Hopeling

Right, I'm not saying that pylons are a better metric than farming overall. I'm saying that neither pylons nor farming are good examples of normal gameplay. Farming is maybe less dissimilar, but still not very similar: you get a fuckton of inspirations, you have an insane saturation of enemies, you have no disruption from teammates. That's why we're trying to come up with another metric.

I don't understand what there is to be confused about. People think TW is better than Spines in normal gameplay, but not better at farming. This is not a contradiction. It's exactly the same argument as you saying that TW is great on pylons but not in normal gameplay. We're trying to get metrics that more closely correspond to normal gameplay, rather than only using extremely limited metrics like pylons or farming. The meta farming build used to be SS/Fire, which doesn't have a second aura and isn't especially user-friendly because you have to manage crashes. With the rage crash -def now unavoidable, Spines seems to be the next-best thing. Its popularity over, say, Rad/Fire can also be partly attributed to the fact that some people wrote very popular guides about it shortly after Homecoming first appeared.

Part of the reason I don't want a thing that involves a second player is that it means I probably can't contribute to testing. I don't have a reliable second player, and my schedule is erratic. I'm not sure no inspirations is a fair condition if we're trying to represent normal gameplay. For example, Super Reflexes and Shield Defense look a LOT worse without inspirations, because they have absolutely no healing or endurance recovery, rather than having it in limited but nonzero amounts from consumables. But I can go with no inspirations as the default test condition if that's what the consensus is. No, but you do face AVs. I mean, you don't face them often, but the subset of people who came up with pylon tests did. The whole reason that pylon tests became popular is because Scrappers (and some others, but the pylon thread was always a Scrapper board thing) wanted to see if they could solo AVs under challenge conditions, and a pylon has about the same HP and regen as an archvillain, so it made a reasonable measuring stick or dry run for comparing builds and attack chains. Pylons really are reasonably representative of the amount of DPS you can do in an AV fight. Likewise, that's why the default test condition for pylons is "no inspirations, no temporary powers". It's not that nobody wants to know how much DPS inspirations or temporary powers are worth. It's that AV hunting was the thing people did for fun and bragging rights, and AV hunting with inspirations and/or temporary powers was already pretty easy, while AV hunting without them was still a challenge, so that became the standard. If we're measuring under farming conditions, we're only going to find out which sets are the best at farming. I don't think the answer there is going to surprise anybody: it's Spines, followed closely by SS and Rad. But that doesn't tell us much about regular content.

I'm not sure that's a better test condition than an NPC. It also makes the test harder to construct, since you need a second player, and it introduces an extra variable of how evil that player is. (And it cuts inspiration drops in half compared to the solo condition, which is another variable to account for.)

Well, what you don't do is throw your hands up in the air and decide that game balance is unimportant or impossible. There's almost never going to be a single metric you can point to and say - aha, this is ironclad proof that powerset X needs to be brought down by exactly Y%. You may not have any metrics that aren't confounded in some way. But you take whatever metrics you do have and look at the picture they make in gestalt. A powerset that overperforms in one metric may just be a fluke; a powerset that overperforms in every metric you can think of may be a real issue to be addressed. Determining what's the right change to make is then another difficult problem, but not an impossible one. You come up with possible changes, see how they affect your metrics, and decide on a candidate. Generally, it's better to make small changes rather than sweeping changes. All of us in this thread seem to agree that, if TW is changed at all, it should be something conservative, with maybe further adjustments later if the initial ones aren't enough. Changing every melee set except TW would be an example of an unnecessarily broad change: if people like most melee sets just fine as-is, why mess with all of them when you could instead just change one? I agree that a farm map is probably a bad idea, especially if it's the one on the asteroid map with a bunch of patrols. The whole point of that map is to achieve target saturation way beyond what you can have in normal missions. An asteroid map with ordinary spawns (but not patrols or ambushes) of not-too-dangerous enemies should suffice. We may not even need a custom enemy group; Council or Freakshow may work fine. Steampunkette's idea of using an energy blaster NPC is a good idea; it allows a reasonably controlled test that still incorporates the idea of teammates disrupting Momentum cycles. This shouldn't be the ONLY condition we test under - a teammate spamming Explosive Blast without slotting for KB>KD or knowing how to use positioning is closer to a worst-case scenario than average-case - but it should be one of them. We still need to agree on some conditions for tests before we can get started. We don't have to pick just one set of test conditions, but as a measure of baseline performance, what about something like: SOs only. Slotting can be at the player's discretion, as is the rest of the build, but when in doubt, slotting attacks with 1acc/3dam/1end/1rech is a good default. Run at +0/x4. You have enough targets available that AoE is worthwhile, but not so many that it completely overshadows ST damage, and some spawns will have bosses. Inspirations are allowed, but you have to start with an empty tray. No stocking up on 20 [Enrage]s before zoning in. Level 50 character, but no Incarnate powers.

