Arcanaville

[ QUOTE ]
[ QUOTE ]
Why would it be disastrous for the power to work as designed?

[/ QUOTE ]

Make the -20% defense debuff resistable by the toon's defense debuff resistance

[/ QUOTE ]

Back when it was being monkeyed with, I suggested to Castle that the debuff be changed to be either:

1. -30% resistance debuff, resistable or
2. -20% defense debuff, resistable

and I preferred #1 over #2 due to the issue of mechanics. I also told him how to fix the "stacking preempts the crash" bug at the same time, but I interpret the fact that he didn't proceed further to mean he no longer considered it a priority (this was back when resources were limited and once his time was up, Castle's time was up and it was time to move on to the next issue).

It really is a trivial bug to fix**, so whenever Castle wants to fix it, it'll get fixed.


** as far as I'm aware, somehow, the -DMG and -END were set to Stack, but the -DEF was set to Replace. That means if you can cast Rage before the -DEF is triggered, the next -DEF will replace the previous one and the previous one will never fire. Its not an "Oil Slick Arrow"-like situation, just a typo.

[ QUOTE ]
[ QUOTE ]
[ QUOTE ]
Hey - question.

Why is the Defense Crash the same for Brutes and Tankers? Brutes have .75 Defense modifiers compared to the Tankers 1.0.

When Unyielding from Invulnerability debuffed defense, it had -5% Defense to Tankers, and -3.75% for Brutes and Scrappers. There's precedent. Is there an actual reason or is it just an oversight?

[/ QUOTE ]

That is a good question........Castle?

[/ QUOTE ]

Because its an offensive set not a defensive one. Invul was a defense set and therefore defense mods apply, Super Strength is an offensive one.

[/ QUOTE ]

Actually, the defense debuff in Unyielding was originally** -5% for both scrappers and tankers, because it wasn't set to honor archetype modifiers. Eventually, it was made -5% for tankers and -3.75% for scrappers, but even then I believe it *still* didn't honor archetype modifiers: instead the scrapper version was simply changed to be -3.75% in a hard-coded manner.


** Well, technically speaking it was "originally" higher, but the value almost everyone knows is the -5% value.

The original version of this post got munched by the forum grues, and a number of people have asked me to repost it since. Unfortunately, I forgot to keep a copy, but fortunately I've been able to reconstruct it from google caches and my original data spreadsheets. In particular, it seems to be the only remaining source for the numbers that BaB posted last year, because *his* posts were also expunged.

Anyway, I haven't updated or modified (as far as I know) the original content: I'm just reposted for reference purposes. Note that some of this post has been somewhat superceded by changes to the game since BaB's numbers and my post were posted (such as stalker and dominator changes).

-----[Repost Follows]-----


A while back BaB posted some datamining numbers that I plugged into a spreadsheet and then promptly forgot about. I decided to blow the dust off of it in response to a suggestion that tankers might need attention because of their apparent lack of popularity. I've decided to analyze the given numbers in general, just to see what they say.

I should point out a disclaimer first: even with the numbers BaB provided, the number of data points I possess is limited, and therefore I'm exercising substantial judgement in interpreting the information. But I believe within that limit, my analysis is fairly sound.

First, I'll reproduce BaB's numbers:

<font class="small">Code:[/color]<hr /><pre>
Active since beginning of 2008
Arachnos Widow 1.04
ArachnosSoldier 1.08
Warshade 1.24
Peacebringer 1.71
Dominator 4.8
Stalker 5.77
Corruptor 8.41
Defender 8.87
Mastermind 8.96
Brute 9.08
Tanker 9.71
Controller 10.58
Scrapper 13.71
Blaster 15.04

Last active since Issue 12 golive (may 20, 2008)
Warshade 1.32
Arachnos Widow 1.56
ArachnosSoldier 1.6
Peacebringer 1.82
Dominator 4.62
Stalker 5.5
Corruptor 8.27
Mastermind 8.49
Brute 8.71
Defender 8.96
Tanker 9.52
Controller 11.27
Scrapper 13.48
Blaster 14.88

Logged in since 8-1-2008
Warshade 1.43
ArachnosSoldier 1.76
Arachnos Widow 1.78
Peacebringer 1.95
Dominator 4.54
Stalker 5.24
Mastermind 7.86
Corruptor 8.18
Brute 8.27
Defender 9.31
Tanker 9.5
Controller 12.15
Scrapper 13.32
Blaster 14.71
</pre><hr />


Numbers are all percentages. (there is one more table, which I will get to later)

Now, here's my version of that data, tabulated by archetype and date:

<font class="small">Code:[/color]<hr /><pre>
Archetype Jan 1 May 20 Aug 1
Warshade 1.24 1.32 1.43
Arachnos Widow 1.04 1.56 1.78
ArachnosSoldier 1.08 1.6 1.76
Peacebringer 1.71 1.82 1.95
Dominator 4.8 4.62 4.54
Stalker 5.77 5.5 5.24
Corruptor 8.41 8.27 8.18
Mastermind 8.96 8.49 7.86
Brute 9.08 8.71 8.27
Defender 8.87 8.96 9.31
Tanker 9.71 9.52 9.5
Controller 10.58 11.27 12.15
Scrapper 13.71 13.48 13.32
Blaster 15.04 14.88 14.71
</pre><hr />


Now, a couple of things stand out just from the static tables. The first is that its pretty obvious heroes are more popular than villains in terms of characters logged in (and presumably played). The second is that there appear to be a very distinct set of more popular verses less popular archetypes for heroes: scrappers and blasters appear to be much more popular.

But looking at the population percentages as they change over time, something else becomes apparent: there are definite population shifts occuring. They are relatively small in magnitude, but large enough - and systematic enough - to be unlikely to be random statistically-insignificant fluctuations.

One thing that struck me as interesting is that HEAT/VEAT archetypes all appear to be increasing in popularity over time. That is not inconsistent with the fact that they have unlockable requirements: you have to have a level 50 just to play them, and the devs have stated in the past that across the playerbase as a whole, level 50 characters are not common. Every single epic, even the very long-standing ones, increased in relative popularity over time. I'm not surprised the VEATs are increasing in that manner. I'm slightly surprised the HEATs are still increasing in that manner.

