Board Thread:Debate of The Week/@comment-28112409-20181230235210/@comment-28112409-20190108161707

Addfire wrote:

But the Forger. The Forger is just invaluable. Trying to get an NK or NE out of the way? Forger. Hiding a TI's information, or a Townie's scumreading in general? Forger. The only case where a Forger isn't useful is when there are literally no townies left.

Meaning that your brief little jaunt in starting an argument was pointless and ill-fated, because you just got proved w r o n g.

(but because you haven't offered any counterpoints for when I dunked you on Forger being better I'll let that particular sleeping dog lie) Stop. Just stop. Snarky is tolerable, but snarky and wrong simultaneously is a level of torture this thread did not need to see.

 The "Forger can be good" argument 

The worst role/waste of a slot in the Mafia faction is "invaluable" and being better than Ambusher is quite the bold claim on your part, in addition to extremely lacking in any solid points. As long as Janitor exists, Forger will always be the waste of a slot. It's the new pre-rework Disguiser, so to speak.

Getting the common things out of the way: They both have 3 uses, they're both countered by a town player simply claiming a role and pasting the will, and they're also both countered by Medium and Ret. Let's say they don't get countered by either of these, since both of them getting countered by the same thing equalizes these points of the argument.

For a Forger to be successful, the will needs to be convincing and coherent, obviously. If we check at the TIs, the Forger has a 25% chance of succeeding if he was to randomly forge a will without knowing the role. You can't write a will with visits and call it a day, every TI is expected to have a will format with specific information. Investigators need to actually show results, Sheriff needs to do so too, Spies need to show mafia visits and bug results, Lookout needs to show you who the targets were visited by. Janitor? He doesn't give a fuck, clean that son of a gun and hide his information from the entire town, while creating claimspace for your own mafia and having little to no worries about your chances of success.

Forging TPs, TKs, and TS is not quite useful. What exactly would you attempt misleading them into? TPs and their targets doesn't mean anything to the Town, Vet is a pointless target for forging and getting a use out of Vig requires very specific conditions to work and is still pretty unlikely to succeed. Forging a Mayor/Ret is pointless, and a forged Transporter/Escort will is easily disproved. At best, there's the Medium, who rarely offers any possibility of misleading the town worth a forge. Janitor? He doesn't give a fuck, clean those sons of bitches and make claimspace for the mafia, hell yeah.

Spies can counter Forgers by simply seeing the visits happening. Even if you succeeded in forging a will, the Spy pointing out that there was a 2nd visitor with the mafia kill alone can insinuate for the Town to be wary of what the will says, and they might divert to using more reliable information like collecting claims, other trusted TIs, etc. since there's a 50/50 chance that the will is forged, and what the forged will says can shift that balance more in the forge failing such as visits that don't make sense or results that were already pretty obvious. Janitor? He doesn't give a fuck what Spy says, because it's worthless since the Janitor's deed is public after the kill and no information can be recovered by seeing him be the 2nd visit. Spy is essentially a null role against the Janitor, and definitely not a worry.

Worst case scenario utility are not in favor of Forger either. At its worst, Forger can just clean a will, preventing the Town from knowing the information in there but also preventing the mafia from accessing the information of that will. Janitor? He doesn't give a fuck, he'll clean the will AND keep it for the mafia to see AND clean the role on top of that, clearly being superior to Forger in a worst case scenario because of a better chance of contributing with more utility.

For Forger to gain a significant unique edge that makes it even equal to Janitor's potential, it needs a Consigliere. It relies on another non-killing mafia member to have better chances of forging successfully without a Spy to keep him in check. The Janitor, as always, does not give a flying fuck. He doesn't need a non-killing mafia member to be significantly useful, and holds its own ground in utility better than the Forger can. You could say that public claims can also be used by the Forger to identify a role, but claims are often accompanied by pasted wills, which counters both deception roles anyway.

Anything that Forger does, Janitor does better. Anything Forger does that Janitor can't do is on a negligible and an abysmal scale that is insignificant and inconsistent. Forger is a worse Janitor, and therefore a waste of a slot. It is not good in any capacity when there is a superior version of it in the alignment.

With Framer and Disguiser being below Ambusher in terms of contribution, and yourself admitting that Ambusher can "probably" beat out those two, the argument that Ambusher is better than the majority of the Mafia Deception alignment still holds and is absolutely valid. I don't care if Bet doesn't argue it any more, I'm arguing it now.

