Planned Changes: Newbie Game Deadlines

[Papa Zito]

(and the information they leak)

Examples of how bad this is

• If a player who loves/hates playing scum and has >rand reads replaces in, the slot is that much more likely to be scum/town
  
• A newbie replacing out is significantly more likely to be scum
  
• A replace-out considered "bannable" for a scum slot is town-indicative assuming most people don't want to get banned

etc.

Modkills would be way more punitive to the remaining players than a replacement.

Well yes

I think if you commit to playing a team game and don't play, you deserve high loss equity, you deserve to get flamed by the remaining players, and if you do it a non-negligible portion of the time you should also not be allowed in future games

I think this comes from a difference in "competitive"/"for fun" mindset. My mindset is highly competitive and like I said I don't expect most people to agree with me on it. (Also modkilling AFKs probably isn't practical in newbie queue in particular because of how many newbies flake and I realize that too)

But I think the rest of my post still applies anyway because replacements are pretty objectively bad whether or not modkills are subjectively better

You should check out the BaM ruleset. Wouldn't work in the newbie queue tho.

edit: Nevermind, Zor already linked it boo.

[Papa Zito]

The first newbie game with deadline/prod changes has been assigned.

Deadlines have been updated to 10 days for Day 1, and 7 days for all further Day phases. Prodding now occurs after 30 hours of inactivity. (7/2/2018)

I will be watching winrate and replacement stats comparing old and new systems.

This looks solid to me. I've signed up to IC a game and give the new system a whirl.

[Papa Zito]

The first newbie game with deadline/prod changes has been assigned.

Deadlines have been updated to 10 days for Day 1, and 7 days for all further Day phases. Prodding now occurs after 30 hours of inactivity. (7/2/2018)

I will be watching winrate and replacement stats comparing old and new systems.

This looks solid to me. I've signed up to IC a game and give the new system a whirl.

My first game just finished and I'm here to report my findings.

1. The deadlines proved to be largely irrelevant because of replacements. The faster deadlines also didn't seem to spur additional activity in the game, things still tended to drag right until the the last second.
2. The prod timer turned out to be really obnoxious and lead me to being replaced. I'm not usually around on weekends and I had a busy Monday so I didn't get to the game in time to post. I expect many of our older (geriatric) players will have similar experiences.


With that being said, I have some suggestions.

1. Given that we now that we have daytalk for the scumteam, and given that the number of power roles is small and the interactions between them are well documented, I don't see the point of 48 hour nights at all. I think those can be completely eliminated in favor of instant nights. This will help speed up games and keep people engaged.
2. Replacements really are killer. Not only are games getting dragged out because of deadline extensions to let replacements catch up but we're also draining players from filling new games. That second bit is a huge problem I think because newbies don't want to sit around waiting for games to fill, they signed up because they want to play now. I see newbies pop into sitechat every once in a while asking when a game is going to start. I think we'll continue to lose players and have start-of-game flakers if we don't fix this. I have 2 possible solutions.
a. Follow a BaM-like system. No prods, no replacements. If a slot doesn't post in a game day, modkill it.
b. Have a replacement queue just like we already have for ICs, SEs and newbies. The replacement queue have players ready to step into any slot at any time. Don't worry about trying to match experience levels with the player replacing in.
3. We need to stop being so deadline obsessive. We drag out day phases far too much. This is a site-wide cultural issue and not something to solve here but I thought I'd throw it in as an observation.

All in all, in my giant sample size of 1, I didn't feel there was really any change at all except for the prod timer, which was profoundly negative in my case.