Conversely, [u]all[u] villain archetypes decreased in popularity over time. Now, that might be seen as a potential problem: it could suggest an unhealthy diminishing of the red-side game. And in fact, in some senses (zone population, for example) it probably is. But there is another perspective that might be valid. First, here's a table that shows EATS vs Villains vs Heroes:

<font class="small">Code:[/color]<hr /><pre>
Archetype Jan 1 May 20 Aug 1
EATs 5.07 6.3 6.92
Villains 37.02 35.59 34.09
Heroes 57.91 58.11 58.99</pre><hr />


Intriguingly, Villains are not being scavenged by Heroes, but rather by moreso EATs: Heroes are relatively stable as a percentage of all CoX characters, but players are shifting from Villains to EATs. What's interesting about that is that the net effect is not villains to heroes, or villains to VEATs, but also villains to HEATs, which is interesting because that represents an unusual cross-over: red-side to blue-side, but not to actual heroes. I'm theorizing here that heroes are the "base" starting point for the majority of CoX players, and that the population shift is mostly shuffling around the alternatives to hero archetypes: in a sense, villains are psychologically a form of EAT that doesn't need to be unlocked: its an "alternative" choice. The percentage of players choosing to play "alternatives" is roughly stable.

If my theory is correct, then its possible that the devs themselves undermined villains by creating the EATs (although that would be a retrospective statement in the case of the HEATs): Villains, HEATs, and VEATs might be all competing for the same pool of players (or more generally, player attention span across all players). Since January 1, 2008, Villains have given up 1 percentage point of their population ratio to Heroes, but 2 percentage points to the EATs (a word of caution here: these are normalized numbers, and do not account for actual total numbers of player-characters increasing or decreasing over the time period, so when I say Villains gave up 1 percentage points of the their population ratio, I'm saying nothing about actual numbers of players giving up playing villain characters, which could be a totally different percentage number).


Now, because EATs, Heroes, and Villains are all changing in relative population, its not easy to see directly how the individual archetypes are changing relative to each other within their types. Here are two tables, one of which shows the trend in hero archetypes relative to each other, and the other showing trends in villain archetypes relative to each other:

<font class="small">Code:[/color]<hr /><pre>
Defender 15.32 15.42 15.78
Tanker 16.77 16.38 16.10
Controller 18.27 19.39 20.60
Scrapper 23.67 23.20 22.58
Blaster 25.97 25.61 24.94

Dominator 12.97 12.98 13.32
Stalker 15.59 15.45 15.37
Corruptor 22.72 23.24 24.00
Mastermind 24.20 23.86 23.06
Brute 24.53 24.47 24.26</pre><hr />


Here we can see the relative population trends on the blue side and the red side. On the blue side, Defenders and Controllers appear to be getting more popular, while Scrappers, Blasters and Tankers less popular. On the red side, Dominators and Corruptors seem to be getting more popular, Masterminds, Brutes and Stalkers are getting less popular.

One thing the numbers can't tell us is just exactly how players are shifting around: are they dumping blasters to play defenders or controllers or both, for example. But there are a few things we *can* say. If Tanker populations were going down and Scrapper populations were going up, we might suggest that the data at least implies a shift from Tankers to Scrappers. But that's not what's happening: Scrapper percentages are dropping faster than Tankers are. While its still possible that players are shifting from Tankers to Scrappers, that could only happen if Scrappers were actually *less* attractive to players playing them than Tankers are to players playing them, causing an even higher percentage of them to be shifting out of them to other archetypes. That seems unlikely to me. What seems more likely to me is that the "support" classes of Defenders, Tankers, and Controllers are gaining overall relative to the primary offensive classes of Blasters and Scrappers, and there is rotation going on within the support classes, with Controllers gaining ground.

But if Blasters have *always* been popular, and Scrappers have also tended to be popular for a long period of time, how could they be in decline? Its possible that this is a new phenomenon: Blasters and Scrappers just happened to start losing popularity with players in the last year. But I find that to be an unsatisfactorilly ad hoc explanation, and I can think of no good reason to explain Blaster/Scrapper exhaustion relative to all other archetypes.

There is another possibility, and it has to do with BaB's last posted data table:

<font class="small">Code:[/color]<hr /><pre>
Archetypes created since Issue 12
Warshade 1.05
Peacebringer 1.44
ArachnosSoldier 1.55
Arachnos Widow 1.63
Dominator 4.69
Stalker 6.36
Defender 7.82
Corruptor 8.4
Mastermind 8.57
Brute 9.11
Tanker 9.46
Controller 9.62
Scrapper 14.16
Blaster 16.14</pre><hr />


Once again, my way:

<font class="small">Code:[/color]<hr /><pre>
Archetype Jan 1 May 20 Aug 1 Created
Warshade 1.24 1.32 1.43 1.05
Arachnos Widow 1.04 1.56 1.78 1.63
ArachnosSoldier 1.08 1.6 1.76 1.55
Peacebringer 1.71 1.82 1.95 1.44
Dominator 4.8 4.62 4.54 4.69
Stalker 5.77 5.5 5.24 6.36
Corruptor 8.41 8.27 8.18 8.4
Mastermind 8.96 8.49 7.86 8.57
Brute 9.08 8.71 8.27 9.11
Defender 8.87 8.96 9.31 7.82
Tanker 9.71 9.52 9.5 9.46
Controller 10.58 11.27 12.15 9.62
Scrapper 13.71 13.48 13.32 14.16
Blaster 15.04 14.88 14.71 16.14
Defender 15.32 15.42 15.78 13.67
Tanker 16.77 16.38 16.10 16.54
Controller 18.27 19.39 20.60 16.82
Scrapper 23.67 23.20 22.58 24.76
Blaster 25.97 25.61 24.94 28.22

Dominator 12.97 12.98 13.32 12.63
Stalker 15.59 15.45 15.37 17.13
Corruptor 22.72 23.24 24.00 22.62
Mastermind 24.20 23.86 23.06 23.08
Brute 24.53 24.47 24.26 24.54</pre><hr />


Notice: the number of Blasters and Scrappers created since Issue 12 (May 20) is higher, as a percentage of the total, than the number active since Issue 12. I'm going to presume here that any character that is created is "active" for the day that it is created, whether it is played out of the tutorial or not. This means, in effect, that although more Scrappers and Blasters were created since Issue 12, less of them have been played relative to other archetypes between May 20 and Aug 1. The most logical explanation for this is that Scrappers and Blasters are abandoned more frequently than other archetypes. You can't keep making more but playing less unless a lot of them are ultimately being deleted or idled.