 Ambusher's Math 

The argument in question is that the "Ambusher being visited by more than one person is not that common", just to keep this clarified.

The math is wrong, with assumptions either out of laziness or ignorance, reference to an outdated source, absurd logic, and also wrapped up nicely with an asinine conclusion.

After getting that 0.01% for N1, you should've just stopped there. It's not that common in the game, argument's over. However, the result doesn't please your side of the debate so you made an entire pointless extension of the calculations reeking with bullshit logic.

The 0.01% is not even the correct value, so I'm going to redo the math myself and show a more accurate result that ironically would've supported your point. We'll go with the Mafia Returns rolelist, the only balanced one where Ambusher can be fairly tested.

The formula that goes "Percent of people the Ambusher would attack x Percent of people who could also visit the target of the Ambusher, and could conceivably reveal the Ambusher = Chance of Ambusher being outed by ability on given day" is missing a 1/11 in there where the chance of an Ambusher visiting the same target as 2 other visitors is represented, and is also worded like a Mathematics degree was stuck on a 5 yr old's forehead and then the 5 yr old was carried and smashed onto the keyboard until we got this. Consequently, your formula is wrong.

A more proper formula for this, worded correctly, would be [Chance of a non-mafia visiter role visiting a non-mafia target in the Mafia Returns rolelist] * [Chance of a 2nd non-mafia visiter role visiting the same non-mafia target as the first one in the Mafia Returns rolelist] * [Chance of Ambusher visiting the same target as the aforementioned 2 roles].

First is first, we gotta calculate the chance of a non-mafia visiter visiting a non-mafia target, before adding the Mafia Returns rolelist factor. Very quickly, we have 1 out of 14. This value includes the 4 mafia roles that the Ambusher cannot visit, and it would be inaccurate to keep it that way. Subsequently, we multiply the whole thing by 10, then divide it by 14. Instead of 1 out of 14, it becomes 0.714 (3 significant figures) out of 10. 0.714 out of 10 is the chance for a single non-mafia visiting role to visit another non-mafia role.

Now we gotta know how many visiting roles there are on average in a game of Mafia Returns to multiply with the chance of a non-mafia visiter visiting a non-mafia target. You assumed that all of them are visiting roles, which is just inaccurate and lazy. Quick math ought to do the trick.

Sheriff, Lookout, Pirate, and Plaguebearer are all guaranteed visitors, so we got 4 visiting roles so far. It should be pointed out that Plaguebearer alone can reveal an Ambusher since he has Basic Defense, but we will not add that possibility. Once again, we are acknowledging the fact that there is a human player behind the screen and not a dog holding a keyboard with its mouth. There is very little reason for a PB to reveal the Ambusher when no one else was killed or attacked by an Ambusher, because you would just attract suspicion to yourself as to how you lived. The only defense you can use is claiming a Doctor helped you, which is very unlikely due to only 1 TP slot that has a 25% chance of having a Doctor, and 3 RTs with a much lower chance than that. The Doctor would not reveal or support your claim even if he was present, you don't exactly have allies willing to communicate and fake for you, and hence the risks greatly outweigh the rewards of revealing an Ambusher. The lone PB possibility will not be considered.

We come to the TP slot, where 3 roles are capable of visiting on N1 and Trapper isn't capable of doing so. Since we are calculating the values for N1, Trapper is excluded and is not considered a visiting role on N1. As a result, the TP slot has an average value of 0.75 roles that can visit on N1, and one could even argue that Crusaders wouldn't go randomly visiting on N1 without direction because of the risk of killing a townie. We'll still place the value at 0.75 in favor of the Town. Adding it with the previous 4 we got, we have a total of 4.75 now.

Now we got 3 RTs, so we gotta calculate all that shit in. 5 TIs out of 6 can visit. The 3 out of 4 TP from before is used. None of the TKs can visit on N1, so that's 0 out of 2 and VH is excluded from the calculation completely. The TS is 2 out of 5 but I'm turning it up to 3 out of 5 because Transporter visits twice per night, which doubles the chances of him getting hit by an Ambusher. Your overall total for an RT slot is 11 out of 17 roles. I know there's 3 unique roles in there, but I can't factor that in. That's a whole probablility tree for something with very negligible effect, so we'll just leave it in the air. 11 out of 17 translates into 0.647 (3 s.f) average score for roles that can visit resulting from an RT.

