I am going to reply to this - but I have to first get into a broader discussion, because I wanted to correct a defect in my analysis and it led me down a broader rabbit hole.

It didn't make sense to me that The Night Elf interest curve went down in 2009 of all places, and frankly, their interest SHOULD have gone down in 2006/2007, but it increased. So I had a thought - what if the downward trend in Night Elf interest was due to a downward trend in interest for Warcraft? I then had the idea of comparing the two datasets out of google analytics using the same timeframe that Krainz had provided to me earlier.

A lot of data is about to follow, but before I get into that - Google analytics doesn't give you a raw number when you run comparisons, and this can get to be problematic when you go to run comparisons between very large search terms and not-so-large ones. "Warcraft" as a search term, is immense in comparison to the others I'm looking at, and it does no good to compare "38, 37, and 42" with "<1, <1, <1". That's why I pulled each term independently, downloaded the data into excel, and started to run comparisons based on their relative performance.

NE vs. Warcraft Plot by Kyalin Raintree, on Flickr

You might be wondering what the grey line is about. That came out of a regression which asked how much of the change in the Night Elf trendline is due to overall trends with Warcraft. Changes in Warcraft overall can explain roughly 69% of the movement.

Warcraft regression by Kyalin Raintree, on Flickr

With that in mind, I pulled data for Illidan, Maiev, Tyrande, and Malfurion to see what the composite data looked like.

All factors by Kyalin Raintree, on Flickr

The P values on Maiev and Malfurion were too high to indicate significance. It should be below .05. Here is what we have with only the significant factors:

P greater than 0.05 by Kyalin Raintree, on Flickr

I want to specifically point out Illidan as a factor as well, of course. As you can see, he DOES explain 21 to 20% of the variation:

Illidan by Kyalin Raintree, on Flickr

I will come back to that, but I looked at the other factors by looking at residuals after just Warcraft, after all of the characters, and after the significant characters, and the results are less surprising than I initially thought.

Residuals comparison by Kyalin Raintree, on Flickr

As a legend, the green line is for the Burning Crusade release. Red is for the Cataclysm release. Purple is for the Legion release. The red lines are the ten-month moving averages.

So, in the full residual or the P < 0.05 residual datasets, we can see what I would either call the "Illidan effect" or the "Blood Elf" effect. In either case it's irrelevant because we see the numbers ascend to the highest they've been after that, before they begin a long trend of continual decline starting at the Cataclysm release, when the decade of humiliation kicks off. But the recovery of these numbers taken with the stable trend indicates that the draw of Blood Elves from the Night Elf population didn't render a sustainable impact on the interest in Night Elves. They surged to their highest levels of interest AFTER the blood elves were introduced, in fact.

For some historical data, here is what the census numbers looked like back then. Night Elves dipped to 16% from 22%, but that is within expectations: https://web.archive.org/web/20090414...ervertypeid=-1

I'll just put a pin in this as well - Wrath was a quiet expansion for Night Elves. There's no reason for this increase, but it happens anyway.

In Legion, I do see the decline flatten - which in part might speak to Krainz's point, so I felt that it was important to take post-2015 data to highlight potentially significant trends, because there are two identifiable possible spikes. Illidan's numbers shoot through the roof just after the Burning Crusade, and his and the numbers of the major Night Elf characters improve in Legion. But notably, Tyrande and Malfurion are flat through the Burning Crusade, because they largely weren't present. You therefore have to look at data in a narrower time horizon, and include demon hunter searches. So I did just that:

Post 2015 by Kyalin Raintree, on Flickr

If that looks like a mess to you, you're not alone. That would be because trying to find a single factor that explains the trend is a mess after 2015:

Post 2015 regression by Kyalin Raintree, on Flickr

Here's what you can say: Tyrande, Malfurion, and "Demon Hunter" in the post 2015 data bear no relation to the trends. Their p values aren't just not below .05, they're above .5. Maiev and Illidan are at least below .2, but they still aren't below the .05 threshold where we can say that they are significant. It is also at this point that I'd like to bring up correlation and covariance, as demonstrated below for the 2004 to present dataset:

04 correlation and covariance by Kyalin Raintree, on Flickr

As we can see, Tyrande, Maiev, and Malfurion are negatively correlated. Illidan is positively correlated. Now let's look at the post 2015 figures:

Post 2015 corr &amp; covariance by Kyalin Raintree, on Flickr

There is positive correlation and covariance here. That much we can't deny.

