Shanba wrote:Look, I accept we're going to have differences of opinion on things like these, and I can accept that she gets nommed for collecting the data - ultimately the judges will make the decision as to what they think is most important. But I don't see what hoopla is adding as that notable.
Don't get me wrong, I really appreciate it and I hope she continues to get data for analysis, but what does the data actually tell us? I mean, it's just a list of win percentages relating to different setup types. It's a great starting point for discussion about balance, and I hope people take up that opportunity (I'm going to make some comments in a bit). But this kind of thing has been done before: there are countless people who have run stats on newbie games win percentages, and I know that I myself
have done something like this before, and a google search reveals that you have too, Adel. What makes this particularly noteworthy?
(As to why I think the things I nominated to are particularly noteworthy, as someone is inevitably going to throw this in my face: I think that both JD and shea are raising topics that aren't often discussed yet have relevance to the way that people play mafia. JD's is more self serving and shea's is more notable imo, but I'm interested in seeing what JD adds in the continuation of his series.)
This is intended to spark debate btw. If you disagree, please please please say why and don't just blow me off. Adel, I'm looking at you here.
I certainly don't want to minimize the contribution of others writing essays and the like. But I think you're wrong to dismiss the contribution of data.
I think comparing Hoopla's data to newbie data is useful, but newbie data is more easily collectible as there's a limit to the number of setups possible and each gets reported and cataloged together, making compiling the data relatively simple if a little mind-numbing. Hoopla's, on the other hand, not only shows whether town or scum won but the actual setup. Having tried to do this on a limited number of games before, it is really tough stuff to do because mods aren't always great about including what the setup was on the front page. That can lead to a lot of hunting for information.
Obviously the data does not show anything in and of itself. But it gives the opportunity for analysis that did not exist before. It helps fight bad conventional wisdom. It helps people more rationally think about the balance of the games they're creating. 63% win rate for scum is HUGE. In the NFL, the owners were convinced to change playoff rules because the winner of the coin toss had gotten SUCH an advantage in recent years... and that was something like 59%.
.
