Jump to content
  • Sign Up

Constructs and Mechanics that lead to the Death of the Game.


Recommended Posts

IMO this is one of those things that are common sense. Back when woodenpotatoes was doing his no-map challenges, he remarked something to the effect of "GW2 has such a rich environment, yet most players only ever interact with it once for map completion." To put this in comparison, back when I played Runescape one of the few stand-out moments I had with it was showing a newbie the blast furnace. Particularly, how the whole journey got waylaid for a bit once we hit the dwarven city of Keldagrim. The kid had no idea there was a gigantic metropolis just under his feet, and he had to spend some time running around just exploring the place when we got there. He actually typed out his amazement in chat. That is the sort of thing that I've never seen once happen in GW2.

In a very broad sense, all of the obstacles and challenges you have to overcome in a videogame are inconveniences. The story is locked behind a boss, the treasure chest is at the top of a treacherous mountain, limited build points forces you to pick one ability over another, etc. If there wasn't some kind of problem to solve, then it wouldn't be much of a game. I can understand simplifying things like tedious inventory management or UI elements, but it is possible to make things too convenient. The waypoint system is one of those conveniences that has a negative trade-off. While it relieves you of the burden of having to fight through enemies and trudge through jungle to get to your destination, those lost things are gameplay elements. All of the interaction with the environment is sacrificed. No more plotting paths, getting caught up in unexpected events, learning to navigate dangerous terrain, carving economic niches from the environment, or striking up unlikely friendships. Instead, you just warp from big shiny-giving event to next big shiny-giving event, which have their rewards balanced assuming they'll be zerged in mass.

I wouldn't fret too much about the negative feedback you're getting. All of the people who are still playing the game are the kind of people who don't mind all of these design features.

Link to comment
Share on other sites

  • Replies 117
  • Created
  • Last Reply

Top Posters In This Topic

@JusticeRetroHunter.7684 said:

@"Julischka Bean.7491" said:One question though. How do you know people are in complete silence at banks and such? They might be engaging in guild chat, or whispering to the person next to them.

So i wrote this as a response to someone I'm currently messaging whom I'm having a more in depth discussion on the topic, so I'll just paraphrase what i said over there, over here -

----Unfortunately, these hidden interactions can't be realistically tested to any reliable accuracy. The assumption is that anything that could be considered outlier behavior (Extremely chatty guilds, to extremely dead guilds.../extremely chatty players in hidden channels, to extremely silent players in hidden channels) are treated as outliers, and thus the frequency of guild chats, whispers or other 3rd party communication services lay somewhere in between, which we would expect to be distributed as a bell curve with some average frequency.

The reasoning behind just observing interactions we can see, and using that as data we could use, is that it should be representative of interactions as whole by a similar proportionality, which is again based on how we aren't treating interactions we can't measure as outliers, but instead as an average. This is the same tactic that's used in statistics, where for example,
, (based on the 80/20 rule). In other words, the people that use reddit or forums will be some proportion of the population. We are using this tactic but in reverse, taking the total population and using measurable interactions to determine proportionality of hidden interactions.

So in a general sense, the amount of interactions you can measure in chats you can actually observe, is assumed to be some proportion of all chats that you can't observe. So if there are 100 people in Lions Arch, and you have 10 Interactions per hour, than it implies that may also have on average 10 interactions per hour that you can't observe.

Again, because you can't reliably measure these hidden interactions, you can't make a case for it because it could be anything, ranging from 0 to 100 to a thousand. It's unfalsifiable.

Wouldent the 80/20 rule mean that 10 is the 20 you can observe, it would then be 40 the 80 you cant?

Why is it 10 and 10 in this post?

Link to comment
Share on other sites

@"Seera.5916" said:You can't have it both ways. You can't use unfalsifiable data to prove your hypothesis is correct if you're saying we can't use it to say that it's wrong.

Counting the actual number of interactions you can actually measure is not unfalsifiable. It's empirical... because you can actually measure it. This is scientific method 101...

It's like saying that because we can't see the bearded man in the clouds, means that god exists we just can't observe it therefor everything we know about science is wrong. The correct conclusion is that because you can't measure it, you have to exclude it, and if you wanted to be generous, assume what it would be if you could measure it if it were there (they don't even do this in scientific method either...it's just being generous to such a position)

In the scientific method, things you can't measure in theory or practice is considered unfalsifiable and essentially useless.