Even so, it still failed to resolve the overall decline, and that's understood in the context of WoW's overall decline as a factor.

So, in reply to Krainz's most recent points, which are a reiteration of the previous points: no, I don't believe that the mere presence of other elves in retail explains the decline in classic. I also don't think the demon hunter effect is anywhere near strong enough to talk about given that there is no significant impact in the numbers as demonstrated. There are certainly expected effects in the data, but the unexpected ones, and the extreme lack of significance among demon hunters in the Legion Data as it relates to Night Elves demonstrate the other factors that are at play.

Which leads me to classic racials, and the arguments surrounding them. Krainz brought up the highest echelon of players, and I think there is a tendency to assume that they drive overall playerbase statistics....

Pie by Kyalin Raintree, on Flickr

I don't think it holds up... and further from that, minmaxing wasn't born yesterday. It existed in 2006 and 2005, and yet the numbers are what they are. Not everyone minmaxes or is going to minmax. I'd in fact argue that it's a small proportion of the population give the above, and it certainly can't explain the trends I've mentioned previously. As for what can in the context of the other data I've mentioned? Well, you know what I feel and what I've said.

__________________

EDIT: nevermind the Warcraft / World of Warcraft difference I brought up. They both follow a same pattern.

EDIT2: it seems that you were already on it by the time I realized it. Oh well

__________________

Sure. In my excel sheet, I abbreviated the search term "World of Warcraft" as "WoW":

Night Elf versus World of Warcraft by Kyalin Raintree, on Flickr

Night Elf versus WoW regression by Kyalin Raintree, on Flickr

Pretty similar results to just "Warcraft", but it's slightly worse as a predictor. Notice the lower R2 values.

Edit for edits:

It falls within my expectations. Notice that the Blood Elves' interest spike happens in 2007, and then it drops off. What I think is notable is that, and we see this in the other data, render a substantial effect on Night Elf interest. except for in the "Illidan/Blood Elf" gap that I mentioned taking place in 2007.

So, when we got to ask what a population would act like if Vanilla was released today as opposed to back in 2004, I think that interest curve taken with the other data I shared tells the story. Interest for Night elves that isn't merely explained by a decline in interest for Warcraft itself didn't start to wane until Cataclysm was released.

__________________

What do you say of this? https://trends.google.com/trends/exp...elf,void%20elf

__________________

See my "Edit for edits" section above.

__________________

Even the Nightborne becoming more popular than the Night Elves in a given date?

Anyway, if the evidence I pointed out doesn't faze you - or even qualify as evidence -, then I have nothing to say that could affirm either of our suspicions, aside maybe from a good survey.

But yeah, none of us feel like doing that.

__________________

Two things about the Nightborne

1. They were recently introduced. I'd expect that their introduction would have similar effects to the one we saw in Blood Elves. The important thing for my consideration was whether that interest had an impact on the sustainability of interest in Night Elves, and for the most part it didn't - you can see that especially as we move into the Wrath era.

2. They were introduced after the effects that I'm talking about came into play. For all of the complaining I do about BFA, it's rather clear from the data that if there's an effect from lore (and I believe there is), we saw the lion's share of it taking place due to Cataclysm and MOP, and the effect of things like the War of the Thorns was already priced in.

That isn't to defend the War of the Thorns or BFA, it's more that if the declines are due to lore (having responded to the claim that they were just because there was another elf race, which I think I have disputed) the horse was already dead by the time they came around to beat it again.

__________________

Which you have yet to prove, which remains the big weakness of your argumentation. Data exists. Your narrative exists. But you are not connecting them in any meaningful way.

