Arcanaville

Build decisions aren't usually reasonable to judge in isolation. There are players spend hundreds of millions to slot purple sets primarily for the +recharge. But +10% global recharge is not that big of a deal in the grand scheme of things. However, most players don't usually do *only* that: that specific set is usually part of a comprehensive build strategy (even if an informal one) generally surrounding "grab as much recharge as you can." In the grand scheme of things you're not going to notice one damage proc in one attack power either, so when it comes to deciding if that slot should be put into Hasten, or into Health, or into an attack, what you decide to do may depend greatly on what you decided to do elsewhere. One proc doesn't mean much usually. One slot of recharge doesn't mean much usually. Few decisions "matter" in isolation. The extra slot of recharge might matter to a speed-driven build. It might matter to someone that was trying to minimize the downtime of Hasten because they are in Granite and every bit counts, or because they are Dark Armor and 10 seconds of downtime is significant to Dark Regeneration, or simply because there's no better place to put the slot. It might make more sense to put that slot into a 5-slotted attack and add a proc, if you have a build that can take advantage of that extra proc.

I guess when I build, most slots in most powers are not, by themselves, extremely important. The first forty might be, but the last twenty might be tweaking the build looking for the best of the available options. Cutting the downtime of Hasten is just as valid as adding a percent of DPS or cutting endurance burn by a couple hundredths of a point per second. That's often the choices being made at the end of the build.


Incidentally, another way to look at the third slot is to ask what the average overall recharge buff of Hasten is, averaged over its up and down time. Perma Hasten would be +70% of course. Two-slotted Hasten (no global recharge) would be about +42%. Three-slotted Hasten would be about +45.7%. The third slot is acting in average terms like a +3.7% global recharge buff.

So its worth about half the strength of an LotG. That's true so long as the difference in slotting doesn't make either case go perma. That perhaps frames the choice in better terms relative to the other options players typically choose to take or not take in the invention system. People are willing to put a slot into a defense power as a mule for an LotG, and pay a hundred million inf for the enhancement itself (or use merits for it). If its worth a slot and a hundred million inf, is half the buff worth a slot and less than a million inf?

That seems to me to be more of a personal choice than a question of objective value of the slot.

The devs don't give out base defense anymore because it works against attacks they deliberately want to make untyped. However, even if they allowed this as an exception, Base defense stacks with types, its not overridden by them. In fact, when something debuffs your defense, what they debuff is base defense (so they don't have to debuff all the different vectors separately) and that works because that "negative base defense" stacks on top of every other defense type, essentially debuffing them all simultaneously.

Similarly, an inherent tanker +Base wouldn't affect only tankers with no defense, it would add to all the defense types the tankers might have. In particular it would stack with powers like invincibility, and indirectly powers like RTTC (which debuff tohit).

The difference between 2-slotting Hasten and 3-slotting Hasten with level 50 IOs is about 16 seconds of downtime, not five, with no other global recharge. For the difference to be five seconds or less, net global recharge outside of Hasten has to be +109% or higher, almost precisely at the point Hasten becomes perma when 3-slotted.

For reference, the ED-adjusted difference is about +16% recharge (99.08% vs 83.32%).

This happens more often than most people think, even more often than I'm personally aware of. I wasn't really a part of the Dominator changes, although was aware they were coming, but I think the player community was reasonably well represented by the players that were involved. And perhaps its a good thing they were consulted outside the formal structure of a player board. That means they could not be personally held accountable for unfairly representing dominator issues: that ultimate responsibility still falls to the developers.


My contribution to the backstage Dominator discussion for what it was worth was this: I told Castle that in general, I felt the majority of players would love the changes, but I told him the controversy would center on the fact that some players like the up and down nature of Domination, and some would prefer a more smoothed out performance, and that difference was intractable: he'd have to pick one or the other, and whichever way he went the other group would cry foul. But I think that he already knew that from the feedback from the Dominator players he was already consulting with. I'm sure many of the players that preferred the damage buff in Domination thought the process was rigged by the feedback being given, but Castle knew he was in a no-win situation there, and there was two genuine sides to that issue. He still had to pick one, and he picked the one he felt would benefit the most players.

