We did something crazy. We made a fan game out of FFVII. Here, we talk about why and how. More interestingly perhaps, are some analytics and observations. If you want to see the game, click on this LINK.
Why did we make it?
The idea came about more than a year ago. To us game designers, itâ€™s interesting to try to adapt different ideas into games. One of the more common exercises is to translate an existing game into a different genre. For instance, how would you make Portal into a top-down shooter? What about making that into a RPG? Itâ€™s a really good way that helps you spawn new ideas. In our case, we wondered how a Final Fantasy game will translate into a 2D action side-scroller.
To be honest, we were already in the midst of developing our 2nd title. And we found that there were several elements in our game that could be easily be translated if FFVII was a beatâ€™em up. We were already implementing an elemental system to the playersâ€™ attacks, inspired by God of War: Ascension. We realized that the FF Materia system could be worked in that way. We also have a special attack which we eventually converted to the Summon system in FFVII. It was decidedly more elegant and simple compared to our initial implementation.
But we left the discussion at that. It was nothing more than a temptation to distract ourselves from our main project, Dusty Raging Fist. News came during E3 that FFVII is going to be re-made. Thatâ€™s when we said, letâ€™s do it. After all, FFVII was the game that made me save up and buy my 1st Playstation 18 years ago.
We already have a somewhat well defined process, having developed a previous game and in the middle of our second game. But still, in terms of workload, thereâ€™s still a lot to do.
Cloud and Tifaâ€™s animation
We have a tremendous amount of different animations for our characters, different attacks, idle, run, jump, hit, even miscellaneous ones like lose balance and summon. Cloud currently have about 750 frames and Tifa, 660. Sheâ€™s a faster character, so less frames for her. But even then, this is not much if you compare to our internal project. Each of the playable characters in Dusty Raging Fist has about 2000 frames or more. Itâ€™s because for this FFVII project, we didnâ€™t have theÂ full set of attacks for theÂ Cloud and Tifa. They donâ€™t need it since itâ€™s just a one level demo.
1 x Melee Soldier
1 X Ranged Soldier
1 X 2 Legged Robot
1 x Scorpion Boss
We did 4 new enemies as listed above. The annoying flying robot was recycled from our current project to add variety to the game play. Enemiesâ€™ animation tend to be shorter. All of them are between 300 to 400 frames.
Ifrit and Shiva were chosen to be featured in this demo. Both their animations are about 8 seconds, but they were mainly effects that were previously done, so in terms of art production, it wasnâ€™t too heavy.
In the demo put up, we did Midgar. We wanted to capture the essence of the level without killing ourselves. So it was important to get the mood right. The lighting and color scheme were vital. We decided to capture a few important parts right. For instance, the beginning of the game where Cloud alighted the train, the warning sign and the reactor. Also, thereâ€™s the boss fight area.
To shorten production time, we recycled some of our old graphics and rearranged them. I think we were able to get away with it due to the relatively fast paced nature of the game. No players were idling at one place for too long.
As you can see, although itâ€™s a 1 level playable demo, there are quite a lot of art assets involved. It took slightly more than 2 man months. And that is only because we recycled some of our existing assets and we have a pre-established workflow. I estimated that if we had to do all art assets from scratch, weâ€™ll probably need 6 - 8 more man weeks.
This is the part where we got away with most things since we have an existing framework. Most of the player controls were up, but we still had to customize quite a lot of details. All hitboxes had to be redrawn, parameters such as hit force, hit angle have to be meticulously keyed in and tested.
A lot of the AI needed to be done from scratch though, especially the Scorpion Boss.
We wanted to make sure that we incorporated some elements from Final Fantasy VII so that itâ€™s not just another generic hack and slash.
We ran through some notable features in FFVII. Materia, Limit Breaks, Summon, and your default RPG leveling stuff.
We were able to fit the Materia system easily into the genre by looking at them as elemental attacks. But we had a hard time with Limit Breaks and Summons as we could only come up with an idea for 1 more feature. At the end of the day, we felt Summoning was more iconic and memorable, after all, who could forget watching Knights of the Round being played in full each time you decided to summon them?
This is a short 1 playable level demo. You get to play as Cloud or Tifa, or co-op locally (you will need at least 1 controller to go with your keyboard).
Barret works as a fire support for you. He's on a timed recharged system. You can only call for him when there is at least 1 blue bar. This is based off a key mechanic in our 1st game, Dusty Revenge.
The green gem in the game allows you to cast your Summons. You charged it up by inflicting damage on enemies. For simplicity sake, Cloud is equipped with Ifrit by default, and Tifa with Shiva. Rightfully, if this is a fully developed game, players will be able to equip their Guardian Forces as they acquired them in the game.
As for Materia system, we built it into the players' attack. Ideally, they should be collected as players progress in the game. Again for simplicity purposes, both Cloud and Tifa comes pre-equipped with Fire, Ice and Lightning. At the end of a 4 hit combo, the Materia will be effected. Ice will freeze the enemies for X seconds for players to continue hitting them. And when the freeze ends, an extra damage is dealt to the enemies. Certain enemies areÂ immune.
