Anyone with a vested interest in Project M would likely be aware that while Roy and Mewtwo were added into Project M, more characters were meant to follow later. Knuckles eventually made it into Project +, but there are several others who simply never made it in.
When we were selecting potential newcomers, there were many considerations, and I will list some of the more important ones:
1. Moveset:
An interesting concept and playstyle was an obvious thing to think about when implementing a new character. What would the character bring to the roster? Does the character fit an under-represented niche? Did we have some inspiration for a potential moveset? There was little point in adding a character if they were not fun to play as. Patching up missing playstyles in the roster, or creating a brand-new playstyle was something that could influence our decision.
2. Popularity / Series representation:
Does the character hail from a series that has no / few reps? Would people be happy to see the character in the game? Was there any demand for them? For example, there was a large demand for Isaac in the Smash ballot for Smash 4. That was a big reason for a serious consideration.
3. IP already in the game:
We had no desire to gain the ire of overzealous publishers who would strong-arm us into ceasing development. As such, we limited our choices to characters already in the game, such as assist trophies, stickers, etc... More companies involved, means more of a likelihood of painting a target on us. As such, we tried to stay within IPs already featured in the game, since we had not been targeted by any of these IP holders at the time.
4. Feasibility:
When the clone engine was first being used, there were many limitations that aren't quite as prominent today. We also were considering feature-creep as a potential roadblock. The game could not be in development forever. There needed to be a definitive release one day. We needed to consider how long it would take to make said character and how difficult it would be to make them and balance them. We also had to consider limitations at the time. If we had an easy model to port it over, (such as Roy over Marth) that had all of the essentials, like articles (projectiles, effects), physics-affected features, etc... In other words, the more reasonable to do, the better. None of these characters were easy per se to add, but some were easier than others.
There were other considerations, but there were especially prevalent. Roy was quite easy to implement compared to say... Mewtwo. He has a similar body type and animations to Marth, we had the idea of taking what made him unique and expand on those features, he was relatively popular, and Fire Emblem was under-represented at the time. He also did not require any projectiles, and adding a fire effect was quite a reasonable ask.
Mewtwo was more demanding. He was cloned over Lucario who had everything required to make Mewtwo work, but it demanded a lot more editing from the source character. The one advantage was Melee's Frame data and animations actually existing. The reference material did help in terms of feasability.
Knuckles was a semi clone of Sonic, and the moves that were added were more physical in nature. They were simpler, and definitely reasonable to implement. Knuckles also had a lot of motivation behind him, and I would take a gander in saying that this relative ease of feasability compared to other options under consideration very likely helped in propelling progress on the character. It also helped that Sega approved of the addition of Knuckles through their social media. As for being an existing IP in the game, he was a sticker. As for the moveset being of interest, the potential of having a Sonic-style character that is slower, has more of a bruiser moveset, and that has unique mechanics such as gliding and wall-climbing.
Lyn was actually considered for many reasons. She was essentially the Marth of North America. For many people, the character of Lyn was their first introduction to the Fire Emblem series. While we were not trying to make a political statement, there was the feeling that there were not many female characters in the roster. Furthermore, the female characters in the game tended to be very slow and floaty in general (Notwithstanding Zero Suit Samus or Sheik), not providing many alternatives for people who may have wanted to play as female characters. Some ideas for the character were to use aikido as inspiration to make Lyn into a waller type character that could also sheathe her sword between attacks. Wallers tend to have precise attacks that can neuter approaches. For example, if you try to approach from above, your character can slash diagonally upwards, if they try to attack from below, you can slash diagonally downwards, etc... think Johnny from Guilty Gear. Do remember that I was not in charge of this character, and I have no idea if the members behind this character had other ideas they did not share with me, but these were ideas put forward. She would have had a command arrow shot influenced by where you tapped your stick as well. Tapping a button at the right moment would sheathe her sword, and the next attack would be stronger. If she stopped attacking for a bit, the sword would sheathe itself. Command-Sheathing the sword would require an animation, so it would interrupt potential comboes, but in exchange, you would get power for the next hit.