Example: It's reasonable to expect that Discord is where a large number of conversations happen.

No it's not reasonable. Again you can't measure it to within any accuracy.

We can estimate forum/reddit users to active players because there have been enough companies who have released active player data to base those calculations off of. I've yet to see any official data for number of public vs private chats and they would always be lower on private since they wouldn't be able to account for third party chat systems.

You can actually figure out the proportionality by averaging out how often you see say chats in proportion to map chats, in proportion to guild chats in proportion to whisper chats etc.

If in a 24 hour time period, observe 500 map chat entries, 100 say chat entries, 50 guild chat entries and 10 whisper chat entries, you can with ever increasing accuracy show that as a common proportionality between the different mediums and as a proportion of all entries as a whole...That's not something that only a company with dev controls can do...anyone anywhere can figure that one out...in fact I'm inclined to do it myself.

Link to comment
Share on other sites

@Linken.6345 said:Wouldent the 80/20 rule mean that 10 is the 20 you can observe, it would then be 40 the 80 you cant?

Why is it 10 and 10 in this post?

The 80/20 rule would imply that the number of people interacting is representative of the population as a whole via a power law. If 100 people are in a map, if such a thing obeys an 80/20 power law, than 20 people on average should be chatting out of those 100 people. This is why i said 10 and10.

Of course we don't know if it obeys a 80/20 power law, we simply can't acquire that data. What we can say, is that we can count how many people are interacting via what we can see via say chats, map chats, emoting, being in a party or whatever interactions we can recognize, and use that as a proportion to the total population size we are sampling which will obey some kind of proportionality. The proportionality I'm presenting is one where it treats hidden interactions as an average of the other interactions (weighting them as equally likely to happen) which also gives you a 10 and 10 in that example.

Link to comment
Share on other sites

@Blood Red Arachnid.2493 said:In a very broad sense, all of the obstacles and challenges you have to overcome in a videogame are inconveniences. The story is locked behind a boss, the treasure chest is at the top of a treacherous mountain, limited build points forces you to pick one ability over another, etc. If there wasn't some kind of problem to solve, then it wouldn't be much of a game. I can understand simplifying things like tedious inventory management or UI elements, but it is possible to make things too convenient. The waypoint system is one of those conveniences that has a negative trade-off. While it relieves you of the burden of having to fight through enemies and trudge through jungle to get to your destination, those lost things are gameplay elements. All of the interaction with the environment is sacrificed. No more plotting paths, getting caught up in unexpected events, learning to navigate dangerous terrain, carving economic niches from the environment, or striking up unlikely friendships. Instead, you just warp from big shiny-giving event to next big shiny-giving event, which have their rewards balanced assuming they'll be zerged in mass.

Well said.

Link to comment
Share on other sites

@"JusticeRetroHunter.7684" said:For years, I study a field of science called "Complexity Theory", which is just the science behind complex systems dynamics. Complex systems are essentially how non-linear systems where things interact with each other create higher orders of organization, patterns, and chaotic behaviors. Complexity Theory is used in almost every scientific field known to exist in which the systems are non-linear. This also includes gaming, and many aspects of it. But what I'm going to focus on is more along the lines of Social and societal complexity, and how societal interactions lead to complex behaviours, and how some behaviours are good and how some are bad for the longevity of the game (maintaining healthy player population and interaction)

I'm afraid I don't see how this links to any of the rest of your argument. Which results or techniques from complexity theory have you used to draw your conclusions? This might be clear to you, but since you haven't explained it, it's not at all clear to your readers. Without explaining your methods, your argument is unlikely to be persuasive (which, perhaps, is why so many of the responses disagree with you).

To look at the above scenario in a mathematical sense, one would create a network diagram that would show the density of interactions in particular places on a "network map." In the game in which there exists waypoints, the density map would look something like this.

In a game in which way points did not exist, the density map would look something like this.

