The Trajectory and Velocity of 5.0
We're deep into the planning process for 5.0, which will be the most significant version of the game since 4.0's release, nearly 20 years ago. Planning out a set of changes of this size and scope involves examining all of our existing systems, thousands of changes which have been proposed over the decades, and imagining things that we've never had the werewithal to attempt before. From there, we're seeking to refine our strengths that make Materia Magica fun, shedding ideas which haven't realized their potential, and modernizing Materia Magica for contemporary players who are looking for different things out of games than we might have when we were younger. With that in mind, we plan on really digging into 5.0 ... Soon. To minimize surprises and focus on stability and long-term happiness, 5.0 won't be one enormous launch of years worth of work, but will instead take the form of a ~year's worth of reboots, rolling out changes as they're ready, and refining them over time. Our goal is to let you see changes as they're developed, to surface issues and integrate feedback in realtime, so we don't have an avalanche of new stuff, followed by an immediate avalanche of reactive updates. Over the many months that this rollout occurs, we want to release changes in such a way that you get exposure to new ideas and mechanics in a semi-stable environment- as core mechanics within Alyria change, you'll get to experiment with new mechanics in unique ecosystems. Maybe you'll get to play with new Threat abilities before the new Boss logic takes advantage of it, and that will have novel, temporary implications for how combat works that reboot. This will reward exploration, imagination and cleverness, and we're excited to see what people do with it.
Threat & Tanking
Tanking in combat, and specifically determining who will be the next to tank when something happens to the current tank, is obtuse and not particularly fun to discover or work around. With that in mind, we plan on implementing a new Threat stat, which will effectively be the inverse of the Combat-Inconspicuousness stat. One of the design rules around stats is that more is always better- you might not know exactly what a stat does in the heat of battle, but you'll at least know whether a given affect is helping you or not. Threat and tanking and avoiding combat will be more value-neutral soon, and having more or less of it won't be implicitly better than the other, depending on context and the intentions of you and your form.
Threat will be generated in a bunch of ways, and will be able to be modified by players in a number of ways. Specifics aren't there yet, but it's likely that warriors will have options for generating threat and thieves will have options for reducing it.
Second Defense Isn't Changing
This is the part that everyone's going to read first, but read these things in order.
Valks' second defense has posed a large design challenge for the majority of 4.0. Being able to reduce melee damage taken by 2/4 instead of 1/4 is obviously huge, and makes second defense one of the most powerful skills in Alyria. Over the years, there have been a number of balance regimes- the rock-scissors-paper dynamic of web/slow/2nd defense was sort of silly, and was tweaked many, many times, but mostly in ways that papered over the problem instead of actually addressing it. Later tweaks, such as the splitting of slow into slow/hinder, and then each into slow/hinder (regular)/magic, did little to fix the issue. Changing combat tempo, in either direction, also did little.
At the end of the day, second defense's functionality is very straightforward due to its lack of variables- tweaking it to be more powerful or less is almost exclusively a question of "how many defenses do you get". Does 1 defense give you 100% chance of defending, and 2 defenses give you 200% total? What if we nerfed it so that 2 defenses gave you 170% total? What if each successive defense consumed an exponentiating amount of stamina? What if, what if. None of these is the right solution, but we keep looking for a solution because this has been an insurmountable design problem for all of 4.0- valkyries are just disproportionately amazing at tanking, to the detriment of all other classes' balance.
With the advent of threat and tanking, we have a neat solution: valkyries stay the same. Second defense is okay, and we just ... Leave it alone. Maybe cavaliers can take someone else's threat temporarily- maybe lances generate a disproportionately high amount of threat, temporarily- maybe barbarians can force aggro on themselves. And maybe valkyries aren't very threatening, and are disinteresting for bosses to attack, because they're so good at tanking.
This is an elegantly lazy solution in that valkyries can continue doing what they do best, tanking while solo, but have fewer options to force their opponent to actually attack them. Those hypotheticals are all still strictly hypothetical, and we're not even yet to the point where we're ready to play with the devilish details, but hopefully this provides an idea for the directions we're going, and the scope and novelty of the problems and solutions we're examining.
Weapons have 3ish static characteristics, and 2 dynamicish characteristics. A weapon is defined by its weapon type, and each weapon type has a default amount of damage that it deals and accuracy it grants, respective to its level. These three values are tightly correlated and basically immutable. A weapon's damage type and its special weapon attributes will vary, but each weapon does have a default damage type based on the characteristics of the weapon- a dagger will always pierce, by default.
