As much as I do agree that it’d be fantastic to have more content on the FEU channel in general, I think that this kind of “regular random content” idea is a bit of a non-starter. I absolutely do not have time to curate every submission, nor does the rest of the staff (and no, I have no intention of handing over the keys to anyone who is not on the staff). I’m talking with the others right now about how we can expand this without creating too much more of a manpower burden.
4 Likes
Thanks for saying the same thing I did, but in a different way and a few slight adjustments. I don’t really have to say I agree, but I will.
In regards to anything that takes time, however, there’s a lot that can be done with heavy planning. I think it’s very very important that viewers have an idea of what they’ll get into with just a brief look. Perhaps a simple community graphic or, again, weekly teasers. If we can all collaborate as a community and not let everything fall onto the shoulders of one person, we won’t have the problem of burnout so seriously. This would of course be after everything regarding scheduling is finalized.
Hello all,
You may recognize me as the “fire emblem the last promise 2” guy. While I did this as a an inside joke
and something I could chuckle about with friends (there was a user last year in the premieres who,
without context, spammed the phrase “fire emblem the last promise 2” verbatim in the chat in multiple
videos; this is the origin of the joke, it’s completely unrelated to the ongoing TLP2 project), I didn’t really
anticipate starting a movement of sorts, or creating any real annoyance for anybody. So I apologize for
that.
However, the reason I did end up posting the now-notorious phrase in the comments of every single
video was to let the project creator know that I was indeed watching. As of now, less than 24 hours
after the last video was released, I have watched every second of every FEE3 video. With quarantine,
I do both my research and my coursework entirely from home, which grows rather glum; however,
FEE3 has been a very welcome bright spot in the doldrums of my life. So thank you all, both
project creators and volunteers. You are appreciated, and at least one user has watched your entire
work in its entirety.
As far as qualitative observations, I don’t have much to add that hasn’t been said already. However, as
I am a statistician by trade, I thought I could present a short analysis of FEE3 viewership, as this is
something quantitative that we can analyze. Therefore, I present an updated version of this report:
A few notes before I break down the results:
- Views/likes/dislikes (and duration, but this won’t change) are scraped from the YouTube webpage
at the generation time of the reports. Obviously, views can still pour in, and I plan to rerun the
report both about a week and a month from now to see what changes. - Sample sizes for many attributes are miniscule, and even overall is small. So take much of this
with a grain of salt. - There are strong outliers here that buck trends, notably Sacred Echoes (popular project, very long
length, very high viewership), Sacred Radiance (tail-end of event, but high viewership), and
Midnight Sun (posted on a different channel). - I didn’t know what the heck to do with the extra Hiraeth uncut video, so it’s excluded here as a
special case. - Lastly, as a statistician, I’d be remiss if I didn’t remind you this is observational data - we didn’t run
a controlled experiment here, so we can’t make statements about causation, only association.
So, without further ado, some preliminary findings:
-
Videos average about 600 views, with a median closer to 500. Again, views are still coming in,
but this is notably down from last year. -
Viewership decreases as the event goes on. Viewership is stellar Day 1, but falls of quickly, and
then shows a pattern of either remaining steady or slightly decreasing. Note that this may change
as views for recent videos pour in, but I doubt the strong effect of days 1-4ish will disappear. -
Mean video length is 32 minutes, with median 28 minutes. This means that 50% of videos were
28 minutes or less. The Interquartile range (middle 50%) of projects is 18-43. With 75% of
projects at or below 43 minutes, people seemed to a decent job in general of staying within the
length recommendation. -
Duration of video does not seem to hurt viewership. Views don’t seem to decrease with video
duration, but in fact may increase with it. However, Sacred Echoes seems to be driving this effect,
and when I regress views on duration excluding the SE datapoint, the effect is much further from
statistical significance. But it’s still slightly positive, not negative. Thus, duration may not inherently
hurt viewership - an engaging longer video can perform well. -
The 12 PM EST timeslot may give a slight bump to viewership. It’s not drastic, but this category
has a noticeable effect on increasing viewership. -
SRPG Studio projects and new FEE3 projects seem to get fewer views than the rest of the field.
These two categories each seem to be associated with pulling marginally fewer views. -
Effect of LPer seems non-significant; however, effect sizes suggest LPs by the project
creator may give a small boost over other LPers. Some LPers have greater average pull than
others (MK404 benefits from SE as his only showing), but the sample sizes are too small and
dependent on the project given to make inference here. However, other than MK404, creator LPer
seemed to have the most positive effect on viewership, although the effect is still non-significant
(effect still positive when excluding Sacred Radiance).
