42.9%) {7.78%}[*] 1 vs 3 <1> (45.7%) {1.46%}[*] I delete scenarios with chances below 1%
Night 6 - 1 vs 1 (40.5%) VICTORY {49.4%}[*] 1 vs 2 <1> (40.5%) {24.7%}[*] 1 vs 2 <1> (42.2%) {11.3%}[*] 1 vs 2 <0> (45.0%) {2.11%}[*] 1 vs 2 <1> (42.9%) {7.78%}[*] 1 vs 2 <0> (45.7%) {1.46%}
Day 7 - 1 vs 1 (40.5%) VICTORY {49.4%}[*] 1 vs 1 (27.0%) VICTORY {24.7%}[*] 1 vs 1 (28.1%) VICTORY {11.3%}[*] 1 vs 1 (30.0%) VICTORY {2.11%}[*] 1 vs 1 (28.6%) VICTORY {7.78%}[*] 1 vs 1 (30.5%) VICTORY {1.46%}

32-34% chance of victory with this tactic. Not terribly much lower than the SK when the SK playing suboptimally, and the Arsonist has a chance to bend votes away from doused people to bend the chance more toward 40%, which is right about exactly what a SK might get in this set-up. However, this is not the best way for an Arsonist to play.

Arsonist Strategy: Douse people at random until you have a chance to win by making a single burn.
Night 1 - 1 vs 9 <1> (100%) {100%}
Day 2 - 1 vs 8 <1> (90.0%) {88.9%}[*] 1 vs 8 <0> (90.0%) {11.1%}
Night 2 - 1 vs 8 <2> (90.0%) {88.9%}[*] 1 vs 8 <1> (90.0%) {11.1%}
Day 3 - 1 vs 7 <2> (80.0%) {66.7%}[*] 1 vs 7 <1> (80.0%) {31.9%}[*] 1 vs 7 <0> (80.0%) {1.39%}
Night 3 - 1 vs 7 <3> (80.0%) {66.7%}[*] 1 vs 7 <2> (80.0%) {31.9%}[*] 1 vs 7 <1> (80.0%) {1.39%}
Day 4 - 1 vs 6 <3> (70.0%) {38.1%}[*] 1 vs 6 <2> (70.0%) {51.4%}[*] 1 vs 6 <1> (70.0%) {10.3%}[*] 1 vs 6 <0> (70.0%) {0.198%}
Night 4 - 1 vs 6 <4> (70.0%) {38.1%}[*] 1 vs 6 <3> (70.0%) {51.4%}[*] 1 vs 6 <2> (70.0%) {10.3%}[*] 1 vs 6 <1> (70.0%) {0.198%}
Day 5 - 1 vs 5 <4> (60.0%) {12.7%}[*] 1 vs 5 <3> (60.0%) {51.1%}[*] 1 vs 5 <2> (60.0%) {32.6%}[*] 1 vs 5 <1> (60.0%) {3.60%}[*] 1 vs 5 <0> (60.0%) {0.0331%}
Night 5 - 1 vs 1 (60.0%) VICTORY {12.7%}[*] 1 vs 5 <4> (60.0%) {51.1%}[*] 1 vs 5 <3> (60.0%) {32.6%}[*] 1 vs 5 <2> (60.0%) {3.60%}[*] 1 vs 5 <1> (60.0%) {0.0331%}
Day 6 - 1 vs 1 (60.0%) VICTORY {12.7%}[*] 1 vs 4 <4> (50.0%) {10.2%}[*] 1 vs 4 <3> (50.0%) {53.9%}[*] 1 vs 4 <2> (50.0%) {21.7%}[*] 1 vs 4 <1> (50.0%) {1.47%}[*] 1 vs 4 <0> (50.0%) {0.00661%}
Night 6 - 1 vs 1 (60.0%) VICTORY {12.7%}[*] 1 vs 0 (50.0%) VICTORY {10.2%}[*] 1 vs 1 (50.0%) VICTORY {53.9%}[*] 1 vs 4 <3> (50.0%) {21.7%}[*] 1 vs 4 <2> (50.0%) {1.47%}[*] 1 vs 4 <1> (50.0%) {0.00661%}
Day 7 - 1 vs 1 (60.0%) VICTORY {12.7%}[*] 1 vs 0 (50.0%) VICTORY {10.2%}[*] 1 vs 1 (50.0%) VICTORY {53.9%}[*] 1 vs 3 <3> (40.0%) {5.43%}[*] 1 vs 3 <2> (40.0%) {17.0%}[*] 1 vs 3 <1> (40.0%) {0.739%}[*] 1 vs 3 <0> (40.0%) {0.00165%}
Night 7 - 1 vs 1 (60.0%) VICTORY {12.7%}[*] 1 vs 0 (50.0%) VICTORY {10.2%}[*] 1 vs 1 (50.0%) VICTORY {53.9%}[*] 1 vs 0 (40.0%) VICTORY {5.43%}[*] 1 vs 1 (40.0%) VICTORY {17.0%}[*] 1 vs 3 <2> (40.0%) {0.739%}[*] 1 vs 3 <1> (40.0%) {0.00165%}
Day 8 - 1 vs 1 (60.0%) VICTORY {12.7%}[*] 1 vs 0 (50.0%) VICTORY {10.2%}[*] 1 vs 1 (50.0%) VICTORY {53.9%}[*] 1 vs 0 (40.0%) VICTORY {5.43%}[*] 1 vs 1 (40.0%) VICTORY {17.0%}[*] 1 vs 2 <2> (30.0%) {0.246%}[*] 1 vs 2 <1> (30.0%) {0.494%}[*] 1 vs 2 <0> (30.0%) {0.00055%}
Night 8 - 1 vs 1 (60.0%) VICTORY {12.7%}[*] 1 vs 0 (50.0%) VICTORY {10.2%}[*] 1 vs 1 (50.0%) VICTORY {53.9%}[*] 1 vs 0 (40.0%) VICTORY {5.43%}[*] 1 vs 1 (40.0%) VICTORY {17.0%}[*] 1 vs 0 (30.0%) VICTORY {0.246%}[*] 1 vs 1 (30.0%) VICTORY {0.494%}[*] And at this point the chance of the last scenario happening is far too insignificant to consider

