Arcanaville

Not to mention all those poor SR players that committed suicide between I4 and I6.

Just keep in mind, I'm not an employee of Paragon Studios. I'm just the most comprehensive tester of tohit willing to answer questions. The devs don't like touching this subject with a ten foot pole because it tends to draw them into other discussions about the whys of game implementation which they often cannot answer. I'm slightly less restricted and much more credibly deniable than, say, Television or pohsyb, and unlike them I can't be drawn into a debate about source code (because I have no direct access to source code).

But just to put my usual disclaimer, its against the rules to impersonate a dev, and also against the rules for devs to post anonymously. Not a dev.

As far as I can see, most of the nitty-gritty of combat mechanics seems to be not well documented in actual in-game documentation or printed manuals for most MMOs, and in particular mechanics that are added to the game after launch tend to be discussed but not always well documented officially.

And the accuracy of the attacks matter to prevent overskewing of the results. You want miss streaks to be broken quickly when the chance tohit is high because the expectation is for shorter miss streaks when net tohit is high. But you can't break similarly short miss streaks when net tohit is low because that would then increase the number of hits to unacceptably high levels for those cases.

Long ago there were a couple of independent calculations done on the streakbreaker to determine its overall impact on average tohit for NPCs. These calculations were more or less in agreement: they differed in certain assumptions about how long the average continuous attack streak was likely to be, which affects the streakbreaker's ability to see and thus truncate miss streaks. I believe the worst case scenario was somewhere around 30% net tohit which could, in some cases, experience something like 25% more hits in very long duration circumstances. As a practical matter, though, because few things actually live long enough to generate such long miss streaks, the actual increase in incoming hits was much lower than that. An AV could live long enough to trigger the streakbreaker often enough to approach the worst case scenario, but then again I don't think many players attempt to tank AVs with only 30% defense and no other meaningful protection.

Its combat logs, not simulations. If it were a simulation that I wrote, it couldn't possibly be worth anything in terms of checking the actual game server's random number generators.

During I18 beta I constructed an AE mission designed to test the limits of Fury generation. It occured to me that I could use this same mission to generate millions of tohit rolls, because I was generating several attacks per second in the combat logs; over half a million tohit rolls in a twenty four hour period.

To supplement that, I also analyzed all of my chat logs going back a couple years which are all from normal play. The normal play logs give me hundreds of thousands of tohit rolls which allow me to do some basic statistical analysis, and the AE mass combat rolls allow me to do much higher resolution analysis of the random generator itself over a short period of time.

In the Options menu, there is an option to log chat. Turn that on. Make sure you are displaying tohit rolls in a chat tab, and the tohit rolls should show up in the chat logs which will show up in \Program Files\City of Heroes\logs\game.

You should be able to see something like this in those chat logs:

As to whether the problem is that people aren't being believed, trust me when I say if I there really is a problem, I'll find it. But to be candid, the odds are not in your favor. Its almost always something else weird, or observer errors. But when a real problem existed, I've always been able to confirm it eventually. Its just that I can count all of them in and out of the tohit system on one hand.

You may have a better case there than on the concurrent week numbers. An analysis of the daily numbers suggests that not only was Scott Pilgrim's screens reduced more heavily in week two (although random spot checking movies suggests SP is not extraordinary in this regard) theater operators were also more willing to yank SP for mid-week releases. The American opened on Wednesday, September 1. That day saw a significant intra-week drop in SP screens, and no change in VS screens. Eat Pray Love and Inception both lost only marginally (but given those movies' numbers, that's entirely understandable). Piranha 3D and Nancy McPhee lost no screens (but that would have only been week 2 for those releases).

Come to think of it, if it wasn't for Vampires Suck, there would be reasonable explanations for Scott Pilgrim's screen logic compared to all other movies. Its specifically VS that seems to suggest that either theater owners disliked SP more than normal, or liked VS more than normal, or both.