Multiply 0.647 by 3, the number of RT slots, and we get 1.941. Add that to the previous 4.75 and you get a total of 6.69 (3 s.f). There is an average of 6.69 visiting roles per Mafia Returns game on N1. We would multiply them here with 0.714 out of 10 and get done with it, except we have a Pirate so fuck you for thinking we're done. We need to calculate the average non-mafia visiting roles roleblocked by Pirate because that affects the average visiting roles per Mafia Returns game on N1.

The Pirate can visit 1 out of 14 roles. We can't use the previously calculated 6.69 since the RT slot value is different because Transporter and Escort are RB immune to a Pirate. The TS score is dropped into 0, resulting in 8 out of 17 roles or 0.47 (3 s.f) visiting role per RT slot that can be roleblocked by Pirate. Multiplying 0.47 by 3 and adding it to 3.75 (which now excludes Pirate since Pirate can't visit itself), you get an average of 5.16 non-mafia visiting roles that Pirate can roleblock on N1 in a Mafia Returns game.

5.16 out of 14 equates to an average of 0.369 (3 s.f) non-mafia visiting roles roleblocked by Pirate on N1 in the Mafia Returns rolelist. Subtract this number from the previously calculated 6.69 and you get the result of 6.29 (3 s.f). The final result with consideration of Pirate's roleblocking is an average of 6.29 non-mafia visiting roles on N1 in a game of Mafia Returns.

We then multiply the average number of non-mafia visiting roles and the chance of a non-mafia visiting role to visit another non-mafia role in the Mafia Returns rolelist together, meaning we're multiplying 6.29 with 0.714 out of 10 to get the final result of 4.49 (3 s.f) out of 10. 4.49 out of 10 is the chance of a non-mafia visiter role visiting a non-mafia target in a game of Mafia Returns.

Calculating all this again for the 2nd non-mafia visiting role is easy, you just gotta make a few adjustments this time round. This is probability without replacement, and we're also not considering Pirate for this one since we already did for the first one. Considering it in this one too will create impossible scenarios and is inaccurate. This means we are taking the 6.69 we previously calculated and only subtracting 1 from it for a result of 5.59, since the first visiting role counts as a whole one on its own and thus cannot be considered for this one since it has done the visit.

5.59 is the average number of non-mafia visiting roles when the 1st one has chosen the target. The chances of visiting a non-mafia role are still identical, meaning the 0.714 out of 10 is used here too. Multiply 5.59 with 0.714 out of 10 and your final result is 3.99 (3 s.f) out of 10. 3.99 out of 10 is the chance of a 2nd non-mafia visiter role visiting a non-mafia target in a game of Mafia Returns, without Pirate's harassment.

3.99 out of 10 is multiplied with 4.49 out of 10 for a result of 17.9 (3 s.f) out of 100. That's the chance of 2 non-mafia visiting roles visiting the same non-mafia visiting role. Then we multiply all that by 1 out of 11 because that's the chance of Ambusher visiting a role. 17.9 out of 100 multiplied by 1 out of 11 results in 17.9 out of 1100. That's the chance of an Ambusher getting outed by its ability on N1. 17.9 out of 1100, when changed to percentage, becomes 1.63% (3 s.f).

1.63% is the average chance of an Ambusher being outed by its ability because of 2 visitors on N1 in a game of Mafia Returns, not 0.01%. It's actually higher, because of the possibility of 3, 4, 5 and more visitors but that's just a whole bunch of tedious calculations I don't have the time to go through and isn't significant to this debate. The chance would still be somewhere less than 10%, which still means Ambusher being outed is not common. You can go ahead and calculate it yourself, there's not much to gain.

The math I've done is not perfect, there are several factors and elements that I cannot factor in, there might be a few mistakes I'm not aware of, or the factors are too negligible of value to waste so much time making the probability tree of shitassfuck just for them. This is still multiples more accurate than the half-baked steaming garbage Add has put out.

Let's talk about other elements that you completely disregarded because of your assumption that every player is a degenerate chimpanzee munching on his mouse. The risk is already very low, as the math proves. But if an Ambusher ever feels that they don't want to take that risk and would rather plan thoroughly when there are leads to utilize, they can do this amazing thing that you can also do with Escort, Vigilante and Crusader. Don't visit. It's not that hard.