It's great that your coalition is having a good time, but again, that doesn't seem to stop it from bothering other people. It's great that making melee sets stronger won't affect your friends, but of course it won't if they don't even play melee. I've pointed directly to threads full of people that such changes would upset. When the proposal was changes to TW, you (very reasonably) insisted that the burden of proof was on me to justify why a change should be made; it's not enough for me to say that I personally don't mind a reduction. That still applies the other way around: if all melee sets should get buffed, that still needs to be justified. It's not enough to say that you personally wouldn't mind it.

That does not seem to prevent other people from having problems with it. Like I said, if the difficulty settings went higher, this would not be an issue. Game too easy, just turn up the difficulty. But we're currently at a point where at least some decent fraction of the playerbase finds that turning it up all the way is still easier than they want. Across-the-board buffs amount to reducing the difficulty further, in a way they can't opt out of. Meanwhile, if you feel that current difficulty levels are fun and challenging, why exactly do you want to reduce them?

Buffing powersets across the board amounts to making the game as a whole easier. This is not necessarily a popular thing to do. For example, see the power creep thread in the general forums right now, which is not the first of its kind and probably will not be the last. Homecoming has made the game easier-as-in-less-tedious in many ways, like reducing merit costs or allowing double XP, but they haven't literally made enemies easier to kill across the board. Such a change would be largely unprecedented. If the difficulty scale went higher, this would be less worrisome. And maybe I'm overestimating the impact of buffing melee sets. But it is at least a potential concern. Since nobody ever answered this: yes, every TW attack has a chance to crit, and the area attacks get a separate chance to crit on each target. This part works the same as any other scrapper set. Like most sets, it has one power with an increased crit chance (15% instead of 10%), in this case Arc of Destruction.

Define "better". The scrapper will deal more damage, and benefit more from the +damage buffs. /Shield doesn't have enough resists for the lower scrapper cap to be super important most of the time. Clobber crits are beautiful. You won't need to build fury to deal good damage, which is especially nice when you want to lead with Shield Charge. The brute will be tougher, and can viably tank for a team thanks to the taunt component in all their attacks. 90% resists on top of /Shield's defense is really really good, even if you can only get there during One With The Shield. I think /Shield tends to be slightly better on scrappers than on brutes, but is still very good for brutes, and War Mace doesn't really favor either. It's basically a question of taste.

Eh, this is only slightly true. Typed defenses come in pairs, so there's still only three things to chase: S/L, E/N, and F/C. Psi is an issue for typed defense sets, as is Toxic (there's not even such thing as toxic defense). But these aren't common enough to be large problems.

Regen ticks don't cause 200-DPS swings. Regen occurs in 5% ticks; one regen tick for a Pylon is about 1500 HP. At the levels of DPS we're talking about, it takes 2-3 extra seconds to deal 1500 damage. A 60-second run is 766 DPS, a 63-second run is 736. Not a huge confounder. There's always going to be randomness; no test can remove it. Crits are random, procs are random, misses are random. All you can do is smooth it out by using a larger sample size. Longer fights do that, but so does averaging multiple fights together. The only issue with averaging multiple fights instead of one long fight is that it can overcount things like Hybrid whose cycle time is longer than a pylon test. But Hybrid is basically the significant factor that has a cycle time of multiple minutes like that.

Sure, but doing five pylon tests and averaging them smooths out variance too.

Yep. Battle Axe is essentially a clone of War Mace, with the powers in a slightly different order. Besides slightly tweaked numbers on the cones which are basically a wash, the two main differences are: Lethal damage instead of Smashing damage Swoop instead of Clobber And both of these things are drawbacks! Lethal is resisted more than Smashing, and Clobber is much better than Swoop. Yeah, Mace was widely considered pretty bad before the Clobber change. I know MA got some buffs late in the game's life. I haven't played it or heard a lot about it since then, so maybe it's OK now? Dunno. I've only played one KM character, but I was kind of underwhelmed, and that seems to be a common feeling. I don't really know where it stands in objective terms though.

