Mini 804 - The Resistance - Mod Abandoned

[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.

[Knight of Cydonia]

At this point the best thing we can do is let FL make a triple die roll, random it, and see how things fall tomorrow. The only reason I didn't do that vesterday was because I'm a massive paranoid idiot.

[forbiddanlight]

Alright, how about this. Let's start simple.

I'd like everyone to vote between random proposal and guided proposal

We'll work from there. Only cast one vote please. And explain your choice.

I support a guided proposal

I think we'll get more information out of a guided proposal as well as lower the risk of losing bases early as opposed to randoming.

[Knight of Cydonia]

Random Proposal

for the reasons outlined above.

[ortolan]

What reasons? You didn't actually give any

[Crazy]

I think I'm finally agreeing that a guided proposal is best... not necessarily for getting the best chances of whatever... but for getting more information from the person giving the plans.

Also, veerus, the 56%/44% odds aren't the win percentages; it's literally just the percentage of players that are town/scum.

[veerus]

hm, i didn't realize that... I'm still for a

random proposal

however. If we hand-pick people we will likely WIFOM ourselves to death and be forced to doubt all of our reads based on the results, especially if they are not what we expected. If we randomize the plans, we can evaluate the results in hindsight without a preconceived notion of who will do what.

[Fishythefish]

guided proposal

My aim today is to minimise the risk of having a ticket with precisely one scum on it. This outcome lets that scum choose exactly the outcome of the night. I believe that by picking 3 scummy players for the ticket, we can reduce the odds below random.
Randomising the plans also removes the important element of who accepts and rejects them- useful information for later.

To be totally clear on my take on the scummy ticket idea:
- 1 scum on the plans is bad.
- Any other number is much, much better.
- If we pick a scum ticket, we can get a read wrong and still avoid 1 scum.
- If we pick a townie ticket, we can't.
The best of all possible worlds is multiple sabotages. But even if the scum choose not to sabotage for fear of this, that's good.

[Albert B. Rampage]

Let's go with random.

[forbiddanlight]

Guided: forbiddanlight, Crazy, Fishy
Random: KnightsofCydonia, veerus, ABR

And ort, you didn't choose one?

And KoC, just for convenience could you link the post that you explain why?

[Albert B. Rampage]

I think I'm finally agreeing that a guided proposal is best... not necessarily for getting the best chances of whatever... but for getting more information from the person giving the plans.

Also, veerus, the 56%/44% odds aren't the win percentages; it's literally just the percentage of players that are town/scum.

Do you realize that in a guided proposal you will not get any plans because of the suspicions placed on you?

[PsychoSniper]

I vote random

[ortolan]

I have always wanted guided

[forbiddanlight]

I have always wanted guided

I figured so, but I wanted it directly out there.


Guided: forbiddanlight, Crazy, Fishy, ort
Random: KnightsofCydonia, veerus, ABR , Psycho

And it's up to...NabNab.

[Fishythefish]

hm, i didn't realize that... I'm still for a

random proposal

however. If we hand-pick people we will likely WIFOM ourselves to death and be forced to doubt all of our reads based on the results, especially if they are not what we expected. If we randomize the plans, we can evaluate the results in hindsight without a preconceived notion of who will do what.

This is a very odd post. You appear to be saying that we should evaluate night 1 independent of day 1. That's just throwing away information. I fail to see how there could be more WIFOM if we pick people.

[Crazy]

I think I'm finally agreeing that a guided proposal is best... not necessarily for getting the best chances of whatever... but for getting more information from the person giving the plans.

Also, veerus, the 56%/44% odds aren't the win percentages; it's literally just the percentage of players that are town/scum.

Do you realize that in a guided proposal you will not get any plans because of the suspicions placed on you?

If I don't like FL's proposal, I can still reject it. Just because I support a

guided

proposal doesn't mean I support

forbiddan's

guided proposal.

[forbiddanlight]

If I don't like FL's proposal, I can still reject it. Just because I support a guided proposal doesn't mean I support forbiddan's guided proposal.

You do realize by guided I mean ask the town who I should put on there. We'll see what Nabby says.

[Albert B. Rampage]

I think I'm finally agreeing that a guided proposal is best... not necessarily for getting the best chances of whatever... but for getting more information from the person giving the plans.

Also, veerus, the 56%/44% odds aren't the win percentages; it's literally just the percentage of players that are town/scum.

Do you realize that in a guided proposal you will not get any plans because of the suspicions placed on you?

If I don't like FL's proposal, I can still reject it. Just because I support a

guided

proposal doesn't mean I support

forbiddan's

guided proposal.

You might as well switch to random because no one is going to follow your guided proposal.

[NabakovNabakov]

I definitely support a guided proposal.

And I think Fishy's approach is a good way to go. We had established enough theory by voting time for the first proposal to determine that there is no plan that is really neutral in the eyes of the scum; they have no incentive (other than WIFOM) to vote for a proposal that is not 1-scum, so I think we definitely stand to benefit by drawing today's proposal entirely from either the pros or the cons of yesterday's proposal.