These are not networks in the mathematical sense. They look more like heat maps – that is, plots of 2-variable functions – but they look to me like generic illustrative pictures of heat maps rather than actual plots of actual functions or data. When you say "the density map would look something like this", what has led you to that conclusion? What data collection and/or modelling did you do? I know that you gave this explanation:

Now why would there be such a difference in interaction density with and without waypoints? It's because in by taking a waypoint, potential interactions that could have happened in between traversing between point A and point B simply do not occur because it doesn't exist. Therefor, player density becomes more tightly packed into denser and denser regions, and thus, interactions that could have happened elsewhere do not occur, and the ones that do occur are now confined to a smaller region, because interactions between other individuals, will more likely be found in these major hubs rather than outside of them. We see this behavior in game currently, where the majority of people...perhaps at least 80% of the population inhabit major cities at any given time.

But this seems both vague and conjectural. It certainly doesn't give an indication that you looked at this "in a mathematical sense", as you claim. And where does this "at least 80%" figure come from?

Note: I'm not saying that you haven't used proper mathematical/scientific methods here – just that you haven't explained your methods sufficiently.

Link to comment
Share on other sites

@Tommo Chocolate.5870 said:These are not networks in the mathematical sense. They look more like heat maps – that is, plots of 2-variable functions

Good post. I explain at some point in this thread to address what you said in this comment. It's true, that it's not a traditional network map, it basically a contour map. The reason it's like this is because of lousy paint skills.

It is a 2 variable, but each variable is a density which requires 2 parameters (mass x area), rather than a scaler like an actual heat or contour map, in order to define clearly which is why it's not particularly easy to draw in paint.

There is the Density of Population (The number of people within a certain radius) and the Density of Interaction (The number of Interactions that occur between people, within a radius). The easiest way to make an illustration of the above in my opinion at the time and, given my skills with paint was a heat map. And was just meant to illustrate the differences of one behavior to another behavior. In hindsight I should have spent more time making something that takes that information to make it easier to understand based on experimental data to support the hypothesis better...but it's just a forum post that explains the behaviour... I'm not trying to write a thesis.

Anyway, the math's is somewhat trivial and what I probably mistakenly did was trying to dumb down in order to make it into a forum post... For example, if I'm talking about densities that should suggest both a volume and and a mass, which so long as those things are defined I shouldn't have to draw anything for someone to understand the numerical differences between the two densities of the images in the OP.

In conclusion, I'm not trying to give people a thesis paper in this thread...nor am I really required to in order to present an idea (which is essentially what it is as I said earlier, a hypothesis since it's based on observation not experiment)

Would my post have been better if it were a thesis backed by experiment? Of course it would. I'm inclined to actually do the experiments because of how painfully stark the observational data appears to support the above...but this is a forum post that uses a more scientific approach to a problem (it's actually more of a logical analysis) to present an idea... it's not any more complicated than that as an answer to your comment.

Link to comment
Share on other sites

@"JusticeRetroHunter.7684" said:Counting the actual number of interactions you can actually measure is not unfalsifiable. It's empirical... because you can actually measure it. This is scientific method 101...Ignoring things you can't personally see and assuming they don't exist is not part of a scientific method though.

Specifically, when you make a theory based on observed data, that theory applies only to the part you can observe - you can't automatically assume it will hold true when you include the parts you couldn't observe. You'd need to make a reasonable explanation why the unobserved data would not impact you results first - and you didn't do that. You claimed that this unobserved data is irrelevant, without explaining why it is so. That's highly unscientific.

No it's not reasonable. Again you can't measure it to within any accuracy.Actually, no. Discord does exist, and conversations do happen there. That's a fact, not a belief. Sure, you can't measure them easily (which, by the way, doesn't mean you can't measure them - it just means you, personally, don't have easy access to the right tools for that), but they do exist, and you can't simply ignore them. At least not when you're talking about social interactions.

We can estimate forum/reddit users to active players because there have been enough companies who have released active player data to base those calculations off of. I've yet to see any official data for number of public vs private chats and they would always be lower on private since they wouldn't be able to account for third party chat systems.

You can actually figure out the proportionality by averaging out how often you see say chats in proportion to map chats, in proportion to guild chats in proportion to whisper chats etc.You could do that, if you had access to that kind of data - but you
don't
have that for anyone that is not yourself.

