Arcanaville

Another way in which the concept of reward thresholds comes into play is that critters (even standard ones, not just custom ones) aren't valued solely on the basis of how dangerous they are, nor on how difficult they are to kill, NOR on a combination of the two. Rather, offense and defense are both considered "separate but equal" requirements for critters. Critters must have some minimum offense to be considered valid, and some minimum defense to be considered valid. So Bosses automatically have more health, and are almost always designed with more attacks. Put simply, instead of saying bosses must reach 10, we say they must have at least 5 offense and 5 defense (massive oversimplification). 8 offense and 4 defense is too weak, as is 2 offense and 12 defense. But 7 offense and 7 defense is not necessarily worth more (cf: Dark Ring Mistress). That same basic principle applies to the custom critters in the AE.

So a safety valve to prevent an all fire-offense farm from becoming *too* exploitive is that its supposed to still be hard to kill them, even if they are of little danger to you. What we *can't* do is allow players to make the critters *weaker* than normal in exchange for making them stronger offensively, because that's gameable. And that's also partially why its not always easy to make full value bosses that don't have at least *some* defensive or utility powers: most (at higher levels) do have them. Its not perfect, but its in the ballpark.

The point of a twisted reality is to see the twists: reference points make sense in that case. Otherwise, you could argue that having a Marcus Cole in Praetoria is also unoriginal, and the least they could do is have some other guy end up being the ruler of Praetoria.

Breaking out the thesaurus just to come up with synonyms for *everything* in Praetoria would be laughably transparent to me personally, so that's not necessarily a safe option for the writers even if they were not intentionally trying to reference the Seers directly. Its practically a form of trope. So this is somewhat a matter of personal preference.

But in this case I think its very deliberately trying to connect the Fortunata Seers with the Praetorian Seers as alternate twists on fortune telling. Recluse believes in Fate, and uses the Seers to try to alter it. Tyrant believes in Control, and uses the Seers to try to impose it. By keeping the Seers constant and changing the world around them, we get a sense of how the world has pivoted around this one commonality.

If Hickman's comic run involved all those radioactive Dominatrix backstories we're contemplating, the comic sales could fund half the moon zone.

"The really, really bad kind. And trust me I've seen the really, really bad kind."



A former Trade Federation employee breaks her silence on MMO addiction.

The main difference between bind on pickup (thanks for correcting my language there: that's what I meant) and unique is that a unique item can be farmed by one character and given to any number of characters, so those items have to be balanced around the worst case scenario, which might be a level 50 exemplared down to farm the items. With bind on equip you can make something that is worth a lot to a level 20 but less to a level 30 so in effect the earlier to try to go for it, the harder it might be to get but the higher the benefit will be to you if you get it.

You can also grant rewards based on the earner: I could give a special bonus Res/Def/Heal enhancement to a tanker or scrapper that completed Positron Part One and a Dmg/Rech/End enhancement to a blaster that completed the same task, knowing those cannot be traded around.

That's *exactly* why soulbound items are tagged such. To prevent you from earning them from multiple sources (however easy or hard that is) and then stacking them into a single ultra-powerful character.

Think about those three big enhancements they give us in the tutorial. Think about what they could be if they were bound to the player and couldn't be traded or exploited by making tons of new characters just to get them. How about instead of a large respite, an enhancement that boosts recovery for an entire mission? Or an awaken that works like Soul Transfer? All possible with binding, all impossible without.

Hey Best Buy. Please stock our new boxed expansion in your stores and market it. If you do, we'll let you tell your players that with their purchase they can get things everyone else can get anyway without even having to get in their cars! See: that's the beauty of it: you get to market our game for us, and we'll advertise the fact that no one actually has to buy it from you. That's a win/win for everyone, right?

EvilRyu? No, no one by that name on the marketing team, why do you ask? Hello? Hello?

And you know they didn't do this because...


Even if they did think of a way, they would have to be either completely stupid or graduates of the EvilRyu School of Loyalty to announce that fact before the promotion is actually over.

Forgot to mention: personally I think this is a good idea. I know some people don't like the concept, but for me it allows for much more freedom in the kinds of benefits the devs can give out when they know they cannot be shuffled around, stacked beyond what a single player can acquire, or exploited in other ways.

If the choice is between bind on equip and not bind on equip, the choice seems obvious. But if the choice is between bind on equip and all of these bind of equip goodies we can make, or we just don't ever get them at all, that choice is a different one, and one I come down on the side of diversity of power options.

Same with ED. ED or no ED seems obvious. But that's not the choice, the choice is ED with inventions, or no ED with no inventions. That's a different choice (some would still pick no ED with no inventions, but at least they are making an honest choice).