I hope this is of use to the community, or at the very least some entertaining tidbits. I thought it would
be fun to put my profession to use here. Again, I plan to generate the report again after some time
has passed and re-evaluate, but I thought I’d get this out so we can look at immediate impact.
If you are interested in the source code for this, feel free to message me. It’s somewhat of a mess and
I’m far from a computer scientist, but if you want it, you can have it. Additionally, if you’d like me to
scrape the data and give you a .csv of current data, let me know and I can do that.
Thank you,
fire emblem the last promise 2
38 Likes
Thanks Warpath for sharing. Diving into the data is fun and it’s great to see that we can observe the event in this analytical lens and find more degrees of objectivity. While I agree it’s important to remember the sample size and outliers create some inconsistencies, there are some bits here that are fascinating to me at least, especially since some of it is counter to the suggestions/feelings expressed throughout the event.
This chart shows that length isn’t the full story, at least if we look at total views as a success metrics (should we? hard to say. this is a broader question of what defines success for FEE3).
I don’t think longer videos are necessarily hurting, but they seem to be the biggest extremes one way or the other, compared to the shorter ones which seem to cluster more around the mean. Perhaps it is a stretch to say that some projects do better in longform, while others do not - regardless, we should use this as a guiding point to help give people direction on what they should consider when creating their videos.
This data will be helpful. Thanks again for doing this analysis.
11 Likes
PROFESSOR WARPATH
8 Likes
I’ve probably expressed this elsewhere, but bless you for this Warpath, as well for watching every single video. 
5 Likes
Is there a form or something to rate our FEE3 experience?
No form, but this thread is where you can share your feedback and opinions.
Okay, thx
Would it be a good idea to make one? 
I’d rather discussion and feedback be shared publicly so we can iterate on each others’ ideas and share experiences in one place so it’s easier to refer to when proposing changes.
4 Likes
I wanted to address the topics of content exclusion and potentially getting rid of FEE3 as an event from my personal point of view. Note that I don’t believe I am a common data point by any means; I just wanted to share my perspective.
On the topic of content exclusion, I believe that most off-the-cuff metrics could have absolutely cut my project off if I had finished it a few months later. I didn’t heavily participate in the community at all before my project concluded, and in fact my first post ever on the actual FEU forum was, well, my project. Again, I doubt that’s normal, but hard-and-fast rules will likely exclude some interesting stuff.
However, soft rules are even worse. Let’s be real, impartiality is not a thing that exists on this forum. Even if every judgment was made fairly by a committee of people who decide what gets in and what doesn’t and the event was the highest quality it could possibly be under the circumstances, people would still get salty. And that’s a humongous big walloping if. Mistakes would absolutely made, and no one would ever admit it, while other projects that should get cut would get cut and those people would assume that they had been treated unfairly even when they specifically weren’t.
With that in mind, I would argue very inflexible and objective exclusivity rules be implemented if that’s the approach going forward. I do think this will unfairly give weight to recurring projects, but that’s probably inevitable anyway so it’s the lesser of two evils.
On the topic of scrapping FEE3 entirely: I am not a very internally motivated individual lol. I became aware of the hacking community through Ragefest, and a large part of what motivated me to finish my hack was to get it presented on what I perceived to be the auspicious occasion of FEE3. People say you should do things because you want them to be good and for your own satisfaction and yadda yadda, but let’s put aside the pure art and get real for a moment. When you make a game, or really any form of media, it is inevitable that you will have your audience in mind. Ultimately, people strive for acknowledgement. We are not hermit martial artists punching trees on an island somewhere to perfect our bodies. If anything, I would say I came the closest to that description, and my motives, as I am laying clear here, are entirely “impure”.
I believe FEE3 might serve as motivation for some people, and I am a proponent of keeping some version of it around as an annual event purely for that fact. If that means it has to get more exclusionary purely because of workload reasons, then I would begrudgingly say I am for such exclusions, whatever they may be.
12 Likes
So, I’ve also been working on my own set of data for the event (though not nearly as detailed as Warpath, holy crap, all that data makes the statistician in me giddy, great work man) and I’d like to add a couple small things from my own set of notes.
