Mini 804 - The Resistance - Mod Abandoned

[ortolan]

/confirmatory

[ortolan]

I think Fishy is a Spy

It's a pathetic chance we manage to get no scum on the plans. I want to get a number of scum other than one on the plans, so I'd advocate giving plans to scummy people. I do not want a set of plans.

It's true that if you choose randomly you don't have a particularly good chance of getting 3 Rebels:

5/9 * 5/9 * 5/9 = 17.2%

However as with other mafia games, it is assumed you can increase the odds above chance by using logical deduction/induction etc.

Furthermore, do you know whether the "chance" odds increase substantially on subsequent days anyway? If we play it like a game of chance we're more than likely going to lose.

By choosing randomly you not only confine yourself to these odds, but you also deny us the opportunity to form opinions about your motivations in selecting the players you did.

I do not want a set of plans.

What is your rationale for excluding yourself? (Please no-one else answer for him).

If more than one scum receives plans, we have a chance of one of two things happening:
-More than one chooses to sabotoge, and we get closer to identifying the Spies. (Remember, we're going to need a solid idea of who's who when it comes to distributing the last 5 plans to all 5 townies.)
-They WIFOM into inaction, and we win a round.

I don't see how this is a good outcome. Even assuming there are say 1 or 2 spies out of 3, and they WIFOM themselves into not sabotaging, it's likely to cause chaos on subsequent days when we think we have a confirmed group and in fact do not. The potential for these "sleeper cells" is not a good outcome (although potentially unavoidable).

We may as well get a bit of information before starting the discussion.

The information comes at a price though. It's exactly the same as random lynching day one, complete with no reactions to the day one lynch to judge people on, as Naba said.

Also, the longer the day goes on, the more the scum can try to daytalk- which is an immensely powerful tool for them if we don't spot it.

Daytalk in the thread is the only way for scum to communicate, and there's nothing concrete for town to discuss until we see the results of the first porposal, so short day benefits town today.

Is the 48hr time limit mandatory? Or can it end prematurely?

So does this mean the less talk from town (just for Day 1), the better?

You're kidding right? If any rebels have the attitude "we must make the day as quick and random as possible because scum might day-talk and we might not be able to catch them" we may as well give up now. It's no different to any normal mafia game in this regard.

[ortolan]

Put simply- regarding the argument that we should "randomly lynch" or "randomly attribute plans" (for the two are analogous) for information, the response is that it is better to simply have discussion prior to this event- which both increases the chances of success and/or gives us more information for subsequent days. It's quite simple, really.

I suspect forbiddanlight also.

Also, the longer the day goes on, the more the scum can try to daytalk- which is an immensely powerful tool for them if we don't spot it.

Do you mean in thread or by PM? This is important.

If you were indeed asking a question which was supposed to catch someone in a slip, why would you signpost it so obviously with "This is important."?

There is also the fact that contrary to what she seems to be implying, there was no opportunity to "catch someone in a slip" here. Scum cannot daytalk, it says so in the rules.

Fishy could either have replied "no, they cannot day-talk, it says so in the rules", or "they can day-talk". Neither of these constitutes a "slip" though. The former is simply true, and the latter is hardly a slip on his part- it is just wrong, but doesn't prove his towniness because it's still subject to WIFOM like every apparent instance of "not reading the rules" is. I therefore question the basis of her FoS on PsychoSniper.

[ortolan]

Also, the longer the day goes on, the more the scum can try to daytalk- which is an immensely powerful tool for them if we don't spot it.

Attempting to incite paranoia about the scum day-talking or communicating via code is a tactic I've seen previously used by scum in a nightless game:

http://www.mafiascum.net/forum/viewtopi ... c&&start=0

See Corvuus' 368

It basically amounts to just blindly declaring "well scum might find a totally secret way of communicating which we won't be able to spot and somehow the benefits for them will severely outweigh the risk of being caught; so we'd better just do everything randomly and without discussion." This ignores the fact, of course, that random lynches/ plan distribution is always the worst possible manner of playing. I ask you how you propose the scum are going to be able to day-talk while keeping it totally invisible to us, having not had the chance to co-ordinate themselves at any point prior to the game. It's exactly the same thing I asked Corvuus (who was scum) in that game, and received no satisfactory reply (because he was scum).

I don't think it's so hard to understand why Psycho and ort are scummy, but apparently people's brains on in the off position tonight.

