GU50, Tunso Testing, and Player Testing

Attacks can be vulnerable to block at any range. Being melee has nothing to do with that. If you're talking about interrupts, that's different. You can be more vulnerable to interrupt by certain enemies that have shorter range attacks that interrupt.

I think what you're suggesting is that all powersets be given equal vulnerability to interrupt at close range. Am I understanding correctly?

Just to comment more on MIDRANGE. Here's the highest level options I see the devs doing...

OPTION 1: REMOVE MIDRANGE
> Step 1: Shorten all 12.5m attacks to 7m attacks
> Step 2: Make sure all MELEE results are balanced
> Step 3: Make sure all RANGE results are balanced

OPTION 2: EMBRACE MIDRANGE
> Step 1: Make sure all MELEE results are balanced
> Step 2: Make sure all RANGE results are balanced
> Step 3: Make sure all MIDRANGE results are balance (may require revamps)

OPTION 3: IGNORE MIDRANGE
> Step 1: Make sure all MELEE results are balanced
> Step 2: Make sure all RANGE results are balanced
> Step 3: Continue to hear complaints about how some powersets get more reward with less risk at 12.5m

In either case, you need to make sure MELEE and RANGE results are balanced. No matter what choice is made, that has to happen. So why no do that first. Take our results here, verify them, and make adjustments. Then, decide what to do with MIDRANGE. It buys time and stays productive. Keep in mind if they balance MIDRANGE first while still having MELEE or RANGE imbalance, later on balancing MELEE or RANGE could throw MIDRANGE back out of balance. So no matter what they do, again, balancing MELEE and RANGE should be the first step.

However (in my opinion) testing methodology is crucial. If everything is balanced to Tunso Tests (knowing it has serious scaling issues) and then later on they find out their balance efforts don't translate to balance on the live servers, they may very well come around to the Live Testing methodology after all. If they do, they're just going to have to give everything another balance pass. That second balance pass will be redundant. So while I think balancing MELEE and RANGE is the right first step, I'm really concerned about getting them to acknowledge the issues with Tunso Testing sooner than later. So focusing on methodology and scaling is what I'd really like to be the first step. And I think nothing is more constructive or objective than what we're doing here to try to support that position.

That is why I like to ensure we...
A) Have adequate Tunso Test coverage (to ensure they'll listen)
B) Have adequate Live Test coverage (to show scaling concerns)
C) Have adequate Live Test coverage (to show a significantly more accurate state-of-balance)

But also if you look back at my Rage Test post, there's the extra benefit that we can also more objectively show the effects of any improvements they do make to powerets on PC Test. Granted people have tested before. Some have posted numbers too. But this gives such a great before-and-after picture in apples-to-apples circumstances. It's very conducive to graphs and charts. It also has a baseline to compare against. As long as we don't lose track of that baseline, we can keep track from here forward.

The thing I dislike about his test is that it's set to a small timeframe. The community "standard" - 30 seconds in my opinion showed better where powersets stand against one another. Sure Tunso's test shows some level of accuracy but in those 10 seconds it's basically burst damage. Which for electricity was adjusted but in the greater picture the dots are still lacking in damage.

Also take into account that there are 4 powers that depend on weapon damage to be effective. So "balance" under these guidelines is b.s. because I cannot effectively use a melee combo against Stompa. So therefore I can not use my power to dps because rng refuses to give me an acceptable weapon to range with. Either make weapon drops selectable to which weapon you want or remove weapon dependant ams.

30s only hows better if you're eyeballing results. While I am capturing Low, Mean, Median and High for the results...I am ONLY looking at the Mean (average). I capture the others just for curiosity. I do agree though. If you are going to eyeball results, 30s is better than 10s. But that also implies 60s is better than 30s. But all that is moot because from this point forward, charts and graphs like I posted here should not be based on someone solely eyeballing their numbers. I was guilty of eyeballing in GU50 testing as were pretty much everyone else. It should be a calculated average every time. Eyeballing needs to be done.