Sami, I knew the least about. I know that she was going to be heavily-inspired by Snake, but would be more up-close and personal. A machine-gun from the infantry units in Advance Wars was considered as a special. The idea of a rocket jump for a recovery was also thrown around. Sami was considered over Andy mostly due to her role as a commander in the games. She was an infantry specialist, so it was surmized that she would be at home fighting on foot. One important idea in using the clone engine was to provide alternative playstyles, and a commando CQC movest that was less reliant on putting distance between yourself and the foe fit the bill.
Now, finally, we get to Isaac. I cannot say that we had completely agreed on a direction for him considering how we did not get too far into his development, but I can at least attest to the ideas I had been putting forth.
His codename was Vlad the Conqueror, as a reference to his name in French versions of Golden Sun. He was going to be, by far, one of the most ambitious project we would undertake, and there was no guarantee that Isaac would even get finished. The reason is that at the time, Isaac was going to be a big time-sink. His moveset was going to be completely different from anything present in Smash so far, he would have required a brand new model since the assist trophy model was far too low-res to use (The same could have been said about Lyn as well). We were intending to add subtle sun-inspired designs to the seams of his clothes, and detail him further. The sword he came equipped with was merely a placeholder, and we had intentions to make him use the Sol Blade, albeit redesigned as we felt that the low res reference material might not translate to a game that is far more-detailed.
These are from an early design doc I had written. Those are opinions I held at the time on how we could proceed. When proposing new characters, we were asked to find simpler, more reasonable solutions to more ambitious ideas. That is because there was no guarantee that they could be achieved. This is why many moves shown below have multiple options to choose from. This was more of a… proof of concept more than anything. Characters change massively during development, and there was no guarantee that any of these ideas would have even been used at all. Those were my interpretations of the character, but I know other members had their own.
---
Isaac:
Isaac’s core concept is reminiscent of traditional fighting games, heavily inspired by the likes of Ky Kiske and Jin Kiraragi from Guilty Gear and Blazblue, respectively. His core physics are similar to those of Link, but he has a wide range of magical-based attacks that can be delayed in order to capitalize on an opponent’s habits. He also has a disjointed grab, which will require a mechanic to offset his weakness to pressure. It will not necessarily render his weakness null, but will instead be a tool for him to fight it off differently in spite of his weakness to it.
Delayed Projectiles :
Isaac will have access to various forms of psynergy. As the default class for Isaac tends to be heavily Earth-inspired, I suggest focusing on Earth psynergy as the majority of players have a tendency to use the default elemental builds on their first playthrough of Golden Sun. While not mandatory, I think that the essence of the character ought to represent the experience of the player with said character.
When using a special move, Isaac will have options to delay the activation of a projectile. The key reason that would drive Isaac to want to delay his projectile is to set up traps, as a fake out, and to potentially set up comboes by controlling a small area of the stage.
Initially, I thought to use a Smash versus tilt input, not unlike Samus’ missiles, but this kind of input would leave neutral B without an alternate version. Instead, we can make the input dependent on how long the button is held. Tapping the B button lightly, much in the same way as what we do when initiating a short hop would initiate a faster version of the move (I will refer to it as a light version of the move), while holding it any longer would initiate a delayed or different version of the move (which I will refer to as a heavy version of the move).
Examples :
Neutral B : Ragnarok / Odyssey
Light : Ragnarok : Summons a fiery earth blade over Isaac’s head. Falls diagonally. It will pop the enemy upwards, slightly toward the user for aerial comboes.
Heavy : Odyssey : Summons a magic blade where Isaac stands It is delayed in place, and will only have a damaging hitbox when it is on the move. After a brief pause, it will travel forward, horizontally. It can be dash-cancelled at the end of the casting animation. The delayed projectile will allow Isaac to run alongside the blade, let it past him, or be used as a pressure tool.