- As Warpath said, the viewership as the event went of drops drastically, but I want to show just how drastically. Here’s a chart of the days averages for anyone who may not want to look through all of Warpath’s own data
now, to be fair as possible, a lot of Day 1’s views comes from the Nuramon animation showcase, as it is the highest viewed video of the event as it was last year, but even when I excluded it, 
day 1’s numbers still double every day except days 2, 3 and 5. With it included, it is easily doubt the value of every day and even triples the viewcount of days 7, 8 ,9, 10, 12, 13, 15, 16 and 17. The only videos at this moment (1:54pm AST) that pass 1k views are the day 1 posts, the Nameless Heroes, and Sacred Radiance, though Endless Nights is close behind at 911 views as well as Path of the Midnight Sun on Alfred Kamon’s channel with 987, though it should be noted Alfred Kamon has around 17k subs with the FEU channel has just under 500.
If you look at the top 15 most viewed videos, 
only 3 are from beyond day 6: Pokeemblem, which has the recognition of being a unique concept attached to another Nintendo series, Sacred Radiance, a demake of one of the most popular FE games (this isn’t meant as an attack on either project, both look like true passion projects, it’s just the reality that such concepts will naturally draw more views, certainly know I was intrigued by them) Path of the Midnight Sun, which as said before was on a channel with over 16K more subscribers then the FEU channel and, funnily enough, my team’s own project, Starlit Souls, which I can confirm we were busting our butts off to try to signal boost and promote so I’m not sure how fair it is to compare since I’m not sure how much other creators were actively working to signal boost their own projects.
In short, I feel like more needs to be done to promote the event, especially later on in the event when people may be starting to get burned out.
- With regard to videos that got the least amount of views, looking over the videos and thinking back on my own experiences with viewing the event, two things come to mind for me:
a) SRPG Studio projects. I won’t speak on this much as Warpath went over it, but it is true that SRPG Studio projects are all lower in terms of view counts. I don’t want to name names because it would make me feel bad, but 4 of the 5 lowest viewed videos are SRPG Studio projects and the highest one reached was 43/73 in terms of raw view count. If the event is going to include SRPG Studio projects going forward, keeping this in mind might be worth something.
b) thumbnail quality and direction. While I am grateful that we got proper thumbnails, I’ve noticed that a decent number of the lower viewed videos have weaker thumbnails, though this might just be correlation not causation. Again I would prefer not to name names to avoid making creators uncomfortable. I know this year the thumbnails were somewhat rushed, so perhaps next year if something similar is done next year with thumbnails there should be some direction and guidelines given out to make sure people provide the best quality thumbnails possible since a good thumbnail is an excellent way to boost interest in the video, as I’m sure was discussed in the FEE3 thread when thumbnails were brought up.
Those are the main two things I want to go more in depth on since like I said, Warpath did an amazing job with his data.
Edit: I realized that I forgot to include Path of the Midnight Sun in my Top 15 data, I’ve updated it appropriately.
11 Likes
So I kinda just lurked through this whole FEE3 but I’ve seen a few videos and have been following along with the thread and discord.
Overall I enjoyed the event, but there was just one issue I had.
I don’t really know whats happening, but in the recent year or so, people have been really antagonistic and there’s been a lot of controversy and fighting going on and its just annoying. People are getting way too upset over small things, and blowing small issues out of proportion. Like even if you have an issue with something or someone, you could do it privately, and if you want to make it public at least try to be respectful and civil. The FEE3 thread had so much unnecessary drama that could have been avoided. To me it’s just really disheartening to see our community be torn over something that should really bring everyone together. FEU is like a second family to me and I don’t like what’s been going on in recent times and I hope that everyone can get it together. From what I’ve been seeing on this thread, everything has calmed down and I hope it stays that way.
Pretty much everything else that I had thoughts on has already been discussed in great detail by others, so that’s pretty much all I have unless something new arises.
14 Likes
I think viewer retention has to be mentioned as something that was not measured here. Many people click on something to try it out, but often leave after a minute or two.
A specific video might be clicked on 4x more than it’s actually viewed, or 10x, or worse.
My youtube channel is very small, but it’s very easy to see that shorter videos on it have higher retention.
Percentage viewed (on average) for some of my ~30 min vids
Percentage viewed (on average) for some ~5 min vids
Considering the huge amount of feedback that suggests that even dedicated FEU members won’t actually watch vast majority of the FEE3 content, I would presume that many videos have rather low retention. I know I have clicked some videos and given them a like 
just to try and be supportive, with no intention of actually watching the video. I don’t know if others do that at all, though.
4 Likes
This is a good point. My script just looks at interaction count on the webpage to get views, so I don’t
have access to average length viewed and things like that. I could get this through the YouTube
Analytics API and query their database, but it was much easier to avoid all that and just scrape
basic things from the webpage itself.