Fire Materia creates a fire ball that moves slowly to hit enemies along its path. An explosion at the end deals extra damage.
Lightning Materia stuns the enemies. Any other enemies that comes into contact with an electrocuted enemy will get electrocuted as well, somewhat like a domino effect, but with a "taper off" to prevent abuse. Unfortunately, this feature is not fully developed for this demo. We may update in future if we come around doing it.
In the spirit of any RPG games, these Materia can be upgraded. Players accumulate experience points and can choose to upgrade various stats, such as Materia Levels, number of Summon Gems and Fire Support Bars, and even the playable character move sets.
I think with the Materias, Summons, enemies, boss fight and level, we made a pretty good homage and fan service.
So what happens after the demo went live?
To be honest, we never expect so much attention with the demo. We shared bits and pieces of the work in progress on Facebook, Twitter and Instagram, but it never really gathered much attention. And we thought that it would be the case and really, itâ€™s fine with us. It started as a homage and a fan service, and we were looking for a distraction from our main project.
But we were so wrong. Once we released the footage on Youtube, we can see the number of views rising steadily. We hit 30,000 views within a day. And that was when sites such as Venture Beats were writing about us. Our site â€œwww.ffvii-reimagined.com; went live shortly after. The spike really came in when someone posted a thread on Reddit. And once Reddit started talking about it, mainstream gaming sites started writing about it. It could be a coincidence, but that might show how influential Reddit is.
In less than 3 days, we have an astonishing 75,000 thousand users. Â The biggest hourly spike came when Kotaku put them their article. And considering that itâ€™s 2 days later, I would say Kotaku has a tremendous following.
Close to 60% of traffic to our site is by referral. Itâ€™s apparent that kotaku.com has a good reach. And I suspect itâ€™s partially due to the layout of their site. Every story gets a time limited front page treatment, while some other sites, stories are thrown into different sections. In some sections, the visibility is not there.
After about 3 weeks, the traffic is almost gone. We have less than 1,000 visits and itâ€™s constantly dropping. But we had a good run.
Social Media Presence
Social media, marketing, PR, the one thing we really sucked at. Our Facebook page has about 690 likes before the page went live. But we did gain a good amount of followers after that.
As mentioned, we didn't do much in terms of disseminating the information of the game. We shared a few images on FB, Twitter and Instagram, but there wasn't much of a response. But once we release the video of the game on Youtube, it kinda got viral without us knowing.
We previously only have about 230k+ views on all of our other videos combined. This FFVII reimagined video alone trumps them all.
As expected, views were the highest on the 1st day. With so many sites covering the news, it's Â expected. But the steady and swift drop of views is a big caution. This goes to show for your game, you really need to capitalize on the initial launch. Any screw ups at the initial launch could prove deadly.
Traffic sources to the YouTube video bears some similarities with the sources to our site. Kotaku is high up among the referrals.
In terms of geography, the top 10 countries isnâ€™t too far off between Youtube and Google Analytics. United States is a strong number 1 taking up more than a quarter. To be noted is that some of the non-English speaking countries contributed to a high viewership. Maybe itâ€™s time to seriously consider localizing for certain markets.
The one very inconsistent data between Youtube and Google Analytics is gender. On Youtube, 96% of the viewers were male. But for visitors to our site, only 54% were male. The male to female ratio was almost on par. Honestly, we find it quite odd.
So What Now?
The team was psyched to work on the FFVII: Re-Imagined project, despite being unofficial. We would pore over the original game and discuss what should be in our version. It was highly motivating and the team was more than enthusiastic. But we knew right from the start that is must be a fan homage and nothing more. It would be wrong to profit from the project. Yes we are crazy enough to build the game, but we are not stupid.
The initial idea was to have a bit of fun and distraction from our project. In the end, we achieved the objective and honestly, it re-energized the team for our own project.
Soon after we posted the video trailer, we started to see comments with regards to the IP issues, even though we clearly stated it is a fan-made and non-profit project. The comments made us nervous. At one point, we even considered taking the fan site and game down.
But still, the overall experience is a postive one. We are now re-energized and more focused on our own game. And because of the feedback received, there will be some tweaks and new implementations. We are replacing our convoluted and confusing Special Attacks with the Summoning System in Dusty Raging Fist. We think itâ€™s going to work much better and at the same time reduce workload.
We also realized that having a playable out there is a fantastic way of not just getting user feedback, but a way to catch more bugs. We watched many of the Youtube Letâ€™s Play, and honestly we cringed at some of those bugs players encountered. But itâ€™s a good way to do some user testing by releasing a demo.
So right now, we are hard at work on our internal project and are striving really hard to release our own demo soon.Â
We may update the fan project when we find time, if you are interested, be sure to follow us on our Facebook Page