If in a 24 hour time period, observe 500 map chat entries, 100 say chat entries, 50 guild chat entries and 10 whisper chat entries, you can with ever increasing accuracy show that as a common proportionality between the different mediums and as a proportion of all entries as a whole...That's not something that only a company with dev controls can do...anyone anywhere can figure that one out...in fact I'm inclined to do it myself.Those proportions will be different for each player, though - you can't "average" that data based only on your personal experience. Well, you can, but it will apply only to you, it won't be something you could extrapolate on the entire playerbase.

Link to comment
Share on other sites

@"Astralporing.1957" said:Specifically, when you make a theory based on observed data, that theory applies only to the part you can observe - you can't automatically assume it will hold true when you include the parts you couldn't observe. You'd need to make a reasonable explanation why the unobserved data would not impact you results first - and you didn't do that. You claimed that this unobserved data is irrelevant, without explaining why it is so. That's highly unscientific.

Ehh...i think you need to go back and understand how scientific method works. Sorry not trying to be rude. Firstly, you never hold ANYTHING in a science to be 100% true ever because there will always be a margin of error in experiments (Experiments are done via orders of accuracy, where the only way of getting 100% accuracy is by doing an experiment an infinite number of times) .

Secondly, you do not need to explain away things you can not measure. In science you don't explain away the theory of the bearded man in the sky everytime we talk about protons and neutrons. The theories are based on what you can measure...and that's as far as the theory will go to address a phenomenon Things you can not measure in theory or in practicality is what's called unfalsifiable...and can't be USED to prove or disprove a theory...its inconclusive information. This is not my words here...this is something you go on to google and nod your head and accept that this is the reality of scientific approach.

No it's not reasonable. Again you can't measure it to within any accuracy.Actually, no. Discord does exist, and conversations do happen there. That's a fact, not a belief. Sure, you can't measure them easily (which, by the way, doesn't mean you
can't
measure them - it just means
you
, personally, don't have easy access to the right tools for that), but they
do
exist, and you can't simply ignore them. At least not when you're talking about social interactions.

We can estimate forum/reddit users to active players because there have been enough companies who have released active player data to base those calculations off of. I've yet to see any official data for number of public vs private chats and they would always be lower on private since they wouldn't be able to account for third party chat systems.

You can actually figure out the proportionality by averaging out how often you see say chats in proportion to map chats, in proportion to guild chats in proportion to whisper chats etc.You could do that, if you had access to that kind of data - but you
don't
have that for anyone that is not yourself.

Those proportions will be different for each player, though - you can't "average" that data based only on your personal experience. Well, you can, but it will apply only to you, it won't be something you could extrapolate on the entire playerbase.

Not exactly true. Because you can go looking at a streamer's chat logs can give you data to reference. But, in general the only person you have to make a reliable measurement is yourself, ...one can think of clever ways to get non-bias data , and or you of course can collaborate with others to get that number, which realistically you can do. (Send picture of your chat box, send over for analysis. Done.) Since you're only finding a universal proportionality, you don't have to interview complete strangers, you can go about and ask your friends for snaps.

Anyway, the scientific methods on measurement are very precise and clear for a reason. Reliable measurement means you are supposed to do an experiment over and over again, and what this does is confirm accuracy of the measurement. You can take data that is just yourself, but you have to include a margin of error due to the fact that it's just one experiment and not 5 sigma's of experiments. Here in gw2 we don't need 5 sigma's of accuracy to talk about a forum post...you just include margins of error based on how rigorous the experiment was.

Link to comment
Share on other sites

@JusticeRetroHunter.7684 said:

@"Astralporing.1957" said:Specifically, when you make a theory based on observed data, that theory applies only to the part you can observe - you can't automatically assume it will hold true when you include the parts you couldn't observe. You'd need to make a reasonable explanation why the unobserved data would not impact you results first - and you didn't do that. You claimed that this unobserved data is irrelevant,
without
explaining why it is so. That's highly unscientific.

Ehh...i think you need to go back and understand how scientific method works. Sorry not trying to be rude. Firstly, you never hold ANYTHING in a science to be 100% true ever because there will always be a margin of error in experiments (Experiments are done via orders of accuracy, where the only way of getting 100% accuracy is by doing an experiment an infinite number of times) .

