Arcanaville

It might be fixed now: I'm not sure. If I become absolutely certain it is fixed, I will write up something for the forums**. Until I'm certain, however, its always possible someone will be able to read what I write and figure out alternate ways to exploit the issue other than the ones I know about (even after that, someone could still outsmart me and figure out a way, but that's always a possibility you can't eliminate with absolute certainty no matter how much testing you do).

This is not intended to be a tease: if I thought the chances were low I would ever be able to mention this ever, I wouldn't refer to it now. I *think* the day will come when I will be able to describe this in detail, which is why I'm willing to refer to it now. Whenever I *can* describe something to the players, I generally try to, and the devs generally give me enormously wide latitude to do so.


** This is a case where I need to be certain its unexploitable myself, independent from whatever the devs think, because I'm the only person that has tested it to an insane degree.

By "the devs" you do mean, of course, all of the MMO devs in the industry combined and not just Paragon Studios devs. Because no dev team I'm aware of could or would give Werner the guarantee he's referring to, and every dev team I'm aware of would, if they were being honest, give the converse guarantee that everything that exists at the very highest levels of performance would eventually be reviewed, and possibly altered in some way.

Min/maxing in MMOs is a blood sport. If you are good at it, but don't treat it as one, you'll burn out eventually, because no MMO is going to let you consistently do it with immunity.

Could it be a lot more stable? Sure: its well within the realm of human ability to get five, perhaps ten times more stable by much more numerically careful design. I'm not giving the devs a pass for making what are clearly avoidable errors**. But no MMO designed by human beings will ever be able to grant min/maxers guaranteed performance stability at the high end of the game. Its not going to happen here, or anywhere else.



** Fundamentally, the problem is that the design tools aren't designed for transparency or auditability, nor do they have a trivial way of performing validation checks on anything. As I've said before, if I could add anything to the game by snapping my fingers, it would not be new zones or powersets or task forces: it would be better design tools.

Lots of people are asking me if there was or is a problem with the streakbreaker. To the best of my knowledge, it was and is still working fine.

A bug in the streakbreaker wouldn't be something that would lend itself to be particularly exploitable either (any more than it already is, which is not really), and if it wasn't exploitable I would have written an article about the change to inform players.

Strike two. That would neither buff nor nerf player performance in the game by any significant amount.


Microscopically warmer. But probably observer error, unless your definition of "few" has double-digits.

I'm sure.

And no, I don't know that the devs have been monkeying with the bad guys power levels without saying anything for months now. Moreover, I can tell you what the bad guys power levels have been for almost every calendar day from this moment all the way back to I11 beta, on test and live. I can, in fact, give you that information faster than Castle can. So if you have some evidence that the devs have been "monkeying around" with bad guy power levels for months, post it. I can with 100% certainty confirm or deny that evidence. Give me a specific example.

By the way, the devs can't really change the power levels of critters without pushing a patch to the game, so the fact that the live game had minimal patching recently until I17 went live kinda calls that entire thesis into question.


The part where I17 and GR buff things as an alternative to nerfing them when there is a power imbalance. In other words, I'm failing to grasp the relevancy in anything you've said.

Relevancy is usually important. Like, if I were to say that the "buff not nerf" philosophy was something the devs have thought about enough and rejected firmly enough that Positron decided to write one of his first blog articles about the subject that would be relevant to the question of whether the devs themselves think they have any practical way to sustain a buff not nerf strategy.

I'm really not sure how to respond to this statement:

Buffs and nerfs are ways to generate 'New Shiney' in existing content.

I don't know many people that would classify a buff as "new shiney" and no one that would classify a nerf as "new shiney" content. Its just as irrelevant as everything else you've said about whether all buffing and no nerfing is sustainable but its in a completely different class of weird.

I get the feeling this is going to be one of those cases where the poster is having a different conversation in their own head than he is actually having in real life, and this is going to turn into one of those situations where the reason why I just don't get it is because I've made the mistake of reading the words, and not telepathically extracted the actual point instead.


