124 replies to this topic

[Ashnod]

Convergence is already tied to your locked target, just I believe ballistics are currently bugged with it.

[KinLuu]

The convergence system does not work instantly. The weapons converge slowly over time, weapons in the torso converge slower than those in the arms.

Use this to your advantage. Shoot the enemy with lasers, while doing this your crosshair will be on target and your weapons converge on it. After the laser-dot ended, lead with your crosshair and fire your ballistics asap.

Does this work always? No, not really... but with some practice it should work most of the time.

[HugeGuns]

But you have to lead the target with your lasers as well.... so miss with your lasers so you can hit with your Ballistics?

[Wolfways]

xenoglyph, on 12 November 2012 - 10:24 PM, said:

Math mostly. Your comment adds a lot to the discussion, btw. Good work.

It was a question, not a comment. Your post was a comment...which added nothing to the discussion btw. Good Work.

Anyway, Sawa963.
You say "Snipers do not experience this issue", yet my K2 with ERPPC's experiences exactly the same issues.

[KinLuu]

HugeGuns, on 12 November 2012 - 10:49 PM, said:

But you have to lead the target with your lasers as well.... so miss with your lasers so you can hit with your Ballistics?

Since the latest patch I no longer need to lead with lasers. The only exeption may be a 250+ ping Jenner.

[Nullzero]

Great idea...

But there exists one glaring problem that could be abused.

Once a target is locked in, and convergence is set, you don't even need to have a visual on the target to hit them. What I mean is:

1) You target a mech somewhere down range with R key.
2) Park your mech behind cover (such as the corner of a building), with only your right arm (ala Centurion) poking out from cover.
3) Place arm reticule over red target square, and fire. As long as there is no obstruction between your arm's muzzle and the target, you hit every time.

If you can't quite visualize what I'm saying... just picture Gears of War blind-fire mechanic, only perfect accuracy on your reticule over the locked target.

Locking convergence to target basically lets you fire safely from behind cover. Yes, your firing arm is exposed, but all you would have to do is fire, and twist away between cycle times to minimize that exposure.

[HugeGuns]

Let's just make the Artemis system work with Gauss Rifles, guiding the Shots to any target over 45m

[Straylight]

xenoglyph, on 12 November 2012 - 09:54 PM, said:

re: Straylight: That's great and all, but the masses of people we need to make this game successful don't care about the lore. They just want to hit what they're aiming at.

I know, but the lore IS a factor that can't be ignored; otherwise we wouldn't be playing a Mechwarrior game.

Lore discussions aside, though; yes, there are some technical considerations that need addressed (lag and hit detection, firing delay), but "fixing" weapon convergence isn't one of them. Ignoring aforementioned technical considerations, a pilot who can't hit what he's aiming at isn't aiming correctly. Sometimes, aiming correctly means placing your point of aim above, below, or to the side of your intended point of impact. Some of this is target-leading, some is plain old kentucky windage, and some of it is compensating for parallax issues created by the huge distance between your gunsights and your muzzle (tip for Dragon pilots: aim high and left).

Quote

Are you suggesting that in the future they forgot how to do simple math? Those are among the simplest technical problems to solve in the BT universe I would expect.

No, I'm suggesting that in the future they teach their pilots how to do the simple math themselves because their targeting computers can't do it quickly enough to be effective in battle.

Funny how that idea translates directly into the game we're playing...

Nullzero, on 13 November 2012 - 12:08 AM, said:

Great idea...

But there exists one glaring problem that could be abused.

Once a target is locked in, and convergence is set, you don't even need to have a visual on the target to hit them. What I mean is:

1) You target a mech somewhere down range with R key.
2) Park your mech behind cover (such as the corner of a building), with only your right arm (ala Centurion) poking out from cover.
3) Place arm reticule over red target square, and fire. As long as there is no obstruction between your arm's muzzle and the target, you hit every time.

If you can't quite visualize what I'm saying... just picture Gears of War blind-fire mechanic, only perfect accuracy on your reticule over the locked target.

Locking convergence to target basically lets you fire safely from behind cover. Yes, your firing arm is exposed, but all you would have to do is fire, and twist away between cycle times to minimize that exposure.

First; why is that a bad thing? And second: how is that any different than fighting in darkness, bad weather, through concealment or in any other circumstance where you have to "fly on instruments"?

A 'mech that's lurking behind cover and firing out of it is fighting smart, taking advantage of the EWF/infowar capabilities provided by his teammates in order to maximize his terrain advantage. This is NO DIFFERENT than firing LRMs from behind a hill, except that the LRMs are guided projectiles and whatever gun you've stuck around the corner is not.