The numbers don't seem to support that conclusion. Both movies maintained their screens for two weeks, and both began to unwind after that second week when numbers dropped to similar levels. Once a movie starts dropping off of screens, you can't keep comparing per-screen revenue across movies in different release weeks, because the act of contracting the screens affects the per-screen numbers. That's why Scott Pilgrim's weekend numbers actually blip upward in weekend four compared to the previous weekend, because it was now in about half the screens it was on in weekend three. That's a smaller number of people compacting into an even smaller number of screens. But the per-screen numbers are already small enough that expanding the number of screens wouldn't significantly increase box office materially.

A movie like Eat Pray Love is an even worse comparison. That movie demonstrates significantly stronger legs than Scott Pilgrim, holding its numbers above the threshold a week longer and suffering only a small per-screen drop when its numbers were cut by only a small amount - suggesting that the interest in the movie was dropping slower than most films do at that point in their release. There's no way to argue that somehow EPL is stealing screens from SP. EPL earned those screens with much stronger overall numbers. And when you factor in the fact that SP's screens had to be cut in half just to achieve those per-screen numbers, that all but proves to me EPL has the stronger legs. SP wouldn't make better use of those screens than EPL even though SP's per screen numbers last week were slightly higher. They would have just diluted those viewers.

Basically, SP's per-screen numbers for last weekend are only higher than VS's and EPL's *because* of the large reduction in screens that took place prior last weekend. My guess is that an examination of the other films would show a similar pattern.

The numbers suggest that the theater operators are collectively employing the same strategy for both movies based on the revenue numbers. I don't see significant bias between those two movies in terms of deployment strategy except for the fact that Vampires Suck and EPL opened on 10%-15% more screens initially. But the screen allocations seem to be entirely consistent with the strategies I see most everywhere else I've looked so far.

Well, when I talk about comics I'm talking about what I read and know about, which includes core superhero titles but also includes the kinds of titles I mentioned. The industry is large enough that "comics" is almost as broad a term as "music." To pick the core superhero titles of Marvel, DC, and Image as the representatives of all of western comics is rather like picking tentacle hentai as the representative of Manga.

Also, Transmet, Preacher, Enigma, Planetary, 100 Bullets, Y: The Last Man, and League of Extraordinary Gentlemen were all published by either Wildstorm or Vertigo (the first two books of LoEG anyway) which means all of them are DC titles. That was kind of a deliberate choice on my part, to pick titles that were mostly "independent" from the core superhero titles of Marvel and DC but still published by one of the two bigs (and without meaning to, I ended up with nothing but DC titles, which unfortunately doesn't surprise me). Which is why I avoided obvious non-Marvel or DC titles like Sin City (Dark Horse), or ambiguous titles like Fallen Angel (which started at DC but went indie). Even the bigs publish lots of titles worth reading (and while I don't tend to like most of the "event" stuff dominating core superhero titles at the big 2, that's not to say the core superhero stuff isn't often worth reading as well).


I think public perception is starting to change with regard to the place classic superhero comics has as the exemplar of western comics industry. They are still the big money makers, but when we have movies mining comic books like V for Vendetta, Watchmen, The Crow, Wanted, Kick ***, 300, From Hell, Constantine (ouch, but still), Hellboy, Sin City, League of Extraordinary Gentlemen (ouch again, but still), and Road to Perdition (movie based on a comic based on a manga), and of course Scott Pilgrim, people are increasingly being exposed to the other side of the street.

I'll bet 90% of the people who went to see those movies didn't know most of them were based on comic books, but that awareness is improving. And as it does, the notion that comic books are mostly represented by the classic superhero titles is beginning to shift.

Importantly, the comic book buying audience mostly knows this to be true already. By the time the comic book buying audience can drive to the comic store, I think most know classic superhero comics are only a small fraction of what is out there.

I'm not sure about that. Collectively they might account for more published issues or actual printed copies, but what really matters if you're going to judge an industry I think is number of titles. You wouldn't consider Archie and Jughead to be representative of the entire industry if they happened to single handedly print a billion of them.