Secondly, you do not need to explain away things you can not measure. In science you don't explain away the theory of the bearded man in the sky everytime we talk about protons and neutrons. The theories are based on what you can measure...and that's as far as the theory will go to address a phenomenon Things you can not measure in theory or in practicality is what's called unfalsifiable...and can't be USED to prove or disprove a theory...its inconclusive information. This is not my words here...this is something you go on to google and nod your head and accept that this is the reality of scientific approach.

No it's not reasonable. Again you can't measure it to within any accuracy.Actually, no. Discord does exist, and conversations do happen there. That's a fact, not a belief. Sure, you can't measure them easily (which, by the way, doesn't mean you
can't
measure them - it just means
you
, personally, don't have easy access to the right tools for that), but they
do
exist, and you can't simply ignore them. At least not when you're talking about social interactions.

We can estimate forum/reddit users to active players because there have been enough companies who have released active player data to base those calculations off of. I've yet to see any official data for number of public vs private chats and they would always be lower on private since they wouldn't be able to account for third party chat systems.

You can actually figure out the proportionality by averaging out how often you see say chats in proportion to map chats, in proportion to guild chats in proportion to whisper chats etc.You could do that, if you had access to that kind of data - but you
don't
have that for anyone that is not yourself.

Those proportions will be different for each player, though - you can't "average" that data based only on your personal experience. Well, you can, but it will apply only to you, it won't be something you could extrapolate on the entire playerbase.

Not exactly true. Because you can go looking at a streamer's chat logs can give you data to reference. But, in general the only person you have to make a reliable measurement is yourself, ...one can think of clever ways to get non-bias data , and or you of course can collaborate with others to get that number, which realistically you can do. (Send picture of your chat box, send over for analysis. Done.) Since you're only finding a universal proportionality, you don't have to interview complete strangers, you can go about and ask your friends for snaps.

Anyway, the scientific methods on measurement are very precise and clear for a reason. Reliable measurement means you are supposed to do an experiment over and over again, and what this does is confirm accuracy of the measurement. You can take data that is just yourself, but you have to include a margin of error due to the fact that it's just one experiment and not 5 sigma's of experiments. Here in gw2 we don't need 5 sigma's of accuracy to talk about a forum post...you just include margins of error based on how rigorous the experiment was.

There you go again stating that unfalsifiable data can't be used to prove or disprove something, yet you've constantly used it your attempts to prove your theory.

You can't determine the ratio of hidden conversations to ones you can see because there are not tools for that and you wouldn't get a large enough sample size from friends where you could be sure you could reliably eliminate duplicate chat snaps. Not to mention friends wouldn't be random enough to make sure enough different play styles were captured in order to get the widest chat types.

Therefore any argument that hinges on using the 1:1 ratio you claimed is derived from unfalsifiable data and therefore based on your own words unusable to prove your hypothesis. Especially when there's good reason to suspect that there are more hidden conversations than ones you can see.

Link to comment
Share on other sites

@"JusticeRetroHunter.7684" said:Ehh...i think you need to go back and understand how scientific method works. Sorry not trying to be rude. Firstly, you never hold ANYTHING in a science to be 100% true ever because there will always be a margin of error in experiments (Experiments are done via orders of accuracy, where the only way of getting 100% accuracy is by doing an experiment an infinite number of times) .And yet you do. You take the hypothesis based on your observed data, and claim it holds true for all data.

Secondly, you do not need to explain away things you can not measure.Things that cannot be measured. This is not the same as "things that you cannot measure". And it just so happens that things you ignore are measurable. The data to do so exists. We simply lack access to that data.

In science you don't explain away the theory of the bearded man in the sky everytime we talk about protons and neutrons.Sure, but the existence of Discord (and other third-party channels) as well as in-game private channels of communications is not a matter of belief - it's a provable fact.

The theories are based on what you can measure...and that's as far as the theory will go to address a phenomenon Things you can not measure in theory or in practicality is what's called unfalsifiable...The things you were unable to measure are measurable in theory however. You can't ignore them simply because you, personally, lack the data.

