I've been working on an indie 3D Action RTS or "MOBA" and it's still in its design phase. I've done minor testing to see if the things I'm imagining are remotely feasible with my resources, and, while everything is adding up, I'd like to know based on the design do you think that this a game that you would like to play. I've likened it to a hero shooter because I want for it to be like the bridge between Mortal Combat and Overwatch. Essentially, a team vs team fighting game.

I've been interested in it purely because of how much I enjoy fighting games and overwatch though I don't much enjoy first-person-shooters. And yet I dont see many 3D team vs team fighting games unless they're the standard League of Legends MOBA. In which case it's more of a point and click game. According to my research, it doesn't seem like this genre is inherently unsuccessful but it seems companies are more interested in the standard MOBA.

In shooter games, I find what makes them fun as opposed to fighting games is the fact that you do not flinch. While in overwatch there are times when you are disabled, in a losing engagement you still have full control over your character to maybe make that one critical headshot or run away whereas in fighting games you'll be flinching or juggled where you can't do anything. It can be so bad that you just have to put down the controller and wait.

I've taken two approaches to flinching. Not all attacks cause you to flinch (like small projectiles), but when you are flinching you'll have two options. There's the Limit Break, based on the limit break of Senran Kagura Shinovi Versus and the Guard based on "Instant-Transmission" from the Dragon Ball Z games or "Subsititution" from the Naruto Ultimate Ninja Series. These are some of the most interesting responses I've found to flinching.

Limit break requires no skill. It's a highly reliable "Oh Crap!" switch. It's designed so that when under attack especially against multiple enemies you can use the limit break to flinch them (with an area attack) and maybe escape that engagement or counter attack. The problem is this attack does damage to you as well, and it does more damage to you than to your enemies. The limit break also has a painfully long cool down. Some up to 10 seconds.

Guard is a highly skilled ability. Whereas if you memorized or predicted when a certain attack will land you can pull off a quick guard that will stop the attack and usually lead into a counterattack. This is an ability you can use while flinching or just before an attack lands. Either way it requires strict timing within a 0.3 second window. This does not do self damage and usually has a 0.5 - 1 second cool time depending on the character. This means even while flinching you have something to put your mind to.

Some attacks cannot be interrupted by flinching and those attacks will tell you and will also have a visual cue that you have "super armor".

Combos are universal and I've designed it for PC gameplay. So the mouse controls combos which is like basic attack - basic attack - smash attack or basic attack - basic attack - basic attack - even stronger smash attack much like Vindictus Online's controls.

You have 4 health / energy bars to manage. I promise it's not as complicated as it sounds. You have Armor, Health, Mana and Stamina. Armor and Health are parameters for whether or not you're dead and Mana and Stamina are parameters for casting skills or combo attacks. Armor acts as insurance against being instantly killed by anything. Attacks will reduce armor first and then after armor is reduced to 0, attacks will reduce health. Most characters have 2000 HP and 2000 Armor, so if an attack does 4000 damage it will break your armor and then you will still have 2000 HP. Some attacks have qualifiers like they deal 10% damage through armor which may be important to know when fighting against certain characters. Poisons and damage over time effects like them will deal damage through armor because clearly you can't poison armor. Mana is used for casting skills. So summoning a fire ball requires mana, not necessarily stamina.

There are 5 classifications of heroes. Defense, Damage, Support, Ranged, and Flank. I should mention this is medieval fantasy.

Defense Heroes are the Front-liners, the tanks, the Knights. Damage heroes, are the melee damage dealers, the swordsman. Support heroes are the healers, the clerics. Ranged heroes are the ranged damage dealers, archers, mages. Flank heroes are melee damage dealers designed to disrupt formations and win 1v1 engagements and escape losing engagements, the theif. These roles merely define the skillset of your character, not necessarily the play style. My goal is to give characters abilities that allow them to approach many situations, and I'd rather not make characters that are direct counters like Genji to Bastion on overwatch such that it will be entirely your fault if you lose an engagement.