The problem with longer samples is that they have more opportunity for mistakes. Most of the time I just run my test until I make a mistake. Some people are better than others and I'm certainly not perfect. If I can go for 3 minutes then I would either have 18 10s samples or I would have 6 30s samples. But let's say I flub right near the end of that. Now I only have 5 30s samples whereas otherwise I'd have 17 10s samples.

I do think the longer you test and the more samples you get is generally better. But I'm also seeing that sometimes there isn't a whole lot of difference in your results between 12 10s samples and 18 10s samples. The average starts to establish itself pretty early. So the nice thing about 10s samples is that I can cut the test whenever I feel I've got a consistent result. And I can see that as I go. I get that feedback earlier and more frequently.

That is a MAJOR benefit if you're testing a lot of variations. In a given night of testing, it can be the difference between testing out 5 different loadouts vs. testing out 3 different loadouts. Because I've firsthand experienced that bigger sample sizes, 30s or 60s or killing the old 3-targets, means testing fatigue kicks in earlier for me. I start procrastinating before tests. I stop testing earlier in the night. It has a mental effect. I'm certain we'll get less results from the community with 30s samples and even less with 60s samples even if they were to test for consistent durations every time.

There's a couple things you're communicating here. I want to spend some words really separating the two.

Not being to use melee combos against certain bosses.
If mid-range gets removed, that problem would be consistent for everyone. That said, maybe Ice would have a benefit because they could stand at 6m away stationary using powers while weapon combos continuously pull people towards 0m. But there is very similar risk between 0m and 6m. But maybe there's an argument for all melee powers to slowly pull you forward to make that equally unfair. But let's stick with the focus that some bosses can't be melee'd, well you'll have to switch to a range rotation like everyone else. And all will be fairly balanced as long as they balance melee, balance range, and make a decision regarding midrange.

Having bad luck getting your weapon of choice.
I personally think weapon-dependent AMs are a good idea because they force variety. I know other people wish every AM could be "play-from-the-tray" only. I see their opinion. I respect it. I also disagree with that opinion strongly. But we can leave the difference of opinion there and let the devs have the final decision. So yes, play-from-the-tray powersets only need a weapon for CR where you need a weapon that performs. One that is balanced and that you're capable of making good use out of. And sometimes that means needing two weapons. One for your Melee armory and one for your Range armory. Where the play-from-the-tray players don't have that problem. They don't care. They'll equip Dual Pistols if they have to. I think it's a fair concern. I think vendor weapons help. I think hunting drops can make for a compelling experience (YAY...I FINALLY GOT IT....SQUEEEE) but it's also not a fair experience between the two. It's a tough one, for sure. Selectable weapons would be nice.

Amen. I used DW on my Earth took because the other options I ever got were Staff and Brawling. I do have a pair of unattuned Brawlers in my inventory. Hmmm....

I highly doubt that OPTION 2 will be taken into account, too time/resource-consuming, since 'quantum' get its revamp we live in
OPTION 3, OPTION 1 seems to be right tho and I would see it on test BUT imagine how many 'crying threads' would be created after that (who cares about the bigger picture?!), all that people can do these days is respecing from one fotm to another and Devs act like everything is fine. I have high hopes with this thread, keep up the good work.

Ideally, I would like to see the devs embrace a mid-range damage category. So damage dealt increases as you approach the enemy:
range < mid-range < melee

As far as testing methodology goes, I am in favor of adopting what Sore has laid out for his Live Test. However, if the devs truly want player testing from the masses and for the sake of consistency, a Test Room instance should be created. This idea was suggested in another thread and I know it would take some effort on the devs' part.