Not exactly true. Because you can go looking at a streamer's chat logs can give you data to reference. But, in general the only person you have to make a reliable measurement is yourself, ...one can think of clever ways to get non-bias data , and or you of course can collaborate with others to get that number, which realistically you can do. (Send picture of your chat box, send over for analysis. Done.) Since you're only finding a universal proportionality, you don't have to interview complete strangers, you can go about and ask your friends for snaps....you do realize, that what you're doing now is create a heavily biased set of data - and without even knowing that it's biased, or the direction of the bias, so you can't even try to correct it. Based on this kind of data, depending on the bias i could probably "prove" anything about social interactions, including things that would be completely contradictory. This kind of data gathering is completely worthless.

You've just failed at the very basis of data analysis. That's an F right here. Try harder next time.

Here in gw2 we don't need 5 sigma's of accuracy to talk about a forum post...you just include margins of error based on how rigorous the experiment was.So, how big do you think the margin of error would be in this case? (hint: it may be bigger than you think it is)

You can't claim you follow the scientific method, when you build your theory on biased data and guesstimates (especially ones that are based solely on your preconceptions).

Link to comment
Share on other sites

@Astralporing.1957 said:Sure, but the existence of Discord (and other third-party channels) as well as in-game private channels of communications is not a matter of belief - it's a provable fact.

Nobody said anything about belief... it's about measurement. like I said again if you can not measure it in theory OR in practice, it is unfalsifiable. We obviously know there are people out there that use discord...but we can not to within any accuracy show how many that number of people is. It's that simple. I'm not explaining this any further go look it up...again not my words.

Link to comment
Share on other sites

@JusticeRetroHunter.7684 said:

@Pockethole.5031 said:What i'm saying here is that if you choose to play alone, then your lack of interaction with other people, means those people also lose an interaction with you. If everyone was like you, then everyone would lose interaction with everyone else. That's the macroscopic picture of such a behavior. Interaction leads to growth of a society, while no interaction leads to degradation of a society.

Hey that's fair ... But that has nothing to do with convenience. Targeting convenience features is a bad strategy if you want to push the social interaction aspect of an MMO ... at least in my opinion.

The idea behind the attack on having too much conveniences is that it causes individuals to be self-sufficient. Self Sufficiency is an isolationist behavior, and the thing about convenience is that it is an optimal strategy for accomplishing goals...so despite it being bad for the overall growth of a society, a convenience is often the most optimal choice to have on an individual level.

OK ... but I don't see why that's a problem. I mean, it would be ridiculous to remove some elements of self-sufficiency just to force people to interact don't you think? So removing waypoints ... are players more socially interactive if they have to run around all over the place instead of just using waypoints? No, they aren't ... that's silly. SOME places, convenience IS the answer. The trick is putting the social interaction in places that make sense ... and NOT putting it in places where it don't. Removing travel options is NOT a place where it makes sense to remove it because people can't socially interact if they are running around in a map anyways.

Certainly, some of these conveniences you talk about .. they are not the death of the game. That's absurd. These kinds of ideas you have assume that the game can only exist and survive with high levels of social interaction. That's not true at all. Some elements of the game that might be true ... but for the game in general, the idea that not forcing players into some minimum threshold of social interaction is it's death is nonsense.

Link to comment
Share on other sites

Your whole theory is busted. You posses 0 amounts of data or measurements that suggest the game is dying due to conveniences, or that waypoints, mounts, and convenience items are problematic...

You even attempted to draw a parallel for your argument between newer maps and less waypoints, but didn’t realize that was due to the introduction of mounts... Not because of your speculation that Anet was recognizing waypoints were bad for the game...

And again, what’s the point? Anet won’t remove the things you are complaining about. If you don’t like certain things about the game then don’t use them. Make your own guild of convenience avoiders.

And speaking of “social” and claims of the “death of the game”... Do you have your own guild? How many members? Do you plan event runs in game and lead them? Have you advertised your guild on the forums? Have you planned events for the community on the forums? Do you tag up at events and bring people together to play? Do you tag up in wvw on your server? Do you lead teams for fractals, raids, dungeons, spvp? How often do you use waypoints? How many mounts do you have? How many mount skins did you purchase? What fast travel items do you use? And which fast travel items did you buy? What other convenience items do you own and use?

