tsunamayo / Starship-EVO

Welcome to Starship EVO bug tracking repo !
116 stars 17 forks source link

[Feedback] Weapon ranges are too low relative to ship speed [20w17e] #1867

Open Dwarf-LordPangolin opened 4 years ago

Dwarf-LordPangolin commented 4 years ago

[You mentioned wanting in-depth descriptions of why we were giving the feedback that we were, so this is my attempt at that. Let me know if you need more details! I will keep updating this as I test other scenarios.]

Feedback In a nutshell, weapon range is too short compared to ship speeds. Ships are fast enough and guns are short-range enough (and in the case of the laser unit, slow enough) that most combat ends up with the ships running into each other.

Part of this is the fault of the AI, which tends to use (pretty good, actually) fighter tactics when flying much larger ships that ought to stay as far away as possible while still in range to use the majority of their DPS, and which ought to take advantage of their turrets by turning such that as many turrets as possible can shoot their opponent (you might want to look at the old Star Trek game "Bridge Commander" for an example of that; the AI in that game was very good at moving its ship to maximize the number of weapons that could fire on its target).

But the main reason this happens is that the distance between "what the ship can shoot" and "what the ship is about to run into" is very, very small. Most of the weapons only have a range of <2km, while the ships themselves regularly fly at 200m/sec or faster. What that means is that much of the time during combat, the ships are at most 5 seconds away from potentially running into each other, and often less, since much combat takes place closer than the extremes of weapon ranges.

Scenarios Manta v3 vs Correlian shuttle The Manta won. The beam turrets were the most useful weapons by far. The sniper-barreled laser units were 2nd most useful. Even at very short ranges of ~300m, the gatling laser unit could barely hit the shuttle; the projectiles just fly too slowly to hit without leading the target by a huge amount, even at very short ranges. Also, the burst mode the gatling weapons are currently using means that it is impossible to "walk" fire onto the target; by the time the player has gotten a feel for how much they need to lead the target to hit it, the gun has stopped firing. Ironically, the sniper-barrel guns seemed to have a higher rate of fire.

Manta v3 vs Manta v3 Draw. I gave up, it was taking forever to go anywhere. The beam turrets could hit the ships, although there are some accuracy issues going on with the turrets; nevertheless, they were hitting their targets reasonably regularly. The sniper-barreled laser units could also hit the target (I don't think the AI used anything other than the beam turrets though); this was harder than against the Correlian shuttle, since the Manta is faster, but was still possible. Again, the gatling laser unit was mostly useless ... unless the AI started a head-to-head attack, which it did fairly regularly. Collisions were frequent. 20200426021004_1

Part of the reason for the draw is that the beam turrets the Manta currently has are very weak; I definitely need to upgrade these. However, the other weapons were not weak, and it was hard to hit the Manta with these, unless it was flying straight at me. The sniper-barrel guns actually felt pretty good; the difficulty of shots felt similar to how guns do in Elite: Dangerous, for example. It was challenging to hit the target with them sometimes, but in a good way. On the other hand, the gatling-barrel gun's projectiles were just too slow to be practical, even at close ranges.

Whirlwind v4 vs Whirlwind v4 Draw. The AI jumped to lightcruise and disappeared. The v4 is currently testing different weapons like we talked about, and has beams in the small turrets, and 6 beam guns in the bow; the rest of the weapons are the same as the v3. It also has significantly stronger shields (thank you, Sabre ;) ). The beam turrets were very effective, and reliably hit their target. The gatling laser turrets did not fire during most of the scenario; some bug must be preventing them from firing (it most closely resembles this bug); however for some reason, once the AI jumped to lightcruise, the gatling turrets "woke up" and started shooting at it. The fixed guns were hard to aim without a crosshair, but using a gun-camera made it possible; because of their hitscan nature, the fixed beam guns were easiest to hit with. As with the Manta, the AI kept making head-on attacks; if I kept my ship stationary, it would sometimes pull up in time to avoid crashing, but even then it sometimes ran into me. If I tried to approach the target to attack it, collisions were even more frequent. 20200426010801_1 20200426014054_2 20200426014052_1

Summary There are two reasons why ships keep running into each other in combat: the AI, and the current range/speed balance. Beginning with the first reason: the AI's fighter tactics are actually pretty good. It should do head-to-head attacks less often, since that is the most dangerous dogfighting tactic, and it should try to get behind the player harder, since that is the best place to be in a dogfight, but when using fighters it is becoming a decent opponent. That's awesome! Unfortunately, it uses the same tactics when flying much larger vessels, as seen above. Collisions between fighters should be rare; collisions between big warships should be extremely rare. The second reason is just down to the math, because not all the collisions were the AI's fault. If trying to point your ship at the enemy while moving towards it, it is frequently hard not to run into them. The reason that is hard is because weapon ranges are very short compared to ship speeds. Only a few seconds elapse between "I am now close enough to shoot the enemy" and "I am now running into the enemy." This narrow window of time between entering range with the enemy and being in danger of colliding with them tends to result in the scenario portrayed in this dramatic reenactment.