I think that the players that the devs reach out to tend to be players that give that kind of feedback: they might be advocates, but they want people who can see all sides and who can provide objective and calm feedback. Impassioned, but reasonable, and situationally aware. Its not a bunch of players all trying to get their personal pet project done. The devs wouldn't keep going back to players that kept doing that.

I also said "hey, those numbers get kinda close to blaster numbers don't they?" but that's a story for another day, and that story isn't over yet.

The counter-example are WSTs. From the perspective of Incarnate progress the Notice of the Well and the component drop upon completion are far more valuable than the shard drops you're likely to get from even a plow. And yet players run them all the time, both fast and slow.

Also, the devs stated that shards allow players to make "progress" in the incarnate system already through defeats. Having NPCs in Dark Astoria drop threads instead of shards but at comparable rates to shard dropping fully satisfies the devs promise to provide progress for street sweeping. It does not imply that the rewards for defeats within solo incarnate missions is likely to be significantly higher than that.

And as I mentioned, that's not the important reward from the perspective of pacing. Once your slots are unlocked iXP no longer matters, and you'll unlock them generally far faster than you'll earn the components to craft the powers that go in them. If component earning is going to be slower in the solo path than in the iTrials, its highly unlikely the thread drop rate will be appreciable enough to make defeat rewards comparable to end of mission or end of arc component or merit-type rewards.

I find it difficult to believe the devs will not backload the rewards primarily at completion.

The arc itself doesn't have Statesman in it, so it can't really alter your opinion of Statesman. I mention the arc only to demonstrate that the Smoke and Mirrors events are canonical, which implies the events in the comic book are likely canonical as well, and not easily dismissed as "not in the game."

Another touchstone is the Cyrus Thompson statue in Kings Row, which is dedicated at the end of the Smoke and Mirrors story arc in the comic book and was added to the game presumably as a canonical connection to that story, or possibly because they tricked Hickman into accepting that instead of actual payment for his work.

And yet stalkers tend to take Build Up.

If people don't want to be good most of the time, but all of the time, shouldn't the critical rate be 100%? 40% is not even half the time.

Its impossible for the devs to regret the cottage rule. If they ever regretted not making a change due to the cottage rule, they'd make it now. The cottage rule only requires certain changes to be made only as a last resort and only when the devs feel the change is both important enough to justify, and can only be done that way. They can't regret the cottage rule, because if they feel the rule isn't in the best interests of the game the actual rule says to override itself.


The question "what do the devs regret" depends on which devs you're talking about. For example, I have no real first hand knowledge of what geko regrets, although I suspect there are a lot of things he does. I know Castle had some retroactive regrets about things that happened before he got there, like the original design of how buff strength stacked in some cases. The fact that Champions Online changed those mechanics suggests geko had the same regret. But in terms of things he actually did, I know he regretted some things, although I'm not sure I should be precise there.

I don't think the current designers have been at the helm long enough to have too many regrets: I think what we're seeing now in terms of things like powersets and critter design is an expression of how they want to do things, which is a bit different than how Castle might have, or Geko before Castle. Castle, for example, was I believe a much more conservative designer in general than Black Scorpion is.

If there is a person for whom the question is most meaningful, its probably Positron. He's been making executive decisions for a long enough time to have regrets, and sufficiently distanced from day to day implementation that its not specifically within his power to simply change those things he regrets (Positron by his own admission is a delegator: he prefers to give his designers lots of latitude and not micromanage them, which is probably also why he wanted to get his hands more "dirty" working on the end game and incarnate system initially).

