342 replies to this topic

[MischiefSC]

So to facilitate accurate stats I've been screen-shotting my end of round stats and team stats. Again, this isn't about e-peen stuff, I'm a pretty average player. My Elo is pretty settled I think and especially in heavies I'm about as high as I can get. I drop reasonably often as filler for top-tier teams so I can't be terrible. Normally I rank pretty average, ergo the following stats are part of what's driving my interest in LB10Xs. This last dozen I've capped have been pretty indicative of my experience with this Jag build. Approached from the perspective of maximum performance for the weapons I played each match until I died or we won before someone could kill me. My KDR is skewed accordingly. Over the last 12 matches I've got -

8 wins, 4 losses.

6 out of 12 I was the highest scoring player between both teams.

Of the 6 I was not top score, 4 of them I was in the top 3 - this includes matches where my team lost, I still scored higher than most the winning team. There were only 2 out of 12 matches I was not literally leading the match in damage done, kills, assists, etc. That's very much atypical for me.

In all 4 matches I lost I was the highest scoring person on my team.

I scored 24 kills, broken up in ranges from 0 to 5, specifically it was 4, 4, 2, 3, 2, 1, 1, 4, 0, 4, 0, 5.

I scored 48 assists, they were 7, 7, 6, 4, 0, 2, 1, 4, 6, 3, 5, 3.

I destroyed 54 components, my highest two matches were 7 and 8 respectively. One of them I only got 1 kill.

Average damage was 572; breakdown was 605, 706, 889, 686, 551, 133, 458, 734, 422, 462, 549, 679.

That's why.

That's some very a-typical performance for me in a Jag and distinctly different from prior LB10X performance.

While it's not 12mans peak, for pugging it's not bad at all and more consistent than LRMs or most other builds.

[Bishop Steiner]

mwhighlander, on 30 April 2014 - 10:00 PM, said:

Well that is just a reflection of the spread mechanic the LBX has.

This test is pointing out that with a pilot having 100% accuracy, the AC10 is strictly better in quickly destroying components.

As many have mentioned, those who are more likely to use the LBX10, don't have the sharpest grouping in terms of accuracy, so the cone mildly benefits them as it is easier to get SOME damage on target.

The LBX becomes progressively worse as pilot skill raises, and oddly enough becomes more effective as the fewer number of players in a game remain. Less players in a game tend to gravitate towards better brawling conditions and generally more beat up mechs with exposed internals for some bonus damage.

Yup, I'm a bad shot.

[80sGlamRockSensation David Bowie]

Bishop Steiner, on 30 April 2014 - 10:11 PM, said:

Yup, I'm a bad shot.

That is a self-implied label.

[Bishop Steiner]

mwhighlander, on 30 April 2014 - 10:13 PM, said:

That is a self-implied label.

Oh no, not at all. "The LBX becomes progressively worse as pilot skill raises", so since I am having very good success with it, conversely, my skill, and that of those I face, must be low.

[ShinVector]

LBX is weak at opening armour which is what it needs to be a good killer with its critical hit ability.
LBX also has uniquely strange problems like not being about to hit mechs barely hidden by cover.. You in cases when a bit of the top of the mech is visible when say.. AC10 round could probably hit... LBX rounds just explode all over the top of the cover...

[monk]

The LB 10-X is situationally fantastic and horrible. Given the right circumstances it will tear up your enemies and garner you some seemingly hard earned sheaves. However, it's just as likely to perform poorly if the circumstances are not good. I have been playing some of the builds in question for a while now, and I have to say I really enjoy them due to their "in your face" brawler attitude, but they do have drawbacks. They do not give you sufficient long range capability, so you more or less have to pray the enemy is foolish and will allow you to close range. Sure, you can fire them and do some damage at a pretty great distance, but your spread is so bad at that point that many of the projectiles will miss.

If I see another LB 10-X build, and I'm not in mine, it's often a simple thing to just keep range and pound them to death. I treat them much like dual ac20 builds. If they press, you simply kite them.

Is it an overpowered build? Not really, in my opinion. It can be amazingly good, but my experience shows that it's also just as likely to produce nothing for you. It all depends on what the enemy brings to the table, where the varied players go, and who you end up matching up against in combat.

Before I seriously began running my more recent builds I was one of those who said the LB 10-X was garbage and inferior. I fully subscribed to that camp. I think anyone who still follows that simply does not get how to make them effective. When you have LB-10X builds capable of pushing 10dps, you start to understand where they excel. I can assure you people start to drop a brick when they come under the assault of the LB 10-X augmented by proper accent weapons. It's terrifying. And if they think it's just flies on the windshield, they can enjoy that fallacy for the next 4 seconds before I tear a whole through their CT.

