Building a better strategy game engine
PostPosted:Mon Feb 17, 2003 2:45 am
<div style='font: 12pt ; text-align: left; '>If you've been paying attention to my general comment on strategy game you'll probably notice I think most of them are hopeless broken/suck/etc. I've been brainstorming some ideas and I'm interested what you guys think.
1. Strategy game should ever be technology based. Otherwise it's not a strategy game. It's a game of whoever researches the fastest. Strategy game should be fought between units of comparable parameters. Master of Orion is probably the classic example where a fleet of arbitrary size and less than say about 50% of the tech tree has no chance of even defeating Antaran Raiders just from inability to hit the Antaran's Interphased Drive-powered ships.
2. The combat model should be simple. Anything complicated just makes it less likely to be balanced.
3. Any kind of healing, whether you want to call it magic, repair, or whatever, is a very very very very bad idea, especially in any game where experience matters.
4. Focus fire on a grand scale battle should yield no appreciable advantage over random firing. If anything, it should yield less. Focus fire is the lifeblood of RTS game strategy and it's wrong. It makes absolutely no sense how your 8 marines can do telepathetically communicate and all hit the same target at the same time. Besides the communication issues, your units should be taking extra damage if all they're ever doing is focusing on just one unit. Too many games mistake something like say they have 15 and we have 15, but we cleverly manuever so 15 of our guys is attacking 5 of theirs to mean that the ability to select all your unit to attack the same unit is a valid form of strategy. The easiest way to solve this is to give all units unlimited retaliation attacks.
5. This goes with healing, but units with experience should not be able to recover to full strength. We're not dealing with cloning here. You start with say 10 rookie guys and through the battle 5 of them becomes veteran and fight better. That's fine. But what's not fine is that those 5 veterans gets repaired/healed/whatever and then becomes a full strength 10 unit again. When you have two of those super elite units the point is that even though you only have 2 of them, they can function as say 4 units, not so that you can heal this 2 strength super elite to a full strength 10 so it can instant kill any rookie units.
6. Battles needs to be *slow*. Also slow battles must not be effected by focus fire or any type of healing otherwise you end up with something like Warcraft 3 where a few number of big units can stay alive nearly forever. This is because if battles are too fast-paced, the first attacker has way too much damage. It's not uncommon to see a console SRPG game where you can destroy half of the enemy army on your first turn. Rock-paper-scissors need not to be the only way to go. A tank should not always stomp on poor infantries or resupply vehicles becuase you can have a 95% kill rate. Sometimes a tank ought to duke it out with enemy tanks so that their tanks can't attack your weak positions. Usually rock-paper-scissor wipeout is also due to lack of zones of control, allowing the player to attack almost any units at will.
I'll refer to one of my favorite strategy games, Super Daisenryaku on the Genesis. One of my favorite map is Battle of Five Lakes which is a map with 4 players connected by bridges. Each bridge position is typically fortified and the tanks invariably are in the front with support units (anti air, artillery, etc) in the back. You can either assault the tanks head on with your own tanks, or you can use Bombers/Heliocopters but because units do not do sufficient damage in that game, you will rarely be able to say just move all your Helicopters to attack their tanks in heavily fortified positions, and then move all your tanks to trample over the defense. The typical result is that they'd have a few badly injuried tanks left but the support will be relatively intact, and next turn your air force will be bombarded by enemy air force and anti air, and since your tanks are assaulting from a bridge, they will get completely annihilated by enemy Heliocopters. Although you can clearly model Super Daisenryaku as a rock-paper-scissors games, actual fights are rarely fought this way because it exposes your unit too much and is liable to a counterattack.
Okay, that's all I can think of right now... maybe I'll think of something later.</div>
1. Strategy game should ever be technology based. Otherwise it's not a strategy game. It's a game of whoever researches the fastest. Strategy game should be fought between units of comparable parameters. Master of Orion is probably the classic example where a fleet of arbitrary size and less than say about 50% of the tech tree has no chance of even defeating Antaran Raiders just from inability to hit the Antaran's Interphased Drive-powered ships.
2. The combat model should be simple. Anything complicated just makes it less likely to be balanced.
3. Any kind of healing, whether you want to call it magic, repair, or whatever, is a very very very very bad idea, especially in any game where experience matters.
4. Focus fire on a grand scale battle should yield no appreciable advantage over random firing. If anything, it should yield less. Focus fire is the lifeblood of RTS game strategy and it's wrong. It makes absolutely no sense how your 8 marines can do telepathetically communicate and all hit the same target at the same time. Besides the communication issues, your units should be taking extra damage if all they're ever doing is focusing on just one unit. Too many games mistake something like say they have 15 and we have 15, but we cleverly manuever so 15 of our guys is attacking 5 of theirs to mean that the ability to select all your unit to attack the same unit is a valid form of strategy. The easiest way to solve this is to give all units unlimited retaliation attacks.
5. This goes with healing, but units with experience should not be able to recover to full strength. We're not dealing with cloning here. You start with say 10 rookie guys and through the battle 5 of them becomes veteran and fight better. That's fine. But what's not fine is that those 5 veterans gets repaired/healed/whatever and then becomes a full strength 10 unit again. When you have two of those super elite units the point is that even though you only have 2 of them, they can function as say 4 units, not so that you can heal this 2 strength super elite to a full strength 10 so it can instant kill any rookie units.
6. Battles needs to be *slow*. Also slow battles must not be effected by focus fire or any type of healing otherwise you end up with something like Warcraft 3 where a few number of big units can stay alive nearly forever. This is because if battles are too fast-paced, the first attacker has way too much damage. It's not uncommon to see a console SRPG game where you can destroy half of the enemy army on your first turn. Rock-paper-scissors need not to be the only way to go. A tank should not always stomp on poor infantries or resupply vehicles becuase you can have a 95% kill rate. Sometimes a tank ought to duke it out with enemy tanks so that their tanks can't attack your weak positions. Usually rock-paper-scissor wipeout is also due to lack of zones of control, allowing the player to attack almost any units at will.
I'll refer to one of my favorite strategy games, Super Daisenryaku on the Genesis. One of my favorite map is Battle of Five Lakes which is a map with 4 players connected by bridges. Each bridge position is typically fortified and the tanks invariably are in the front with support units (anti air, artillery, etc) in the back. You can either assault the tanks head on with your own tanks, or you can use Bombers/Heliocopters but because units do not do sufficient damage in that game, you will rarely be able to say just move all your Helicopters to attack their tanks in heavily fortified positions, and then move all your tanks to trample over the defense. The typical result is that they'd have a few badly injuried tanks left but the support will be relatively intact, and next turn your air force will be bombarded by enemy air force and anti air, and since your tanks are assaulting from a bridge, they will get completely annihilated by enemy Heliocopters. Although you can clearly model Super Daisenryaku as a rock-paper-scissors games, actual fights are rarely fought this way because it exposes your unit too much and is liable to a counterattack.
Okay, that's all I can think of right now... maybe I'll think of something later.</div>