As such, views should be viewed more as “clicks”, than a full watch. Still a good barometer for
interest, and likely correlated with a full video watch, but not a full watch in and of itself. I suppose
this begs the question of whether we want to focus more on generating “clicks” or close to full views.
2 Likes
Besides the fact that shorter videos will naturally have higher retention rates because there is less content to consume, I also question if we should value retention over total reach via clicks.
Is watching a full 5 minutes of a 5 minute video better than 30 minutes of a 60 minute video? I think retention is worth observing and it is a good callout, but I’d be wary of using it as a metric for success.
Even views and clicks, traditionally awareness metrics and seen as “success”, may not necessarily be what we should strive for - call me a radical, but are there other goals we should look to besides how many people wanted to watch the video? It is a good barometer for interest (retention is as well), but is that our main goal? I want to figure out what we should optimize for and measure as our north star metric to evaluate future FEE3 success. Maybe I am taking this too seriously.
I’m going to keep thinking on this and draft up a more proper proposal later to help guide future decision making. Thanks for the feedback.
I would consider Views, Retention%, and total time watched / length of video as important stats to consider. I wouldn’t throw view count out the window or anything.
If a 15 min video has a cumulative watch time of 6 hours, that is much better than a 30 min video that also has a total watch time of 6 hours. It might be a better measurement of retention than the actual Average Duration Watched chart that youtube generates, in fact.
I know I haven’t really been an active contributor on these forums at all and I may be a bit biased on some things because of submitting a project myself to the showcase. But I guess I’ll say a bit about how I feel.
First off, on the topic of exclusion from FEE3. I really don’t think time is a good measurement to base it on. The amount of work anyone can get done in a certain amount of time varies and excluding hacks based on development time rather than its own merits is too arbitrary and unfair to those who don’t have current long-running projects. As well, cutting out most new projects probably wouldn’t be good for overall attention to FEE3, as the majority of it would become just showcasing existing projects that have been seen before. I know I have a personal bias against this because my hack definitely would’ve been excluded for being too new but it just doesn’t seem like a good idea overall for adding appeal to FEE3. For viewers, or for creators.
Next there’s the topic of video quality and positive/negative attention. I think it should be an expectation if it is used going forward for LPers to confirm beforehand with the creators if they are alright with the way the showcase was handled. I can say personally, that I was not a huge fan of some of the editing choices made in the video showcasing my hack and seeing so much negative feedback on the video about it because of that felt really bad as the creator of the project. It would help with showcases not going along with the creator’s vision for the hack and also it would give an extra level of quality assurance for the videos being posted. I know some people who recorded did this but definitely not all and I think this would be a good step for everyone involved with the hacks. (If LPs continue to be a thing in the future of course, this doesn’t apply if they aren’t)
But yeah that’s my thoughts so uh, tell me how wrong I am I guess c: I am new here after all
3 Likes
I’m kinda conflicted…
As a viewer: I’d rather watch a 5 minute video of somebody giving me their project’s elevator pitch where they list any unique features and design choices.
As a project creator: I’d rather watch somebody play, give live feedback and (hopefully) enjoy what I’ve worked on.
Shifting gears:
I’ve only been around for 3 FEE3s but the event has changed enough over that time period. IIRC the only problem we had in 2018 was the event being delayed. A year later (and almost double the entries?) we ran into the problem of people tearing into projects during their premieres. Another year later (and again, almost double the entries?) we have the problem of people tearing into the staff, jumping on anything at any opportunity and the problem of low view counts.
The view counts being as low as they are kinda make it seem like the views were all internal. Which I guess would be okay if FEE3 hadn’t been such a hostile environment, it wasn’t a celebration, it was a battle royale.
The way I understood FEE3, was that it was first made to light a fire under the butts of the (very few) content creators and get them motivated while also drawing in new members to the community. Hacking tools have come a long way since then and everybody and their Roy has a project in the works (hell, it took me less than 2 weeks of my spare time to make my submission from scratch (I’ve got a full time job and renovations on the side)). There’s simply too much content now, most people aren’t going to watch every video, or even half of them.
This year the staff had a metric boatload of work, hats off to the LPers. But what was their work even for? The view counts suggest that it didn’t pull any significant amount of people in; the only things they have to show for their work are content creators happy for seeing their projects played, knife fights in the main FEE3 thread, and people complaining in the YouTube comments.
Maybe it’s time to retire FEE3 and focus efforts on fixing the community.
5 Likes