Baldur's Gate : The Sword Coast Chronicles : NWN2 Persistent World

I suggest everyone takes a step back from this until Rhifox makes her announced post.

How one thinks that dynamic scaling of mobs, which allow characters with wider level range to adventure more dungeons is beyond me. Especially when considering how static this server is and how few areas get added. It is so counterproductive (sic) that it doesn’t worth arguing about

Sadly it seems the main point wasn't heard or considered at all

mrm3ntalist wrote: ↑Mon Mar 21, 2022 10:40 am How one thinks that dynamic scaling of mobs, which allow characters with wider level range to adventure more dungeons is beyond me. Especially when considering how static this server is and how few areas get added. It is so counterproductive (sic) that it doesn’t worth arguing about

Rhifox wrote: ↑Wed Mar 23, 2022 5:50 pm viewtopic.php?p=936025#p936025

Thank you for the update Rhifox. It was informative.

Rhifox wrote: ↑Wed Mar 23, 2022 5:50 pm viewtopic.php?p=936025#p936025

How many unique players on every area?

mrm3ntalist wrote: ↑Wed Mar 23, 2022 6:37 pm How many unique players on every area?

We can generate one. A post-rebalance one will probably be a bit too variable after only a few weeks of data, but we can do one for before.

Rhifox wrote: ↑Wed Mar 23, 2022 6:47 pm

mrm3ntalist wrote: ↑Wed Mar 23, 2022 6:37 pm How many unique players on every area?

We can generate one. A post-rebalance one will probably be a bit too variable after only a few weeks of data, but we can do one for before.

Wait, you presented all those statistics and you have no idea of the sample size?

mrm3ntalist wrote: ↑Wed Mar 23, 2022 7:04 pm

Rhifox wrote: ↑Wed Mar 23, 2022 6:47 pm

mrm3ntalist wrote: ↑Wed Mar 23, 2022 6:37 pm How many unique players on every area?

We can generate one. A post-rebalance one will probably be a bit too variable after only a few weeks of data, but we can do one for before.

Wait, you presented all those statistics and you have no idea of the sample size?

We have the sample size for each, but we don't have a viz for unique players per area. Doing per-area unique players over such a short span is going to be too variable to provide meaningful insights; we can only look at the aggregate statistics with this much data.

Before:
Data covers ~106.29 days of 505 unique players from 2021-10-23 07:24:46+00:00 to 2022-02-06 14:39:57+00:00.

After:
Data covers ~27.50 days of 317 unique players from 2022-02-06 14:40:02+00:00 to 2022-03-06 02:20:12+00:00.

Trust me, I'm a professional!

gedweyignasia wrote: ↑Wed Mar 23, 2022 7:09 pm

mrm3ntalist wrote: ↑Wed Mar 23, 2022 7:04 pm

Rhifox wrote: ↑Wed Mar 23, 2022 6:47 pm

We can generate one. A post-rebalance one will probably be a bit too variable after only a few weeks of data, but we can do one for before.

Wait, you presented all those statistics and you have no idea of the sample size?

We have the sample size for each, but we don't have a viz for unique players per area. Doing per-area unique players over such a short span is going to be too variable to provide meaningful insights; we can only look at the aggregate statistics with this much data.

Before:
Data covers ~106.29 days of 505 unique players from 2021-10-23 07:24:46+00:00 to 2022-02-06 14:39:57+00:00.

After:
Data covers ~27.50 days of 317 unique players from 2022-02-06 14:40:02+00:00 to 2022-03-06 02:20:12+00:00.

Trust me, I'm a professional!

I do not trust anyone when it comes to statistics. Especially when i do not know them. The easiest way to manipulate people is through statistics. That is why the sample size is - one of - hte most important part. You say 505 total unique players but for how many areas? How many unique players for each area so that we can understand how skewed and relevant hte info is?

yes, I wouldn't go with "trust me, I am a professional" because if i reported what you just did to my manager in NYC, i would be collecting unemployment the next day

EDIT
Forget unemployment, I would be on Rikers Island.
What does an average party of 1.X mean in real life? There are many areas where there is just one entry. The only conclusion that you can come from the statistics that were posted, is that at the very least there are no relevant to make any kind of informed decision about the system as a whole.

mrm3ntalist wrote: ↑Wed Mar 23, 2022 7:18 pm I do not trust anyone when it comes to statistics. Especially when i do not know them. The easiest way to manipulate people is through statistics. That is why the sample size is - one of - hte most important part. You say 505 total unique players but for how many areas? How many unique players for each area so that we can understand how skewed and relevant hte info is?

yes, I wouldn't go with "trust me, I am a professional" because if i reported what you just did to my manager in NYC, i would be collecting unemployment the next day

We're collecting population-level statistics; all players (and all combat) on the server during the time period in question are represented in this dataset. The data are not skewed due to sampling, but instead represent the choice of players to frequent some areas over others. The reason we don't have a plot of what you asked for (unique players by area) is that there is too much noise for that to produce a meaningful result in unpopular areas in this timeframe. (If I provided a plot like that to my manager in NYC, she'd have some words for me.)

mrm3ntalist wrote: ↑Wed Mar 23, 2022 7:18 pm What does an average party of 1.X mean in real life?