Side B : Move / Force
Light : Move : Sends a magical hand (much smaller than the assist trophy, about Isaac’s height, to push the enemy away. It should not offer so much resistance that someone running at full speed will be completely pushed back, but it should push back someone jumping towards you from neutral. It should be used to hinder recoveries, mess up approach timings, etc…
Heavy : Force : Instead of a push, this is a fist that actually has a hitbox and actually does some damage. Not necessarily good at killing, and weaker than Odyssey. Faster travel time, however.
Down B : Spire :
Light : Stalactite summoned above Isaac, diagonally, when grounded, or in front of Isaac (aerial). Falls almost immediately.
Heavy : Same, but stays airborne for several frames before falling.
Up B : Retreat :
Light : Teleport, not unlike Zelda’s. Short range.
Heavy : Longer range, but not supposed to be very long. His recovery is not meant to be spectacular.
Disjointed grab:
Distance grab that cannot grab at close range. A large psynergy hand grabs the foe, and Iaac interacts with it at times.
Pummel: hand crushes foe, dealing damage.
Down: Pound psynergy. Slams into the ground, then launching foe in the air. If no ground is available, sends downwards. Meteor cancelable.
Back: throws foes at Isaac. He then swings with his blade. Almost baseball-like. Or throw towards him.
Forward: Isaac jumps in for slash or throws away.
Up: Throws up.
This mechanic prevents him from shield-grabbing. However, it will increase his mid range game. I fully expect very unbalanced initial builds.
Burst shield :
Loosely based off the granite Djinn, but also inspired by 2D anime fighters. Isaac will have a long-range grab, and as such, will be very weak to characters that can dish out relentless pressure… perhaps too much so.
2D fighters often have a burst mechanic meant to stave off pressure. It allows players to get out of being comboed to death when in a tight spot, but it is also something that you spend… a sort of trump card.
Isaac's shield system would therefore need to be re-imagined. In Golden Sun, there is a very useful Djinn called Granite, which greatly reduces the damage the party takes for a turn.
The shield would be solid, like Yoshi's. It would be made of stone. The size would remain static. It would crack when taking some degrees of damage (3 stages?), and darken. The shield would have the highest Health in the game, but also recovers its health at a much slower rate.
You will have the option to press B while shielding. This will prompt Isaac to break out of the shield. It will cause a large hitbox to come up, that will do a small amount of damage (or perhaps none). The knockback would be fixed, send the enemy back and a bit upwards (diagonally, preferably) so that it does not gimp the foe and does not kill. Multiple hitboxes could also be implemented so that if you are attacked from behind, it sends diagonally behind you for instances.
Using a burst will drain your shield to near shattering point. At this point, you will lose the ability to burst until your shield has fully recovered. Considering shield recovery is slow, I would not expect Isaac to use it more than once per stock… maybe twice if Isaac manages to survive long enough. Using your shield in the burst recovery phase will of course delay the recovery of the burst, since the shield must reach 100% HP to reset it.
Why the limitations?
A spammable burst would be grossly unbalanced. The point of the burst is to give Isaac an option to fight off pressure, but not in a way that trivializes that weakness. The burst requires resource management. Using it is a calculated risk.
Taunts:
One idea would be to include a Venus Djinn as part of a taunt, similar to Link's Navi taunt.
Another would be to have him cross his arms, and then cause a shockwave on the ground, tripping foes and doing 1% damage. Imagine DBZ style flexing of power.
A third would be interacting with the psynergy hand, a high five, for instances.
The Scarf:
As silly as it sounds, clipping will be a problem. Lyn's hair has caused many headaches, but her hair at least flows behind her. Isaac's scarf is in front of his chest. It will therefore clip into his body with the current physics engine. In the event that fixing this becomes too much of a hurdle, I propose that his intro animation has Isaac adjusting it behind him, a flip. This would be a good compromise between showing Isaac's original design and actual feasibility
Design:
Shadic has already been adamant about wanting to work on Isaac, and he has extensive enough knowledge of the series. Having led Link development, he seems suitable to the task. He has been leading development for characters like Link, who have a bevy of projectiles.