Trends take time. Cataclysm, which you often point to as the big culprit of negative portrayal for nelfs, did not happen until the end of 2010. Cataclysm did not introduce new elves, but it did also introduce new druidic races and other class combinations that were not available previously. Cataclysm shifted my own faction interest, for example, from mostly Horde-side to Alliance-side because it permitted worgen druids and dwarf shamans. This had nothing to do with the presentation of night elves.

This also does not say anything about how many of the raiding guilds began switching increasingly over to Horde from Alliance over time. Is it about lore presentation of night elves? The raiding guilds care little to nothing about that. It was about racials.

You have data and some lovely charts, but you definitely lack compelling or reasonable narratives for the data.

You probably should have noticed that I didn't comment on the reasons for the decline in Night Elf interest after Cataclysm's release this time around, and as a result of not noticing that, you've dodged the entire question.

Krainz raised two possible reasons for the precipitous decline in playing a Night Elf in classic versus the actual Vanilla figures.

a. There are other elf options available in retail.
b. There are demon hunters available in retail, and the interest in demon hunters like Illidan is immense.

Each ends with the implied assertion that if you were likely to play a Night Elf in Vanilla, you are now more likely to prefer retail to the exclusion of classic because classic does not have those options.

I said that I would agree with a. if I could find evidence of a sharp, sustained decline in Night Elves after the other major elf option, Blood Elves, came out. As the data show, there was a decline if you remove Illidan as a factor (which you shouldn't completely do, Illidan is very much a lore-based weighting), but not a sustained one, and after that drop we see the residual interest climb to its highest levels.

I said that I would agree with b. if there was a measurable observable change in interest numbers in Legion. There wasn't.

After Cataclysm I agree partially that there are too many factors to ascribe the decline just to lore, but I would disagree in saying that it wasn't a factor, and again, we can see the magnitude of a lore-based impact when we compare the Illidan-inclusive numbers, and the residuals after we remove Illidan as a factor.

The reason, again, that I did a classic to classic comparison is because it controls for those factors, and mostly isolates lore. The data in the response was only included to respond to those two factors, and to discuss why I don't consider them as compounding the classic to classic comparison.

__________________

Maybe if we want to measure the impact of the Night Elf portrayal we should take a look at max-level Night Elves in Roleplaying realms versus other realms?

US:

(percentage of level 120 Night Elf characters in comparison to the Alliance total in each realm)

Argent Dawn: 26.8% (second most played)
Blackwater Raiders: 27% (most played race in the whole server)
Cenarion Circle: 28% (most played race in the whole server)
Emerald Dream: 26.3% (second most played) (side-note: void elves are the third most played alliance race there at 8.1%)
Farstriders: 27.3% (most played race in the whole server)
Feathermoon: 27.9% (most played race in the whole server)
Kirin Tor: 27.2% (most played race in the whole server)
Lightninghoof: 27.9% (second most played)
Moon Guard: 25.1% (second most played) (side-note: void elves are the third most played alliance race there at 11.5%)
Ravenholdt: 27.2% (most played)
Wyrmrest Accord: 23.3% (second most played) (side-note: void elves are the third most played alliance race there at 10.3%)

EU:

Argent Dawn: 25.5% (second most played) (side-note: void elves are the third most played alliance race there at 6.4%)
Thorium Brotherhood: 26.6% (most played race in the whole server)
Shadow Council: 26.2% (second most played)
Darkmoon Faire: 27.5% (most played race in the whole server)
The Consortium: 27.6% (most played)
Ravenholdt: 27.4% (second most played)
Mithril Order: 28.3% (most played race in the whole server)
The Aldor: 25.2% (most played race in the whole server)
Die ewige Wacht: 23.4% (second most played)
Die Nachtwache: 28.3% (most played race in the whole server)
Kirin Tor: 24.2% (second most played)
Moonglade: 28.3% (most played race in the whole server)
Cenarion Circle: 23.1% (second most played)

Wait, so they're hugely popular in RP realms? Hmmm

__________________

I don't think that this is a viable comparison though for what you are attempting to achieve with the data. So much has changed between games. I don't think that you can simply isolate lore. We have also changed as a community. We have a much more solid grasp now about the game mechanics. There are websites that are much more efficient at explaining the optimal builds. There are guides for everything under the sun. Populations have come and gone. The culture has changed, which amounts to more than what people think about the portrayal of night elves. The factors for why someone would pick a race on the Faction realms does not necessarily amount to lore portrayal.