You are willing to declare us both scum because we "sabotaged" your plan (although you already claimed it was entirely ruined before I at least posted, anyhow) which amounted to an attempt to "confirm" a player which 1) was unlikely to garner the desired response and 2) would have been pretty much entirely worthless due to rule-misunderstanding-WIFOM anyhow and 3) you had already made enough attempts to sabotage it yourself by signposting it so blatantly

[ortolan]

Fishy's reaction makes me feel better about him

[ortolan]

Actually, 5/9*4/8*3/7=11.9%.

You are correct, good point.

And ok, suppose you thought you could raise this, double it even.

Any attempt to do this is arbitrary.

I don't believe there is a method which is substantially better than randomness for not getting exactly one spy today.

Ok, but how is using a non-random method

worse

than random for not getting exactly one spy? Plus, using a non-random method gives us more information for subsequent days.

By being demonstrably random, I remove the worry that I am picking an advantageous set for the scum.

Can you define in definite terms what comprises "an advantageous set for the scum"?

Actually, I just did the maths, and they look bad for Fish:

If you randomly select players, the odds of only one player being a spy are 4/9 * 5/8 * 4/7 = 15.87%

If you exclude yourself and are a rebel, the odds of only one player being a spy are 4/8 * 4/7 * 3/6 = 14.2857%

If you exclude yourself and are a spy, the odds of only player being a spy are 3/8 * 5/7 * 4/6 = 17.857%

So basically, if you actually are a rebel, by excluding yourself you make it 1.3 percentage points less likely that we will have a scenario where only one Spy has the plans. This is hardly a worthwhile difference in exchange for getting no information out of your choices.

However, if you are a Spy, you're actually making it 2 percentage points

more

likely that we will have a situation where we have only one Spy on the plans. Basically, your maths doesn't check out, at all. And having the "worst" outcome (it is not clear how much "worse" it is either than the alternatives) very, very moderately more likely to occur is worth getting absolutely no information out of day one how exactly? Your maths don't check out at all as I've looked at them. Plus your "worst case scenario" only has about a 15% chance of happening

anyway

.

FoS: Fishy

based on current calculations, I urge you all to vote nay to this distribution.

The problem with this game is that one tiny hint about intent to sabotage or not sabotage, perhaps in the form of thoughts about scum tactics, could do a huge amount of damage. We must all look out for these, and treat them as enormously scummy. Anything that could be construed as a hint is very dangerous, and townies should be careful not to make them.

This is partly true, but "anything that could be construed as a hint" is not very dangerous, it has to actually

be

a hint to be dangerous, for obvious reasons.

In the game you reference, there was no way in which the scum could make an enormous gain with minimal communication. Your appeal to history is totally unjustified.

Please name a conceivable way in this game scum could attempt to co-ordinate with minimal risk this game. No, it is not scummy to ask for this (as was again suggested in that previous game). I do not believe there is one.

[ortolan]

I will also make the point that assuming we are going to do things non-randomly from day 2 onwards, and assuming that it is always undesirable to have a situation where only one person sabotages; who's to say by choosing non-randomly day one you are not decreasing the likelihood of this outcome on subsequent days due to the information it provides (especially when the odds give your position such minimal benefit to begin with)

[ortolan]

You are correct, I should clearly remember to walk before I can run with my somewhat "limited" maths education

0 = 5/9 * 4/8 * 3/7 = 11.91%

1 = 4/9 * 5/8 * 4/7 = 15.87%
Or 5/9 * 4/8 * 4/7 = 15.87%
Or 5/9 * 4/8 * 4/7 = 15.87%

2 = 4/9 * 3/8 * 5/7 = 11.91%
Or 5/9 * 4/8 * 3/7 = 11.91%
Or 4/9 * 5/8 * 3/7 = 11.91%

3 = 4/9 * 3/8 * 2/7 = 4.77%

This puts the odds of there only being one spy in the group if drawn entirely randomly at 47.61

If you exclude yourself and are a Rebel then chances of one scum are:

4/8 * 4/7 * 3/6 = 14.29%

4/8 * 4/7 * 3/6 = 14.29%

4/8 * 3/7 * 4/6 = 14.29%

Overall 42.86%

If you exclude yourself and are a spy then chances of one scum are:

3/8 * 4/7 * 3/6 = 10.71%

4/8 * 3/7 * 3/6 = 10.71%

4/8 * 3/7 * 3/6 = 10.71%

Overall 32.14%

I can see your reasoning now.