This is significant because if Blasters and Scrappers have a higher idle rate (maybe they are played by trial accounts more, or short-term subscribers, or are impulse creations more often) then that means their population numbers are skewed upward to begin with: they include a higher percentage of characters that are not played consistently or at all past a certain point. This could also explain why it appears that Scrappers and Blasters are dropping in popularity: they might not be. One peculiarity of the numbers BaB posted is that they are cumulative numbers: the Jan 1 numbers represent the percentages for all logins from January 1 to about September 1. The Issue 12 (May 20) numbers are for all logins from May 20 to September 1. In effect, the Jan 1 numbers *include* the Issue 12 numbers, because anything logged in from May 20 to September 1 has obviously also logged in from Jan 1 to Sept 1.

This also means those data points are structurally different: the Jan 1 data point represents a larger window of time, the Issue 12 a smaller window of time, and the Aug 1 data point the smallest window of time. If there exist characters that, say, are only logged in once, or only for a week, or only for a limited time before being shelved, they are more likely to show up in long duration windows than short ones: there exists a subtle bias towards "testing" characters in the Jan 1 numbers relative to the Aug 1 numbers. Its statistically possible for that to explain the apparent reduction in popularity for Scrappers and Blasters. If that is true, its possible that if we were to do a more complex datamine of recurring character logins to determine popularity, Blasters and Scrappers could end up being slightly lower than the numbers suggest.

Conversely, anyone suggesting that the increase in Controller popularity comes only from recent powerset proliferation and new characters would appear to be contradicting the numbers. A relatively small percentage of all new characters created between Issue 12 and September 1 were Controllers - less than their overall play percentage. So while a burst of new characters are almost certainly helping Controllers, what matters more is that Controllers appear to have a lower attrition rate. They are gaining ground on archetypes that players are creating more of: that's extremely strong evidence that Controllers, for some reason, retain players much better than average. The same statement can be made for Defenders, and to a slightly lesser extent, Dominators and Corruptors.

Tankers and Brutes, meanwhile, seem to be relatively stable. Their creation rate is very slightly higher than their play rate for the same period of time, which implies that they have a very small (relative) attrition rate. And that small attrition rate ought to lead us to conclude that the visible decrease in population percentage is slightly overstated. And *that* leads me to conclude that, as a percentage of their respective factions, Tankers and Brutes have basically stable population percentages (at least during the period in question). They are neither losing ground nor gaining ground significantly, even though the raw numbers show a slight decline.

If I had to grade the archetypes in terms of their ability to *retain* players, and not just convince them to create characters, as a metric for intrinsic attractiveness, I would rank them as follows:

For Heroes, best to worst:
Controller
Defender
Tanker
Scrapper
Blaster

For Villains, best to worst:
Corruptor
Dominator
Brute
Mastermind
Stalker

And in particular, the archetypes that have *exceptionally difficult* times retaining players appears to be Scrappers, Blasters, and Stalkers.

My take on this: Blasters and Scrappers are the most intrinsicly closely identified with superpowered characters in beginning player's minds, and there is a more subtle but still measurable bias towards them conceptually even for experienced players. They are more likely to be rolled purely on the basis of evocation of the name, without regard to how they've actually been designed in the game. They are thus much more likely to fail to live up to expectations. On the other hand, Defenders, Controllers, and Tankers do not trivially match most players mental genre images: there are less visceral connections to them. They are much more obviously MMO constructs, and the touchstone for them are MMO exemplars. As a result, players who roll them are statistically more likely to base their expectations on MMO conventions, not evocative concepts (I say "expectations" because even if someone creates characters based on character concepts intrinsic to their own imagination, that's a completely separate issue from whether or not they believe they will be able to represent that concept to a certain level of fidelity: character concept-driven players are not necessarily unrealistic in their expectations for the game). As a result, they are much less likely to be surprised or disappointed with their design or performance.

If this is true, then there isn't anything structurally "right" or "wrong" about Blasters and Scrappers that causes them to have very high initial attraction and character creation, and then subsequently lower retention rates and higher drop out rates. Its simply that their evocative concept is more attractive to the very people who would have intrinsicly low retention rates no matter what they played (because their expectations are more likely to fail to match the game design). In other words, the moment the two archetypes were called "Blasters" and "Scrappers" they were probably already locked into this situation. To determine if the net overall play rate of Scrappers and Blasters is unhealthy (unhealthily high or low) I would need more data: at this point I've exhausted what BaB's statistics can tell me in this regard.

That is hero side. I believe its much less likely that neophytes will jump first into City of Villains before City of Heroes (statistically less likely, not overwhelming certain - we're talking about skews on the order of just a few tens of percent). They are much more likely to set expectations based on the design of the archetypes as presented than external expectations. And in general, the character creation rate never veers very far from the play percentages; heroes are much more widely dispersed, which supports that contention. That makes Stalkers stand out as a very strong exception on the red-side. In a faction that is statistically much more realistic about its general expectations than average, Stalkers have a relatively vast retention rate gap. This suggests that there is a problem intrinsic to Stalkers: they are more attractive on paper than in the actual game among a playerbase that almost always correctly judges how attractive something will be in-game at character creation time (or rather, they get it wrong equally often for every other archetype but Stalkers, where they get it wrong a whole lot more often).