The one thing I will say is that I think if bind on equip is *only* used for the highest possible strength gear, then I think its a partially wasted effect. If it is also used to make low and intermediate stuff that players can work towards that don't involve the absolute highest levels of end game, then I think its an effect that can be well worth it for the entire playerbase and not just a small segment of it. For example, if Positron's task force(s) granted the player a significant powerful enhancement that you could only earn once per character, as a bind on equip it can be more powerful and interesting than if its can be farmed, traded, and gifted. It can be a reward balanced around the average effort of acquiring it, rather than balanced around the best case farming scenario.

The bind technology behind these perk enhancements can be used for good or evil: the technology is neutral itself. I hope they are mainly used for good.

How do you know when this deal was even signed? So far as I know, these enhancements didn't even exist at the time the original preoffer was made.

In fact, right now NCSoft might be talking to Amazon. They may, or may not come to a deal which may or may not involve perks that are or are not the same as the Gamestop one. Deal.

There are people that hate Amazon and Bestbuy also. What makes your hatred more special than theirs. Also, as far as I know Amazon doesn't have actual stores in actual places people can walk into and see the box.

Personally, I have no problem with the bonus enhancements. If it acts to attract new players to the game, and the net benefit is something that mainly assists them with the lower levels of the game when we most want to make a good impression, count me in. Not even a tiny little bit am I jealous. Heck, if it would attract a significant number of new long-term subscribers I would probably donate my inventions to give to them.

Us old timers have got to think in terms of helping NCSoft attract newer customers, not stomping on every attempt to do so as an affront to our "loyalty." Loyalty is faithfully supporting the game because you believe in it. Supporting the game because you believe you will get perks before anyone else and the game will always treat you better than anyone else is not loyalty, its bribery.

That would, unfortunately, only work solo. Once you try to do this based on who's on your side, you start having to figure out what team mates are likely to do with powers like taunt, force fields, heals, etc. And you'd have to make intricate tables of how things like burst damage is more dangerous to defense and regen and less so to resistance, things like that. That would be, as I tend to say, a non-trivial problem.


It depends on what your goal is. In the absolute sense, probably not in the practical sense. But it is possible to do so in still interesting ways. For example: the current system assigns XP values to critters. A synergy-aware system might attempt to assign XP values to factions. So if you make a faction composed of X, its worth something. But if you put X, Y, and Z in it, then spawn points that are allowed to pick randomly from your faction could be worth more because of synergy factors. That sort of thing is at least something that I can conceptualize how the numbers might work.

If you're saying the system should be better documented, I agree completely.

I agree completely. Its an unfortunately historical quirk of implementation. In fact, the system was originally intended to have *three* values: alpha, beta, and gamma. Alpha would be the current Alpha, and Beta and Gamma would be the two values that Alpha/Beta powers have. And it was intended to be possible for a single power to have all three values set.

Why? Two words: Follow Up. Something I hope can be revisited down the road (along with some other issues).


That's not quite how I would put it, but that's essentially correct. The way I would put it is that a critter with score X is not necessarily X dangerous, the system only (attempts to) guarantees that its at least as dangerous as X. So no matter how you make an 80% XP critter, it will always be at least 80% as dangerous as the average 100% XP critter of that rank and level, that sort of thing.

And very specifically, its intended to say that for the 100% mark. The farther away you get from 100%, the more margin for error the system allows itself. Because if it turned out that a critter worth 20% XP was actually only 16% the difficulty of standard critters, the devs are ok with that because that's not exploitable.


You've hit the fundamental design principle of the system. That's exactly the mindset under which the system was created: what's the best possible value the players can be given for each power that is not exploitable.


I hope not too badly.

Ok. Its a little more complicated than that. Basically, there are two kinds of powers. The first class of powers has just alpha scores, and beta is zero. These are powers that are basically "worth" their alpha score. The second kind have alpha and beta scores: these are worth different values dependent on both the number of and value of your first class of powers.

Lets say you take a 100 point ranged attack, and then a 60 point melee attack. And lets say this is for a minion at level 4 where the threshold is 150 points. So you have 160 points, and the XP meter would be pegged at 100%.

So you remove the melee attack, dropping you to 100 points and try to add a defensive power. Lets say you add a 3/0.067 power of some kind. That will be worth 6.7 points when you try to add it.

Ah, but suppose the XP meter actually went above 100%, and you had added the secondary power *first*. In that case, that secondary power would have been worth 10.72 (100 * 0.067 + 60 * 0.067), not 6.7. And perhaps not intuitively for some, if you then removed the melee power, the score would drop from 170.72 (or 115%) to 106.7 (or 71%) which is a drop of 64.02, not 60 points like the power is "worth." And that's because the type 2 powers are multipliers. So judging what will happen when you add and remove powers based solely on looking at the score can be confusing and misleading.