Any reference to potential scum night tactics would be the most obvious example.

There are no "night tactics" in this game though.

[ortolan]

I suspect Crazy, think Naba is town and have no read on ABR. I would not like to send them in.

Also people are forgetting that we get more info out of letting a few people have a go in the driver's seat and nominating who they want before we go with one of their proposals. That way if ever we have evidence or suspicion someone or other is scum we can look at the people they proposed and analyse.

Even Fishy himself has expressed doubt about his random method now.

In order to defeat it we need to ensure at least all 5 rebels vote against it.

At some point in the near future I will give my reads on people and exactly what gave rise to those reads.

[ortolan]

Why?

FL, who do you want to send in?

[ortolan]

Also Crazy, ABR, Psycho, why did you vote yes?

[ortolan]

ortolan: Clear no vote from 29 onwards. (Incidentally, maths is still slightly wrong, in a way which reflects well on me- if I was a spy, random method would be advantageous to me.)

Actually afaik my maths are correct as of my latest revision. This isn't a criticism of you though. I would probably support a me - Fish - Naba as the first three to get three town on the plans at first at present.

[ortolan]

I endorse any plan that ends in me being given a plan. ortolan's massive wall of math could either be intended to drown players who don't have time for that shit (me) or give the semblance of effort to make less active players look bad.

Nice false dichotomy where either conclusion makes me look bad. Or, it could have been to attempt to resolve the argument I was having with Fishy at the time.

Incident'ly, I cannot remember a single Veerus post all day of any substance. THerefore, I would suggest a me/Veerus/ortolan Proposal, although obviously I would be willing to compromise on one of the three in order to get at least two of them up in the spotlight.

What is the rationale behind your selection of players?

THat would be acceptable, certainly. As I say, I would like to get at least two of the three players I mentioned above into the spotlight.

Why would you like to "get these players into the spotlight"?

It's hard for scum to post content in the ways of scum-hunting

No it isn't

but talking theory is not hard... and thus doesn't need to be rewarded by having plans given.

Yah except we're not even necessarily using "talking theory" as a device to try and pick out the scummy players, we started using it in the first place to discuss and conclude what the optimal play actually is; not to try and get scum-reads from it.

1. He believes that we should try to get plans of all town.

Having all town on the plans (or at least having no-one sabotage, for that's all we'd know at that stage), is amongst my optimal results. It is not

the

optimal result obviously, that would probably be having three people on the plans all of whom sabotage. However if it is possible it is certainly a desirable outcome, and I was considering those players I nominated as having a good chance of all being town (obviously in my case I am certain).

Scum at least want

one

of their people on the plans, most likely, so they'd try to look pro-town.

Their win con is to sabotage three times. They are going to try to look pro-town regardless of any other considerations, in line with every other game.

2. Yeah, well that's me... and I'm not really going with the grain here.

Do you mean you are "pro-town"? If so, why? If not what did you mean?

3. It's because ort picked you two on the plan, and Nab. Nab didn't matter to me because he was on the plan to begin with.
4. But you changed your mind.

I'm not saying ort is obvscum, but I don't agree with his philosophy. I'd much prefer ABR's proposal.

I don't understand any of this. I am not judging players on the basis of "posting more content", I am saying scum have little motivation for suggesting non-obvious pro-town things to do or consider, which both Naba and Fish have done.

Quality of content is the most important, and if they have posted more quality content than anyone else then yes, I am going to conclude they are more likely to be pro-town.

And I don't even know why you'd need to suggest "I'm not saying ort is obvscum", you have given no reasons to suggest I am scummy at all.

Why have a fast D1 when you can have a leisurely day with a copious supply of feedback at absolutely no charge? The way the rules are structured, we could adhere directly to deadlines and still have an infintely long day. Of course, we're going to want to agree on a proposal sometime, but why not make use of the fact that we now have solid information (the voting record for last cycle's proposal) on which to build a better plan. I rejected the random assignment because there was a better path with no extra charge.

QFT. This is what the people saying "we should just go randomly" are persistently failing to pay attention to (even though it was already brought up previously, at least by me I recall).