I got news for you. They didn't really keep anything secret. The game mechanics were there for anyone to know, if they just asked the right questions. The problem was that in the past, the devs didn't have the time to explain everything fully, and didn't want to spend the time building the requisite common language to explain it. And in some cases they didn't even know the correct answer themselves.

When people asked Geko about mez mechanics, he wrote an entire article about that. If you were paying very careful attention, they divulged the entire system of archetype modifiers back in I1 when Invuln was being looked at - but without context, the players didn't fully understand what they were being told. Even the mechanics of accuracy were released by the devs, but because the devs didn't all *know* the correct mechanics, different devs released different and contradictory information.**

What I discovered when I started talking to the devs is that for the most part, the devs were reluctant to discuss game internals for two reasons:

1. Some players refused to take a hint when the devs said a question entered territory they couldn't discuss.

2. Many players were asking technical questions for the primary purpose of being able to engage the devs in a technical debate.

When I separated questions from debate, I almost never had a problem getting technical questions answered. Statesman even told me his design philosophy on respec, something he never did on the public forums, just because I asked nicely.

Whenever the devs were asked a *specific* question, they often ran the calculations for us, exposing the game mechanics. They were not as forthcoming about their decision making processes (see #2 above) but frankly in terms of the basics we knew everything except tohit mechanics more or less correctly by I1.


Strike one. If its something you've heard of before, I'm pretty sure its not that. I am pretty sure that on the day I discovered the issue, I was the sole person aware of it, and to the best of my knowledge it has not been discussed or even mentioned on the forums from the day I discovered it to this day. You are not going to get it by random guessing.

This is a change that could have saved you, or killed you. From virtually level one to level 50. You just wouldn't know it was responsible.



** These inconsistencies are what prompted me to work out, once and for all, the defense, tohit, and accuracy mechanics with pohsyb. And I asked him to verify my description against the actual code. The defense description added to the Guide to Defense circa I6ish is essentially a source code-confirmed description of accuracy and defense, which is the one we use today. At the time, though, I couldn't specifically say "this is a source-code confirmed description of the tohit mechanics which I discussed with a programmer." Cryptic wasn't *quite* that open back then.

Have you ever known me to place a bet on something about the game I didn't already know I would win? This has already happened once.

I just can't discuss it at the moment because it involves a potential exploit. Rest assured there's a really good game mechanical article coming one day about it. But to be honest, there are players out there sharp enough to have caught it in theory, so I am slightly surprised it has not been detected to this day. But you would have to have been looking right at the issue to have seen it, and ironically the very players most likely to spot the problem were also least likely to experience the problem often enough to notice. And if you tried to look for it now, I doubt there would be a way to detect it because nothing would appear to be wrong today. In terms of the issue, the game looks probably exactly the way most players assume should. It just never actually did until it was changed to fix the issue in question.

It originally did, but was changed to its current form.


Worse. Originally the power did 1.0 damage at the center and 0.7 in an AoE. That meant the power did 1.7 at the center and 0.7 everywhere else. When it was changed, someone took the 1.7 at the center as the base damage and *stacked* the 0.7 bonus damage on top of that, turning it into a 2.4 scale damage power which now did that damage everywhere.

My memory on the precise order of events for Shields is a bit hazy, but this might have happened at the same time they were having issues with the mechanics of the power, and players were complaining the power's damage was ineffectual relative to the difficulty of getting the power to work. It might have been buffed when the mechanics were still being monkeyed with, which partially masked the in-game effect of the damage buff.

Its still a mistake that should never be made. Its basically a translation error: someone tried to mentally translate what the power did into simpler numbers for analysis incorrectly, then back-translated it to the power spreadsheets with that error. Its a mistake that is easier to make than people might think due to the nature of the powers database, but its still a mistake that should not be made.


Basically yes, although not quite so straight forward. But yes.