In terms of actual titles, I don't know that the Marvel and DC core titles actually swamp out the Manga titles, or even necessarily all of the independent titles and Marvel and DC published non-core titles combined. I'm not even sure how you'd go about trying to judge that. Would you count completed titles that are no longer running, but still in print as a Manga (or comic) title? Would you count what shows up on a comic book shop's shelves (where Marvel and DC have an significant edge most of the time), or what you could buy from Amazon, or some combination of the two?

And to be honest, I think most people (honestly, myself included) aren't good judges of what's crappy manga. And we know that, and thus don't judge as harshly. There's a cultural distance to a lot of it (and I'm actually Japanese, but not nearly enough so) which means for many people in the United States manga either connects, or fails to connect. If it connects, its "good" and if it doesn't connect its "weird." Its rarely "bad" because its hard to judge the bad within its proper context. I know the rules and I know the culture: I can judge for myself whether a western comic book is good or complete crap. Its beyond my experience to declare a manga a complete waste of paper most of the time, because I'm never sure if, and often suspect, I'm missing something and making a culturally inappropriate judgment.

I know what I like when it comes to manga. I don't believe I'm fit to judge what's objectively good or bad except at the extremes. I'm sure there's a Joe Q in Japan somewhere screwing something up as we speak. I just can't tell what it is. I'm sure someone in Japan can, though.

You're conflating "comics" with what I call the Big 2 Continuity titles: the DCU and the Marvel Universe core titles. I think the independent titles and independently published non-manga comics industry as a whole is the equal of the Manga industry. For every great Manga title, I can point to a non-Manga title that is its equal within its cultural context. Transmetropolitan is the equal of any cultural evisceration in Manga. Invisibles is the equal of any mind-screw manga with at least a vaguely coherent story. I would put up Preacher, Enigma, Planetary, 100 Bullets, Y: The Last Man, and League of Extraordinary Gentlemen against any Manga title. And even the big boys occasionally lose their minds and give us a Sandman, or a Marvel Zombies, or a Lucifer, or a Nextwave.

Yeah, what Marvel and DC tend to do these days to their traditional cash cows tends on average to make me want to vomit. But Comics as an artform outside of Manga is doing quite well, and is in no way inferior in practice to the Manga titles that are out there. Factoring out the cash cow Marvel and DC titles, I don't even think Manga is better (or worse) on average than the rest of comicdom. Its just different.

And even more luckily, many of us lived long enough to read Planetary.

Technically speaking, that's true. But an analysis of the box office numbers suggests that SP is dropping off of screens too quickly, and probably because its per-theater averages, while fairly stable, are also lower that the theater operators want. Looking at Kick-***, as soon as the per-theater numbers dropped below about $3000, the number of screens plummeted which held the per-theater numbers stable, while dropping overall revenue proportionately. Jonah Hex had the misfortune to *open* under $3k per theater, which caused it to be dropped from screens almost immediately. Scott Pilgrim held its screens for two weeks, and Kick *** for about three weeks with only a small drop in week three.

That seems to be the trend. Coraline, a movie I mentioned earlier, seems to have held its screens for five weeks, and started to lose them when its weekly theater number dropped below $3k.

In any case, if the pattern holds true, theater operators will continue to replace SP with other fare and it'll be playing in about half the screens as it did the previous week, and even if the per-screen numbers hold or even rise a bit (which happens when screens condense) its box office will still drop proportionately, which is pretty much a predictable spiral.

The thesis of the article is that you can't judge by opening week, and that seems to be true. If all you saw was the opening weekend for Kick ***, Jonah Hex, and Scott Pilgrim, and for that matter Coraline, it would be difficult to predict their true success. Coraline had legs, and incredible word of mouth given its week two numbers. Kick *** charted kinda like the average movie in terms of week to week descent, but it was cheap to make and had decent overseas numbers. Jonah Hex, well, apparently that was predictable from week one numbers if my $3k rule holds generally true. Plus, everyone was betting on that one being a disaster. Scott Pilgrim actually seems to have slightly better legs than Kick *** did in terms of holding its numbers, but those legs are coming after the point where theater operators are unwinding its screens: at that point its basically too late to have anything more than a minor impact on your box office unless there's no competition and you actually start getting put back onto screens, which seems unlikely at this point. But Scott Pilgrim wasn't predictable from week one. The critical week seems to have been week two. Had word of mouth brought people into the theaters to see SP in week two and held its numbers at or near $3k, it would have, I believe, bought SP an extra week in wider release. To do that, though, it would have had to do what Coraline did: have a week two almost as good as, or better than, week one. It doesn't happen often that word of mouth can completely reverse a movie's fortunes, but it happens. **