Conclusions.

Overall, I would have to say that the data supports the contention that at the moment hero-side archetypes are probably fine, relative to each other, in terms of player popularity and long-term attractiveness. The highest percentages also have the lowest retention rates, and vice versa. But that's a relatively coarse conclusion: more detailed numbers could demonstrate more subtle population problems, or turn one over that the current numbers are unable to present clearly.

Tankers seem to be the most stable of the five. It might be better if they were stable and higher in representation, but the real gap is probably much lower than an at-a-glance estimate of the numbers would suggest without more detailed analysis (because of the slight overstating skew of Scrappers and Blasters), and its unclear if Tankers are less attractive than Scrappers, or less attractive than other support classes. Without additional data, its impossible to know and subsequently, what changes could affect the situation properly, assuming you wanted to.

Controllers are just plain doing fantastic.

Villains are slowly losing players, which is - or should be - a significant faction concern. But curiously, they are losing only partially to Hero archetypes: they are losing more to EATs of *both* factions. This suggests a strong psychological signal: the devs may have failed to create a competing faction to Heroes when they created Villains: they may have actually accidentally created the largest EAT group of the City of Heroes game. That has long-term implications for the Powers That Be, when deciding how to handle the red-side moving forward, *and* in terms of any future plans they might have for a City of... expansion.

Stalkers need ... something. I'm not sure what they need, and I'm not sure how much of it they did, and without additional data I can't even say with absolute certainty that they even need help. But given only the data I have, something seems off on Stalkers with regard to their design. I'm sure there will be no end to the competing theories here. And perhaps we need more time to let the latest batch of Stalker changes to sink in. But somehow, I doubt they will have a very strong impact on these numbers.

EATs are on the way up, even HEATs. But VEATs are, by any measure, a much more successful venture. The question is, does the quick adoption rate of VEATs imply that they will continue to grow at a significant rate, or do they have a soft cap like HEATs appear to have. And how long it took for HEATs to reach current levels is an unanswered question that would require more historical data.


A note on the analysis: I glossed over much of my analysis to save time and space. These numbers are, by their nature, non-trivial to derive conclusions for (the overlapping nature of the intervals of measurement are a significant complication unto itself). There are often implications that numbers lead to conclusions more directly than they do. I would recommend caution to anyone attempting to duplicate or extend this analysis, and for that matter given the limits on the amount of data I have, I would also be careful claiming stronger certainty than I am. These are judgements not proofs of anything. I believe they are sound and reasonable judgements based on a lot of experience with such analyses, but this does not have the certainty of mathematical calculation. Others could reasonably come to different conclusions. They'd probably be wrong if they disagreed with me of course, but not certainly wrong. Seriously, anyone quoting from this should not say I "proved" anything here: it'll only force me to post a correction.

And there is a big elephant in the room regarding any attempt to analyze population data: they don't distinguish between PvP and PvE activity. That is potentially very significant, but would require more data to determine.

[ QUOTE ]
"Actually" makes it sound like a contradiction and then you proceed to just expand on what I've been saying in this thread.

[/ QUOTE ]

The "Actually" refers to the notion that Fire was considered superior on paper. It always had been up to that point, but in the months leading up to the Fire changes that significantly reversed. It wasn't so much that experience trumped the numerical analysis as it was the numerical analysis agreed with experience (by that point) in this case.

My recollection is that I was somewhat surprised to discover that while Fire was always assumed to be the AoE king, it seems no one actually tried to compare Fire to AR because the presumption was that AR's performance was so low it wasn't worth bothering. Furthermore, while Fire did have very good DPA no one (that I'm aware of) did an analysis of just how much recharge was required to fill a chain with Fire's top DPA attacks. It was always presumed that attack chains for reasonable builds were always "full" so there was no need.


[ QUOTE ]
I personally think blasters as a whole overstated their problems and whined themselves right into new defiance. In any event the dev's did not agree that Lrn2ply was sufficient advice. At the time most people were arguing that fire was fine.

[/ QUOTE ]

The I11 changes were datamining driven. When the devs ran their performance stats across the game, they discovered that when you look at the average levelling rate and reward earning rates (influence, drops, etc) of all players for each level range, and compare that average to the average levelling and earning rates of each blaster powerset combination (i.e. all energy/energy blasters, all fire/dev blasters, all ice/elec blasters) it was discovered that *all* of those combinations were lower than the overall average. By double-digit percentages in all cases. Even Ice/Ice, even Fire/Energy: all of them. I was also told that that situation was *unique*: there was no other archetype for which you could make that statement. That is what made blasters the high priority action item at that time.

Blasters had been "whining" about various things since I1, and their complaints peaked around I5. By I11, they were actually practically non-existent (at the time that blasters were being worked on, there were more complaints about Kheldians, Tankers, Stalkers, and Dominators in the public forums).

Whether Blasters overstated their problems is somewhat moot, as Castle determined on his own that a) blasters of all powerset combinations were underperforming the playerbase as a whole by significant amounts and b) that underperformance was partially due to being killed and under debt more often, and c) defiance 1.0 was probably encouraging more risk-taking than the archetype was designed to handle for the average CoX player. All of the player feedback related to the changes were details fit into that basic framework already established.

[ QUOTE ]
[ QUOTE ]
You're saying the system should be terminated because you can't stand other players receiving a reward earlier than you.

[/ QUOTE ]
I'm saying it should be terminated because I can't stand other players receiving a reward I can't. Continuing the system until the game closes creates just that problem.

[/ QUOTE ]

That's mostly a theoretical problem and not a real problem for any of the current rewards, and furthermore I addressed it already:

[ QUOTE ]
I say "to the same degree" because there is a loophole never acknowledged in the iterative argument that an indefinite veteran system prevents players from "eventually" getting all the rewards, simply due to the fact that one day the system will be shut down and prevent players from achieving all of them. The devs could simply state that if the game were ever to be shut down they'll autogrant all the veteran rewards to the remaining players at least 24 hours before the servers close.