Either walk-in or queue for the instance and automatically get the suggested CR, stats, and pre-determined skill points for damage testing. Everyone would have the same baseline. No consumables, no trinkets. Your weapon would be whatever you equip, but with the suggested DPS baseline. Temporarily turn the three damage league buffs on if a player doesn’t have them. Might build or Precision build would automatically be chosen based on power set. Test whatever loadout/rotation you want.

It could be a very basic room with three areas for the one-, three-, and eight- sparring targets. Not much else needed.

I've had some discussions regarding this issue with leaguemates lately, and being a nerd, I've had some discussions with Zhouz on it. I do think the Range/Midrange/Melee discussion needs to be had. But to me, that's not the most concerning issue at this point.

The graphs Sore posted a few pages back make me especially uneasy. Most notably the ones that show deviations/scaling.

I tested with Light, at it's the rotation I'm most comfortable with, at both Melee and Range. I did this in 126 gear (a Tunso test and a test with full mods, though no proficiencies). In other words, a Tunso test and a Live test, as you cannot promise leagues have all proficiencies. I did the same thing using 116 gear. The scaling differed.

In other words, here's what I worry is happening: The dev team is making minor tweaks so that power sets do similar damage at CR 126. However, as the power sets are not perfectly balanced, this causes further deviation at other CR levels. This will be an oversimplification, but to put it as mathematically easy as I can, if the power sets each do damage based on a linear equation...it's like Power Set A is doing y = 2x + 1 damage and Power Set B is doing y = 3x damage. You want them both to cross at (0, 0), so you tweak the equation on Power Set A to be y = 2x. However, Power Set B, as CR increases, will completely blow past Power Set A because you didn't really balance the equations, you just put a bandaid on the first one.

Sore im a long time fan your guide along with anierna taught me to heal with nature several years ago. My problem currently is these gu's on test and the way balance is being ignored. Rage on test is so far from balance at range its not funny. Buffing melee isnt viable unless you give us a buff similar to what the tank side gets. i post this here not to hi-jack your thread but im hoping this gets read by someone who can make a difference. i think its ultimately this that if dps is balanced you dont sell respecs thats the only reason why previous powers have been released op. I have used every powerset in game except mental. really burned my biscuits when we had a gu for sorc which is my mains power and mental came in and dominated the gu. What can we as players do to see to it everyone gets the love they deserve?

Having only come back recently, and is rather on-and-off about it, I still can't understand the problem with midrange. Anyone care to explain without me having to dig up the thread for a simple explanation?

From my initial understanding, though, this only affects a certain power-type. I'm not talking about a powerset. I'm talking about the power's AoE shape. It's not just a ranged issue but a positioning issue. Using Earth powers as an example, long-ranged single-target abilities (Pebble Blast) and radial AoEs (Rumblecrush) shouldn't be affected much by any ranged options and should also be perfectly fine in a melee situation. Point-blank radial AoEs (Localized Tremor) should be excelling at melee range, although some of those abilities may have a larger AoE (Tectonic Break). Therefore, the only issues here should be concerning abilities that are either ray or cone types, whether it has limited range (Shards), or long range (Upheaval). Anything else in the powerset should not matter.

I had forgotten to put Max Damage in for my testing. I got about 1k more DPS in my Tunso test. Here's the data:

Might = 8255
Precision = 2796
Weapon = 580.6 DPS
SP = all DPS crits, but no other boosts

Loadout = Retribution, Smite, Plague, Anoint, Benediction
Rotation start = RcW/Anoint/Benediction>PcDL/ScH/RcW/PcDL/ScH/
Rotation continue = Retribution/Anoint/Benediction>PcDL/ScH/RcW/PcDL/ScH/
I was using the fraction of a second after the subsequent buffs to reposition my toon.

20326 AVG 18511.61
19171 MEDIAN 18529
17714
21116
17853
23382
16958
18895
19442
13004
22017
19477
18281
15943
21356
18550
20044
20146
17552
19453
18529
17781
13818
15992
15996
19731
17998
18291
20031
18426
16587

I dropped the ramp-up parse, but it was 10957, FWIW. Still, fairly close to Penryn's runs, but I thought I should include it. I'll try and put together a live test in a few.