I see this build much like the AC2 builds. Are they the be all end all? Of course not, but only a bad player will stand in front of them and just let the AC2 rain feast on their armor. The same goes for LRMs being called over-powered. Sure, if you stand out in the field with no cover and just soak in the damage, you'll think they are the cheapest weapon of all time, but if you learn the game, take advantage of positioning, and make use of your team, you rarely find yourself in a bad situation that you didn't create (or that occurred simply because your team has more or less lost and you're the lone straggler being hunted by superior numbers).

Could they use a nerf? Sure, if you want to make sure nobody uses them. They are already almost non-existent on the battlefield. They are situational. That's the thing. Great in the right circumstances. Horrible in others. I really like that. I wish more weapons had better reasons to use them.


TL:DR version: The LB 10-X can score you 6 kills in a match. The LB 10-X builds can score you 13 damage in match. It's a fun build that will both delight and frustrate.

I hate that this thread is going to make more people run them. I enjoyed being a bit different in a sea of meta. Oh well. =) The game evolves.

Does anyone know if there are stats on what the % breakout of weapons being used in the game is? I'd love to see which weapons are the "garbage" choices by the community whether they community is right or not.

[YueFei]

Don't discount *defensive* piloting by your target, either. Even if you're an ace-shot with the AC10, if your opponent is twisting / juking well, maybe some of your shots end up hitting his side torso, his arms, or he jumps and you hit his leg, whatever. Even amongst top players, when they fight each other, rarely do they end a fight with just the CT cored out and everything else untouched.

Nobody's perfect. Even NBA stars miss free throws, etc. So under imperfect conditions like that, maybe the LB10X sanding off armor on all 3 torsos isn't such a horrible thing, and it overtakes the AC10 in performance once internals are exposed? Also can't discount the extra 2 tons it saves you if you're running 2 of them, which maybe goes into something else that helps give you a bit of an edge, like 2 extra medium lasers.

In the end we all got theories. It's *important* to come up with theories. But coming up with theories for the sake of having theories is pointless. Gotta test them.

Edited by YueFei, 30 April 2014 - 11:35 PM.

[monk]

YueFei, on 30 April 2014 - 11:33 PM, said:

In the end we all got theories. It's *important* to come up with theories. But coming up with theories for the sake of having theories is pointless. Gotta test them.

Totally agree with this. I come up with some "amazing" builds on paper that perform horribly. The opposite also goes for some of my "joke" builds. Sometimes you come up with a combo that just shouldn't work, and yet it does. It's delightfully entertaining when that happens.

Edited by monk, 30 April 2014 - 11:38 PM.

[MischiefSC]

So after 60 drops I can say the following:

LB10X isn't bad. At all. Not used in a build that works well for brawling and can carry it in pairs. 2xLB10X, 4xMG and 2xML is a beast that's easy to put up 2-6 kills/match in and rarely less than 300 damage, generally 500 or more. It's a pure brawler and one of the best I've played since the SRM changes.

Honestly I find it a better overall performer than 2xAC10s, which I did 12 drops in after I finished 60 in LB10X. The AC10s required me to either drop the MGs or drop engine weight, I chose the later. Pinpoint damage is great - for sniping and long range combat. Given the projectile speed nerf the AC10s got and range reduction (none of which the LB10Xs got) the AC10s are about a 200-300m weapon, maybe 400 if you're in a sniping spot. They do not synergize with anything else that build can carry though at that range; to get all weapons working on target you need to be at 120m or less.

At brawling range the perks of pinpoint accuracy vs less speed/mobility were dramatically reduced. I got less overall damage, less assists, technically less kills though at 12 matches that may be sample size difference.

For brawling the ability to just hold the trigger down on two weapons and fire lasers for precision as opportunity permits is not insignificant. It's almost impossible to overheat even on Terra Therma and churns damage very quickly.

It syncs terribly well with arty. If I use arty to soften that armor up and send a couple spots red I can chew them with MLs, then close to blow a side torso off. The high ROF and constant peppering between LB10Xs and MGs keeps return fire accuracy low. The ability to disarm and disable mechs at opportunity can also not be overstated - while the goal is, always, to kill mechs, if you're degrading their fighting performance significantly then you're improving the odds overall of a kill. If you're focused on the kill alone then you're leaving that target at full performance until you get the kill opportunity. I found I had more viable opportunities to damage and cripple opponents with this build than I do with precision kill builds. This impacts the overall statistical impact of your performance on odds of winning, total kills, assists, everything else.

