Lance

quote: Basically, I want some simple description of what the behavior is meant to be, and I couldn't even see one after poking around in the unit script. As far as I can tell the Lance should shoot at a Ravager orbiting at 200 elmos because it is able to track it. Maybe something is going on with script threads. I'm not happy with code being the bedrock of design, I need an abstract design goal which is then meant to be implemented in code.

I don't know how it was implemented and what are edge cases here, that's something to look into. Maybe Shaman can shed some light here.

I can describe how I picture it as a model to implement:
- Lance now has tracking weapon which deals no damage but increases its internal resource "Tracking".
- It only charges after consecutive hits of the same target with tracking weapon.
- Each shot/pulse (could be every tick like beam weapons) contributes to the Tracking counter.
- Switching targets resets tracking counter
- Tracking counter decays over time (about half as fast as constant tracking increases the counter)
- Optional: Turning turrets decreases tracking counter proportionally to the radians turned. If this is not added then it would be balanced by turret turn speed which could be adjusted.
- When tracking reaches 100%, swap weapon for original Lance's laser, decreasing turning speed, acceleration and turret turn speed by 95%. The lance's main weapon should have slightly higher range than tracking laser to help with edge cases.
- After 1 second or when Lance starts reloading, bring back its stats back to normal. Swap the guns and put the tracking beam on cool down of main weapon (0 if didn't fire, reload time if it did). Reset Tracking counter to 0.

I added here a suggestion for Lance to slow down/stop to fire, to help with team killing other Lances and the riotness of its gun.

quote: I think walls are a fine counter.

Well they countered BB as well. But BB was nurfed anyway..

quote: This sparked an interesting idea though. If the goal is to make Lance more visible and susceptible to counter battery, why not make it actually visible? We could add an extra game rule that reveals enemy units to teams with vision on the beam laser projecting from the unit. This would many affect Lance and Lucifer, but would technically affect all beams. It makes sense from a unit AI perspective, but it does run the risk of being too visually messy.

Another cost of making beams reveal their shooter is the reduction of fuzzy information. When you give people incomplete information it leaves room for them to make predictions. The gap between making the prediction and finding out the answer creates anticipation. If you get it right you feel smart for predicting correctly, but get it wrong and you feel surprise. Either way, this seems like a more valuable emotional journey than just having a Lance shoot and then instantly seeing where the Lance is, and that it is a Lance rather than a Lucifer or Paladin.

Mobile radar dot speed is a good example of fuzzy information. In theory a widget could track radar dots, estimate their speed, and then output a short list of likely unit types. But I feel like we'd lose something if this was done.

• Identifying radar dots from their speed is a skill that is built up in stages. Fairly new players might be able to tell a fast raider from a slow assault, intermediate players may tell skirmishers from raiders, and the best players might be able to guess at specific raiders. Noticeably getting better at something feels good, regardless of what it is.
• The best UI for identifying radar dots is for a player to observe a radar dot, then for a hunch to arrive in their brain. Any UI that tries to replicate this will make the screen messy and mostly provide redundant information.
• It ties into game-feel skills around knowing what sort of units the enemy might have. Multiple units are more likely to be cheap than expensive. Knowing the enemy factories rules out many units unless they've built an extra factory.
• It creates a lot of short prediction-anticipation-reveal loops. The skill is used and improved upon every game.
• Overall, identifying radar dots doesn't feel like a skill that has a massive impact on the outcome of games. I don't expect anyone to need to actively train this skill to improve at ZK.

It looks like the cost of replacing the skill with a widget would be far greater than the benefits. The UI would become messier, players would lose a skill to improve at, and we'd lose the enjoyment of the prediction-reveal loops. Sure, ZK is fundamentally not about estimating the precise speed of moving dots, but keeping it as a secondary skill seems fine if the enjoyment of using and mastering it far outweighs any impact it has on games. It would feel bad to lose games to secondary skills such as this, but I don't see it happening for this one.

