EPTWe are designing a race for an RTS game. RTS games generally fall in one of three categories: Human vs. Human, Human vs. Fantasy Monsters, and Human vs. Sci-Fi Monsters. For our purposes, we'll choose a Human vs. Fantasy Monsters universe. One of the fantasy races will be an evil armies race (no reason to worry about the name of the race when we're doing such a simple model. For now we'll Warcraft it up and call them the Horde).
Each player can design five unit models over the course of the game. In this case, they have a goblin, a troll, an ogre, a dragon, and a demon. More dangerous models are more expensive, as well as requiring more tech to produce.
At the start, all models are bare. Each has base characteristics, which include base hit points, base mana points (if applicable), unarmed attack damage, unarmed attack speed, and unarmed movement speed. No units start with spells, and with the exception of the dragon, all units begin with ground unit status, and all attacks are melee. At the goblin tech structure, though, there is research for a battle axe, bow and arrow, and magic staff. The battle axe has slightly longer range than unarmed and does much greater damage, but also reduces the movement speed of the goblin. The bow and arrow is longer range, and costs less, but has less dps. The magic staff costs a great deal more, but enables whatever spells are available to the goblin caster unit.
The research costs some money on its own, but also increases the cost of building each new goblin. Goblins that have already been produced when the research completes do not automatically gain the benefit of the research, although there may be a way to upgrade already-built units to take advantage of new research for some cost. Additionally, researching one weapon locks off the others, so the player must choose between goblins with battle axes, goblins with bows and arrows, or goblins with magic staves.
Weapons are not the only research available to the goblin. The goblin can also research plate metal armor, which greatly increases hit points but reduces mobility further. Light armor improves hit points less, but also increases the cost per goblin less and allows for greater mobility. Magic armor increases the mana points of the goblin in addition to an increase in hit points comparable to the light armor.
The goblin also has mobility research. Some fairly cheap leather shoes improve base movement speed. Spiked shoes enable the goblin to climb cliffs, or to climb a tree and perch there, making it immune to melee attackers. Winged shoes give the goblin flying status, but are obviously the most expensive of the three.
Each of the five models has its own weapon, armor, and mobility research available to it. This research does not update all units on the field, but rather makes all subsequent units of that model cost more, and only updates existing units by some mechanic that costs money (e.g. they enter your main base structure for a brief time, pay the difference between their original cost and the new research time, and after some build time, emerge updated). Each research also locks off the other research in its category, so a player with battle-axe winged heavy armor goblins cannot have bow and arrow tree-climbing goblins with light armor.
All upgrades to units will be visible in some obvious way. Your opponent can tell by looking at a unit of yours whether it has a battle-axe, bow and arrow, or glowing staff, as well as what kind of armor it has. Barefoot goblins are noticeably barefoot, and goblins with leather shoes will move faster than goblins with spiked shoes, so scouting them will reveal what mobility upgrades they have. Wing-shoed goblins will obviously be scoutable as having flying status.
Now that a model has been constructed, we can draw a few basic conclusions. How many possible units are there for the "Horde" (FOR THE HORDE!)? Well, 5 models * 4 weapon types (including unarmed) * 4 armor types * 4 mobility types means 320 different possible units. Aren't too many units a problem? After all, Dustin Browder often talks about how Starcraft II is approaching the number of units BW had for each race, and going much beyond that would wreck the game.
Well, no. Dustin Browder might be right about too many units being bad for Starcraft, but that is partially because Blizzard aims to make every Starcraft unit distinct and serve a different function. Our model doesn't have too many units, because there is so much overlap in their roles. If anything, our model has too few (which is fine, because it's a model, and not intended to have the complexity of a fully fleshed-out game). In addition to overlapping roles, Many units cannot exist simultaneously; at any given time, a player can only have 5 models, and any units on the field that are not their most updated model are less-upgraded versions of their model.
The beauty of models is that even simple models can give a lot of conclusions. As such, I won't attempt to exhaust the information that the above model can provide. Rather, I will attempt to deal with some of these conclusions in future posts on this blog.