"Perfect accuracy" is meaningless in this case, because the perfection of your accuracy is dependent on your ability to actually track a moving target, and in this case is functionally identical to tracking a moving target it any other circumstance.

HugeGuns, on 12 November 2012 - 10:49 PM, said:

But you have to lead the target with your lasers as well.... so miss with your lasers so you can hit with your Ballistics?

Only if you fire them both at the same time, at which point you're either generating heat needlessly or wasting ammo pointlessly. Keep weapons with different ballistic profiles tied to different triggers. If you don't have enough triggers, get a better mouse.

Wolfways, on 12 November 2012 - 11:18 PM, said:

Anyway, Sawa963.
You say "Snipers do not experience this issue", yet my K2 with ERPPC's experiences exactly the same issues.

That's because PPC bolts are slower than cold molasses. A sufficiently quick 'mech is actually capable of dodging them from a standstill if the firer is far enough away. PPCs are medium-range weapons, and the real advantage of the ERPPC is that is has no minimum range.

Edited by Straylight, 13 November 2012 - 01:01 AM.

[One Medic Army]

Seeing as I suggested the same exact thing in a (largely ignored) suggestion thread pre-open beta, I heartily agree.
Seems the post in question got wiped with the beta forums, oh well.

Would make all projectile weapons (everything other than lasers) function much better. Heck even lasers would work better vs lagshielded lights since you're aiming ahead of them.

Edited by One Medic Army, 13 November 2012 - 01:07 AM.

[Mordekai]

I totally agree with the OP. I am heavy ballistics user myself and its just really hard to get right currently. Lag overall is a big problem and URGENTLY needs to be adressed by the devs.

[Skirich]

Straylight
Thanks, for menting paralax compensation.
At this moment the main issue is unpedictable fire delay after peuling a triger. I have ping about 160-200 and in addition to random fireing delay somtimes I think there is an additional multiplyer there: I have delay between 0,7 to 4 seconds. Add lagshilding to this - at some mathes i unable to hit the moving atlas, by standing still behind him.

The highest delay i registred was about 5 seconds. I puled a trigger, start evading enemy fire and have friendly atlas killed via CT backshot with AC20 slug from back-left (as my temmates says) of my hunchie cockpit.
That was exception, of cause, but 1-3 second delay still in game most of mathces.

So the solution either target-based convergence (that will not compensate paralax) and do some compensation on moving targets, but add some dispersion based on AC type or to make weapon predictably fire after puling a trigger. The +- 0.3 second dealay will easily ruin your shots on same size moving > 40km/h targets, and now we have much more unpredictability.

Edited by Skirich, 13 November 2012 - 01:56 AM.

[Elizander]

I'm having a lot of problems with this when I use the YLW. I twist my left side to my opponent between AC20 cooldowns, but it never hits when I twist back to shoot at them. I've seen my shots go off on wild angles and there is no indication if/when my weapon has adjusted to my current target (normally in FPS games this is indicated by a shrinking crosshair or something).

[Sickocrow]

Some sort of server sight would be a blessing. It helped a fair bit in WoT when I played on NA.

[Roadbuster]

I agree with the OP.
ACs are really bad the further your target is away.
The only times hitting a target is pretty reliable is when you're playing facehugger, or if your opponent is running straight towards, or away from you.

I'd love using ACs in more builds, but they are too unreliable to be a good choice.

[Rogue Phoenix]

Straylight, on 12 November 2012 - 09:35 PM, said:

Two points:

1. Light ACs and Gauss Rifles are not brawling weapons to begin with,

If a projectile comes out the end of a tube, it's a brawling weapon. I'm not going to leave a powerful weaopn idle just because I'm at close range.

Quote

and their convergence issues are irrelevant when being used in their intended role where targets are distant and have low transverse velocity.

fine.

Quote

Big ACs are meant to be used at short ranges where time-to-impact values are effectively zero and leading targets becomes unnecessary anyway (and if your target is 55m away and you DO need to lead it, you probably shouldn't be firing an autocannon at it in the first place.

I will hapily lead a target with an ac-20 in this situation. In fact, I will be ecstatic that my opponent is accomodating enough to come inside my AC20's range.

Quote

This is what lasers and SSRMs are for), meaning that if you are finding convergence and leading to be an issue, you're probably using the weapon incorrectly.