StarWaffle is already working on a model. He will be based on Toon Link's file due to his access to multiple projectiles, but we will use Ike's bone structure, made shorter. Ike's bandana tassels will be reworked to serve as a base for Isaac's scarf.
As for the sword he would use, the Sol blade would be the most iconic, followed by the Gaia blade as a potential second choice. More detailed models can be found in Golden Sun: Dark Dawn. My understanding is that StarWaffle is going to give Isaac a placeholder sword for the time being.
We cannot use the Assist Trophy model since they are far less detailed than character models.
We were thinking of adding subtle fabric details to upscale his model. Refer to FireballStars' recent drawing.
Other possible moveset references:
Jab could be two sword slashes followed by a magic palm strike. Earth shattering crack.
Down smash could be inspired by the earthquake psynergy, and act similar to Wario's Down smash.
Forward Smash could reference his jumping crit animation.
Another idea is to have Flint or even Echo Slash, which would work similar to Link, but faster, less damaging on first hit, but more damage on second (the speed is admittedly a way to save frames of animation).
Up Smash… I have fewer ideas for this one. Perhaps something inspired by Lucas. Perhaps Isaac summoning rising stones from the ground. Gaia might be a good psynergy to reference this on.
Forward air could be echo slash. One forward slash similar to Roy's or Marth's, but with a less powerful hitbox. A second magical sword trail would appear as an echo and deal damage right after.
Down air could be Gravity. Isaac slamming his palm down and creating a ball of dark energy that spikes or meteors downwards. Perhaps more powerful if hits in middle of hitbox
Back air: I imagine a hand behind him and an explosion of magic. Powerful, a kill move at reasonably high percent. Not very fast, but good if timed.
Neutral air should be good at staving off pressure, and hit on all sides. Horizontal spin slash? Rocks floating around Isaac in a spin? Sex kick might also be good.
Side tilt: a magic palm strike to distance enemies to reset neutral and put in mid-range?
Alts and colours:
For colours, I would go…. Default, and costumes based off party member colours schemes. Garret, Ivan and Mia would conveniently serve as good team colours too. Then one based off Felix with brown hair, and one based off Saturos with blue skin. For alts… perhaps the Dark Dawn trench coat, or Matthew's clothes?
---
As you can see from that design doc, I did not include a djinn system and put it in the moveset. I had considered an entire summoning system that would essentially let you use Djinns to cancel a move into a stronger attack, possibly using A+B as an input, each being tied to a specific Djinn. One would have high damage (Flint), but would lower your damage output after use. One would poison the enemy (Bane) but also make you slightly slower, one would heal you a little (Sap), but also reduce your launching power, and one would ground or meteor the enemy if they are airborne (Petra), but would reduce the max height of your jumps.
Now, the reason I thought this was hard to even pull off is because of the idea that came after. Using an A+B input a second time would have then summoned one of the four Venus elementals, depending on how many Djinns were used. This would have required an insane amount of work, so I did not even deem it worthy of presenting to anyone. This was not thought through very much as a result.
The only other element to this concept was that using the summons would put you on a timer and then your stats would gradually return to normal in the order the Djinns were used. This was meant to replicate the in-game RPG battle system.
However, there were so many potential ideas involved, that it may have been unreasonable to ask that if the team at that point.
As for my personal motivations, they were very similar to the rest of those who were involved. They liked Golden Sun. A small part of me hoped that by adding Isaac into Project M, it would help spark some discussion on the series, and perhaps create enough buzz for a revival, but that was just a little bit of wishful thinking on my part. Of course, liking a character wasn’t quite enough. We were making this game for an audience too. Still, I am sad that we have not gotten to finish the character, as I am sure that with current tools, we could have done something great.