I doubt we'll ever squeeze that list out of Positron, because the obvious follow up question would be "if you regret them, why not change them?" Which is another way of saying the things the devs regret most are probably the things they can least change now because of precedent, player expectation, or the sheer amount of work it would take to alter it after the fact.


If I had to stick my neck out and guess at a really big non-obvious thing the devs regret and can't trivially change, it would be deciding originally to make City of Villains a stand alone game. That decision cascaded into a number of other decisions, and those decisions cascaded into even more decisions, almost all of which are now far more problematic than beneficial. From that one decision, we have the split markets, the dilution of the players into more zones than necessary, failing to design the red side archetypes on the assumption they had to coexist with the blue side ones, all sorts of powerset proliferation issues; the list is endless.

If the decision had been made to make City of Villains embedded in City of Heroes, and all that work went into extending and revamping Paragon City zones to accomodate villains, if the red side archetypes were designed to mesh with the blue side ones, if all that work was concentrated on a smaller but deeper footprint, we'd almost certainly have a much better game now.

May?

Theoretically speaking, you could quit an iTrial at any time and still earn iXP from defeats, but iXP is the easy half of incarnate progress. The components to craft incarnate powers is the more critical half, and that requires you actually complete tasks. If you can make progress everywhere by simple defeats without actually having to complete anything in the solo path, I'll be gravely disappointed. And I'll probably express that disappointment by exploiting the heck out of that reward structure until Marty has a stroke.

You say rewarding on task complete instead of intra-mission progress is a recipe for disaster, but the reverse is a disaster without the need for crafting.

Err, Smoke and Mirrors is not an AE guest author story. Smoke and Mirrors was originally a story arc in the Image Comics. It was canonically put into the game as an Ouroboros arc given by Twilight's Son. In the arc, the player takes the role Statesman himself played in the comic.

My interpretation of the arc is that the events of the comic book are canonical, but for a timeline that has been altered and Twilight Son intends to partially restore. "Canonical" is a bit slippery when it comes to Ouroboros and the Menders.

The combination of these two alone would very obviously be broken:

a. +Def increase % for tanker and for each team member within the tanker's taunt aura range.

c. Tanker's taunting aura causes -to hit for all foes in range of the tanker and those team members in the tanker's taunt range.

I'm assuming, of course, you didn't intend to suggest actually debuffing your own team mates and that's just a typo. But just the consequences of granting Invuln and Willpower tankers these two abilities would be enough to disqualify them as practical.

Except that is a hypothetical assuming a change to the turnstile that I'm currently unaware of any plans to implement. *If* that occurs at the same time the solo incarnate path is introduced, *then* the asymmetry I mentioned would not exist in the form I described. But its the very fact that trials have to be organized at the moment - or rather they almost always are - while the solo path can be started at any time that generates the strong asymmetry I mentioned.

However, even if that change occurs, its not clear what happens then. People can now leave their solo path for the trials, but if the solo missions take significant time, it isn't necessarily very likely that someone would bail out of it near the end to join a trial. A significant percentage of players would commit to the solo missions once they started them, and may not be willing to exit them to join a trial at a random time.

Keep in mind why we can't queue in missions in the first place. There is the belief that if you don't want to join because you're busy, you don't have to. But if too many people do that, it could disrupt a trial trying to start if too many people drop or fail to join. The attitude of "I'll queue, because its entirely optional if I join or not" is precisely the mindset the devs are trying to avoid. They want people to commit to the league or the queue, so when it does launch they are far more likely to agree to go.

There's a huge difference. Badges aren't just things they decided to add to the game, and then decided how you'd get them. Its not like they said "lets add a Task Force Commander badge: now how should we allow players to unlock it?" The mode of awarding a badge isn't arbitrary: it exists *solely* to award for a specific activity. Each badge is bound to the activity that awards it. Allowing players to buy their way out of that activity damages the entire badge system in a way that is completely impossible for the costume system.