If I were to play WoW Classic, I would probably roll an orc so that I could re-experience the old Horde zones that had received fairly large updates with Cataclysm. I would be curious about whether the nostalgic charm of "Barrens chat" would still be present. A lot of my nostalgia about the old game is more oriented towards "Pre-Cata Horde" gameplay. I don't feel that the starting zones changed as drastically between Pre-Cata Teldrassil and Post-Cata Teldrassil as they did with other zones: e.g., Durotar, Barrens, Dun Morogh, Tirisfal Glades, Silverpine, etc. And that certainly would contribute to what I would choose to play for Classic. If I wanted to relive those dark days when hybrids were shit, then I could play a night elf druid, which was the main reason why I played a night elf back in the day.

@Krainz

1. Please link your source.
2. Presenting % of Alliance data only next to whole-population comparisons is misleading
3. This is an interest-over-time discussion. You presented me with snapshot-in-time numbers, which tell me nothing in terms of the impact of the lore since classic.
4. Even if you were to perform the analysis in 3, you'd be stuck comparing numbers on retail with numbers from expansions with vastly different systems, making the comparison useless. There's a reason I'm favoring a classic to classic analysis.
5. You didn't even present these next to non-RP server data. I couldn't even compare them to what the interest is on retail outside of RP - exactly what conclusion is there to draw from this?

Nothing here demonstrates a trend, although if you are interested on the impact on RP realms, go back to my original comparison, and look at the realm types where the decline in the Night Elf population playing classic/vanilla that I've discussed is hardest hit.

@Genesis

I've had an opportunity to dismiss some of these claims already.

The claim that the community is now a bunch of minmaxing supergeniuses who just know more about the system than the primitives in 2004-2006 doesn't make a ton of sense to me. For that argument to stand up, we should start to see this effect during Classic as people became more familiar with the systems involved. As I demonstrated earlier though, Night Elf numbers do not change year over year, and the residual interest after accounting for WoW's popularity does nothing but increase.

Other than that, you don't think people talked about or weren't concerned optimal builds or used guides back then? What would you call Thottbot, just as an example? (https://wow.gamepedia.com/Thottbot)

As for the rest, I'm not sure in what way you're arguing that your personal experiences explain the variations in the data that I've described above. Barrens chat is now the reason why people are flocking to Horde? What's going on with Tauren and Troll numbers then? Why are undead numbers increasing? While we're here, on RP realms, why are dwarves, gnomes, and humans increasing by 60%, 39%, and 22% respectively, while Night Elves and Trolls are decreasing by 53% and 33%, respectively? Does Barrens chat explain that?

It really can't and I wouldn't expect it to. But further, why should we assume that Barrens chat holds supremacy over other aspects of the game that people find nostalgic? The only way you can really backdoor your way into this is by pointing to the relative appreciations for the different factions and races changing over time - which is a lore based point and circles right back to what I'm saying.

But so long as we're using lived experiences as a guide to explain variances in the numbers, then it's only fair that I get to bring in mine. As I've said before, I've watched scores of people quit, walk away from Night Elves, or declare that they would play a Night Elf and/or invest in them if not for their recent presentation. As much as I hated it: "I used to love them" (the Night Elf achievement in A Little Patience) was an apt meme that captured the feelings of a lot of us quite well - even if it thumbed its nose at our pleas to fix it.

Don't take this as equivalence though. The difference between mine and yours is that mine explains the trends we're seeing in the data that I've been presenting.

__________________

Realmpop.

I can present a percentage of both factions, it won't change that Night Elves are the most played race at 120 in many RP servers.
Well, it tells me a lot of the current perception of Night Elves after so many years of "degradation". Especially when put side by side to the intention of creating a Night Elf in Classic.
Wait, so a 2019 to 2019 retail to classic comparison isn't relevant for the subject matter?
I won't present the Night Elf population server by server, sorry.