[/ QUOTE ]

[ QUOTE ]
It was fueled by players specifically saying Fire was not performing well enough above the other sets relative to its total lack of mitigation.

It was a situation where feel and experience trumps paper because even on paper fire was still better. I'm inclined to believe that even though the game has changed, it hasn't changed enough to make the reason for the fire buffs invalid.

[/ QUOTE ]

Actually, it was right around that time when people started looking at blaster primaries on paper again. I adapted PeakDR for blaster/defender ranged sets around mid '06 on request, and my Ignite post dates from that period as well. PeakDR showed that Fire's performance reputation was exaggerated, which partially knocked the legs out of the notion that Fire's mitgation was balanced by Fire being the absolute top performer (damage types aside, AR exceeded Fire's performance under the PeakDR metric).

Its also possible a post I made about DoT around that time influenced this, although I'm not sure. While adapting blaster primaries to my PeakDR calculations, I noticed that the average DoT estimates of sources like CoD and others were assuming flat percentages, and not factoring in CancelOnMiss. CancelOnMiss significantly reduces the amount of average DoT a power will generate, and I think its an interesting coincidence that the true average tick value of a five-tick cancelonmiss DoT (2.69) is about 50% lower than the estimate you'd get with flat percentages (4.0).

Basically, five ticks is the maximum DoT for fire blast (Blaze has five) so if you wanted to adjust Fire's DoT damage upward so that the average DoT was at least as high as the flat percentage would imply, you'd increase the DoT by about 50% in magnitude.


Nevertheless, I think Fire blast's problems were overstated then. Or rather, they were overstated relative to blaster problems in general: the entire blaster archetype was subsequently discovered through datamining to be underperforming (which was the trigger for the I11 blaster revisions). I think Fire's expressed problems in I8 were really endemic of blaster problems in general for most of the player population. And in a possibly fortuitous twist, Fire ended up benefiting slightly less than most blaster sets in I11, because it was already a fast set (Flares notwithstanding) and thus didn't gain quite as much ground as other sets did by being cast time adjusted. So its entirely possible that Fire ended up right where it ought to have been in the end by getting a DoT buff and less benefit from cast time changes.

[ QUOTE ]
I've gotten the same reaction to "stop giving out new rewards" as I have to "give newer players a chance to get more rewards". One is not the same as the other, but is treated with the same disdain.

Stop giving out new veteran's rewards. Let there be an upper limit to how many rewards there are. That will make it "fair" for everyone; everyone puts in five years, everyone gets five years of rewards.

[/ QUOTE ]

That's how the system already works now. Quoting "fair" implies there's any reasonable grounds for debate as to whether the system actually is fair. It proscribes the same requirements for the same level of reward for all players. It allows no player to circumvent those requirements. That's essentially the definition of a fair reward system.


[ QUOTE ]
Let me turn it around: Why do you need more than that? Isn't five years of reward enough? Why are you demanding more? Why do you feel entitled to your continued rewards? What makes you so special?

[/ QUOTE ]

I'm not asking for more rewards. If the devs ended the system because they felt it had run its course I wouldn't ask for an extension of the system. The question is whether anyone else is important enough to dictate what others should get from it and when.

You're saying the system should be terminated because you can't stand other players receiving a reward earlier than you. I'm not saying the system should not be terminated because I deserve more rewards: I'm saying the system should not be terminated because that's a crappy reason for terminating it.

[ QUOTE ]
You want to know what I hate? I'll tell you!

[/ QUOTE ]

I hate it when people ask me a question when the answer is irrelevant to them.

[ QUOTE ]
How, if you're so good at the game, could you possibly confuse the Cape and Aura missions? As a followup to that, how can you suggest that either are hard to solo at the earliest level possible?

[/ QUOTE ]

A curious twist to the cape mission is that when it first came out, a lot of people *thought* it was hard, and so tended to invite more and more people to "help" with it. Ironically, the more people you added, the harder it got to save the time capsule before it was destroyed, which caused people to add even more people to try to make it easier. I didn't know it was supposed to be hard and soloed it with all of my alts at the time without any difficulty.

Today, though, although the cape mission isn't necessarily trivial for neophytes, experienced players should have no difficulty soloing it with most archetypes. It is still actually more difficult to do it with "help" than it is solo, unless you're running a completely offense-less defender or something.

[ QUOTE ]
I agree with your overall point that this is cartoonish violence and not realistic, I just find the statement that having hair makes you vulnerable to be a bit over the top.

[/ QUOTE ]

In Paragon City and the Rogue Isles, choosing to have hair is not nearly as dangerous as choosing not to be bullet proof, so I don't consider it to be an especially critical defensive choice.

[ QUOTE ]
unlikely, that's still 20k+ HP you have to chew threw even if he's grey to you.

[/ QUOTE ]

16556 health, and at -15 to you your damage would scale up by a factor of 2.6. The net result is that you'd need an intrinsic damage output of about 446 dps at level 50. That's within the realm of possibility for a player with buffed temp pets (I don't think any singular player character without pets reaches that level of damage output with any realistic build and slotting for a long enough period of time). At least on paper, I think a player + shivan + heavy all buffed to the damage cap can do this (it may not require that much buffing either if you position the heavy correctly). I don't know if the pets will cooperate well enough to do it in reality fast enough, though.

Its worth noting that unless you're playing the CoH version of trivial pursuit, this is also completely impractically worthless advice. This is the sort of thing TopDoc and I might debate on the forums while BillZ and Werner check our math, but I don't see the practical application of claiming this is possible under these extremely contrived conditions.

[ QUOTE ]
The devs hate me. It's a fact. Partly because I'm a crusty forum regular.

Here's the proof:

They nerfed claws by removing weapon redraw and increasing overall DPS.

They nerfed soloing by introducing IOs. Game content became too easy. Now they're going to give us the ability to set mission difficulty based on virtual team size thus increasing the difficulty of standard missions.

They nerfed SR after fixing it. Once upon a time we could have permalude. They ED and the GDR ripped that out from under us. SR was hurting so they gave it passive scaling dam-res. Then they gave us way to many positional defense buffs with setIOs and Arcana's critter accuracy change and now we're back to permalude levels of performance.