Now, in terms of actual non-damaging defensive powers, you're not likely to see a *big* difference here, because their values are inevitably low. They won't shift your score by more than about 50 points unless you are *loaded* with ultra-high value attacks (like fire blast attacks). And on a minion, a power that increases your score by 50 points is, even at level 4, only worth about a 17% swing in minion value (and for any other rank, and any other level, its going to be even lower than that).

Only two kinds of powers have the ability to really move your score by more than 20 percentage points or so: single attacks, and damage buffs. Only attacks generally have values above 50 points and can move an XP score by more than 20 percentage points. And only damage buffs have beta multipliers high enough (more than 0.2) to generate total multipler scores high enough to swing XP scores by levels comparable to (or sometimes exceeding) attack power scores.


I should also warn you that some powers are highly problematic to give to critters, but have relatively low scores. A low score is no guarantee that the power is of minimal importance to you as a critter designer. Endurance drain powers, for example, can be very dangerous to give to a custom minion, if you then intend to fill an entire mission with those minions. But the system doesn't - and can't - factor that in. It values those powers as if a single minion walked up to you and hit you with the power. And a single burst of endurance drain just isn't powerful enough to be worth much in this system. Similarly, rezzes are worth practically nothing, even though at the moment they can be highly problematic to use. They aren't worth anything because when a critter rezzes, it can be killed again for more rewards. In effect, the rez contains its own reward: it can't also make the critter itself more valuable.

The score is not there to help players learn which powers are "dangerous" and which are not: they are there to protect the system from exploit. If the best way to do that mathematically was to make all of the most dangerous attacks have zero score and all the trivial powers have high scores, that's what the system would have done.


That's something I'm thinking about, but as I mentioned in an earlier post and implied in this one, the big complication to that some powers are multipliers against the scores of other powers. So no power has an unchanging "intrinsic" value relative to the total. A 100 point attack will increase the value of a critter by 100 points if it has no alpha/beta powers. But if that critter has build up, that power can increase the value of the critter by 180 points, because BU will in effect amplify the value of that attack. But even if the critter has BU, BU is only allowed to multiple against the top two attacks. So if you have a 120 point attack and a 110 point attack, adding a third 100 point attack will add 100 points to your total, not 180, because BU is already buffing the 120 and the 110 attack.

That's what makes it tricky to state simply what the value of a power is, and why the uncapped number can be misleading. It might be multiple powers multiplying together that are responsible for the high score, and losing either could drop the score substantially. Or not. And unfortunately, while I recognize that this makes the system less transparent than a system that just adds up numbers, such a system would simply not work correctly to do what the system's primary purpose is to do: prevent the creation of exploitable critters while still giving critter authors reasonable XP for their creations if they are at least close to or higher than the difficulty of standard critters.


I will, however, think about better ways to show the incremental "what if" information for the powers. Let me mention one more example, though, of how the system sometimes has to choose between giving a power a high score because its dangerous, and giving it a low score to prevent exploitability, and how those two objectives can be at odds.

Take Nova. Now *that's* a dangerous power. But its value is actually rather low. Why? Because Nova has two catches that make it potentially exploitable. First, its AI-dependent. Critters don't just pop off Nova whenever they want. The power is designed so critters never use it until they reach low health. For a boss or AV, that's usually not a problem. But for a minion or LT, its possible that the critter actually dies before it can trigger the power. Very possible, in fact. So while the power would be dangerous to most players, you'd probably barely see it in a farming situation because the team would steamroll everything before anything could get a Nova off.

Second, its a PBAoE on a very long recharge, which means if you can recognize when the critter might use it and simply get out of or stay out of its radius, you won't be affected by it. And once its used, its basically gone for the rest of the fight.

So Nova is dangerous, but potentially exploitable, so it gets a lower score than it might otherwise have if it a) could be used at any time, and b) used often enough to be impossible to game. And that's why you shouldn't rely on the score to suggest dangerousness. According to the system, Power Bolt is three times more valuable than Nova. But I don't think you believe Power Bolt is three times more dangerous to add to a critter than Nova. So you have to be careful using the numbers for any other purpose than their intended one.

Fair enough. I'm leaving it.

To be blunt, the former is probably a lie and the latter is definitely the truth. Power color customization basically required new code to be added to the game and for the color-tintable things to be often replaced with things that are actually designed to be tintable. Even so, there are still many things that are not tintable because the work hasn't been done to redo them to be color customizable yet.