Costume parts are not normally bound to the method of acquisition. That's the exception to the rule. Badges are normally bound to the method of acquisition by design: that is the rule not the exception. Tampering with that in the badge system runs the risk of making the entire system pointless. That's not true for costumes.

The canonical television example are the various Columbo series, where the murder and the murderer are nearly always revealed at the beginning, and the whole point of the story is to watch the noose slowly tighten around him or her. In fact, not only do you know who did it, you almost always get the notion that Columbo himself had a pretty good idea who did it from the start as well, and the story focuses not on figuring that out, but seeing if and how he'll be able to prove it.

In any event, there are lots of reasons why you'd signal what's going to happen in a story ahead of time, most of them revolving around the fact that the story isn't about that thing, but about the events surrounding that thing.

I'm only an extremely casual marketeer these days, but the converters have me looking at market numbers and economic calculations very closely.

I have to be honest when I say that getting me to think about marketeering closely is not something that should normally be considered a good sign. Converters are going to do a lot of good things I think, but still, I'm looking at my computations, and I keep thinking I've forgotten to carry a two or something. Its a little too good to be true if you know what you're doing. And I have indeed noticed a shift in activity exactly in line with what I think smart marketeers should be doing right now.

Well, if you're going to slot it in a toggle, recharge is not that useful, and the full set overslots damage far over the ED cap, so a logical option is to replace the Armageddon Damage IO with something like the Scirroco's Acc/Dmg/End IO. That reduces the amount of damage slotting but its still over the ED cap, and you gain more accuracy and endurance reduction in the damage toggle, both of which are useful.


Also, its a bit late but Degenerative is not the best option for Interface unless you plan on teaming a lot with other players that all slot Reactive. Reactive is, at the moment, in all respects better than Degenerative. For reasons not worth dumping a ton of math to explain, -Maxhealth is *exactly* identical in overall effect to -Res, except in numerical strength calculations (-res stacks differently with more -res than -maxhealth stacks with more -maxhealth, and the actual calculation for the strength of the effect is different - but the "flavor" of what happens is in all respects the same). And Degenerative has two things working against it. Its effective strength is lower than the -Res in reactive, and it has been explicitly gimped against AVs and Monster class critters (in a way that it shouldn't be in my opinion, but the devs haven't said they intend to change it so far).

For a solo player, Degenerative will never be better than Reactive**. Degenerative makes more sense in a teaming environment, where Reactive is so common it often stacks to its maximum stacking ability, which would make any more Reactive in the team have no effect. In that situation, Degenerative would do something where Reactive would do nothing.

But except for Monsters and AVs, while Degenerative is not as good as Reactive, its not too much worse either, so its not the end of the world, you'll only notice a major difference for those two critter types. For reference, each stack of Reactive debuffs -2.5% resistance, so four stacks is -10% resistance. In terms of effective overall effect, one stack of Degen debuffs for an equivalent of -2% resistance, and four stacks debuffs for about an equivalent of -7.4 resistance, all assuming Degen still debuffs -2% maxhealth per application (I haven't checked recently).



** Except in one case. If you are facing something that has 100% resistance to a damage type, that's impossible to debuff with -Res: the debuff would also be resisted 100% and have no effect. There aren't many things that have that level of resistance and would be vulnerable to Degenerative.

Setting that question aside for a moment, there's a related thought I've been thinking about off and on regarding the trials, particularly in light of my presumably well-known general animosity for the design of the Underground trial. I've been thinking about a metric that measures the average number of minutes expended per reward earned for each trial. Basically, you'd keep track of the number of player-minutes each trial takes, and the number of threads, merits, and component drops it awards. Simple enough. We could figure out how many minutes it takes to earn a component drop in Lambda, or Keyes, or the Underground.

But what about failures? Well, we could simply add those minutes to the bucket and say they generated no reward. So the average number of minutes per reward would go up. And we can make sure the metric can't be exploited by making sure only player-minutes for players that have full participation scores are counted. If a player doesn't participate, they don't count. So you can't fill a trial with dummy accounts and let the timers run out while doing nothing just to make that trial look bad.