Wait a minute.... this post makes no sense at all. The devs LOVE ME! wheeeeeeeeeeeeeeeeeeeeeeeee

and the love burrrrrrrnnnnnnnnnnnns... proof that they hate me.

/confused now

[/ QUOTE ]

Bill, I'm not sure if they keep luring you back because the devs love you or if its that they hate the rest of us.

[ QUOTE ]
ask yourself the simple question.....What things do i need that i can acquire to overcome Rommy in 15 seconds.

[/ QUOTE ]

You'd need The_I_Win_Button and access level 9.


A mastermind whose six pets were all yellow mitos might also have a chance, with the appropriate secondary.

[ QUOTE ]
If your problem is "impatient players", how about they just stop adding new Veteran's Reward after the 72-month mark (or some other, near-future cap)? Then newcomers and returning players can eventually earn all the rewards, after staying subscribed for a period of time. That alternative appeals to me as well.

Your alternative does not, because the issue is not wanting a reward sooner, but wanting all of them eventually. So long as new rewards are added, the current system will never allow that.

[/ QUOTE ]

This is true to the same degree for all rewards, not just Veteran Rewards.

I say "to the same degree" because there is a loophole never acknowledged in the iterative argument that an indefinite veteran system prevents players from "eventually" getting all the rewards, simply due to the fact that one day the system will be shut down and prevent players from achieving all of them. The devs could simply state that if the game were ever to be shut down they'll autogrant all the veteran rewards to the remaining players at least 24 hours before the servers close.

Under those circumstances, players have the same chance to receive all the veteran rewards in the same sense they have the same chance to eventually receive all the other in-game rewards, in fact they have a higher chance because that chance is 100% under that policy if the player plays indefinitely. That's not true for all the other in-game rewards.

[ QUOTE ]
[u]3-Step Plan for Appreciating Dimension Shift[u] according to LW

1) Find smarter teammates.

[/ QUOTE ]

Completely coincidentally, I made a suggestion to Castle regarding detention powers about the same time this thread was started specifically to address the fact that when used by critter allies it can often be a net buff to foes as anything else.

I was specifically thinking about detention field and not dimension shift - although I was thinking about foe intangibles in general - and the suggestion was basically to apply -regen/-recovery to the target (so it doesn't actually get stronger while shifted) and a fast-ticking low-damage DoT (so a target at the edge of being defeated can get eventually knocked out by the power), and a new "detention" animation so its obvious which targets are affected (the rapid low damage DoT ticks above their heads should also help there).

That doesn't affect anyone who is using it for some tactical purpose now, it just increases its useability a tad for players and makes it a bit less annoying when allies use it. I can even think about an AI tweak that might make it more tactically intelligent: mark the power as a heal: critters will then try to shift your foes generally only when you get low on health, and if it targets you with the power it might even target through to your target, which means you could theoretically select the target that is shifted (although this is less applicable to dimension shift).

[ QUOTE ]
[ QUOTE ]
[ QUOTE ]
[ QUOTE ]
[ QUOTE ]
If I can't change out the geometry for Spines then forget about it.

Just say no to banana melee!

[/ QUOTE ]

Apart from Spines being able to use the geometry from Thorns, and Thorns being able to use the geometry from Spines (both of which can be colored), there will be at least one other set of alternate geometry/FX.

As for Martial Arts, the alternate animations we have right now are mainly punches. The idea was to provide you guys with a way to make the classic, fights with their fists and feet, super-hero that many have been asking for. That includes a new animation for assassin's strike (Assassin's Blow?).

[/ QUOTE ]

Any chance on the alternate animations that you can set it to cycle between the two (or more), or choose at random at each time the power is used?

[/ QUOTE ]

Right now, the tech doesn't work that way. It's more an issue with the VFX not syncing up with random animations, as we've had the ability to randomly or sequentially play multiple animations for a while.

It might be something we could do in the future, but very unlikely to happen for I16s launch.

[/ QUOTE ]

Wouldn't that be similar to how you can spam Shadow Punch or Swipe, one time it uses the left hand and one time it uses the right hand? it's two different animations for the same power (that we currently have)

Also, first after a redname: yay!

[/ QUOTE ]

I think BaB is saying that if the two different animations have different timings (a punch and a kick might both take the same total number of frames to animate, but they might strike the target at different moments during the animation) then the visual effects associated with the animation (kapow) have to be configured with the correct timing so they are in sync. If the system is given two different animations to play for the same power activation with the same priority, the system will randomly cycle between them (i.e. left punch-right punch) but in those cases the two animations generally have the same timing so the other surrounding effects stay in sync regardless of which animation plays.

Customized animations in I16 don't work that way: the power is given two or more completely different animation "configurations" and in effect the player is given a setting to select which one of them is operative for their power. Theoretically speaking I think it might be possible to give the player a "random" setting in the power configuration screens (whatever they look like) but I don't know if there are deep game engine limitations that would make such a setting break stuff (maybe swapping animation settings like that cannot be done on the fly, for example, but only when the character is "idle" and in the Icon screens).

[ QUOTE ]
[ QUOTE ]
The real purpose of VRs is to entice players to keep paying and playing.
Every player knows that in less than 3 months, there is another little perk waiting for them, if they stay.

[/ QUOTE ]
Actually, the fact that I will never get all the Veterans Rewards, that there will always be another reward out there that I'll want and will never get to makes the Veteran's Rewards completely ineffective as incentive.

[/ QUOTE ]

For some. But I suspect not for the majority of players.

[ QUOTE ]
If you want to see a good range of challenge-focused enemies, check out Arcanaville's Scrapper Challenge mission arc.

Personally, I can tell you from experience that Sonic Resonance on Extreme on higher-ranking enemies can be pretty nasty, especially if you're playing a resistance-based character. This is doubly true if the Sonic Resonance enemy is supporting some minions with a damage or defense-debuff focus.