Close. If you assumed that the original power implicitly assumed 1.0 damage modifier (because that's what pets basically do) then scrappers were supposed to go up slightly, and tankers and brutes were supposed to go down moderately, from the original numbers (which is what the numbers Castle posted in this thread also imply).

1. You're overpaying for words. This costs only $5.95 and includes all of those words plus many more.

2. The only way to consistently buff the players without the PvE game becoming ludicrous is to compensate by buffing the critters proportionately. The net result is that rather than one thing being nerfed, everything else gets buffed and that thing still drops in relative performance, which should not impress many people. But if you are explicitly interested in impressing the people that could be impressed by such a numerical shell game, or are one of them yourself, then my original post isn't snark, because that's exactly the sort of numerical shenanigans (now on sale for only $0.15) that is AWESOME.

3. That's why the psychological shift is unsustainable, and probably also deleterious (SRP: $0.18) in the long term. You could get lucky and join the game just as a global buff cycle was going on, but very quickly you'd find yourself in an even larger nerf cycle to compensate, and because these changes affect more things than targeted nerfs they would affect a far larger number of players every time they occured.


About the only way that the "buff not nerf" philosophy has any chance of working is ironically to do most of it in secret, because this would leverage the simple fact that most players - even most forum posters - cannot tell if they are being buffed or nerfed unless they are explicitly told, or it affects a number they explicitly check. Implicit buffs and nerfs, which occur as a result of the situation changing rather than a big floating number changing, almost always slip detection for long periods of time, and are sometimes never detected except when a player happens to stumble over them while testing something.

In fact, I bet I could personally buff the players by 30% or more in a wide range of normally encountered situations, then several issues later nerf them by an even larger amount in the same situations, and no one would notice either change. The question is whether the devs would accept such a development philosophy just to manipulate the psychological state of the playerbase to their advantage. I suspect generally not, although I also suspect that every dev team does it to at least some slight degree in certain corner cases.

can I haz buffz 2? give buffz 4 all every1 b happy. den other stuff buffz make every1 sad. not supr anymore. devz hates us.

The "buff not nerf" philosophy is a vacuous fiction with no rational numerical consequence due to the fact such changes ultimately require compensating alterations which when normalized have an immaterial differential result. Its only game design benefit is a transient psychological perspective shift which is unsustainable.

Next issue, the devs are buffing everyone's damage by 10. They are also buffing everyone's health by 10. They are also buffing the critters by 10 to make them a real challenge.

There might be a decimal place error in the displayed numbers in your chat windows and in the console in general, but don't worry about that. Especially because in Issue 20 they are buffing us again by 100.

To resolve some balance issues, some things won't be getting the full 1000x buff. Some might be getting only 990x, or even 850x. But really, what's the difference between being eight hundred fifty times stronger and being one thousand times stronger. You're still awesome.



See also: Normalization.

Thunder kick did too much damage for its recharge. When they tried to buff MA in I6, they also corrected that bug. Unfortunately, the devs really didn't have the numerical tools to properly buff offensive sets at the time, and their buffs didn't so much improve MA's damage as it did make ridiculous powers a little less ridiculous (CAK, EC). So the net result was that MA's single target damage went *down* after the buff to CAK and EC and the nerf/correction to TK.

Unfortunately, although I was able to prove the changes would have a net negative effect, I didn't possess the credibility at the time to make that charge stick. It wasn't until the Claws "hyperchanges" a couple issues later when I made the numerical prediction that the Claws changes would "turn Claws into a warp-speed buzz saw" and the devs were predicting only a small 7% increase in damage that I think numerical analysis started to drag itself out of the gutter of offensive comparisons (numerical analysis had already started to assert itself in damage mitigation by then, with me on this side of the Pacific and players like Dr. Rock on the other side of the Atlantic, but most numerical analysis on the offensive side of the equation was little more than brawl numbers turned into a sudoku).

I deliberately haven't discussed this topic with Castle since this thread started, except with my public posts, so that I could continue to comment freely. So I'm saying this within the context that I haven't actually asked Castle what he thinks about the situation directly, except what's publicly in this thread. This makes the following statement pure speculation and credibly deniable by the devs.