Given both the duration of each trial, and the failure rates, I wonder what the average minutes per thread, per merit (both types), and per component drop would be for all of the trials. I wonder if they are remotely close. And if they are not, I wonder to what degree failures cause one trial to net far less rewards than others (normalizing based on drop types would be an extra bit of complexity for such a metric).

If Lambda was the fastest trial but had a 15% failure rate, that might be acceptable. Players would still run it, because on average they would still come out ahead. But if the longest trials also have the highest failure rates, that basically implies the average length of the trial per reward is even longer than the average successful run, and its entirely reasonable that trials that both offer poor rewards per minute *and* have high failure rates would be less palatable.

But from a computational perspective, I wonder if that is a useful metric to analyze the trials with. Its ok if Underground or TPN offer a lower reward per minute if they are intended to be harder, but does anyone know *how much* lower they are, factoring in both duration and failure rate? Lower is one thing, but are they lower than they need to be?

Which is specifically what I said, and specifically what I said you didn't say. You said, by way of trope, that the devs are deliberately trying to make the players fail. I said that wasn't true, that instead the problem was that they were taking the difficulty level to too high an extreme at times, and judging overall playerbase success by what the best players can accomplish, not what the average player groupings can accomplish.

Setting aside the "never be satisfied" comment, I agree completely with that statement above. And its exactly what needs to be focused on, not a most-likely false interpretation of the developer's motives.

People have been talking about the inside ball at Amazon and Google for years, just not in public. The noteworthy thing about that post was that it wasn't brought down soon afterward. But I remember talking to someone years ago that put it this way: Google attempts to monetize services (primarily through advertisement). Amazon has decided to attempt to monetize infrastructure through what the plus post calls platforming. And the work necessary to make that happen started a long time ago, when it wasn't remotely obvious to most people that was a good idea.

My contribution to the conversation was to remind him on the surface, Amazon is a bookseller and Google is a search engine. But that's incidental. Amazon is a clustered computing service provider that happens to primarily offer one computing service, and Google is the best internet advertising company in the world.

I will say this also: Yegge says Google "does everything right" but in my experience that's not true, nor do I think Yegge means that literally. What Google tends to do is treat their employees right, and one way they do that is to place trust and responsibility in them to invent solutions to problems. There is no singular "Google Way" of doing things per se: Google engineers can, within some limits create their own solutions to problems. That has good and bad aspects, but when you allow your employees to take responsibility for, and work out the solutions to problems, they tend to do so: if the solution fails, Google didn't fail, *they* failed, because they've taken the responsibility given seriously. Google did "the right thing" in letting them devise a solution, and they just let the company down by not delivering. So Google gets most of the credit for everything that goes right, and engineers tend to blame themselves when things go wrong.

From the outside, Google sometimes does the right thing and sometimes does the wrong thing. From the inside, Google always does the right thing. Because from the outside, Google is a black box that spits out internet services. But from the inside, "Google" is the rest of the company besides me. I don't always do the right thing, but "they" do.


On a related note, I have always found it somewhat amusing that there is this strange code involving inside dirt. I was talking to someone a few months ago who was dishing some really juicy dirt on a chief executive of a very large company. He was giving me the entire sordid details of a project that jumped the rails and the ludicrous story behind the acts of the players. And after telling me all this, I asked him a question about a specific aspect of that project's technical details, and he looked me right in the eye and said "I can't tell you that."

We've all been there.

You also need people to unironically call themselves his "filthy assistants."


I think the canonical characteristic of Frank Miller is a complete disregard for proportionality. If Miller wants to show "doughnuts are bad" on the comic book page, a character will eat a doughnut, choke on it, stagger into traffic, cause a car to skid off the road, carom off a telephone pole, crash into a Dunken Donuts, and explode. While two hookers standing outside catch fire.