It seems to me that ABR is trying to put together an assignment with the three players who have contributed least so far. I'm not sure exactly what the point of this is (maybe he would like to explain), but I think it goes against both the objective of information gathering (the more a player has said, the more we have to draw conclusions and directions) and the objective of ideal scum distribution (why put our money on unknowns?).

Again, QFT.

I was leaning putting ort in. I wish people would stop self nomming. It makes me nervous and more likely to reject you.

Emotion has no place in this game. Attempting some horrid reverse psychological WIFOM when you are in fact town does not seem optimal to me.

Nabakov probably her russian liaison. Definitely no plans for both of them.

?

A disturbance in the force, you have felt. If thrown in a cage with veerus I was, safe would I feel.

ABR's "gut reads" based on one line posts are of "questionable usefulness".

ABR is not doing anything to make me more comfortable about giving him plans. At all.

Well he has, you know he has kind of been calling you obv-scum the whole game, so this isn't much of a surprise. (For reference I'm inclined to view the ABR-FL thing as distancing because I don't think either are pro-town at all).

ABR is fitting his meta - oh so very angry and jackassish.
Unfortunately, this is his meta for EVERY role. So I have no idea.

What was the purpose of making this post?

ort- appears to be putting too much effort into checking Fishy's math and such. It's WIFOM, and he's probably who I'm least solid on, but I still feel he's town.

Why do you think I am putting too much effort into doing something "useless" but still feel I am town? Also what is WIFOM?

veerus- I have gotten no scum vibe from him, and overall felt he's town.

Re-read veerus' posts and tell me if you think this assessment is genuine (I don't think it is).

in me Nab and veerus I would expect veerus to sabotage, especially with FL's support for this combination.

That's...actually acceptable, with my bet on Psycho being the scum if we get a sabotage act.

So you think someone is likely to sabotage but still support the idea? Interesting.

Honestly, the BEST way to proceed is to have NO DISCUSSION, and group THREE SPIES TOGETHER.

Best case scenario they all sabotage and we have three CONFIRMED SPIES.

HOWEVER, if they can COORDINATE who is sabotaging and who isn't, then we are SCREWED.

DISCUSSION HURTS THE TOWN. FL AND NABAKOV --> EXTREMELY SUSPICIOUS.

As I said earlier, that is indeed the best outcome but realistically it is almost entirely impossible. You need to get 3 scum on the plans to begin with, WHICH IS EVEN LESS LIKELY THAN GETTING THREE TOWN, then they need to all be stupid and sabotage. Optimal play if they are all on the plans is for all not to sabotage then they end up with no saboteurs they all get confirmed town which leads to subsequent chaos.

DISCUSSION HURTS THE TOWN. FL AND NABAKOV --> EXTREMELY SUSPICIOUS.

ABR seems to have actively ignored me already arguing this is definitely not the case. Scum want to incite paranoia about coded daytalking. For your information, even after we decide on the first set of three, on all subsequent days we're going to have extensive discussion over the results anyway, which still leaves us open to the perils of "scum daytalking" on all days but this anyway.

Remember, we will know how many acts of sabotage have occurred after each plan is carried out. Scumhunting is MEANINGLESS at this stage in the game.

No.

119 is just "silly".

123 I agree with.

"It is much more of a logic puzzle than Mafia, though traditional scumhunting skills should still help."

Logic > Scumhunting.

Yes but even treating the game as a logic puzzle the approach you propose is not optimal.

And I'm really liking ABR now... I didn't consider that multiple sabotages was the best option for town, but he's absolutely right.

Incorrect. Also, if we get two scum on the plans how do ya know both will sabotage anyway?????? Same with three. Seriously this logic is horrible.

ABR: 137, it says in your pm you are not allowed to communicate outside of the game thread at any time with the other spies, just a heads up.

[ortolan]

Psycho, you're forgetting that we would have an extra set of plans to distribute every time in that scenario.

Look, this all comes down to the fact that there is NO optimal play since there's always an optimal counterplay. According to boardgamegeek's website, odds of rebels winning are 56% and spies 44%. Results threads on there seem to confirm that. So it's pretty close to 50-50 no matter what happens. Best way to go is to randomly distribute the plans and evaluate results based on those people's posts. In fact, based on this logic, I would now support the original proposal since that was our first true random.

^^ That does not mean the result of the game is necessarily random, it just means that the game is balanced. I do not see how this supports that we should go for random distribution, and as soon as possible at that.