Here's my live test. I only have 140 SP allocated on my test toon, so my stats aren't as high as on my live toon. I also did not have League Hall affinities or Base affinities. I did have Core Strength and Escalating Might equipped, in addition to Max Damage.

Might = 8441
Precision = 4441
Weapon = 580.6 DPS
SP = all DPS crits, tier 2 precision, 6 tier 3 prec/might

19453 AVG 23810.77
24629 MEDIAN 23539.5
16985
25529
30493
22883
25629
24573
27175
23871
18584
23255
22401
22908
24835
19347
28440
28968
22972
22536
23824
26341
29362
20962
21825
27478
17726
22062
22737
26540

My DPS values are about 500 lower, but I think my precision was also 236 lower than Penryn's.

The problem is midrange isn't a separately recognized entity by the Devs. There are some powersets that have abilities that only damage from 7 m in, but others are 12.5 m in. The Devs only recognize the 25 m and 12.5 m ranges. So, everything that produces damage from 12.5 m in is considered melee. Everything outside of 12.5 m is considered ranged. The midrange phenomenon is actually a consequence of the 7 m abilities creating a closer melee subset.

Sore early on noted that how we perceive conal AOE is different than how Tunso perceives conal AOE. Some abilities, like Time Shift, have a vary narrow cone, but it only hits four targets regardless. In the greater scheme of thing, not an issue.

The issue can be found in powers like Totem that is PBAOE, but in order for it to obtain the most damage, it must be within 7m of your target. True, it will also tick within 12.5m, but you will lose that second tick. In a nutshell, imagine that Totem was a cone. Imagine if you got the same damage at 7m as you did at 12.5 m. Well, you would never have to play at 7m. Ice is like that. It can push it's melee damage loadout out to 12.5m with no reduction in damage, whereas other powers, notably Rage, Light, and the update to Nature Canine Form, can not push the range of their melee mechanics out to 12.5m. Balance, in Tunso's mind, is obtained when all powers do the same basic damage at melee, or at full range, regardless of how far they can push their melee mechanic out to.

The best thing we can do is test range as both a Tunso Test and Live Test and post your numbers. It either lines up or it doesn't. We have baselines now. Melee is Melee. You either can or can't do it. If you can't, nobody can. If you can, everyone can. The only asterisk there is MIDRANGE and that's not getting addressed for any powerset in GU51. It will be tremendously imbalanced until it is addressed.

Live Tests are meant for a 160 SP toon. I'd also make sure you have league hall proficiencies and generator mods. I posted details about the spec a couple pages back.

In a nutshell, midrange is a compromise players make.

Devs view Melee as the old T4 days of using 1H Spin Chop on everything at point blank distance (0m) on your topic. You may be able to hit additional targets up to 7m or 12.5m away.

Devs view Range similar to the T5 days of being as far away as possible (20-25m) while spamming HB Solar Flame.

Players decided that Melee was too risky and not viable, courtesy of the T5 days. So they stay at range. Well, they learned that Melee-oriented abilities hit harder. Some of them hit targets as far as 12.5m away. They decided to use these Melee abilities in a ranged manner. They use 12.5m strong Melee abilities with Ranged weapon combos. They invented MIDRANGE. A distance where you can get some (or all) of your Melee-strength damage without having to be point-blank (0m).

Players kept asking for 12.5m distances for abilities. Devs assumed we wanted that to hit more targets as we were point blank. They didn't realize we wanted that to be able to scoot back to midrange. So players started thinking of balance only in terms of midrange performance. Based off our feedback, certain powersets got to take advantage of that oversight more than others. Ice and Fire can do the same damage at 0m as 7m as 12.5m. Most powers sacrifice something as they scoot back. They sacrifice nothing. So it is a topic needing to be reconciled.