Equipment upgrades: Simpler or more choice?

My current project is a top-down space shooter where the player commands a capital ship with increasingly powerful armaments and alien artifacts. Primary ship components would include: main turret, missiles, armor (health), shields, flak (auto-defense weapons) and possibly engines.

I'm considering two different models for purchased upgrades:

1. The player spends money upgrading items along a single path. Turret I, Turret II, Armor V, etc. The player could choose to upgrade from Turret I to Turret II for $5000 and get a faster, more powerful weapon, or the player could choose to upgrade from Armor II to Armor III for $6500 and have better protection.

2. The player spends money purchasing upgrades that enhance one property of a component. For example, the player could spend $2000 increasing turret speed, or $1500 increasing turret damage, or $1000 increasing his shield's recharge rate.

The second model gives more choices, the first is less confusing.

Which would you go with?

PRadvan will probably also jump on this one, because he's also making a rpg'ish space shooter game. For now, my 5 cents.

In my upgrade system I have several ammo types. Plasma is good against Hulls but crap against Shields. EMP is good against Shields but once broken they're crap against Hulls. Missiles are moderate against both and Torpedos are good against both. This gives the player some choices.

Then in designing my weapons I think in close and long range. Bombardment versus Dogfighting. Bombardment weapons are longrange, powerfull and a slow rate of fire. Dogfighting weapons are the exact opposite. The one relies on a few heavy hits while staying outside of the others range, the other for a lot of hits in close range.

I believe this system is more diverse then just upgrading turrets (although I do have stronger and weaker versions of all turrets). Some ammo is more natural to dogfighting (plasma) and others more to bombardment (torpedos and missiles). Some shipframes (or chassis, I call them frames) are fast, nimble and small, making them more ideal for dogfighting, others are slow and bulky making them ideal for bombardment. But... that's just my idea of how the game should be played. If a player would like to use his smaller fighterships to break a shield with missilefire so a big ship with heavy plasma turrets (relatively close range) can finish it off when it gets there, that's a valid strategy.

In short: I'm pro choice, if the difference can be easily defined and explained to the player. Make the choices stereotyped.


Your second system can be used in conjunction with the first, but I think they would serve better as general upgrades, which apply to all turrets. Think of it as upgrading your targeting software or training the crew to better handle them. Ofcourse this should be a bit expensive as the bonus gets bigger with more and heavier turrets.


OFF-TOPIC

Defensive fire eats op processing power. I use a grid to do collision detection among other things, which I could easily use for defensive fire. The actual defensive fire isn't what eats up the cpu though, it's finding targets. Even with a grid to eliminate possible targets, I still need to do a distance check to quite a number of ships. If you have turrets with limited rotation that's even more of a hassle, as targets can be inside the defensive perimeter, but not in the possible arc of fire. I chose to dumb down my defensive fire and just check if it was inside the defensive perimeter.

Interesting points about different weapon types. In my game the player's ship will primarily be attacking with a single, upgradeable turret, plus "smart" missiles.

How does your game handle switching between different weapons types (or does it?)

I'm currently leaning to having equipment upgrades + modifiers. So the player upgrades armor along a single path, but can purchase or find special artifacts that give bonuses, like 10% damage reduction, etc, 20pt damage thresshold, etc.

Regarding performance, so far I'm not really hitting any bottlenecks in terms of the actual game mechanics. The player's FlakController (defensive fire) averages ony 0.04ms per frame. For comparison, updating enemies is about 0.1 ms per enemy per frame.

What does hurt are sounds (8 ms per sound.play call) and graphical effects, which can be adjusted to match the user's machine.

Quote:

---

Originally Posted by justkevin

How does your game handle switching between different weapons types (or does it?)

---

Actually it doesn't, as it's more of a strategy title. The turrets have one or more "Tubes". They're named after torpedotubes in submarines, but I also could've called them barrels just the same. These Tubes contain variables like maximum burst, cooldowntime, ammotype etc. If a turret locks onto a target and is aimed at the intercept vector, it calls the trigger function of the first tube that is ready to fire. Basicly a turrets can have different ammotypes by giving it tubes with different ammotypes, but I chose not to do that in my game. Switching would be as simple as to reassign which tubes are used.

Both methods are fine. I think you should pick the one that best fits with the scaling of your AI. In the first one, you jump higher up in power, but less often. In the second method, your strength increases more gradually. So think of how your enemies behave, and pick the one that fits most accordingly.

Quote:

---

Originally Posted by TOdorus

In my upgrade system I have several ammo types. Plasma is good against Hulls but crap against Shields. EMP is good against Shields but once broken they're crap against Hulls. Missiles are moderate against both and Torpedos are good against both. This gives the player some choices.

---

[geekmoment]Kind of like ion torpedoes compared to missiles in star wars...[/geekmoment]