2 KPN for the mafia to utilize is a great weapon. Doctors and Bodyguards suppressed completely without any harm and without having to look for them, investigatives caught off-guard and taken down, and thinning down the Town faster. It is far more beneficial than you give it credit for, and can yield results more than 0 KPN (The minimum results for an Ambusher) in the right hands. Skill ceilings should not be disregarded for any role. If a good player can make the most use out of a role's potential, then it must be considered because all players are capable of reaching such a standard by simply playing the game and gaining experience.

If assuming every player is braindead and that they will randomly do things was a valid basis for deciding how good a role is, Vigilante would've been nerfed way back when random shooting was a lot more controversial and commonly seen. It hasn't been nerfed, because you can actually not visit and pay attention to chat, is what players found out.

Even if your math was right, the logic behind appraising the role is wrong and the aspect of the player and what they can do with a role was completely disregarded, which is unfair and very inaccurate when judging any role.

Next is the unnecessary extension of calculations that contributed nothing to the argument.

First off, your source is very outdated. It's a steam post from 4 years ago, when the Coven DLC didn't even exist and people barely knew what they were doing. Secondly, you've failed to acknowledge that only a portion of players own the Coven DLC but not everyone does. Thirdly, you failed to acknowledge that people with the Coven DLC could be playing other game modes and not all of them are playing Mafia Returns. We're just in the first sentence and there's 3 fatal mistakes already.

Let's ignore all of that and say you just wanted a sample size, thought the source would've been completely unnecessary for this.

I don't get how 3 games out of 333 (according to your flawed calculations) having the Ambusher instantly revealed on N1 when he visits is not great. Unfortunate events are something that happens to every role, that doesn't affect their value however. Retributionists get killed N1 in Ranked games before they can do anything, didn't stop the role from being the most brokenly overpowered Town role in all of ToS.

The town alternating between getting mafia every day is also a pretty big assumption that is inaccurately favoring the Town and going against Ambusher. I know the confirmed Psychic is a thing, but that doesn't mean the evil is the mafia member when there's 3 other non-mafia evils, nor does it mean the mafia are guaranteed to die when a Psychic says their name along another two.

For your N2 calculations, you've made the big assumptions of Town performing a D2 lynch when there could barely be evidence and the Executioner is a deterrent for the TIs, and also that the Ambusher gets a kill despite the large chance that he might have been on a target with 0 visits, both already downgrading the reliability of your calculations since those 2 assumptions significantly tilt the results in one direction. You also used the same flawed math that you've performed for the first night.

You then got a number of games where Ambusher is revealed instantly on D2 and D3, despite the hefty assumptions and the flawed calculations. 6,972 games only looks bad, that's all it can be. It isn't actually bad for Ambusher, because in its face stands a whooping 186,501 games where the Ambusher isn't revealed in either days according to your (flawed) results and your (flawed) calculations. That's 1 in every 27 (rounded up) games, which is actually not common but rather rare.

This whole section is based on the flawed assumption that frequency and rarity are one and the same, which they are clearly not. Common is a value for rarity, not a value for frequency yet you attempted calculating everything as though common is a frequency value. 1 in a 1000, 10 in a 10000, 100 in a 100000, 1000 in a million, they all simplify to the same thing. They are different frequencies but their rarity is still the same.

If an item has a very rare drop rate, and few people play it, then getting the item will be very unlikely and quite uncommon. If a lot of people play, the drop rate and the unlikelihood of getting the item does not change. I don't know what you're trying to prove with this section.

Whether you use my results or your results, Ambushers getting revealed by visitors is just not common. The statistics are pretty clear on that when you do the math, and the assumption that an Ambusher will always visit on N1 and N2 sharply decreases the reliability of the argument, because you're throwing skill out the window.

"Intuitiveness" + Hypnotist

​​​​​​​We have descended into counting words and making that proportionate to the intuitive value of a role. It's like the logically flawed successor of jailor metagaming where you ask someone to read their attributes as proof of being that role.