__________________

Sure, but not well. I'm not necessarily claiming that "the community is now a bunch of minmaxing supergeniuses who just know more about the system than the primitives in 2004-2006" and that explains everything. It explains some. It is one factor among many.

The claim that the community is now a bunch of minmaxing supergeniuses who just know more about the system than the primitives in 2004-2006 doesn't make a ton of sense to me. For that argument to stand up, we should start to see this effect during Classic as people became more familiar with the systems involved. As I demonstrated earlier though, Night Elf numbers do not change year over year, and the residual interest after accounting for WoW's popularity does nothing but increase.

I recall Thottbot. But we now also have WoWhead, Icy-Veins, YouTube channels, Twitch Streams, etc. that all talk about this data with a greater intensity and availability than what Thottbot represents. It's not the same.

It was a formative part of Horde community. There will be people who remember and want to relive it. There will be people who didn't who want to experience it. It is not the sole reason but I suspect that it forms one of many contributing factors.

They must be going down because of Night Elf lore.

I thought that I alluded to this before in my post? Tirisfal Glade and Silverpine Forest received MASSIVE reworks in Cata - not to mention Hillsbrad Foothills - so there will likely be some nostalgia for experiencing Pre-Cata Undead.

I was not aware that I was attempting to explain this with Barrens chat, Kyalin. Though please notice that I did list Dun Morogh as a starter zone that received a fairly large rework. You know what other starter zones received MASSIVE reworks in Cata? Westfall and Redridge. Remember the multi-continent plotline that revolved around the Defias, Lady Prestor, and the Black Dragonflight that mainly involved the humans? Do you not expect that there would be some nostalgia-based motivation that would explain why people would pick humans so they can experience anew the old pre-Cata zones for humans?

Because if you assume that this is what I was arguing, then you will have constructed a convenient strawman position for you to argue against.

Genesis, it's pretty clear, given the inconsistent quote tagging, your ridiculous personal attacks, and arguments that don't even try to address what I'm saying that you aren't here in good faith either. Is there some rumor circulating on the SoL Discord? Or does the anti-elf crowd have its head so far up its ass that it can't see straight?

The only possibly substantive thing you said has to do with revamped zones - and it ignores that Night Elf starting zones, and Ashenvale, were revamped substantially too. I don't know how you can even consider that discussion and ignore that - unless again, you're searching for something, anything, to explain away the obvious factor.

You can leave too.

__________________

Come guys, let's be mature and let Kyalin have his little echo-chamber.

@Marthen

There's a difference between discussion and what's happening here.

Although he frustrates me at times and at times I find his arguments lacking, I respect Krainz and I like him because he tends to engage in the former. He does his homework, he usually addresses my points, and he arrives with data. A lot of the time he forces me to reconsider my position, or to strengthen it as I did with the Google data. You can get somewhere from discussions like that, and both sides can walk away with more information than they came in with.

Genesis didn't do that. He simply reached for any contrary explanation he could throw up regardless of whether I'd addressed it or not, or whether it made any sense or not, and when that didn't work he dragged out Feltongue's ugly caricature, which I suspect is being passed around SoL's discord about elf-fans generally given that the last time I was there, it was nothing but its own echo chamber consisting of "kill elf" memes.

I go to all of this work to provide insight for those interested, and I wouldn't be saying it if I felt I couldn't defend it, but there are some people who aren't here to argue in good faith. They come in with flimsy, evidence-free arguments that violate Occam's razor, and then when they're challenged, they resort to insults, accusations, and non sequiturs - and they are not worth anyone's time after what could pass for substantive discussion has been dispatched. I think it's fair to say that you and Genesis are among them - and they have no place in a thread like this one.

If you want to prove me wrong, next time, be like Krainz.

__________________

This is rude and diminishes you.

__________________
Brought to you by Sanguine Enterprises.

My Worldbuilding:
http://bloodinkworlds.tumblr.com/

I hope you feel better after saying that, but if your point was to make me feel bad for sounding harsh towards someone who just outright told someone else to leave, that's not going to work. But worry not, I am done here one way or another. I was done before, but what was said to Genesis felt so absurd to me it prompted a response. But beyond that no reason to linger.