Not much, right?! The number itself doesn't mean much in a vacuum. Especially for areas with low traffic, you'll have a hard time figuring out if that's meaningful or if they lucked into a party of 6 for a little bit. (I usually look at the distribution of combat XP by party size for a specific area when I have an interesting case we want ot know more about.) When looking at stuff like that, I usually also look at the variance in party size, spread in party levels, etc. but I didn't want to overwhelm people with a lot of information when they hover over a datapoint. When you combine that average party size with other data from our server (for instance, that about half of our total combat XP is solo players), it helps give you a picture of which areas are either too difficult for soloing or provide rewards that attract a party.

If you play with the sliders (or look at the color gradient on the second visualization in each), you can see a strong correlation between party size and level, which seems to tie in with another interesting effect, which is that the average weekly playtime per-player-per-level increases with level. The reason for this is that after a certain amount of playtime, players will level up! The later levels take longer to get out of, (for instance, level 10 to level 11 requires twice as much XP as level 5 to level 6) and players never get past Lv30. So in a lot of cases, the duration in time for which other players are in a level range is too small for players to party up effectively. This is obvious in hindsight, but it's something we haven't focused much on in design discussions. I want to do some more analysis later that will let me determine how much of that increase in partying is from incentives (i.e. epic loot) and how much is from availability, but I have to wait on data collection. If you want more in-depth analysis, I started a data blog a while back. It might not see many updates until earnings season is over, but there's some good content so far.

mrm3ntalist wrote: ↑Wed Mar 23, 2022 7:18 pm There are many areas where there is just one entry.

Yep, those areas just don't see a significant amount of traffic; players barely go there at all, or avoid combat there entirely. For some areas this means it's not a place that regularly spawns enemies, but sometimes DMs create events there. In other cases, it's a remote place that nobody knows about. Sometimes it's just a transition area with very few spawns where people don't like to level. In the "before" dataset, it was sometimes hard to tell those areas from actual combat areas just by looking at the data, because players concentrated their efforts so singularly on the best grinding spots. It's very encouraging to see that players are engaging with more content (at least on some level, rather than ignoring it completely) since the update.

Thank you for providing feedback and your rationale.

I was not sure what approach the staff was going to take over the course of the 4 - 6 weeks to bring closure to the many concerns the player base raised.

Candidly, the before and after statistical models are not compelling.

Why? Because every player on the server already knew the Wyverns were the most popular place to level. Then the Staff gimped the spawned rate and the popularity of the Wyverns dropped. This too comes as no surprise.

The good news though is that there has been a lot of dialogue (good, bad, and indifferent) about why players prefer one area over another.

I was hoping the staff was using the 4 - 6 weeks to analyze their alignment with the player's why... and then brainstorming the changes necessary to bring everything into alignment.


Cheers

I’m surprised that it took 6 weeks and statistics to learn what was empirically evident already: players go to where and what is easiest to accomplish the desired goal, that being leveling, a boring process in the context of the BGTSCC Server.

It was also already known that some Areas where better designed than others, for X reason (XP gain, Loot, fun). If you play and played on BGTSCC for more than a RL year, any and every player knows this.

Two things are still missing from this analysis, imho: 1) creating a REASON WHY players shouldn’t powerlevel, as fast as THEY please,using the most efficient means. What does slow leveling, leveling in varied locations, really provide the player? This is what always has been missing from the paradigm; 2) understanding and supporting content for end-Level play, which players will reach, no matter what you do. Players in the majority seek out power because the game mechanics reward power. When you find out how the mechanics can reward role-play, you will then finally succeed.

You can either design for a game where it’s fun for players to try and get all 246 star coins in X hours of play—which is what I think is the current desire by wanting to “spread out players”—or you can make the design of the game about stories and role-play and creativity that is supported and recognized.

Obviously, the real answer is to do both, or, that is what BGTSCC had been trying to build/promote itself for like ever. But as I’ve communicated to the powers that be many times, “fixing” mechanics will never make role-play better.

At the same time, fixing mechanics (and Rhi's announcement makes me very optimistic about that, on the condition that all of that really gets implemented without a crippling dose of burnout) will make the server a lot more tolerable during periods where roleplay is absent.

Furthermore, claims that mechanical fixes are useless seem to purposely ignore just how many ways the mechanics have of influencing RP in the first place: One part of Rhi's post that particularly resonated with me was the notion of being left behind by more effective powerlevelers than myself, in situations where those powerlevelers were my PCs' closest friends and companions. Those friendships were typically forged by the mechanical context of shared adventures; when one side found itself incapable of reasonably participating in future adventures with the other, they tended to die off, having had too little time to be reinforced by other factors.

DaloLorn wrote: ↑Thu Mar 24, 2022 4:49 am At the same time, fixing mechanics (and Rhi's announcement makes me very optimistic about that, on the condition that all of that really gets implemented without a crippling dose of burnout) will make the server a lot more tolerable during periods where roleplay is absent.

Furthermore, claims that mechanical fixes are useless seem to purposely ignore just how many ways the mechanics have of influencing RP in the first place: One part of Rhi's post that particularly resonated with me was the notion of being left behind by more effective powerlevelers than myself, in situations where those powerlevelers were my PCs' closest friends and companions. Those friendships were typically forged by the mechanical context of shared adventures; when one side found itself incapable of reasonably participating in future adventures with the other, they tended to die off, having had too little time to be reinforced by other factors.

Again, pointing problems about players, not the mechanics or design, specifically. But, I’ll concede that to change the mindset of the player, one might have to forcibly limit them, railroad them into a certain behavior.

That too has consequences.