I agree. Working on it.


One improvement that came bundled with the new custom critter system is that critters actually *use* their powers more than they did before. They used to sometimes (often) get into funny states where they had five attacks but only used two, stuff like that. Their brains have been upgraded**, and now if you give them a power they will use it, so long as they aren't too busy using others, and barring some AI flag that tells them not to (for example, powers like Nova are set to only trigger when the critter gets low in health, and some buffs and debuffs are set to only fire under certain situations).

If you give a critter ten attacks, they *should* eventually use them all by picking from their available attacks randomly. But what they can't do is use more than one attack at a time, which means while a critter with two attacks can attack twice as fast as a critter with only one, a critter with twelve can't attack faster than one with ten.

Why five? Because the average "busy time" for the average attack for critters is somewhere around 15%, which means they can't really use more than six attacks efficiently. Factoring AI delays (which can add half a second to the lag between attacks) five is about the saturation limit. Adding more attacks will add more *variety* to critters, because they will try to use them randomly, but it won't add very much more *offense* to those critters, because the powers will get in each other's way.

It was either going to be the Rule of Five or the Rule of Six for that basic reason, and without any further numerical guidance the system picked conservatively.


** Actually, they were made more effective by ironically making them stupider. They think about less things when selecting attacks, and that makes them much less likely to get confused or do something weird or dumb. The AI is now also much, much more difficult to game.

In this example, wouldn't the correct choice be whichever power had the lower value, assuming both reached 100%? You did say you wanted to get to 100%.

If you *didn't* necessarily want to get to 100%, then I could maybe see how you might want to make a judgment call between making a 96% critter and a 108% critter. Maybe you might decide that 96% is "close enough" given that 110% is so much higher.

But see, that's my problem right there. The system isn't designed to accurately portray difficulty above 100%. If you're looking at those numbers and saying well, if the critter is 10% higher than full XP, it must be much more difficult: I should stick with 96% because its only 4% less difficult than full and is "closer" to normal than the 110% critter, the system might lead you astray. It wasn't designed with that thought process in mind. In fact, it explicitly hand-waves that situation away.

The incremental situation is interesting: if I add this power what will it do, without actually having to go back and forth and try it out. But especially given how the multipliers work, even the pre-capped number won't tell you that without doing a lot of math in your head.

Suppose in real-time each power told you how much, in percentages it would change the current number based on the current power selections if you toggled it on or off. And as you changed power selections those numbers changed on the fly. That would seem to be actually more useful than the pre-capped total would be, because it would tell you how much each power was worth at that moment relative to the threshold you are looking at.

That could be very difficult to add, but perhaps not. It does sound like it matches both the intent of the system and the desired ability to guestimate changes far better than the pre-capped number does.

I read exactly what you said. If you're just going to keep saying its a problem and not actually fulfill my request to explain in detail, there's nothing I can do about that. I'm really only interested in working on solvable problems at this moment.


This is not an AI problem. An exaggeration will better illustrate the issue. A critter with ten thousand attacks is not more powerful than one with nine thousand attacks. Once the critter is attacking all the time, adding more attacks of equal strength stop helping. This is important to recognizing why the number can be misleading in many cases.

Separately, cap or no cap a critter worth "150%" would *never* under any circumstances give 150% of normal XP. That's simply not how the reward systems in this game are designed. So saying that would lead people to mistakenly believe that if not for the cap, the critter *should* be worth 150%, when the system is not saying that at all.

The system says a 430 point boss is worth full XP at level 20. It does not say that an 860 point boss is twice as dangerous as a 430 point boss. That's a corruption of the system. Because of that, I'm disinclined to recommend that the system display numbers like "200%." I'm looking to see if there is a compromise that is workable, which is why I'm looking for very specific, down to brass tacks details.

Talk me through this process. I'm not trying to be obtuse: I genuinely want to know step by specific step how you construct these critters, and at each stage what information would be useful to you and in what way. Try not to gloss over anything, because tiny details might be important to me.

See, when I make a critter, I have an idea of how powerful its going to be just by the power selection. I know if its going to be more difficult than the standard critters in the game. If the critter is more difficult than a standard critter, it can't be made to be "worth" its difficulty because the AE itself is capped to 100% maximum.

Now, if you're saying you *don't* know whether a critter will be overpowered or not, and the XP slider could give you a hint by saying the critter is worth "150%" or "220%" then that's something else, but that's a lot more difficult than it sounds.