This is all before we consider counter widgets. In theory a widget could fight the first by telling fast units to move at the speed of slower units. People can already do this to some extent with formation move, but they have to have a slower unit to move. I expect people have actually spoofed each other with ability before. Players can't currently do things like move Glaives at Scorcher speed early in a 1v1. So there would still be some information game with these widgets, but I expect it would be a game of messy UIs and fiddling with widgets, rather than a game of hunches with a rare bit of spoofing.

This is relevant to Lance because the same thing happened to me a week or so ago in a Paladin rush.

• The enemy team rushed a Paladin.
• The first indication of this was a Lance beam from fog.
• The beam appeared at around the same time that my teams Paladin rush was ready.
• This raised the thought "Did the enemy Paladin rush too?".
• A second beam appeared, and my fuzzy feeling of time said "that wasn't 20 seconds". I know Paladin has a 10s reload on its beam, so more weight was shifted to the Paladin hypothesis.
• The Paladin was revealed shortly afterwards.

The period of uncertainty was not long, but I feel like it improved the game. Later in the game there was another Lance beam, but scouting revealed it to be a Lance.

I am also concerned about making the fog messy by filling it with a bunch of revealed units. Revealing firing units in aLOS already looks a little messy.

i think the 2 biggest problems in my opinion are cloaking lances and their excellent AA skills.

a ball of lances dont need AA since they can blast the bombers down them selves.

i don't think the lances need a stat nerf or price increase if you take away the ability to blast down their counter and maybe for all artillery to be uncloaked while reloading.

this means they can still do what they are good at just at greater risking if used without a second thought.

now people spam lances cause they can safely cloak it once it has fired without danger of getting shot by the opposing lances.

what i think will happen if they are unable to cloak after a shot, people will not fire at random radar targets again cause of the chances the opposing lance is waiting for a lance shot and they will have to pick their targets instead of spam blasting everything to hell

Cant find where I said this before, long ago, but my vote is that Lance being powerful is good. Just make its death explosion be more bigger and dangerous(powerful energy weapon and all that reason, please use with care and consideration) and just increase the cost a bit instead of stat changes to see a little fewer of them.

Althou, I would like to kill a full health paladin with spammed lance ball in two seconds again one day. Wouldnt you too, mr Zen? :)

quote: Non-personal recloak delay 6s -> 8s.

Hm, I am a bit concerned about my merlins, but I guess this is worth it.

quote: Althou, I would like to kill a full health paladin with spammed lance ball in two seconds again one day. Wouldnt you too, mr Zen? :)

Even with the proposed changes that should still be possible.

quote: Non-personal recloak delay 6s -> 8s.

What is this number based on? Sometimes even 6 seconds feels painfully long for hiding builders on the front line. Especially under Impaler fire.

quote: A simple question, if a lance had double the price, do you think players would stop building it? The answer is no, because even at current price it regularly makes multiple times its own cost.

Even if I were to grant that the average Lance makes 2k cost (which seems like a high estimate), lots of stuff makes cost. Unit diversity decreases if Lance isn't a viable alternative to other similar but more effective strategies.

quote: What is this number based on?

It isn't really based on anything. It was added to nerf "combat cloaking", that is, using a cloaker in combat to confuse targeting, rather than to hide units. I'm wary about making it too high as it could be annoying.

When I added it I made it double the values for personal cloakers. The numbers are now as follows:

• Personal cloak proximity decloak time: 1.5s
• Personal cloak action decloak time: 3s
• Area cloak proximity decloak time: 3s
• Area cloak action decloak time: 8s

Note that personal cloaking units always use their improved values, even when under an area cloaker.

The turn rate reduction and the turret speed reduction seems like it will make the unit more frustrating to use without actually nerfing it, which I think is a game design direction we shouldn't be going towards.

The vertical aiming speed reduction may be a good idea, but it also could result in increased frustration (goddamn Lance trying to aim at a Swift and now the turret is vertical and it takes 5 seconds to aim at a ground unit)

The range reduction and speed reduction are good I suppose, but they won't affect the balance much.

The goal of those changes is to make getting good Lance shots a bit riskier. I think it would be more interesting to make retaliation against Lances a bit either rather than just make Lances build up slower. Being frustrating is certainly a concern with the turn speed reductions, but I think it previously had unusually high turn and aim speeds so it is probably safe to try out.