A custom punching bag might make a better version of a pylon, but that's still testing solo sustained single-target DPS with zero movement in a vacuum. I don't know what would constitute a more holistic measure of performance. Something like "time to clear a Battle Maiden map" would be a kind-of-reasonable metric of overall (AoE+ST) damage under something resembling real-world conditions. There's movement, there's a mix of bosses and minions, you're facing groups but not ridiculous target saturation like an ambush farm, you have to hunt down runners. We could test that with a variety of powersets, with/without IOs, with/without Incarnates. But we'd have to nail down the parameters (are inspirations allowed? Can you go in with a full tray of reds? What about a full tray of Furious Rages?) And we're still testing solo performance in a vacuum, while part of Infinitum's objection is that the set interacts badly with kill-sniping from teammates. I don't even know what kind of metric to use for that. Since Clobber became a Knockout Blow-level attack with a stun attached, instead of just a stun. It's been quite good since that happened. Proliferating it to Scrappers in i21 made it more popular too. Broadsword, Katana, Battle Axe, and War Mace are all very similar structurally, but with tweaked values. War Mace and Katana do pretty well overall; Battle Axe and Broadsword are basically worse versions of those and probably should get buffed somewhat. Martial Arts, Ice Melee, and Kinetic Melee are also candidates for underperformance, I think.

There's a lot of room between "literally everything is fine and absolutely no improvement can be made" and "the whole game is melting". There is quite a lot of feedback other than contentment about the state of game balance. People routinely complain about power creep, about how too much content is easy, about how new sets seem to be categorically better than old sets. There are plenty of people asking "why play X when Y is clearly superior?", and the Flavour of the Month numbers show that this isn't idle speculation, TW really has displaced other sets to a significant degree. I've personally talked in this very thread about how TW's superiority pretty directly sucked the fun out of several other power sets for me. On the other side, people routinely complain about underpowered sets like EM or Force Field. These are all game balance complaint. It's not that the righteous Bufficans defend our freedom while the degenerate Nerfocrats try to take things away; buffs and nerfs are both tools to keep the game balanced, and balanced games are more fun. You and I have a difference of opinon over whether TW is balanced, but not over whether balance is desirable. You say we need more data, which is also fine. I'm asking you: what kind of data? What metric convinces you that TW is fine, and how could that be supported with something besides anecdote and opinion?

It's funny how all successful games, including this one, get where they are by employing teams of professional full-time busybodies. Almost like game balance is important or something.

Yes, I agree that doesn't have it covered. That was the point of that comment you quoted. I'm asking you to help me come up with ideas for what kind of data would cover it. I honestly don't know what metric to use for performance in regular team play; I have my own subjective feelings about it, but yours are different, and there's no clear way to reconcile them. If you feel that it's fine in regular play, by what measure are you saying that? How could that be made objective?

I'm saying there are chains that use FT only once per cycle and do almost as well as running it twice per cycle. I don't think there are any chains that don't use FT at all and do very well, certainly not while also skipping CB and TS. I honestly don't see how to make a gapless attack chain using just FT, Arc, and RA, so I don't think I can answer your question directly. But for comparison, using your previous chain Rend-Arc-Whirl-Rend-Arc-Whirl, you do 598 damage base during Build Momentum. If your Build Momentum chain is Rend-FT-CB-Arc-FT-Rend-CB-FT, you deal 731 damage in the same time period. So if the same proportions hold, your 295 DPS would turn into about 400 DPS.

Again, what kind of testing? I really don't know what kind of data you would consider worthwhile here. Any plausible data is either going to be in a sterile testing environment like a pylon, or heavily confounded by other effects like the powers of your teammates. So far, the evidence we have is: TW is the most popular scrapper set at 50, and disproportionately popular for Tankers/Brutes once you get past the iconic pairings (Inv/SS) and farm builds (Spines/Fire). TW consistently posts top times on pylons, TF solo runs, and other metrics of solo performance. TW attacks numerically outperform other sets across the board. I would say that gets the "sterile testing environemnt" side pretty well covered. I agree that some kind of data for performance in a team environement would help, but I really don't know what kind of metric to even use for that. Do you have any ideas?