What works with leading is consistency. In a circle, I don't really care what range the weapons converge at, I do require consistency for practice to be meaningful. So either converge at the range the enemy is at, or converge at 400m, or converge at 200m, I don't care, pick one and stick with it. If the enemy is at 50m, 10 degrees of variance because my rangefinder keeps flipping from 50 to 10000m is enough to make me rip the d*mn thing out by the wires so I can fire with some consistency.

Quote

In other words; use the right gun for the right situation, and learn to aim better.

Yes. And, don't put complex obstacles in my way when a simpler solution will work better. In a brawl, a gun is a gun is a gun, and if I've invested the tonnage to put ballistics on my chassis, the dagum things are going to get fired when the heat is on.

Quote

2. This is 3049. The Inner Sphere is just crawling out of a muck of the succession wars and is only now beginning to rediscover technologies that we here on the Aquarian Cusp would consider "advanced". Remember, for most of the last hundred years of BT timeline, the technology level of most of the IS is late 20th century at best. Advanced electronics, in particular, have been hit hard through the destruction of manufacturing infrastructure during the constant fighting and ComStar's jealous, cult-like guardianship of what remains. Given that, it makes sense for targeting computers for not be very "smart". Considering that what passes for sensor packages in the game is cripplingly short-ranged and wholly dependent on direct LOS (nevermind MRI, thermal, radar, lidar, satnav and the dozen other remote sensing technologies we've had for forty years now) it's a stunning technological achievement that our LRMs can successfully track a moving target. Expecting better out of an FCS trying to successfully [a] figure out what you're aiming at, [ b] calculate range and elevation to target, [c] calculate target radial velocity, [d] predict intercept point based on previous data and weapon ballistics and [e] successfully engage the target within an acceptable window of time with an unguided projectile is really a bit much to ask.

I'm not asking the computer to compute leading a target. I'm asking the computer to get out of my way when I do it myself.


There are times when a piece of bubble gum in the windshield works great, and you really don't need 31st century tech to aim a ballistic weapon. WWII fighters had gyros (a mechanical solution) to compute lead (though range on them was selected manually, and the electronics -- if you call a light and a plate of glass electronics -- was mostly the HUD). The actual math to compute lead is algebraic, and quite frankly easier accomplished than, say, setting up current mech's HUD display, or getting LRMs to hit a moving target. If you've got a rangefinder (such as the distance readout on the targeting reticule) and can dynamically factor that into the convergence calculation (which we're already doing anyway), so much the better! The problem here is that the inputs are coming from the wrong place at the wrong time: the rangefinder value that's currently being used comes from LOS along the aiming reticule. When you're circling, the one place the enemy isn't going to be with ballistics is... the aiming reticule. (Lasers don't have this problem because flight time is irrelevant.)



In the field, it's a matter of putting a three-way switch on the convergence acutator's range input: range to highligted target, range through reticule, or manual range setting (and associated dial). 4 sets of wires, a manual range dial, and a soldering iron. Problem solved. Next...

[SPARTAN 104]

It's a good read but everything can be summed up with your first sentence. If pgi can do it right they and we win. If they can't they lose and we move on.

[Urza Mechwalker]

I would prefer If I could turn off automatic convergence. I am good enough that I can manually compensate the range and the lead of a fast moving target. But now with this dumb system interference.

Its impressively more easy to hit a small scout tank in Wot with a tank that has a large circle of precision in WoT than its to nit anything moving in this game. Maisn reason is that there is no crap system interering.

[Slanski]

Solution:

1. Converge on target distance if target selected.
2. Permit [on/off] state for automatic convergence
3. Include a manual convergence override to be set at hard distance of 100, 250, 500, 1000, infinity.

The current system completely destroys projectile game play, as the skill factor of leading shots requires a reliable constant behavior of the weapon system.

Edited by Slanski, 13 November 2012 - 03:20 AM.

[Leetskeet]

tl;dr they ****ed up and anything that shoots a projectile from an arm is a piece of ****.

[Skirich]

Slanski, on 13 November 2012 - 03:11 AM, said:

Solution:

1. Converge on target distance if target selected.
2. Permit [on/off] state for automatic convergence
3. Include a manual convergence override to be set at hard distance of 100, 250, 500, 1000, infinity.

The current system completely destroys projectile game play, as the skill factor of leading shots requires a reliable constant behavior of the weapon system.

If they do it alone - there will be too much crying that AC is OP) Especially AC-20 if they makes it damage even half of basic on 500+ meters (cause high caliber rounds do damage not only by impact of the shell, but by detonation of the shell)

edited.

Edited by Skirich, 13 November 2012 - 03:34 AM.