Gameplay, Units & Balance

LOL No. I was trying.. It started the command, but that's an illegal command, so it didn't allow it. I was trying to troll him, but sadly it failed.

AHAHAHA! I remember that "dirty trick" in TA and was thinking "it's back...."! Too funny.

Just saw it on Meta's stream. Hidden commands huh?

I feel guilty. ;(
https://forums.uberent.com/threads/comnapping-is-back-in-galactic-proportions.51556/
Although you said it were an exploit yourself

Exploit? Or feature...

I just can't wait for those Stargates! Awesome work Uber!

Those are for the continuous build commands.

Well Garat have shown some ambuiguity in the question.

I just want to say that this looks absolutely fantastic.

My only concern is that making artillery ineffective against mobile units because they miss moving targets can create weirdness. SupCom 2 artillery, for example, forces spastic juke micro to make artillery miss. I think just giving artillery a deflection radius is a vastly superior solution. Against large and immobile structures they will be far more effective, but juke micro does nothing.

It's possible there's another game going and being streamed.

Man, I didn't know Scathis was a hot commodity. I hope Brad or Mavor can do a live stream with him along the lines of the live stream done with Sorian; diving into his start and maybe talk about how he balances games.

It would be Awesome.

At what minute of the video was that? I must have missed it? that would be so cool if there was wave animation coming to the game, that would be kool !!!

The short answer is yes, we still are optimizing and have plenty of work and headroom there.

There are actually two separate issues. There is frame rate which is related to how fast we can render and update the client. We have a ridiculous amount of room left there as we haven't even multi-threaded it. However, we are still dealing with basic issues right now like trying to use the correct fastpath techniques on all graphics cards. This is the reason ATI cards run like ****, because we aren't doing it right. Recently they've gotten back to us with some analysis and we are working on improving that. In addition there is plenty of other render perf things we need to do (particles are too slow, units aren't instanced etc. etc.)

Then there is the bandwidth issue. Certain kinds of lag that you see are because we are saturating the bandwidth pipe that we self throttle. There is a TON of continuing work to make this better. This includes sending less data (but the right data), compressing the data better, allowing bursts of higher network bandwidth etc.

I don't see any reason this can't happen. Maybe after we release this stuff so people can play it first.

MINES!!! Far out!!

Mention of nuclear land mines no less. Awesome, guys.

I got an ATI card and don't seem to have these issues, then again I haven't tried a game on anything other than 1 or 2 planets and not more than 10 players.

Really glad to see arty go back to arty instead of super awesome sniping machines. Really didn't like their OPness.

Just a quick question...why are the blue bottles moving so slow? Am I retarded? Or was there a change?
Very possible I'm just retarded, but I swear they are moving really slow compared to latest beta build.

Keep up the good work!!!

It'll be another 30 minutes or so, but.. as I'll be out for the evening, here's the link to my version of the afternoon playtest:



I died a little early on, so got to commentate for most of the game.

Yup, artillery were my biggest disappointment in SupCom. The arch they were firing on was like a mortar. I have never valued mortars in RTS games. There are a lot of issues, and I am probably missing some, but I want to point out some of the nuances related to lobbing units.

1. Land to Ship combat in SupCom2 lacks balance because of the way 'Fortified Artillery' fire too high in an arch. Ships have a massive advantage. If we balance Artillery to fire higher it may make ships even more over the top. I consider the current ship to pelter combat balanced and since land units are so disadvantaged against ships already the pelter is a lynch pin holding land vs. naval balance together.

2. The artillery bots and tanks in SupCom2 fired so high it was ridiculous. They were mortars, but where were the cannons, catapults and trebuche? Please continue to focus on this as PA continues to grow.

3. Pelters would be less over produced if they were just more expensive like punishers from TA. This is a siege unit. It should take some time to build and provide the defender an opportunity to strike.

TLDR: Scathis, please take a strong look at Uber Hack and how it addresses punishers (pelters for TA), ships, T2goliaths vs. T1destroyers etc. I am a huge fan of the projectile archs from the original TA as they were magnified in Uberhack. I think they worked very well.

Hope this was helpful. Thanks for bringing this topic up Ledarsi.

Ledarsi, what's a deflection radius? Please explain if you will.

THIS!!!!!!!!!!!!!!!!!!!!!!!!!!! yes i hate that t2 takes a substantial amount of time to amass any real force with or massive eco to assist those factories. YES

combat fab tho........i see giant medic blobs in the future led by commanders.......... any light on what may reduce this????

I am not opposed to the higher arc. The issue is when you have artillery that is extremely accurate against a stationary target, but which misses moving targets because of the time required for the shot to hit.

This is totally exploitable by rapidly giving your units fast, contradictory move orders. In SupCom 2, you can make your enemy's artillery utterly useless by very rapidly issuing move orders to the left and right of your army. Your troops do nothing but spin in place, but the enemy artillery will miss every shot. This micro trick turns artillery from the dominant late game weapon into a completely and utterly useless unit. It is completely mindless high-speed clicking, and it sucks.

Instead, make artillery imprecise by targeting a random point within a circle around the target. Ideally making it more likely that a shot will land closer to the center. (i.e. randomly generate angle from target unit/point, and generate a deflection distance by making a random number between 0 and the maximum deflection radius squared and take its square root).

What this means is that an artillery unit's shot will land somewhere within a circle, instead of always hitting its target. Against a bigger single target (such as a structure) the shot is more likely to hit. Against more numerous targets a shot is more likely to hit something, even if it isn't the target you were originally aiming at. And against slow targets that cannot get out of the deflection radius in time, the shot is more likely to hit because you will never hit a target that is moving fast enough to escape the entire deflection before the shot arrives unless you lead the shot and the target's course does not change.

However if the shot's deflection is random, random juking has no effect, because you are just as likely to move into a shot as you are to avoid a shot that was aimed where you were before. And even though it is possible to juke using larger moves, such as moving for a few seconds to fool the artillery into leading its shot, and then changing course, you have to know exactly when the enemy artillery is aiming in order to do this. Juking works for the short-range, rapid-fire arty in SupCom 2 because you know that they are within range through visual confirmation, and that they will be firing constantly, so you juke constantly at high APM to make it aim way off to the side of your forces, completely removing the enemy's weapons from the battle by spamming random move commands.

PA should have artillery with longer aim time and shell flight time, and should also use random deflection instead of relying on flight time to miss mobile units. Because of the range and flight time, you can't be sure what the enemy artillery is targeting, or when it is aiming as opposed to when the shots have already been fired. And if you try and dodge an inaccurate projectile by juking, you are just as likely to dodge into the shell as away from it because of deflection. Your best bet is to just move fast in one direction because the artillery has a random deflection.