I prefer attacking the runner-up. I’m ok with picking off low-profile posters. I’d prefer not to pick someone to follow (simply because 1) they may not be town 2) they may be town and wrong.
Another option: have everyone make cases for who should be attacked, explaining their rationale. I think this would be my favorite option, because the second choice of the majority might not be the runner-up in the lynch.
I would like to see this happen, but a) probably won’t participate myself due to class restrictions and b) am going to be very short on useful-participation time for a few days due to stuff I dont’ feel like whining about in public. But I do support it.
In a sense though, they’d kind of have to be. People will already be making cases for who they want lynched so the same cases should apply to who they want killed.
Not necessarily. If 10 people vote MHaye, and 5 people vote you, that doesn’t mean you’re the second choice of those 10 MHaye voters.
Ah sorry, I see your point. The cases will be the same at least - I don’t know if we want to start up a tertiary vote after lynch vote and item vote for attack vote though.
Oh, my bad. Let’s continue to fling our attacks about wildly, then. 
That’s certainly the only other option:p
With no way of policing who attacks who, then voting for who gets attacked isn’t going to achieve anything. Stating who you want to attack is one thing, but if people vote to attack someone you think is town, you’re hardly going to want to attack them.
One last stab at explaining why this is my favorite option:
Let’s say, for the sake of argument, it’s Day 2 again, and we’re lynching me. I had 10ish votes, and the next closest had 3. If we follow the runner-up method, a single scum has a disproportionate amount of influence on whom we attack to death. A single vote or unvote will ensure a townie death most of the time, and because there tends to be a lot of vote mobility as the day draws to an end, it’s largely invisible to the town.
However, by holding a second lynch count, scum are harder-pressed to influence the vote. Instead of being decided by a single scum vote, it’ll likely be decided by 3 or 4 townies. Much better in my opinion.
Do it.
I would like to hear from, like, a dozen people.
The only problem I see is that the attack vote would also have to be near EOD and usually I am in the land of nod by this time as are a few other people. You have all seen what happens at EOD.
I am not totally against it but us Non Americans could look like total twits when the attack vote has changed to someone else. Just my opinion. I wouldn’t like to be voted off just because I need my 8 hours 
Silver Jan makes a good point.
I just had to say this Astral, couple means TWO and you had three ways
I actually had 4! 
I have mentioned before that my Maths isn’t that good 
We’d need to have a decent number voting, say 50%, so there’s some sort of consensus, well before Dusk, say 24 hours, to allow time for defense, claims, and for the attackers to get their orders in.
We can’t know how many townies are willing to attack but even if there are enough to do serious damage, the target is public knowledge: if townie, we just help scum reach their mislynch quota, if scum, they may be able to protect, redirect, watch or interfere who knows how. But I agree that a coordinated effort is likely to be more effective and give more information than random attacks.
That is a really good idea BUT Scum will then not protect their mate (If I can see it so can scum) and the only reveals we get are from a protection, so we will be better off if scum decides to protect but I can’t see that happening, can you?
I would rather do an attack that’s decided amongst your buddies (you all have to agree and you trust them). Coming out in the open and saying we will attack XYZ is just not going to work IMHO.
I’m not sure I like the idea of ganging up on one player. Part of my objection is pure self-interest, since I think I’m quite likely to be a victim, and I don’t like being hurt.
How3ever, I’m going to try and set that aside and look at the idea objectively.
Part of my problem is verification - none of us have executive oversight of the PMs players send ATPG, so we’re never actually going to be sure that players will do what they’re supposed to. While this is partly mitigated by the public nature of the attack (if not the originator) there may be a conscientious objector (someone who genuinely believes in the innocence of the victim) who refuses. There is also the issue that, by going down this path, you give those players who are anti-Town a mandate to attack someone with no scrutiny. That applies to any system, quite frankly, because it is a weakness of the method of submitting actions.
My other bother is that I do not like the method of selection. Selection by vote has several problems, some of which have already been touched on.[ol][li]Killing the second-placed voter has the problem that the person in second place may not be the preferred option of the players on the lynch wagon, and why should they be disenfranchised?[/li][li]There is also the fact that a last-minute vote shift could leave players not US based attacking the wrong player - remember that Day actions have to be submitted before the lynch deadline, but the final vote count isn’t known until some time after. Suppose Lord Phere’s votes push the person we all sent attack orders for into the lynch? Then we’ve wasted the Day’s attacks.[/li][li]Finally, why should Normal Phase and Visorslash have disproportionate influence on the choice of who to attack? Their votes count double in the lynch tally - thus they might, by their votes, change who gets lynched.[/ol]Consequently, I can’t support a “bash the second-placed candidate” scheme.[/li]
The proposal to run another vote for the attack solves many of these problems. Since it’s a player instituted action, Phere has no role and we can discount his votes. Similarly, Normal Phase and Visorslash can be restricted to one vote.
The last minute vote shift problem could be addressed by trying to impose an earlier deadline for this vote, to give everyone sufficient time to collate the votes and submit an order. Problems with that last include[ol]
[li]firstly, trying to get people to vote early is an exercise in futility (it’s been tried again and again, and falls flat on its face owing to the reluctance to vote. (I’m an extreme example, but late voting is very common, until we reach late game.))[/li][li]Secondly, that gap between the attack ballot deadline and the lynch ballot deadline is ripe for manipulation by the secret voter - the one who governs Phere’s triple vote. It gives the player or players time to place the triple vote to most advantage.[/li][li]Thirdly, there’s no flexibility in the event of a convincing claim.[/li]Fourthly, logistics. You have to make sure the vote is easily distinguishable, so that even the most clueless catch it, and either everyone has to do their own vote tallies (meaning it’s possible someone may honestly miscount and attack the wrong player) or we rely on one or two people, which is also a bad idea if they happen to be in cahoots to pull a voting trick.[/ol]All in all, I just don’t like it, and reserve the right not to participate, or make my own independent decision.
MHaye, you posted what I was trying to say a lot more succinctly than I could have done.