Suppose I made a Fire Blast / Ice Blast Boss that took basically everything. We both know that a critter with twelve attacks is literally no more dangerous than one with ten, because you can't make proper use of that many attacks: your attack chain would be full long before then. So knowing the critter was "450%" wouldn't really tell you anything. To make that number not misleading we would have to apply a different set of rules that tried to make that number actually meaningful. Otherwise, while it might be useful to you, it would be dangerous or misleading to others. And sometimes its more important to not mislead than to maximally inform.

But as I said, take me through the process. Maybe its an issue that is addressable, but perhaps not in a way we're currently discussing.

Secondary effects in attacks are worth nothing at all unless they are a hard (attack-stopping) mez. If the power doesn't deal damage, its judged by some rules of thumb for how much to value foe debuffs and other effects.


There wasn't really enough time to hand-evaluate all 955 custom critter powers individually, and that would also inject a lot of subjective weighting that could be different from power to power. So the weights were calculated somewhat automatically by formula(s), and then several passes through the powers were done to make sure the powers weights came out "sane." When a power *didn't* come out sane, one of two things was done: the formulas themselves were tweaked to generate new numbers which were better (the preferred solution) or an exception rule was made that tried to catch not just the errant power but all powers in its class, trying to make that "class" as large as possible to make the number of rules as short as possible.


Trivia: There are about ten rules that govern the weights of the powers, and only about 72 exceptions to those rules out of 955 (more or less) powers, governed by 24 exception guidelines.

I can't list all of the exceptions, but I can give you the ten power weight rules:

1. Powers that do damage get a score based on the damage they do divided by the damage the recharge formula says they should do, times 100. A "normal" single target attack should thus be worth 100 points.

2. If it deals more than Scale 2.0 damage, 20 point bonus.

3. If the power does damage and is melee ranged or is a low radius PBAoE, its score is reduced to 60% of its damage-calculated value.

4. If the power deals attack-neutralizing mez (sleep, stun, hold, fear, confuse) multiply the mez magnitude and the scale duration and add as a bonus. If the power deals multiple mez, pick the one with the highest magnitude. Knockback and knockup don't currently count in the system (they probably should eventually, they just don't count in this first version of the system).

5. Pets such as controller tier 9 pets and mastermind pets get a score of 100 by default. Pets like Phantom Army that behave like these pets but are not up all the time get 100 * uptime ratio.

6. Powers that simulate player defensive passives generally get a score of 10, powers that simulate player defensive toggles generally get a score of 20, and Powers that simulate player defensive clicks generally get a score of 25 unless they already have a score as above.

7. If the power does no damage, convert its Alpha score to an Alpha/Beta pair where Alpha = score/300 and Beta = 3. Powers that do not do damage are *not* allowed to have pure Alpha scores as a rule.

8. If the power is a damage buff or a recharge buff, grant it an Alpha score equal to the scale magnitude of the buff, and a Beta score of Duration/5, rounded up. If its up all the time, Beta = 5.

9. Other powers are given scores between 5 and 50. There are exception rules of thumb to cover these. Rule 7 applies here.

10. Alpha is capped to 150. Beta is capped to 5.


These "rules" are subject to change at any time by the devs. They are not rules in the sense of the Damage/Recharge equation. Testing or judgment can change the value of a power from its computed score to a different one to fulfill the intent of the system. Don't quote me if the devs diverge from the rules I've listed (but if you do find a power that seems to not follow those rules, let me know).

The explicit intent of the system, as long as I'm on the subject, is to allow architect authors the most flexibility in designing custom critters while valuing them as fairly as possible, with the overriding requirement to make sure its impossible to make exploitable critters that are worth more than standard critters while being significantly less dangerous.

That's BaB. Castle I think isn't crazy about a moon zone. Of course, just like Castle can't speak authoritatively about animations and BaB can't speak authoritatively about the powers system, neither has the final say on what zones we get. That's doubly in the realm of War Witch's territory (both in terms of having more zone author expertise, and in being their absolute ruler and all).

Honestly that sounds made up.

I'm pretty certain the term "Guest Author" was first created in I14 beta to distinguish non-player arcs from player authored arcs: the first arc to be designated a guest author arc was I believe pohsyb's imp arc**. It was later focused to be arcs written by authors invited to write arcs.

As far as I know No player's AE arcs were ever turned into canon. Not even Feargas' arcs are canon.


Edit: And I believe that was to test the feature, not specifically to necessarily designate him as a "guest author" in reality.

That is exactly my impression as well, and one of the most pernicious problems with trying to "listen to the players." Half the time, when someone says "listen to the players" what they say next is something I hope the devs don't actually do. If the players were all in anything remotely close to unanimity about the details of what they wanted done and not just the vague generalities, it would at least be easier to "listen to us."