[/ QUOTE ]

Sonic is serving four roles in the scrapper challenge:

1. Provide an advantage to scattering foes, as a counter to the general reward for clustering them.

2. Make the critters incrementally tougher.

3. Make them a stacking threat (usually its the other way around).

4. Interrupt aid self a lot.


Which reminds me: I need to update that. I keep getting distracted from testing some improvements to that mission.

[ QUOTE ]
Isn't that exactly the same thing ?

I.e., if Shield Charge is the power number 7 on your hotkey bar, you'd press 7 and click the location. Here you press shift and click the locationn. I don't see much difference.

(I'm just asking because I've seen this bind used for teleport too, yet I don't understand what makes it different than using the hotkey bar.)

[/ QUOTE ]

With the hotkey method, you are pressing the hotkey to activate the power, then clicking on the location to set the target. With the lshift+button method, you are pressing the shift and mouse button simultaneously and its that act that triggers the power and the location selection at the same time.

Its a little different in that the lshift key by itself does nothing: its just a meta key. I can hold down the shift key and then at exactly the moment I want to cast the macro power click the mouse button. Really, the main difference is that they do not rely on specific powertray arrangement: I don't even *have* teleport on most of my character's trays, which saves a space (which used to be more important before having customizable tray layouts).

Every character I have that has teleport uses lshift+button mainly for consistency. Additionally, some characters use control or alt + button for other things (I have control+button mapped to phantom army on my Illusion controller).

I don't think there's any major advantage either way. Shift-click mainly has historical advantages. I suppose if you're not a touch typist, its also easier to quickly hit shift+button than to hit hotkey+button since the shift key is bigger and you can keep a finger homed over it.

(The only disadvantage I can think of to using hotkeys is that its theoretically possible to doubletap the hotkey and to a quick activate-deactivate on the power; that's basically impossible with shift+click. But I don't think that's the sort of key error that happens often)

[ QUOTE ]
The i16 info actually vindicates me.

[/ QUOTE ]

What a surprise. Let's review:

[ QUOTE ]
Allow me to clarify.

Knowing that:

-You've implied in the past the animation and power systems effects were held together by Elmer's glue and thumbtacks, and that changing animations was cumbersome.

-Cast times for animations are not standard, greatly cutting down what animations could be swapped around without impacting power balance. Unless a system was created to auto re-time them and assuming they still looked OK...or unless you created dozens of versions of the KO Blow animation to fit every conceivable cast time, I wouldn't see the work involved being worth it just for something as minor as every attack having one alternate animation.

-You're already swamped with work doing just the particle effects tinting aspect of power customization, which as far as we know is the extent of the customization i16 will offer(the already existing weapon customization system not withstanding).

-Melee sets without weapons are the minority. There's more bang/buck ratio for putting work in to custom weapons and tinting particle effects.

[/ QUOTE ]

So you assert that:

1. The code that connects powers to animations is fragile enough to make changing it unlikely.
2. Cast times would be a problem or restriction for alternate animations
3. There's too much work with color tinting to allow for the creation of custom animations
4. Melee sets without weapons are a minority not worth working on.


I16 proves you correct that we'd have to be astronomically lucky for I16 to "allow alternate animations for our powers" because in I16:

1. They added relatively simple code to allow for powers to call alternate animation and FX sequences
2. They did so in a manner that always preserves cast time
3. They decided to make time specifically to create custom animations in some cases...
4. ...specifically for a couple of the unarmed melee sets because their color/tinting options are lower, and to introduce the fact that I16 now allows for alternate animations for your powers


Of course, when you look at it that way, its obvious how you could come to the mistaken conclusion that I16 vindicates your position. For at least this definition of vindicates:

Vindicates:
To get revenge for; to avenge

It does look like I16 is specifically designed to take revenge on your posts, but that's entirely a coincidence.

[ QUOTE ]
[ QUOTE ]
However, I can't cut a lot of slack on this one

[/ QUOTE ]

Nothing to do with the guy explicitly saying that your guide is wrong, right?

[/ QUOTE ]

I don't deduct points for that all by itself, since out of context there's no way to know whether a player-written guide is accurate or not. The Guide to Tohit and Defense is written essentially authoritatively, without referencing all of the data that backs up its information (there's simply too much: you either know the historical basis for it, or you don't). So it doesn't have an easy way to check its information independently in all of its details.

The real problem is that while its not trivially easy for someone with no outside assistance to deduce how defense works, its trivially easy for someone that plays Super Reflexes of all sets to deduce that attacks do not work the way the person in question implies, which is to say that attacks check against each power individually. Such a person would have to never have popped a luck in their playing career or have poor observational skills.

It would be different if someone read an analysis of mine that was internally foundationally supported and simply dismissed it out of hand (like, say, the "Arcanatime" post): there I believe I would have more justification in tagging someone trivially dismissive as being unjustified.

[ QUOTE ]
I've been playing a blaster since my first day so I'm new to defense and defense bonuses... Here's the question - let's say I have 20% melee positional defense and 10% SL typed defense, so does that equate to 30% SL defense at melee?

[/ QUOTE ]

The simplest way I can explain this is this:

Find the attack types of the attack: for example: energy, smashing, ranged.

For each of those types, add up all of your applicable defense, in this case to get your total energy, total smashing, and total ranged defense.

Pick the best one of those, and use that.

Basically, attacks don't check against powers they check against defense valus. The defense values are just buckets that the defense powers fill.

In the above case, you would have 20% melee defense, 10% smashing defense, and 10% lethal defense. Your best defense against an attack that was typed smashing/melee would be 20%, and your best defense against an attack that was typed lethal/melee would be 20%. Your best defense against an attack that was lethal/ranged (assuming no other defenses) would be 10%.

[ QUOTE ]
He knew the 1000% defense couldn't be true, so I guess he figured that instead they had to occur sequentially in some way &amp;#8211; 55% of attacks go through his melee defense, 55% of those got through his ranged defense, etc. Or maybe something else, like every melee defense power individually, 93% getting through Weave, 97% getting through the Steadfast Protection IO, and so on. I can't tell what he was thinking. I can just tell that it was almost certainly wrong.