And here, you can use this to verify certain data about your account and show us...

https://gw2efficiency.com/

Link to comment
Share on other sites

@JusticeRetroHunter.7684 said:I can go in game right now and count how many people are in each starter area's town to give you a clue as to how low this density of people are, and i can then tell you if these people are even interacting within a 10-15 minute timeframeVillage of Shaemor = 9 People - 1 Interaction (conversation)Soren Draa = 1 Person - 0 InteractionVillage of Smokestead = 4 Perople - 0 InteractionGate of Horncall = 2 People - 0 InteractionVillage of Astorea = 8 People - 1 Interaction (Party)

@JusticeRetroHunter.7684 said:

@"Seera.5916" said:You can't have it both ways. You can't use unfalsifiable data to prove your hypothesis is correct if you're saying we can't use it to say that it's wrong.

Counting the actual number of interactions you can actually measure is not unfalsifiable. It's empirical... because you can actually measure it. This is scientific method 101...

Sure you can do that. And then you can write down how many "interactions" you have counted per hour-of-day, day-of-week, during a festival and when there is no festival, at patch day, during a pandemic, etc. etc. But first, of course, you have to define what you actually count as an "interaction".

The tricky part starts after you are finished with counting, when you try to analyze the numbers and when you try to find correlations and when you try to draw conclusions based on these numbers. Because: Correlation does not imply causation.

It's like saying that because we can't see the bearded man in the clouds, means that god exists we just can't observe it therefor everything we know about science is wrong. The correct conclusion is that because you can't measure it, you have to exclude it, and if you wanted to be generous, assume what it would be if you could measure it if it were there (they don't even do this in scientific method either...it's just being generous to such a position)

No one can actually prove that god exists and no one can actually prove that god not exists. So, the statements "god exists" and "god does not exist" are both not scientifically proven. The correct statement would be "we do not know." and as long as there is no need for the existence of god in a scientific model of reality, the (non-)existence of god can be ignored in those scientific models of reality.

In the scientific method, things you can't measure in theory or practice is considered unfalsifiable and essentially useless.

Just because you personally can not measure things, does not mean they do not exist or are not important.This is a little bit more advanced than your "scientific method 101" level: "Absence of evidence is not evidence of absence"

You could (in theory, if Anet and Discord would give you access to their servers) count how many players are in every area of the game at any time, how many player-interactions in-game in chat and in Discord happen, what players are actually doing in the game during these interactions and at which locations these players are when the most interactions/chats happen).

Ignoring those interactions, or making very biased assumptions about them, just because you personally can not measure them, but you still want to push your results/theory into a specific direction (so it seems), is bad science.

Link to comment
Share on other sites

@JusticeRetroHunter.7684 said:I can go in game right now and count how many people are in each starter area's town to give you a clue as to how low this density of people are, and i can then tell you if these people are even interacting within a 10-15 minute timeframeVillage of Shaemor = 9 People - 1 Interaction (conversation)Soren Draa = 1 Person - 0 InteractionVillage of Smokestead = 4 Perople - 0 InteractionGate of Horncall = 2 People - 0 InteractionVillage of Astorea = 8 People - 1 Interaction (Party)You should check out Divinity's Reach. Especially around the Minister's Waypoint. Usually quite busy.

Link to comment
Share on other sites

@JusticeRetroHunter.7684 said:

@Astralporing.1957 said:Sure, but the existence of
Discord
(and other third-party channels) as well as in-game private channels of communications is not a matter of belief - it's a provable fact.

Nobody said anything about belief... it's about measurement. like I said again if you can not measure it in theory OR in practice, it is unfalsifiable. We obviously know there are people out there that use discord...but we can not to within any accuracy show how many that number of people is. It's that simple. I'm not explaining this any further go look it up...again not my words.

OK ... but you don't just ignore those things either, especially if you know those phenomenon exist and their manifestations are in opposition to your theory. let's be frank here ... not sure your background, not going to assume anything but it feels like alot of armchair science going on; feels like you read something compelling to you, got hooked and your fishing for something that validates your beliefs ... like you are trying to find a PhD topic or something. I don't see a theory you can prove or disprove (so it's a useless academic exercise) and I don't see what problem you are trying to solve (so it's a useless practical exercise as well) since it's VERY unlikely conveniences are going to get removed from this game as an experiment nor is social interaction needed to be forced in an MMO.

Link to comment
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now
×
×
  • Create New...