Having said that, I think the solution the devs will eventually converge on will be to reinstate area modifiers, so that the base damage won't have to be whacked too hard, but that damage will not be allowed to hit a massive number of targets in a huge AoE. I would also expect the cast time to increase slightly to reduce the power's gigantic DPA.

That would allow the devs to reduce the power without a huge numerical nerf to its base damage. If it were me, I'd propose an increase in cast time to circa 1.9s, a slight to moderate reduction in base damage (probably to at least under 3.0 scale), and a reinstatement of a drastic reduction in the splash damage (to no more than half the current base value).

I would also expect this to not happen anytime before Going Rogue goes live. You probably couldn't carve out enough time from all the stakeholders to discuss this situation. Or collect enough firewood and a strong enough stake to start the meeting.


That's assuming the devs look for a purely numerical way out of this. There is an alternative if they decide its worth asking for tech time to add new mechanics. They could make the damage of SC scale based on the distance of the teleport. After all, its intended to be a charge: presumably its conceptually consistent for the damage to be lower if the player has less distance to build up momentum. Have a base level of damage if the teleport distance is basically zero (hitting a target right in front of them) and increase the damage to a maximum of somewhere not too far from its current value if the teleport distance is at least some distance, say 30 feet. This way, you couldn't use it as easily to min/max damage on a stationary target like an AV or a pylon.

I'm always in favor of adding exotic and interesting new mechanics to get out of sticky numerical balance problems myself. But there's a reason why pohsyb hides in a box and the lead programmer only communicates via paranormal video.

Probably just from the forums. Its perfectly fine to spend that vacation in Paragon City. Maybe with the broadcast channels off.

No player attack that does primarily foe damage is supposed to have an unlimited target cap, ever. If SC originally didn't have a target cap, adding one addressed a bug, it was not a trade-off for having higher damage.

In fact, so long as the aggro cap is 17, being able to hit more than 17 targets with significant damage from a single attack is probably a priori an exploit or a design error or both.


Edit: its weird sometimes discovering which of my posts will be seen as controversial. Apparently at least one person takes issue with this one, even though its essentially a factual statement about the design of the game. But since there seems to be some question over it, I'll tell you what I'm going to do, my anonymous friend: I'm going to compile a list of every single player power that does damage, either directly or via pseudo-pet, that either has no target cap or has a target cap higher than 16, and email that to Castle so we can confirm whether any of them, if they exist, are working as intended. Don't thank me, its no trouble at all.

I had a similar problem once shortly after "upgrading" to Windows 7** and I ran the updater as administrator once, which made it go away. I haven't had to do that since that one time.



** Meaning: after I reloaded the client in my nice clean Win7 installation, because Win7 doesn't upgrade XP. Win7 and WinXP used to hang out, but then Win7 got all like popular and stuff and now pretends like it never knew WinXP and doesn't let it sit next to it in the cafeteria or anything.

MA/SRs that played since release deserve a special veteran reward or something. A machine gun storm kick would have been fine: what we actually had was Flurry for Feet.

There is *nothing* to compare to MA/SR at release. If release MA/SR was a difficulty level, it would award a MasterOf badge.

Here's how you can come close to simulating MA/SR at release today:

1. Replace Storm Kick with Flurry (Storm Kick used to do a lot less damage with an animation time comparable to flurry)
2. Replace CAK with Boxing (Boxing actually does *more* damage today than CAK did back then)
3. Replace Eagle's Claw with Sands of Mu (Eagle's Claw used to do Crane Kick damage with Sands of Mu timing)
4. Take only the SR toggles, no passives
5. Take no power pools
6. Slot no inventions
7. Figure out a way to fight even minions, +1 Lts, and +2 Bosses all at once.
8. Whenever a foe casts a pet, turret, rain, or patch, if you don't have at least two purples just use Self Destruct immediately
(SR at release: 5% passives, 10% toggles, vs even con minions: 50%; Lts: 62.5%; Bosses: 75%; AVs: 90%; turrets/pets: 105% base tohit)

That would probably be in the general ballpark.