My test results are an increase of KDR of 0.41 over 60 matches, win/loss improvement of 0.22, plus as a perk an approximate increase in XP/Cbill generation of almost 40%. That came both from increased damage and the fact that my component destruction over 60 matches average 5.6/match, compared to about 2.2 for either 2xAC20 or 2xAC10,ML/4xMG.

With the current AC20 nerfs and AC10 nerfs I would say that 2xLB10x, when coupled with weapons that synergize (stacked MGs for example or SRMs) is absolutely viable. While brawling is still sub-par compared to poptarting and PPC/AC5 sniping, in the arena of brawling builds I would say that for me, based on this, 10Xs are now viable. Especially for grinding cbills and XP - it's worth noting that my cbill awards with that mech ranged from a LOW of 62k to a high of 212k with an average of almost 120k/match, win or lose. I would routinely lose matches and still make 120 or 140k cbills. Conversely with a pinpoint meta mech a losing match is closer to 40k and a good winning match is more like 160k, though it still obviously peaks over 200k.

Anyone else with some results?

Edited by MischiefSC, 01 May 2014 - 06:26 PM.

[Roland]

Out of curiosity, could you post the actual stats from the combat? Did you record any videos?

[MischiefSC]

Roland, on 01 May 2014 - 06:34 PM, said:

Out of curiosity, could you post the actual stats from the combat? Did you record any videos?

No videos. I screencapped a stretch of 12 end of round stats and posted them prior, can throw up the screen caps if you'd like. Otherwise I just recorded my pre/post test stats for the build and averaged from there.

The 12 match stretch though was a pretty representative sample of my averages though. I don't want to spam screencaps, but lemme see if I can post them under a spoiler tag or something so I'm not stacking 24 pictures.

The straight weapon stats I fear I mudded by playing around with in very non-optimal builds, like on a Banshee and Orion. Even tried 2xAC2 and 1xLB10X. That put the weapons use outside of the viable scenarios I'm calling it viable for and I don't really have a way to extract those matches (about 20) from the weapons otherwise limited composition (66 total matches used).

What data would you like?

Edited by MischiefSC, 01 May 2014 - 06:54 PM.

[MischiefSC]

I'll create an imgur album to link up, that's not a bad option. I also took shots of another several matches, though in those I was in a 4man and thus I'd say the original 12 are a better indication of straight performance. If it's got Vlad on my same lance it's in a 4man.

[Roland]

It's too bad there's no videos, as they're always more interesting than the scoreboards alone, as it gives some insight into what the opponents were doing and what kinds of weapons they were using.

Aside from that, the things like kills and deaths would be useful metrics, as would all the other stats in the end of match screen.

[Khobai]

Quote

It's too bad there's no videos, as they're always more interesting than the scoreboards alone, as it gives some insight into what the opponents were doing and what kinds of weapons they were using

.
Well high damage with LB10Xs and machine guns isnt impressive anyway. It just means you were spraying damage all over and not actually killing enemy mechs efficiently.

One of the major problems with the scoring system is that the game PUNISHES you for getting efficient kills. You kill a mech with a head shot or a rear armor shot. You get less than if you shot its arms and legs and picked it apart slowly.

Damage scoring needs to be weighted based on the lethality of the location hit.

[MischiefSC]

Created an album, can't seem to link it. I've got screencaps of match screen (how both teams performed) and my end of round stats for the last 23 matches I've played, since the idea occurred to me. FRAPS hates me so videos are more of an issue.

Or we could just drop together and I'll show you how it wrecks face

For straight metrics though, here is the post where I brokedown the specifics of the 12 core sample matches.

[MischiefSC]

Khobai, on 01 May 2014 - 07:07 PM, said:

.
Well high damage with LB10Xs and machine guns isnt impressive anyway. It just means you were spraying damage all over and not actually killing enemy mechs efficiently.

One of the major problems with the scoring system is that the game PUNISHES you for getting efficient kills. You kill a mech with a head shot or a rear armor shot. You get less than if you shot its arms and legs and picked it apart slowly.

Damage scoring needs to be weighted based on the lethality of the location hit.

To a degree I agree with you.