Suggestions

  1. Increase weapon range.
  2. Increase projectile speed where applicable to account for increased range.
  3. Add a lead indicator to the HUD to account for increased range (also, just because it's a good quality of life feature to have).
tsunamayo commented 4 years ago

Okay amazing contribution, now I have a clear view of what is needed to improve the situation. 1) yes it was clear already that we need some higher range / velocity (same thing, projectile lives for 3 seconds) 2) Yes I will add a lead indicator. Problem being, which time horizon should I input? I can pick a simple value, or get something more evolve were the player link a weapon to the shipcore and get one or several lead indicator. 3) I need a deeper AI logic that takes into account the size of the ship, and fire power configuration. 4) Yes I do need an indicator or which side of ship have the most turret. Is the startrek game using Player Generated Content? I have some idea of algo on how to do it. Yes one another things to code linked to the building nature of the game! 5) I also want to add more HUD info, like warning and message when your power / heat is low, and and system health indicator.

I will do all the change in my latest post if nobody convince me otherwise and come back with a build soon, I think we will get somewhere.

Battlepixel commented 4 years ago

In my opinion, increasing the projectile speeds of large weapons would be a serious mistake because it negatively affects other combat aspects. Instead, I suggest increasing the damage of large weapons to compensate for the more frequent missed target at greater distances.

Please consider the following: Like I wrote, I think it would make more sense to increase the damage of large weapons rather than their projectile speed to compensate for longer weapon ranges. Otherwise it will hardly be possible for small agile ships to approach larger ships if the projectiles are faster. When approaching a large combat ship, it doesn't matter how slow its turrets are because they hardly have to move as soon as they have their sights on them, because their angular velocity is tiny at great distances. Weapons strong enough to take out a small ship with one or two shots should hardly be able to hit such small targets. Otherwise, large turrets are better in any case and you don't need small turrets against quick targets. In addition, fighter or carrier based attacks would be pointless. It would be cool if large ships with large turrets needed an escort against small agile targets or a lot of small turrets. Large ships have the advantage that they are usually only damaged in battles. It would be really nice if they slowly regenerate themselves again as long as crew is on board (this would make the game easier and more fun and avoid annoying micromanagement -> in many games multicrew ships regenerate through their crew) Small fighters are often totally destroyed. 10 small ships that cost as much in total (have the same number of blocks in total as) 1 large ship should be more effective than the large ship as long as it is not equipped with an excessive amount of Gatling guns. Otherwise the damaged battleship would not suffer any losses after the repair. If the squadron wins, then only with permanent losses because some fighters have been destroyed. Without this mechanism, larger ships would always be a better option than many small ones. This makes them in no way an attractive option in space battles. But they should be, because small, fast fighters bring a lot of fun and action to space battles.

tsunamayo commented 4 years ago

In my opinion, increasing the projectile speeds of large weapons would be a serious mistake because it negatively affects other combat aspects. Instead, I suggest increasing the damage of large weapons to compensate for the more frequent missed target at greater distances.

Please consider the following: Like I wrote, I think it would make more sense to increase the damage of large weapons rather than their projectile speed to compensate for longer weapon ranges. Otherwise it will hardly be possible for small agile ships to approach larger ships if the projectiles are faster. When approaching a large combat ship, it doesn't matter how slow its turrets are because they hardly have to move as soon as they have their sights on them, because their angular velocity is tiny at great distances. Weapons strong enough to take out a small ship with one or two shots should hardly be able to hit such small targets. Otherwise, large turrets are better in any case and you don't need small turrets against quick targets. In addition, fighter or carrier based attacks would be pointless. It would be cool if large ships with large turrets needed an escort against small agile targets or a lot of small turrets. Large ships have the advantage that they are usually only damaged in battles. It would be really nice if they slowly regenerate themselves again as long as crew is on board (this would make the game easier and more fun and avoid annoying micromanagement -> in many games multicrew ships regenerate through their crew) Small fighters are often totally destroyed. 10 small ships that cost as much in total (have the same number of blocks in total as) 1 large ship should be more effective than the large ship as long as it is not equipped with an excessive amount of Gatling guns. Otherwise the damaged battleship would not suffer any losses after the repair. If the squadron wins, then only with permanent losses because some fighters have been destroyed. Without this mechanism, larger ships would always be a better option than many small ones. This makes them in no way an attractive option in space battles. But they should be, because small, fast fighters bring a lot of fun and action to space battles.

Thats why I like to base my decision on actual gameplay feedback. We dont know this yet. We might have do like you said, or not, we cant know for sure at this stage - we need more mileage and progress with the combat system. For now I keep things simple, laser damage scale linearly. Then we will tweak if necessary.