There are 5 game modes I have in mind. Siege maps have an attacking and defending team. Where one team escorts a high HP siege engine to the enemy castle. The attacking team must destroy the defending team's wall. It's best to do this with the siege engine, but anything can deal damage to the wall while the defending team attempts to stop the siege engine or eliminate attackers focused on the wall. Capture maps put players at a more symmetrical setting and require them to capture the same objective by keeping enemies off it and allies on it. Escort maps require one team to escort a low HP person accross a distance while enemies try to assassinate the target. Survival maps have no objective, but merely to eliminate all enemies. In these maps you do not respawn. Raid maps are any kind of PVE I may add in the future.

MOBA aspects are the fact that you can buy powerups as the battle goes on. These are little power-ups that can tilt the battle in your favor to counter a certain enemy or something that's causing you to lose engagements. You could buy a small damage reduction up to -30% or make it so your skills cool 30% faster. You could reduce the duration of croud controls on you for example. Your performance increases the amount of money you get.

I was under the impression that Halo and Team Fortress were the big 3D team vs. team shooter series, so maybe you should look at youtube footage for those.

I was under the impression that Halo and Team Fortress were the big 3D team vs. team shooter series, so maybe you should look at youtube footage for those.

Actually I think Counter Strike and DOTA are the biggest esports titles.

There's one technical aspect that you should really also address at the game design stage -- and that's multiplayer latency.

Fighting games traditionally have been designed around local multiplayer where there's no latency, allowing the competetive scene to go as far as knowing which frame of an animation they should push a button. At an e-sports level, that means there's people who can time their actions within a few dozen milliseconds of the optimal point.

For an online game like DOTA, you'll have to cope with around 100ms of round-trip network latency. If all the players in a match are in the same city as your server, then you could rely on <30ms... or if you want to support players from different countries within the same game then you may have to tolerate >300ms... but 100ms would be an average target to aim for.

This means that if someone launches an attack, they can see their attack animation start immediately, but their opponent won't see the animation start until at least 50ms later (likely more)... If the attack animation normally goes for 0.5 seconds (500ms), you can deal with this by making the attacking player's animation last for 500ms as usual, but speed it up on the defending player's screen so it only takes 450ms. That way both players will see the animation end at the same time (when the damage is dealt)... but the defending player actually has less time to react/counter than you originally designed for. This already creates advantages/disadvantages for each player depending on their ping, which is terrible.

One solution is to host the game on a dedicated server, and force all clients to run at a specific amount of time in the past from it (note that Counter Strike actually does this!). e.g. you can decide to make every player have 100ms of latency, even if they only actually have 50ms of latency.

This removes the advantage/disadvantage for network quality (except for players with >100ms) of latency, and means that defenders will always see the animations take the same amount of time. You can then design your fighting moves so that e.g. the attacker will see an animation take 500ms, but the defender will see it take 400ms. As long as all of your attacks/moves take longer than this standardized delay, then you can hide the network latency within the animations like this, and everyone will feel as if they don't have any lag.

However, if you design moves that are extremely short / instant, then there's no way that you can hide the network latency, and players will feel like the game is laggy when having intense fights.

This isn't as much of a problem for traditional MOBA's because they don't have very much direct control over the player or twitch/instant hits/dodges. However, if you want it to feel like a fighting game, then you're going to have to adjust the design to suit the technology. Having a good network programmer working alongside you as a designer will go a long way here to making something that feels great.

I hadn't thought of these methods for lag compensation, thank you. I did take a brief look at multiplayer server technology and noticed their methods of client side prediction and server side reconciliation to account for lower latency.

Many of the attacks designed are 750 - 1250 ms. I want to do longer attacks because it's a little more convincing from an animation standpoint that it conveys the weight and power of the attack and may allow players to anticipate the hitbox. I studied a bit of the animation in the Monster Hunter series and the way they convey different attacks by slower, weighted movement. And I find that human reaction time is best with 1 second or more to recognize the situation.