I'm going to abstract away WIFOM just for now so I can work things out; feel free to add it in later to the detriment of all of my conclusions.

-Scenario 1: Scum favored proposal. This would mean that only one of myself, ABR, and Crazy were scum. However, this fails to hold up because

both

ABR and Crazy supported the prosposal. If they are both scummy for supporting, than the proposal has two scum. This scenario contradicts itself.

-Scenario 2: Scum did not favor proposal.
--2A: Proposal had two scum: Unlikely. Both ABR and Crazy would have to be scum, and they would have had to have voted for a proposal that did not favor them. This is the other end of Scenario 1's contradiction.
--2B: Proposal had no scum: The most likely scenario in my eyes.

Therefore, a town centric proposal should either be the three who voted for yesterday's proposal OR yesterday's proposal itself, and a scum centric-poposal calls for selection from yesterday's rejectors (KOC, veerus, FL?)

The one issue I have with this reasoning is that it goes quite against my gut. I have misgivings about Rampage, Crazy, and Psycho, so I am apprehensive about assuming they are town based on a contradiction in a model.

[Crazy]

You might as well switch to random because no one is going to follow your guided proposal.

That can change. Just because of some silly "breadcrumb" makes me look scummy now doesn't mean I'll be scum forever. So no, I'm not going to change my entire philosophy just because it doesn't support me personally at the moment.

Nab's last post takes in absolutely no account for distancing/bussing/etc. To assume that the scum would always vote together is

very bad.

[NabakovNabakov]

Nab's last post takes in absolutely no account for distancing/bussing/etc. To assume that the scum would always vote together is very bad.

Well, it would seem to me that distancing and bussing would work in a slightly different way in a game where it is incredibliy difficult to ever discover somebody's allignment for certain, but you're right, the scum might now always vote together. I acknowledged that when I said that I was abstracting away WIFOM (ie. any condition where the scum act against their best interests as determined by game mechanics in order to look better). I'm taking out that confounding factor in search of a decent starting point, otherwise we'll just guide this proposal to randomness anyway.

[Albert B. Rampage]

I'm not crazy about giving Crazy a plan.

[PsychoSniper]

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

Okay, I missed out this sentence yesterday and only saw this in my reread, but I don't understand what this is supposed to mean. I'm assuming this is in response to what I said about the 3-scum-not-sabotaging scenario. Are you saying that if we destroy a scum base we get an

extra

(two) plans the following day? I'm pretty sure that isn't the case (I read the rules again after this to make sure I didn't leave out anything there again). Can you explain what you mean by that and exactly how it contradicts what I said prior to that?

[Fishythefish]

Post 168 is largely true.

Here is how I reconcile events with my gut:
1. The proposal had one scum on it. It had exactly one scum on it- either Crazy or ABR. Pyscho is scum. veerus, who held back on committing himself until it was obvious the proposal was hopeless, is scum. The fourth? Perhaps FL saw which way the wind was blowing and decided randomness looked bad after all. Perhaps the fourth scum just voted no.

2. If the proposal had two scum on it, they are likely to be ABR and NabNab. ABR supported randomness before the proposal was even made, which forced his hand a little. In this case, Crazy and Pyscho both look pretty good, which goes against the gut. One of them may have misunderstood the situation (IIRC, Crazy repeatedly said he didn't know what was good for scum), or just have been doing the equivalent of bussing. I don't much like this answer though- it makes Crazy and Pyscho probably innocent AND NabNab probably guilty.

3. If the proposal had no scum on it. Looks ok. My read on Crazy isn't the strongest, so I could go with this. Makes Pyscho look good- but not that good. I can see "no they've picked a townie ticket" as a possible scum reaction.

I think the model is useful- it shows that multiple scum on the ticket yesterday fit badly. The question that should be asked is "how much does the model need to break to fit in with my scenario?". I don't think it needs to stretch very far to accommodate the scummy players in the "proposal was good for scum" scenario.

Proposals I like atm:
1. Accepting scum: Pyscho, veerus, Crazy
2. Rejecting scum: drawn from KoC/FL/veerus/ortolan/NabNab (in rough order of preference)
3. Rejecting town: Me, NabNab, ortolan
4. Yesterdays proposal

[veerus]

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

Okay, I missed out this sentence yesterday and only saw this in my reread, but I don't understand what this is supposed to mean. I'm assuming this is in response to what I said about the 3-scum-not-sabotaging scenario. Are you saying that if we destroy a scum base we get an

extra

(two) plans the following day? I'm pretty sure that isn't the case (I read the rules again after this to make sure I didn't leave out anything there again). Can you explain what you mean by that and exactly how it contradicts what I said prior to that?

You said that if there are no sabotages, we should give the plans to the same 3 people which will either lead to victory or force scum to show themselves. However that plan falls apart since with each destroyed scum base, one more plan is distributed to the group. So even if the original 3 are all town and the 4th is scum and he sabotages, the argument could be made that any of the 4 could have sabotaged. Which basically goes to my point that there's no optimal play since each variant has a viable response.