[/ QUOTE ]

He is almost certainly thinking (among other things) that the % numbers are expressing percentages of something when in fact in the strictest sense they are actually only expressing percentage points instead. The distinction is critical, because when you think "percent" you think "some fraction of one hundred" whereas percentage points are just points: 45% means forty-five points of defense. Not forty-five out of something, just forty-five. One hundred thirty three points of defense is not unreasonable: 135 percent defense is mathematically unreasonable. When you think this, your intuition will lead you to incorrect places about how defense probably stacks, because "obviously" 50% defense and 50% defense "cannot possibly" stack to 100% defense.

However, I can't cut a lot of slack on this one, especially with real numbers, because even factoring in observer bias, observer skew, and just plain human-limited observation, the defensive powers don't work remotely close to the percentage numbers they imply. I can cut a lot of slack for someone that doesn't understand, but not someone that adamantly understands incorrectly.

In any event, anyone with an incorrect operating theory for how tohit works can only hurt themselves (unless they are handing out a lot of build advice).

[ QUOTE ]
[ QUOTE ]
No, I'm not because whatever the dictionary.com definition of the term, "Rube Goldberg-like" only has to invoke the illustrations of Rube Goldberg. The design of the powers system/animation system interface does. But its not especially complex. Although some Rube Goldberg systems are highly complex, not all are: the defining characteristic of Rube Goldberg systems are that they are more complex than they need to be. That doesn't mean they are very complex as such.

(Example: this is from the Rube Goldberg site in which sitting on a cushion forces air through a tube which blows an ice boat to a lighted cigar butt popping a balloon which causes a dictator to think he's been shot and fall over backward onto shutter bulb. That's characteristicly ridiculously more complex than necessary, but not very complex to understand in general).

[/ QUOTE ]

Actually, I wouldn't consider the animation systems to be more complicated than they need to be. I'd consider them to very uncomplicated, but because of the sheer number of animations that have to be wrangled, that simplicity makes the task extremely complicated and fragile.

The best analogy I can think of is to think of how the internet would work if URLs didn't exist, and instead all web pages were accessed using search tags. Defining an address for "The Official Beekeepers Society" web site would be pretty simple. "Bee" "Keeper" "Society".

However, the first thousand web pages were created using only 100 tags. Then they upgraded the web because there was only so many combinations where you could use "Bee" that made sense anymore, and they allowed for any number of tags to be defined and used. Then added about 5000 more pages into the mixture of old tags, new tags, recycled tags, massively overused tags, innapropriately used tags, etc...and the simplicity of the system makes it complex.

Tags as words isn't a very accurate concept either. Tags being keys that you would have to hold down simultaneously would be a more accurate analogy. So holding down "B", "K", and "S" would take you to "The Official Beekeeper Society." website, while holding down "B", "K", "S", and "L" would take you to the "Barbecued Steak &amp; K'Bobs Lovers" website (great deal on aprons right now).

Then add in a half dozen or so cats who like to pounce on your keyboard while you try to surf the web, and you've got a pretty accurate picture of how our system decides what animation to play and why it's so damn tricky to keep everything working.

[/ QUOTE ]

And then when they decided to internationalize the web, they added the ability to add language search words, so you could look for Bee Keeper Society English and Bee Keeper Society German.

Except if you added a page tagged Bee Keeper Society English but didn't delete the original Bee Keeper Society page, then a search for "Bee Keeper Society" would randomly retrieve one or the other. Editing Bee Keeper Society English but not Bee Keeper Society would cause people to sometimes get the updated page and sometimes get the old page. To force Bee Keeper Society English to only retrieve the "Bee Keeper Society English" page and not the "Bee Keeper Society" page you would need to set a priority on the Bee Keeper Society English page that was higher than the original Bee Keeper Society page.

And if I made a page called just "Bee Keeper" and set its priority very high, everyone searching for Bee Keeper Society English would suddenly be redirected to my page without warning.

To predict what page you'd get if you searched for "Keeper" you'd need to find all pages with that tag, look at their priority, and pick the one with the highest value. If there was more than one, roll dice and randomly pick one. Conversely, to predict what priority to set to a new page called "Bee Society" you'd have to find all pages that use either of those tags, all priorities assigned to those pages, and ensure that your priority was higher than all pages that use one or the other exclusively, but lower than all pages that use both with additional tags.

And then the cats start to mate on your keyboard.

[ QUOTE ]
Sorry about my loads of questions, I just really don't want to suck.
I'd feel stupid if I rerolled a character from a badass to something...less badass.

[/ QUOTE ]

So long as you don't reroll into a completely different archetype, you're not going to suck. No scrapper combination sucks (although I'd be careful about rolling electric armor scrappers: I doubt they will suck either, but they will be tricky to squeeze comparable scrapper-level mitigation performance out of).

I'd say that Willpower has slightly higher potential than Regen, but there are playstyle issues to consider. Regen has strong heals which means it tends to encourage players to push a little beyond the set's capabilities, provided there is a heal available to bail you out of. Willpower has +regen but no heals: if it gets in trouble its options are inspirations, and running. However, Willpower is less likely to get into trouble in the first place all other things being equal.

I emphasize "all other things being equal" because they often aren't: scrappers tend to like to push boundaries, which means what often matters is not how strong a powerset is, but how well it deals with being pushed too far. What I mean by that is that if there were two scrapper secondaries, and A allowed you to take on 12 things at once, but 13 would probably kill you quick, and B allowed you to take on only six things, but if you tried to take on ten it would hang in there for quite a while before breaking down, a lot of players would instinctively suggest that B was far stronger, even though in this case numerically speaking A is vastly stronger.

In this case, I believe given comparable expenditures in build Willpower is going to be stronger than Regen, but a Regen player transitioning to Willpower might die a few times before realizing they don't have that emergency button heal and not pushing as hard as they would a Regen. They'd be soaring at a higher level if they adjust to the playstyle difference though.