And why are you revealing what you consider to be "optimal scum play" for them to see? If you think you know what's best for scum, wouldn't it make sense for you to not bring it up just in case the scum hadn't thought of it themselves?

I don't support trying to get three spies on the plan though. And, even if I did support it, it would be virtually impossible to get three spies on the plan at least randomly. And it's not actually clearly optimal play anyway, as you point out. And again, I think this is unnecessary paranoia about scum day-talking.

[ortolan]

What reasons? You didn't actually give any

[ortolan]

I have always wanted guided

[ortolan]

ABR, is there a reason for your (apparently unprompted) aggression? If nothing else I'm curious

[ortolan]

I think PsychoSniper's opinions are actually quite well argued. I don't think he has scummy motivations for arguing this way.

That said I think it's going to be obvious what we're opting for at this point based on what we've previously stated.

My current suspected distribution of rebels is me, Fish, PsychoSniper, Naba and ABR

[ortolan]

actually that doesn't seem like such a bad idea to me if we get everyone to give a list of who exactly they think is most likely to be town/scum at this point in time. Any dodginess in the lists may well be apparent.

I will also make the point about KoC that his play is indistinguishable from that in Election Mafia where he was scum for me. That's the only game I have for meta reference. Plus I haven't liked some of his posts, will look over tomorrow.

[ortolan]

Fish I meant in the same format as mine. Who would you say, if you had to choose, your five picks for most town are (you know the distribution of the game includes five rebels so who would you wager those are if forced to choose the most likely)? Feel free to also indulge me Crazy and FL and anyone who posts subsequently.

[ortolan]

Thankyou Fishy

now for the other 7 of you

[ortolan]

Just because we don't have a consensus on whether we want a ticket comprising three scummy or three townie players doesn't mean that a jumble of some of each is acceptable.

You think FL is town?

Oh, I like that. You can't read me, therefore I'm a spy. Charming.

FYI, I am only a spy in TF2. Where I spend a pleasant evening strafing behind 4 Red Snipers, knifing as I go.

What a scummy post.

FL is the main player I doubt in that proposal.
Nabakov and Fishy, I have no idea about... so I'll probably be voting Yes on this.

There is no logic here. He is pretty much saying he thinks there's 1 scum and 2 town so he's voting for the proposal.

I have found FL scummy and if we sent the three that she's proposed in I would expect her to defect.

I am getting quite substantial scum reads from KoC, her and Crazy

[ortolan]

Putting yourself up against the two townie players who probably have the most support in the town (with yourself being the possible exception) is foolish if you are going to sabotage, and foolish if you aren't. If I were FL-scum, I would certainly not have picked this proposal.

Well, actually, considering she's pretty much been written off as scum by quite a few people already I see little reason for her not to get in a sabotage and practically confirm herself as scum. Then we still need to work out who the other three scum are anyway, plus she wouldn't really be demonstrating something we didn't already know.

[ortolan]

I am certainly not accepting this proposal after you just told FL to sabotage and any other scum that might be on the proposal to not.

Fine by me

[ortolan]

I really don't get where the FL obvscum contingent is coming from. As far as I can see, it's all based on one irrelevant question she asked.

It was all the stuff about her claiming people were scummy early in the game for "ruining" her questions about scum talking and such. That and her seeming lack of confidence in the fact she herself is town.

Wow O.K., ortolan and FL are scum. Vote NO.

Yay. FORBIDDANLIGHT YOU SHOULD DEFINITELY DEFECT.

I'll probably vote yes because it's a reasonable proposal and because this day needs to end already.

Scummy. The day ending with one sabotage which pretty much proves fl is scum, which we already know, gets the rebels no closer to their win-con.

[ortolan]

When I say I expect FL to defect, I just mean I expect her to be the only spy in three who gets chosen therefore it's a logical choice for her to defect

Crazy, how likely do you think it is that I am scum breadcrumbing?

[ortolan]

This proposal better get nos

[ortolan]

I was a Rebel, if that wasn't obvious

I stand by that Naba and Fish were almost certainly also Rebels

Having a quick look back over the game I'd guess the other Rebels were veerus and ABR

[ortolan]

oh right, damnit

I was tossing up between Psycho and veerus for Rebel, I wasn't sure who was more likely to be

I reckon Spies would win this the majority of the time in forum play

[ortolan]

Having one or two more town than were needed in the final round probably would have evened things out.

QFT for when you run this game again Rishi, imo