On further testing, I think you're right. I have two big gaps trying to run Rend-Arc-Whirl-Rend-Arc-Whirl during BM - one gap between Whirl and Rend, a second gap between the second Rend and Arc. Build Momentum actually expires before I get through it. With more global recharge (I already have +156%), I'd believe you can get through all six attacks, but it doesn't look plausible to get it gapless. I had thought that FT seemed to start recharging before its animation was done, but it must have been in my head, because it lines up just about perfectly now when I'm trying to replicate it. Double FT is not the big offender. Every attack in the set overperforms, even with lower amounts of recharge or in AoE situations. You to run a chain like SlowCB-Rend-FT-Arc-TS-SlowCB-Rend-FT-Whirl-TS with just Hasten and SOs, which only does about 6% less DPS than Rend-FT-Arc-CB-FT. Gaining 6% DPS going from SOs to a purpled-out build is not at all unreasonable.

Bruh, that's what you spent the last 10 pages doing with Follow Through. Meanwhile you've also repeatedly told all of us you don't believe we've even played the set. A little less condescension would be nice.

"Best" is admittedly a loaded word for me to use here. What I mean is: in a decent ST chain, Follow Through is going to do more total damage than any of your other attacks (you can see this directly if you run HeroStats on a pylon test). It doesn't hit the hardest or have the highest DPA (that's Rend Armor), but it does hit almost as hard as Arc (101 for FT, 108 for Arc, 134 for Rend), faster than other attacks (1.188s for FT, 1.716s for Arc, 1.452s for Rend), and recharges faster than either (10s for FT, 20s for Arc, 16s for Rend). It's your bread-and-butter attack. You can put out three FTs for every two Rends, and two FTs for every Arc. Which leads us to... I mean, that's fine. Nobody said that you have to build your character for max ST DPS. It's one thing to not build for pure damage; you don't HAVE to stack purple procs and Musculature and Assault to get decent performance out of the set. But you didn't just not specialize in ST damage; you deliberately sacrificed a significant amount of ST damage for other things. This would be like me slotting Total Focus for stun and then saying that it doesn't hit hard. Sure, it doesn't with that slotting, but that's not the power's fault. Semi-relatedly, even if you don't have FT, Whirling Smash still shouldn't be in a single-target chain; on a single target it has the worst DPA in the set besides Defensive Sweep. Titan Sweep or Crushing Blow would be a strictly better chain. No, in fairness, I think Infinitum is right on this one and I'm the one who made a mistake. TW attacks do weird things with recharge; I think that (in Momentum, at least), they start recharging when the power starts activating instead of when it finishes activating (or something like that; see below). That means that, in a Rend-Arc-Whirl-Rend-Arc-Whirl chain, Arc only needs to recharge in (1.716+1.188+1.452)=4.59s. That may still have a small gap for realistic levels of self-buffed recharge, but nowhere near 1.3 seconds. If I'm right about how TW handles recharge, the chain Slow Rend - FT - CB - Arc - FT requires FT to recharge in less than 2.508+1.188=3.696s when you use it on both sides of a slow Rend, which requires +170% recharge in Follow Through (counting both slotting and global recharge bonuses). In practice, my scrapper has a gap after Rend when doing this without Hasten (99% slotting + 86% global=185% total), but no gap during Hasten (255% total), so maybe it doesn't quite work like that I guess? So the answer to your question is "somewhere between +185% and +255%". Assuming ED-capped recharge slotting, that translates to somewhere between +90% and +160% global recharge. Doing it in a team setting doesn't seem to be much more difficult than doing it solo. Being able to stand still for five seconds in a row isn't THAT rare.

They don't, actually! Sentinels cap at 500% too. Trust me, I'm a huge inspiration whore, hence my avatar 😉

I really like the Sentinel version of /Elec. Its regeneration is kind of ridiculous, and with some defense from IOs on top of near-capped resists to most types, it's very tough. Immunity to endurance drain and high resistance to slows is also a nice perk. I agree that /EA might be a better choice if you're going to fly though.

Just did. Bopper has them here. With Momentum and Arcanatime, they come out to: Defensive Sweep: 1.188s Crushing Blow: 1.452s Titan Sweep: 1.188s Rend Armor: 1.452s Arc of Destruction: 1.716s