The baseline damage variance of weapons is about 1-2x, and accuracy ranges 1-2.2x. Due to the way that accuracy and damage stats scale, it's much easier to make up for 10 missing accuracy at archon than it is to make up for 100 missing damage. This ensures that the high-damage weapon types are almost exclusively the only one used. It could be possible that different weapon characteristics could provide diversity- for example, none of the highest damage weapons can be used to backstab, or martial arts, or be thrown. Unfortunately, those characteristics can never be appealing enough to make up for dealing an extra 100 base melee damage, multiple times per round. This ensures that despite having more than 30 weapon types, only about a quarter of them are commonly used.
For 5.0, we want to shake this up.
Instead of ~2 weapon attributes (damage, accuracy) mattering when choosing a weapon's type, there will be 4: damage, accuracy, range and speed. Every weapon will possess each of those 4 traits, with values correlating to 1 of 4 ordered values (max, high, low, minimum). For example, two-handed axe would have max damage, high accuracy, low speed, and minimal range. A yoyo would have max speed, high accuracy, low range and minimal damage.
Currently range sort of matters: if you're in a formation and your position within the formation is too far from your opponent, your weapon just doesn't work. Ranged weapons have always had complex interactions- your functional accuracy when two forms are fighting takes into consideration respective formation positions, formation structures, the number of people standing in front of you and your opponent, etc, etc- and these interactions were so uncommon, and so poorly communicated, that nobody actually ever noticed that this happened.
Range will soon play a bigger part. Your weapons' range will always come into play, and you will have an effective position relative to your opponent even if you're not in a formation. You won't have to manage your position, as it will automatically be managed based on the weapon you're wearing. If your weapon outranges your opponent's, your opponent will suffer accuracy penalties for every unit of range they're missing.
'Speed' is a misnomer in progress- it's too easily confused with concepts like attack speed and defense speed, and it's rather different from each. Instead, this fourth characteristic represents variance.
In 3.2, combat was chaotic and unpredictable; you could get 7 attacks from nothing but skills, and you could use each of your defenses (up to 6?) Once per round, with each succeeding up to 97% of the time. Each successful hit dealt a huge amount of damage, but that was mitigated by the fact that any given swing would only hit about 2% of the time. This resulted in PK combats which lasted significantly longer than today's combat, even though it only might take 8 hits to kill a maxed hero. 4.0's combat system was a dramatic shift from that- If you deal 200 damage with your weapon, your weapon will always deal 200 damage to your opponent. Lots of modifiers are relatively static- sanctuary will almost always be on, and will always reduce damage by 15%. 100 AR will always reduce damage by 75%. Any variance that exists comes in the forms of skills and affects which don't always trigger- enhanced damage, active skills, item scripts, buffs and debuffs.
The two approaches are dramatically different, and while they hugely influence how and why stats and affects and classpaths matter and differ, one isn't necessarily better than the other. A fantastic game could be designed around either- and has been, around each, over time. For 5.0, 'Speed' will represent damage output variance. A 200-damage weapon with minimal variance would behave just like weapons do now- if a weapon says it deals 200 damage, it always would deal 200 baseline damage. If a 200-damage weapon had maximal variance, it might deal 180 damage 50% of the time, 220 damage 30% of the time, and 600 damage 4% of the time. Instead of combat being predictably orderly based on initial numbers, sometimes novel things will happen, lucky shots can occur, and playstyles will change to accommodate that.
The goal of changing these weapon characteristics is to provide clearer identities to the underappreciated weapon types, make the weapons that different classes are granted more meaningful, and to grant non-fighter classes weapons that are usable in ways beyond pure damage output. Weapon choices will become a facet of playstyle, and many options which currently exist but would never realistically be made, will be able to.
One of the primary goals of 5.0 is to leverage our strengths, and giving adventurers meaningful ways to actually engage with our enormous amount of content. We won't necessarily make more weapons, but make the weapons we have more interesting. We don't need to add more defenses or attacks, but make the ones you have more interesting, easier to learn, and more compelling to make decisions about. Succeed the current tanking model with one that people can use in ways that make sense. It'll be a lot of work, but we're really excited to start working on it, and we're glad that you've stuck around to see it through with us.