I don't even like Krainz but I felt I had to comment because I feel you and Genesis have been consistently uncharitable to Kya and it diminishes you. I say this not for me but for you.

Get your shit together, if you disagree with Kya and you've said your piece, then leave them alone, don't harass them for being upset with the game and taking the time to organize their thoughts, they're just as valid as your worldbuilding or any other detailed endeavor on this site.

EDIT: Piece not peace.

__________________
Brought to you by Sanguine Enterprises.

My Worldbuilding:
http://bloodinkworlds.tumblr.com/

I told Genesis to leave because Genesis flew off the handle and engaged in quick personal attacks rather than addressing the core of the issues under discussion. As I said, I have no issue in actual discussion, but Genesis was not interested in that. Given that, I will happily tell him to leave, because he has nothing to contribute.

__________________

It wasn't about justification (I am not going to open that can of worms), but simply finding the notion of telling someone to leave on a free forum absurd. But really, even though said harshly, I meant what I said, it's best to leave at this point, and like I said to Skytotem, I am on my way. You don't have to worry seeing me here again.

And yet here you are...

__________________
Brought to you by Sanguine Enterprises.

My Worldbuilding:
http://bloodinkworlds.tumblr.com/

I'm not on the SoL Discord. So your accusation is unfounded.

Teldrassil did not feel as drastic of a revamp as other zones. I did say that Darkshore received a revamp, and you are welcome to review my posts where I did mention that.

I don't think that your night elf lore explanation qualifies for Occam's razor. It certainly qualifies for "correlation is not causation."

Flew off the handle? This bit is it:
And you know what? I can retract that. The rest? Hardly. But hey this was only after you strawmanned me. But yeah, you and Mutterscrawl, can claim that I am simply being "uncharitable."

Well, I'm glad to hear that my concern was wrong here, and if we're willing to walk back the attacks, I feel we can continue.

On Teldrassil I agree, on Darkshore and Ashenvale I disagree.

We're slightly running into Hume's problem with scientific observation on this one (one that to this day has not been solved), but the reason I claim Occam's razor is because it is not reasonable to assert that the entire population changed without evidence, rather than that the factors influencing that population have changed. So far, we have discussed the following possible effects:

1. Marketing and Lore presentations of races have affected people's perceptions of them, and hence have altered their popularity. (The point that I am making)
2. The gaming population of 2019 is more informed about classic than actual classic players were.
3. There are more options in Retail than there were in Classic, specifically for playing Elves and Demon Hunters.

With 2 and 3 I set out falsifiability claims. 2 and 3 would falsify my position if in the case of 2, there was a negative trend concerning Night Elf representation and interest during classic, and if in the case of 3 there was either a negative trend concerning Night Elf interest after BC and that there was a positive trend concerning Night Elf interest in Legion. None of these conditions materialized, and instead, we see those trends moving in the opposite direction.

I am again not convinced of the idea, regarding 2, that we have changed significantly on minmaxing, which is something that was and is relegated to a small band of the population in the first place. The data existed then and the data exists now, and the technological developments of the last fifteen years have not unlocked new ways of looking at it. If you want to falsify that claim: explain to me how AI and robotic process automation give us insights that weren't with us in 2004 and 2005 - be specific.

You may notice that I didn't include what I will call the "nostalgic elements" claim - that's because I feel that it loops back to 1. There is a lot to be nostalgic for in classic, but for one set of nostalgic elements to surge above the others, we have to ask why, and what changed to the extent that it didn't produce a change between 2004 and 2005, but IS producing a change now - especially to the magnitude we are seeing, and especially given that as most Horde players can attest, the Horde had an objectively inferior levelling experience to the Alliance. That necessarily turns our attention back to 1, 2, and 3, and given that the trends speak against 2 and 3, we are left with 1.

Now I'm going to put out a similar challenge to you that I put to Marthen on the financial numbers. Before you post again, seeing as you are making a lot of assertions - come back with data to back up those assertions. I have done my homework, I would like to see you do yours.

__________________