Intro: I am arguing in favor of a random lynch. The random lynch, in my view, has one essential strength, and one essential weakness. How you weigh these off against each other should largely determine whether or not the random lynch is suitable for you. The strength of the random lynch is that it has a greater chance of hitting scum than a non-random lynch.. The weakness is that it yields less information than a non-random lynch. Read on for details.

Chance of hitting scum: The chance of the random lynch hitting scum in this game is 25%. In other games this number may be higher or lower depending on the ratio of scum to town. When the number of scum rises in relation to the number of town, it becomes more likely to lynch scum, and as the number drops it becomes less likely. This is roughly true whether the lynch is random or not. For non-random lynches, there are two main factors that influence the odds of hitting scum.

The first is the number of scum. Generally speaking, the town's base chance of lynching scum are a little lower than would be proportional. With 1 scum and 9 town players, the basic odds of hitting scum are a little under 1/10. With 2 scum and 8 town they're a little under 2/10. This offset is to account for the fact that scum players are inclined to not vote for other scum players but town players can't do the same, at least without information.

The second factor which influences the odds of hitting scum is information. The more information the town has, be it soft information like voting records or posting patterns that can be analyzed, or hard information like cop results, the more accurate town votes become. As the game progresses and the town gains access to more information, this factor can overwhelm the scum advantage of knowing their team mates. During day 1 however, the information available to town is usually not significant enough to do so.

Because information is low on day 1, but scum still know their allies, the odds of a day 1 scum lynch will fall below what would be proportional for the game (in a game with 1 scum and 9 town, it would be below 1/10, 2 scum and 8 town, below 2/10 etc.) For this game, the odds of a successful day 1 lynch fall somewhere below 25%. Because the odds of a successful random lynch are always exactly proportional, a random lynch will have a higher chance to be successful than a non-random lynch.

Loss of Information: The odds of a successful day 2 lynch have traditionally been regarded as stronger than the odds of a successful day 1 lynch because of the information that can be gathered during the events of day 1. Information can include things like player interactions and voting stances, and also things like cop results and role information. When a players role and alignment are revealed, that allows people to view their actions under a new light and draw conclusions from their behaviours and voting patterns.

The most useful behavioural analysis arguably occurs when a players life is threatened. This is where players begin to have strong incentives for acting in certain ways. If a scum player is in danger of dying, then their companions have strong incentive to save them. There are other similar scenarios, but it could be said that the way a player, whether town or scum, treats a wagon reveals how they really feel about the players involved in the wagon, and about the game state to some extent.

When carrying out a random lynch, the threat of death on individual players is lessened. Although competing wagons may still yield some incentive for players to act in certain ways, if the random lynch is the popular vote, you lose out on some opportunity to learn about how players would react to an actual player being the popular vote.

Information can still be gleaned from a variety of secondary sources, including voting records outside of life-threatening wagons, player arguments and shifts in opinions, and so forth. Unlike with a no-lynch, a player still dies, allowing players to look back on that players interactions with other players during day 1 and draw conclusions from their behaviours.

In Conclusion: Exactly to what extent the random lynch's strength and weakness affect the game is largely up to subjective interpretation. The odds of a day 1 lynch being successful could be determined experimentally but will likely shift from game to game and from meta to meta, and so a degree of subjectivity is required to determine what one thinks the odds of a successful day 1 lynch is. The amount of information lost by pursuing a random lynch wagon is very much subjective, as information is largely impossible to quantify. These interpretations, in addition to how much you personally value information or increasing the odds of lynching scum, are what will determine whether the random lynch seems right to you.