Oh yeah, and Quickness used to cost endurance (0.01 eps or something like that). Rather unique for a passive.

Its actually that there are two effects, but the first one doesn't stack with itself. So its capped at one, the other is capped at ten, but they both take effect independently.

If my memory serves correctly, Shield Charge (Scrapper) does 3.6 Scale damage. But its a pet that does the damage, and the pet melee damage scaler is 55.6102 (same as Blasters) at level 50. So it should be 200.197 damage at level 50 for Scrappers, if my memory and calculations are correct.

Because Scrappers have a 1.125 damage modifier, and therefore do about 62.56 damage at level 50 for a scale 1 attack, this means Shield Charge (for scrappers) does the same damage as a 3.2 scale attack would do for Scrappers, if the damage was actually dealt by the Scrapper (and not a pseudo pet).

That's what I meant by "normalized to scrappers."

As weird as this sounds, saying a power does "200 points of damage" is something that actually has to get processed by the powers team: even for us players there's all sorts of caveats to such a statement we mentally shuffle through before we put that number into context: what level are we talking about, is that enhanced, etc.

But "hey Castle, SC does 3.2 scale damage and its a ginormous AoE" is going to be meaningful in about 3 milliseconds. Anything above 2 is considered high damage for a single target attack. The mega-hitter Total Focus does 3.56. Nova averages scale 3.875.

For regular humans, its not bad. At its peak, though, its larger than you might think, but in a tricky way. At its peak of ten targets (admittedly difficult to generate except under engineered circumstances), it'll generate a buff of 12.5% + 6.875% * 10 = 81.25%. Not high, as you say, but because its in a toggle, at the high end performance we're talking about that is very significant.

Build Up generates a much higher buff: +100% damage. But at the highest levels of performance, it comes with a significant cost: its cast time. At low levels of performance, 1.17s of offense lost is small, for a big gain. But at high levels of performance, 1.17s of offense lost is big (because your damage is much higher) for a lower gain (because you probably are running with +DMG invention bonuses on top of slotting, which dilute BU's buff). AAO dilutes just like BU does, but it doesn't cost anything (in theory) in offense to get the buff.

Also, because its theoretically possible to have the buff permanently, its going to deliver a bigger punch than BU will. Lets consider a simplified example and compare BU and AAO, just to see the rough numerical difference. We'll assume that the player is running with just +50% damage on top of +95% damage slotting. We'll also assume that they have enough global and slotted recharge to reduce BU's recharge to just 30 seconds. I am going to factor in ArcanaTime, and call BU's total rooted time penalty 1.32 seconds, but I'm not going to adjust the recharge for that (just assuming 30s of recharge for now, whatever value of recharge buff that requires). And I'm going to assume a saturated AAO.

So the player normally has 245% damage (100% + 95% + 50%) and with BU they have 345% damage. Assuming no funny attack chain games, BU will cycle thusly:

10 seconds of BU, 20 seconds of no BU, and 1.32s of casting BU

The damage done during each phase, and total, is:

345% * 10s + 245% * 20s + 0

The average damage becomes (3450%s + 4900%s + 0)/31.32s = 266.6%

In other words, the player went from 245% all the time, to an average of 266.6% when using BU as often as possible. That's a net overall increase of 266.6/245 = 1.088, or about 8.8%.

The AAO player is easier to figure out: they went from 245% all the time to 245%+81.25% = 326.25%. An increase of 326.25%/245% = 1.3316, or about 33%.

Most players aren't going to see this dramatic difference between BU and AAO, but top end min/maxers will.

The cast time penalty of BU tends to be consistently dismissed or underestimated in most damage calculations. But absolutists doing explicit attack chain calculations are, I think, beginning to see the very large penalty associated with BU. Its still a good power to use, but its own cast time neutralizes a significant chunk of its net benefit. Its a big frontloader, but only a small damage booster.