However, I noticed something significant when I noticed my win/loss climb. Spraying damage increases the odds of your teammates getting kills that you don't. Increased component destruction was more representative of being able to exploit an opportunity to cripple an enemy when a kill shot wasn't viable. With PPCs/AC10s/AC5s you'll generally hold your fire for a kill shot, which means you miss opportunities to blow arms off victors and side torsos off banshees and Atlases. Taking the opportunity to wound means potentially missing an opportunity to kill. Spraying fire and taking snap-shots with most weapons is wasted ammunition resulting in either no damage or damage you can not exploit (20 points to a Cents left arm for example, nothing anywhere else) where as with this I'm doing an average of 7.2 points on average *every time I pull the trigger, hit or miss with each LB10X*. It's not just about the damage average when I hit, it's damage/total shots.

With an LB10X loadout and brawling that changes. It's not a tradeoff; the ROF is high enough that you're not going to miss the opportunity and a shot that isn't right on target for your CT hit isn't a miss, doing no damage - you're just doing 14 points to his LT instead of a potential 20 to his CT. That's significant for the overall results of the match - the shot you'd have missed high and to his left, doing 0 damage, or even the shot you didn't take because your bead on his CT wasn't clear, is damage delivered to other locations with the LB10X.

Again - I don't compare it to long range or sniper builds. The poptart meta is still the peak meta.

Compared to brawling builds of any sort however and it's significant. It way, way out-performs SRMs at the same weight in my experience, it measurably out-performs AC10s and AC20s for me now, in the context of a brawling build. I'll still keep the AC20 on my Atlas; I still run 3x5, 2xPPC on my Banshee.

For my Jag though this LB10X build is *significantly* superior to any of the AC40 or 2xAC10 builds I run. As a brawling build I'll take it any time over any chassis of any weight. Within 120 meters I crushed every single mech I faced 1 to 1, often 1 to 2, without exception. When I die in that build it's because I get out of position and get humped by a group or I get dismantled at range. These are issues with any build, brawling especially. Otherwise, in the scope of close range combat, this build that I would previously have said was a throwaway troll build absolutely wrecks face.

I'm not a great pilot - I don't claim to be. That's why I find the results valuable. I never had success with the LB10X before. That's why I found this interesting.

Edited by MischiefSC, 01 May 2014 - 07:20 PM.

[Roland]

Do you have the overall stats from the 60 matches you played?
I'd be interested in seeing the overall number of kills, damage, and deaths that you incurred over that set.

[YueFei]

I am also kind of dubious on the Testing Grounds approach to shooting AC10 vs LB10X at target dummies and counting how many shots it takes to down a target.

If sterile Testing Grounds shoot-em-up environments were the end-all and be-all of testing weapon performance, the old AC2 (at 4 DPS) would look fantastic compared to an AC10. For 12 tons I can get 4 DPS out of 1 AC10, or I could get 8 DPS out of 2 AC2s. Shooting at CT-only on a target dummy, it would seem that AC2 kills it in half the time.

But in actual practice, since pilots move defensively, this often is not the case.

Now with private matches available, more realistic test environments can be setup with actual pilots in the seat. Still not ideal but better than shooting at target dummies.

[MischiefSC]

Roland, on 01 May 2014 - 07:17 PM, said:

Do you have the overall stats from the 60 matches you played?
I'd be interested in seeing the overall number of kills, damage, and deaths that you incurred over that set.

For all 60 matches, only the difference in my stats for the mech. For the last 23 matches I have end of round screencaps. Here's the imgur album - I died almost every match, win or lose, but that was a matter of focusing on testing weapon performance. On wins I could have backed out when I was near death and saved my kdr but, honestly, I never tend to play that way anyway.

I'd take that 23 matches over the total 60 in a lot of ways anyway - there was a distinct learning curve in the first half of my 60 matches anyway. If I did another 60 my stats would absolutely be better now. In a way it reminds me of the old AC40 Cat back when that was new - a pure ambush predator. You catch someone alone or vulnerable and BOOM. They're dead before they have much chance to shoot you. This is a bit more involved and has less range but in application is very similar.

[Ultimax]

mwhighlander, on 30 April 2014 - 10:00 PM, said:

Well that is just a reflection of the spread mechanic the LBX has.

This test is pointing out that with a pilot having 100% accuracy, the AC10 is strictly better in quickly destroying components.

As many have mentioned, those who are more likely to use the LBX10, don't have the sharpest grouping in terms of accuracy, so the cone mildly benefits them as it is easier to get SOME damage on target.

The LBX becomes progressively worse as pilot skill raises, and oddly enough becomes more effective as the fewer number of players in a game remain. Less players in a game tend to gravitate towards better brawling conditions and generally more beat up mechs with exposed internals for some bonus damage.

I'm just providing data points really, I don't have any conclusions to make or an agenda in this.

I might set up a test to remove armor first and then see how many shots on internals each weapon needs to kill.