Quick counterexample to wreck the whole argument of "more words = less intuitive": Retributionist's card has 26 words and Lookout's card has 21 words. By your fucked up logic, you're telling me that an invesitgative that has to predict enemy actions and prioritize targets while constantly updating his will is easier to learn than a button pressing simulator because it has 5 more words in the game card? What sort of ridiculous thought process is that? Airplane mode is for your device, not your bloody brain.

But the strategy page word count difference between one role and another isn't evidence, because you said so. BUT OH GUYS LOOK, THE GAME CARD OF THIS ROLE HAS TEN MORE WORDS, THEREFORE IT'S A LESS INTUITIVE ROLE. You're out of your fucking mind.

Pitting Hypnotist against Framer proves absolutely nothing. Framer is one-dimensional with only one mechanic that doesn't even work most of the time, while Hypnotist has about a dozen choices to pick from and each one of those has their own tactics and situations where they work, which allows for the branching of many strategies. So what anyway? Hypnotist has more words in the strategy section than Framer, so? Any remotely decent role is more complex than Framer, that doesn't mean anything, lol.

Now, you say Ambusher is not "good in a good bit of situations" after horrid mathematics and severely illogical arguments yet Hypnotist gets the lips around the sausage and his hairy ass gets to be covered. In the context of a Mafia Returns rolelist, let me show you how overrated the Hypnotist is.

The Transporter message is easily disproved. Transporting yourself or someone else that wasn't a confirmed townie can be suspicious and is no longer considered damning proof because of Hypno's known presence. Whatever townies are confirmed or not mafia will request for a transport, and you can't do two at once unless that RM happens to be a Hypnotist too. If you actively deviated and avoided transporting trusted or confirmed townies, you get executed or lynched. Very simple. You'd also end up dragging whoever the other poor guy was who tried to help prove you or whoever was claiming Trans from your mafia, if there is one.

The roleblocking works on a very specific set of roles. If that role has an ability, you're instantly disproved, and it doesn't deter townies from continuing what they do in order to attempt getting results so it doesn't work as a deterrent either.

The only roles capable of attacking someone directly which can validate your BG claim is the Godfather, and maybe a Vig if you really wanna stretch it that far. After someone sees this, they expect a BG result. If they don't see a dead BG, and no dead attacker, it's quickly known that there is no BG. The BG message is not useful at all, as a result.

Faking Doctor is not significantly helpful enough. Your GF has to not kill for a night, and such a sacrifice for a single confirmation can mess up the mafia entirely. Considering there is a Sheriff, a Lookout, a Jailor and a Psychic, confirming yourself is pretty short-lived since they would eventually get to you, and there's not much you can answer to 3 TIs that can confirm you being evil. Making the GF claim Doctor might work against Sheriff, but Psychic, Jailor and Lookout are still imminent threats.

The Crusader message is just as bad as Doctor's for the same reason. They see the message, they expect to see a result, your only chance is GF. GF has to not kill so you confirm yourself or the GF confirms themselves, Sheriff/Jailor + LO + Psychic screw you up anyway. At this point, faking Doctor or Crusader has a strong chance of negative utility and there isn't much to control about it.

The witch, the poisoned, the poisoned but healed, and the poisoned but protected messages are all worthless as there are no Coven in the rolelist.

Triggering the trap means you need to identify the role of whoever you're targeting correctly, or you're a Hypnotist. Going for someone who has already revealed, claimed or is confirmed as a role and telling them their role does not prove you as a Trapper when a Hypnotist is known, and failing to identify a role when the Town requests for it puts you under heavy suspicion.

Being attacked and saved by a trap, they will expect results. This time, you can't go saying GF was immune and all that. There needs to be a death specifically by a trap which is extremely unlikely in the same night you decide to give someone that message, and even then the real Trapper would disprove you. It's worthless.

The trap attacking someone is already extremely unlikely to work because the only roles that you can pull that off on is Pirate and Vigilante. Even if those roles were there, where will you get the Trapper/Doctor claim from and why would they believe it? Players are not stupid, they'll quickly understand that Trapper messages are a Hypno's best resort, and they will be more wary about these coincidences between 2 specific protectives. This message doesn't have good chances of success.

More venues and more opportunities my ass, what's the point of versatility when most of it is a magnet for negative utility, has a low chance of success or has no uses at all?

We've done the math for Ambusher, we've seen the low chances of risk, and the fact that it can push the maximum KPN of mafia beyond 1 is very helpful and has better consistency than Hypnotist.