AAO is a damage booster, period. And a good one.


Huh? That doesn't sound right, but I don't have access to my numbers at the moment. I could have sworn Shield Charge for scrappers was *way* higher than that. I think it should be (normalized to Scrappers since it does pet damage) about 3.2 scale (i.e. 200 points of base damage at level 50). Its an AoE with a DS/sec (DPA) rating higher than nearly all single target attacks. That's what makes it always good to insert into an attack chain, even if you are only attacking a single target, if you can activate it fast enough and if you have the endurance to power it.

At SC's scrapper numbers, my Blaster would gladly trade Nova for Shield Charge. It does slightly less damage (on average) in a slightly smaller radius, but doesn't crash and recharges a gazillion times quicker.

Well, it would be a silly way to compare secondaries if anyone thinks a singular test or equation "balances" an entire secondary. Comparing the raw performance of two defensive secondaries, outside of primary power or pool synergy, is one test for design reasonableness. It doesn't mean the two sets are balanced against each other if they pass this one test, but it does mean they *aren't* balanced against each other even if they pass all sorts of other tests, because this test case is significant to game balance.

Usually, I mention this case and the devs mention this case not to state that its definitive, but to remind people that comparing two specific builds, top level or otherwise, is essentially an anecdotal comparison that is highly unlikely to be representative of the playerbase as a whole. They are likely to be corner cases, and corner cases have their own separate balancing requirements.


As to actual mitigation models, the one that I'm currently working on is a combined offensive/defensive valuation model that converts both survivability and damage into a singular score. It starts with a set of "standard candles" which act as proxies for average offense and defense (right now, for scrappers my stand-in for the standard candle for offense is Broadsword without the +DEF of Parry, and my standard candle for defense is Invuln with Invincibility, both slotted with SOs).

Without getting too technical, the really big catch to the model is that ultimately, both offense and defense get valued in terms of the XP they can generate (defense can generate XP by creating opportunties that amplify offense) but they also can get valued in terms of endurance and health: defense is ultimately a way to convert endurance into health. This "dual currency" of the model makes it extremely tricky to make work: it tends to generate surprisingly contradictory statements unless you define its goals extremely specifically.


(For example: suppose you discover that at a particular difficulty level, the critters pose so little threat to you that your offense alone can kill them faster than they can significantly injure you. Its possible then to run through a mission at that difficulty at maximum speed and kill everything without dropping significantly in health. As a result, you're only bound by the amount of endurance you have to launch attacks. In that case, defensive toggles and clicks remove endurance but return only more health you don't need. As a result, in this situation defensive toggles actually have *negative* value. This condition where defensive powers can have almost any value, including less than zero, suggests there are places the game design should never go. Unfortunately, sometimes it does. Most scrappers know that in the early game, prior to SOs, defensive toggles can often seem like they burn more endurance than they are worth. The numbers suggest this is not an illusion, but an actual design issue with the game rooted in the fact that endurance costs for defensive powers were basically just made up, and mostly judged based on the late game.)

There were (perhaps not quite so strongly, but there were) and it was actually adherence to the general principle that caused the devs to goof around with Regen rather than bite the bullet and just make the change they should have made from the start, which was to turn IH from toggle to click.

Instead, they did things which made sense from a certain point of view but not from the player point of view. For example, to try to make IH an only occasional-use power they increased the endurance costs and made the toggle itself have a 60 second recharge. The logic was that the higher endurance costs would make players use it only as long as necessary and then turn it off, and then the toggle recharge would prevent them from flipping it right back on immediately and wait for the cooldown.

Instead, it caused players to find any possible way to avoid the toggle recharge, and slot the heck out of the power to reduce its endurance costs, plus take quick recovery plus stamina to power IH. The net result is that they drove the few players that might have been using the toggle in the intended way to go perma because if you turned it off there was basically no way to get it back for a relatively long period of time.