Overall, the chance of victory with this strategy is 48.9%. This is an incredible improvement over the douse-kill-douse strategy. In any case, this proves that the Arsonist is not actually as underpowered as you might think. You accuse me of being a lousy mathematician but YOU forget the most important part of the game: lynches.

Adding a Mafia group into this equation would make it difficult to model and would merely cause the SK and Arsonist win rates to go down by nearly the same proportion, though the Arsonists' rates would go down slightly more, but to balance that out the SK's rates would lean more toward the middle of his strategic win range due to the unpredictability of the game with all those power roles. The Arsonist has some disadvantage in a game with investigators, but not by such an amount as to make him significantly UP.

[Mysterymeat]

Does sk and aronist get same pts cause I don't think arsonist is too Up, but definetly weaker than sk so arsonist should get more pts like about 10 more I think cause its annoying to take 2 turns to kill someone and by that time they could get lynched or taken out by mafia or sk.

[Faceless]

I actually usually leave Arsonist out. I think it's a cool role but I feel like mafia + vigi + arsonist causes way too many deaths for the game to involve much REAL lying since everything just happens at night.

My neutral roles (almost) always include just a Witch and Jester since I feel those two roles add the most unpredictability to the game (which I feel is the only reason Neutrals should be added, to keep things unpredictable).

Just my two cents.

[Afumba]

It is harder to kill wiht Arsonist... maybe... but the surviving chance is nearly the same.

Only problem ist that he has no allies most of the times while SK will get aid of Survivor or Witch much more often.
But DR already stated that this will be a bit better when Bus Driver will be added... so sounds fine for me.

Some roles are a little more easier some a little harder... but i find the Arsonist fine as it is.
And u can improve his chances a bit with some settings... Give him ability to ignore immunities; make him immune to sheriffs detection & exclude exact role detecting from investigators...

[not happy]

DR, I agree that when you compare arsonist vs 9 citizens or SK vs 9 citizens your reasoning is right. When no one ever dies at night from other sources than the SK or arsonist then arsonist has much more time to douse ppl before they die. But every game of mafia I ever played always contained at least some mafia, which makes Arsonist far weaker since far too often ppl he had already doused are already dead when burning time comes. You made your counting basing on a scenario that doesn't happen in a normal game (who plays 1arso/sk and 9 citizens?). I based my argument basing on assumption that on average 2 ppl die during one night-day cycle which is quite often the case (often average is even higher than 2, which makes arsonist even weaker). Therefore it is my results that have better application to what happens in game setups that people actually play. Btw has any one of you ever played a game where arsonist burned at once 5 or more players and therefore: "drop the game from a lot of players to very few players in an instant"?. And I guess hardly any of you played ever a game when arsonist was able to burn 4 players at once. I personally played already around 25 games with arsonist and never saw arsonist burn anyone, before they got that "critical number of players" doused, they were either dead already (around 23 games), all other ppl killed between each other and they were left alone (1 game) or 1vs1 with a non-killing townie that hasn't been even doused by them b4 (1 game) - I was that arsonist from that game so I know.

