The Newbie 2d3 stats thread (upd. 2019-05-28, 87g)

[mhsmith0]

fwiw I'm actually quite surprised that the column A setups have the highest town win rates so far, as I'd very much expect it to be B over the long term (and then I'd GUESS that C has lowest town win rates though I could well be wrong).

[mhsmith0]

i dont know statistics but is the fact that b column has a ~10% higher wr for town significant enough to need change

I remember thinking there was a real chance the rolecop just wasn’t strong enough for scum
Probably we should think about something altogether different?
Does anyone have thoughts on something like semi nightless, running like a 7/2 format of it (9/2 seems townsided based on limited sample size)?
It’s definitely different from standard approach, but maybe it’d be GOOD to have more vanilla-y stuff that heavily emphasized dayplay and de emphasizes nightplay?

[mhsmith0]

Having the 6 newbies makes it that much more likely to get a scumteam of 2 newbies, and they appear to git rekt most of the time.

This part in particular is VERY worrying. Retention will be hurt if all of the all newbscum teams get clobbered, and I think it’s SUGGESTIVE that more experienced players are particularly effective in terms of gaming the system, to a greater degree than was the case in matrix6.

[mhsmith0]

@BNL: I agree with NK15, also there's going to be a lot more unbalance in a closed type of system, just because ppl have different standards and it's near impossible to perfectly balance stuff over long term. It adds complication that I don't think really adds value.
@NK15: allowing vigs in newbie games is a bad idea. Vig games, more than any other setup, become heavily focused on the competence of the vig player, so there's more pressure and more potential for salt fests postgame. Moreover, it opens up the chance for a scum win d2 which is overly early.

[mhsmith0]

The easiest way to fix the problems with vigs in newbie setups is to avoid having vigs in newbie setups

The essential problem with 2d3 is that high power level type setups are swingy, difficult to balance, and particularly prone towards giving higher win rates to people familiar with the system who know its fault lines and can take advantage, and lower win rates to newbies who don't know what they're doing (the more easily fixable problem is that scum are underpowered on column B, which is pretty fixable by tweaking the power levels, like turning rolecop into a JOAT of some sort [or maybe odd night roleblocker?], or turning the other goon into a 1-shot novice roleblocker, etc).

The main question is whether that's the kind of problem that ought to lead to a different design entirely, or whether that's the sort of problem that is fixable by more minor tweaking (tweaking scum power levels on column B is the simplest fix on a concept basis.

My intuitive recommendation would be to step back, see how win rates change due to removing ICs, and taking that into account when making any changes in the future.

[mhsmith0]

The more role-heavy an open setup is, the more it devolves into setup-related discussions. 2d3 is pretty role-heavy for a 9p open setup (almost always 2 town prs, usually 1 wolf pr), and I think that's probably a decent lesson in terms of it being better to have less rather than more roles in newbie games.

I think a pool of closed setups is a pretty bad idea, because a non-public pool even more heavily benefits people who can setup spec their way into profit, and overly advantaging people who are good at finessing setups seems to be a big downside to 2d3.

There's theoretically some good in terms of getting newbies used to role-jampacked type games (since they are common on ms), but I think the overall downside we're seeing overweighs that benefit by far.

[mhsmith0]

Some examples of opens where there's very little strategy discussion

Cop/6 vts vs 2 goons
JK/6 vts vs 2 goons
7 vts vs 2 goons, but semi-nightless (no nk the first two nights)

Once you start getting into "town gets 2 prs, and each town pr fits in a specific way given other town prs and/or wolf prs" you get into more and more setup/strategy/gaming stuff. 0 wolf PRs, 0-1 town PRs, maybe 2 town PRs in certain specific circumstances but probably it's better to structure it to be 1 town PR at most. That eliminates certain learning aspects, such as how to protect a scum PR and how to hunt for town PRs, but I think the gain of turning the dial way down on setup spec/talk grossly outweighs the loss of exercising certain muscles.