In any case, the cottage rule doesn't prohibit changes to power design. It only forces the devs to make such changes as a last resort, and only to solve problems that are important enough (in their opinion) to override the benefits of adherence to the cottage rule. An example would be the change in powerset order for taunt in tanker secondaries.

If you've played SR from release to now, you've had a rollercoaster ride unparalleled in the history of the game.


But then again, BillZ plays Claws. He probably beat the RCS challenge at least once with Focus and Shockwave while forgetting to turn his toggles on. He calls Elude "my slower backflip." BillZ idea of normal performance is not the same as normal humans.

Every time BillZ talks about performance, you should picture this:


(Actual picture of BillZBubba circa I9)

The riddle of SR? Shall I tell you? It’s the least I can do. SR isn’t strong, boy, stacking make it strong!
What is SR without the slotting that enhances it? Look at the strength in my build, the defense in my powers, IOs gave me this!

Contemplate this in the planner of Mids.

Well, it does have +recharge, which buffs damage output a little. I'd probably take AAO, though.

But this is really based on a misconception. The devs don't balance "the game" around SOs alone. Powersets are designed around the assumption that they must deliver a certain range of performance when slotted with SOs. But the devs do take IOs into account when it comes to powerset and archetype balance in many ways. They obviously factor into datamined performance, because at least *some* SR scrappers slot IOs (as do other players). The devs also have some design constraints on the invention system itself, but they are just much more relaxed than they are for powersets themselves. People have to be able to solo and function on teams when slotted with SOs: that places some absolute value limits on the powersets. But Inventions are incremental buffs above that mark: for obvious reasons they cannot be designed and balanced around the same marks, because all powersets are intended to hit those marks even without Inventions.

However, this is true: if Powerset A has roughly the same performance as Powerset B when slotted with SOs, and when slotted with the best possible IOs Powerset A is twice as good as Powerset B, that's a problem, but a relatively minor one. But if Powerset A has roughly the same performance as Powerset B when slotted with the best possible IOs, and when slotted with SOs Powerset A is twice as good as Powerset B (as the devs define performance), that's basically a game-breaking problem the devs would be compelled to address virtually immediately (essentially, this is what provoked the devs to making sweeping changes to Blasters in I11). In that sense, the balance rules surrounding the SO-slotted data point are weighted far more strongly.

I've been looking at the issue of offensive vs defensive valuation for years now, and even I can't fully predict what will happen to FA under a normalized system right this second. There are different regimes of performance, and offense and defense have different relative values in each. A lot would depend on what your specific balancing target is, and there's no way to hit all of the reasonable ones (the set of all of the reasonable ones is mathematically contradictory).

The large uncertainty comes from the fact that in a normalized system, lots of rules including the one you mention are likely to have to change in some way.

The big question is whether the offense in AAO is "correct" for Shields in a properly balanced powerset. It might not be. And if its lowered to below the offensive power of FA, all bets are off as to whether FA is "protected" from being lowered in a normalized system. It might not be.


Its tricky to predict where something like this would go. Consider this: one really interesting requirements of the current MA custom critter XP system is that *all* critters are actually smarter**. Really: they are (or should be) much better at selecting and using attacks. Even outside the MA. So far as I know, no one has even noticed this fact yet, but its there, and probably having interesting effects even if people aren't noticing them overtly.

I think few people would have predicted that would be a necessity (and no one ever mentioned it in any "quick fix" threads posted on the subject from I14 to now). Normalizing offensive and defensive values in secondaries (and for that matter in primaries) is bound to call all sorts of things the devs currently do into question, and probably force significant changes in how they grant powers to powersets to fulfill their concept.



** In I14 beta, I noticed and reported publicly that it was possible in many circumstances to give a custom critter a power and then engineer the situation so they didn't use it. So long as the critters could be outsmarted in that way, any system that attempted to grant XP based on the powers a critter possessed was doomed to be exploitable. Thus the requirement.