Btw - the counting that I've shown previously was for each night showing how many ppl arsonist can burn that night (statistically) assuming that he was dousing every previous night (and that hes still alive).

I know that I don't take into consideration things that arsonist during lynches may try to vote for ppl that he didnt douse, but SK doesn't have this problem at all - ppl he already killed won't be lynched, and when killing next night he will only choose from those that survived.

I didn't take into consideration that when SK is there, number of ppl drops faster therefore there is higher chance to get lynched or targeted by a other killing role. But sk has that plus that when he identifies some killing role that he thinks he may be target of it, or someone from town that he sees wants him lynched he may just kill him next night and by that decrease his chances of being a victim.

In conclusion you could just add option to arsonist that would allow him to douse 2 targets at once, ppl will playtest it and in the end will tell you whether arsonist is in such situation OP or not. I strongly believe he won't be.

[Dark.Revenant]

2 doused targets is completely and absolutely rigged. He would be the most powerful single role in the game, bar none. Not even the cult leader would be able to compete.

[not happy]

I will just repeat this question:

Has any one of you ever played a game where arsonist burned at once 5 or more players and therefore: "drop the game from a lot of players to very few players in an instant" ?. And I guess hardly any of you played ever a game where arsonist was able to burn 4 players at once (anyone?). I personally played already around 25 games with arsonist and never saw arsonist burn anyone; before they got that "critical number of players" doused, they were either dead already (around 23 games), all other ppl killed between each other and they were left alone (1 game) or 1vs1 with a non-killing townie that hasn't been even doused by them b4 (1 game) - I was the arsonist from that game that's how I know.

If there is someone around that saw arsonist burn 5 people even once then I agree with u DR, otherwise I think you are mistaken. And I don't get why adding that option for ppl just to playtest would hurt (I know it needs some work, if it would be a lot of work then I would understand such an argument).

You also haven't answered that question: if arsonist is covered in gasoline by another arsonist and decides to burn gasoline covered ppl does he also kills himself?

[Clawtrocity]

You also haven't answered that question: if arsonist is covered in gasoline by another arsonist and decides to burn gasoline covered ppl does he also kills himself?

Yes.

[Afumba]

@ not happy
yes, i have seem a few quadkills of Arsonist... 2-3 at least
Also i see triple kills from time to time and double kills happen all the time.

As said before i dont think the chance of Arso to win is that much smaller as from SK.
Also i dont understand y his killing behavior has to be each or higher of SK... Even if it isnt... why should it be?

[not happy]

Ok then, I saw not a single game like that that's why I was concerned that he's significantly weaker that sk. It shouldn't be equal or higher thank sks, my intent was to make it more equal to it.

[Supersun]

If there is someone around that saw arsonist burn 5 people even once then I agree with u DR, otherwise I think you are mistaken. And I don't get why adding that option for ppl just to playtest would hurt (I know it needs some work, if it would be a lot of work then I would understand such an argument).

I have, and he would of gotten more (aka me) if I didn't control him as a witch and forced him to light all his targets.

[Dark.Revenant]

Witch can't make the Arsonist light up his targets, only douse them. That vector of control gets overridden by the Arsonist's choice to burn. Then, if the Witch makes the Arsonist target someone else when doing the Burn command, the action is cancelled. Otherwise (if the Witch makes the Arsonist target himself when doing the Burn command, etc.), the burn goes across normally. The Arsonist lit them up on his own, thereby making your control useless.

[Supersun]

Witch can't make the Arsonist light up his targets, only douse them. That vector of control gets overridden by the Arsonist's choice to burn. Then, if the Witch makes the Arsonist target someone else when doing the Burn command, the action is cancelled. Otherwise (if the Witch makes the Arsonist target himself when doing the Burn command, etc.), the burn goes across normally. The Arsonist lit them up on his own, thereby making your control useless.

Odd, because I know in one game that I forced the arson to target himself and everyone lit up and I'm pretty sure he didn't intend that to happen. He pretty much told me that he was trying to douse me.

[Trenix]

Arsonist is fine, learn to adapt. Obviously if you're not doing so good, maybe you're not doing something right.