Given that, plus SP's much higher costs (at least double Kick ***'s, depending on whose numbers you use), its going to be much harder for SP to turn a profit. Its DVD and rental numbers would have to be very high to compensate for the box office gap. That's certainly possible, so I wouldn't have been nearly as conclusive as the article was. But I think the box office numbers are basically a done deal at this point. Overseas could still be a big surprise there, but I think that's a long shot at this point.


** The $3k rule actually seems to work better than I thought it would. Adjusted for inflation, that would be about $2350 in 1999. Looking at the Matrix numbers above, it started getting pulled from screens when it dropped below $2200 per screen. Fascinating. I wonder how well the rule works across a larger data set.

And that's fine. And if you say you're forced to use a power whose appearance you don't like, or that the powerset is poorly designed to force you to do so, I'm still going to state that is false regardless.

You originally mentioned textures, but I'm going to assume this is theoretically open to all kinds of updates on costume pieces. In that case, I think which option I favor depends on the change.

Changes for which I prefer Option 1:

1. Geometry fixes specifically to address odd clipping or alignment issues. If its genuinely crap, even if some players like that crap I think its justified to replace poor geometry with properly functioning geometry.

2. Higher resolution versions of lower resolution textures that contain comparable features. Self-explanatory.

3. Costume part changes which create supersets of other parts (i.e. in terms of having more texture or coloring options, but can otherwise replicate the original part virtually completely with the right settings). Also self-explanatory. And probably should be the explicit goal whenever its actually possible.


Changes for which I prefer Option 2:

1. All other geometrical changes. If it changes geometry at all, I think you have to allow players to keep the old geometry even if the newer geometry is "better."

2. Textures which add significant visual details that did not exist on the original texture at all. This is a judgment call, but a higher res version of a lower res texture is different to me than a higher res texture that adds significant detailing with that higher res texture that wasn't even implied in the lower res one, even if that was strictly due to resolution limits.

3. Any part which is being revised for the players but for which an identical part is being preserved for the NPCs. If players can see it in-game, its going to only cause complaints if its removed from players, even if the replacement is "better." If you're going to remove it, it has to be gone from everywhere.

I never said anything about what people can choose to call ugly or not. What I specifically said was:

And I stand by that statement. Combining these two things is just going to be fruitless. Most recent example: complaining that buffing CAK in MA "forces" people to take an ugly power. You don't have to convince me its an ugly power. I've been saying that more times and for a lot longer than practically anyone else left standing. But that has no bearing whatsoever on the best way to numerically adjust the performance of the powerset and played no role at all in any of my recommendations. Altering the appearance of CAK is a completely independent issue. But the notion that CAK shouldn't be buffed because it might "force" people to take an ugly power is an argument I find ludicrous on its face, and gave it no weight. It gets buffed because it deserves to get buffed, period. Don't like it, don't use it. Want it to look different, take it up with BaB.

Similarly, buffing Siphon Life doesn't alter the character of the set for those that want to focus on speedier attacks. Those attacks are still there. No one forces people to min/max attack chains. If the min/max chain "forces" you to use a power that takes more than a second to execute, thems the breaks. Under no circumstances would I, for even one second, consider that a power like SL shouldn't have its damage buffed because it might suddenly become advantageous to use. If it should be buffed, it gets buffed. If it shouldn't, it doesn't. But how that buff would affect attack chain min/maxers frankly didn't, doesn't, and wouldn't factor into my thinking.

All evidence suggests it doesn't for the powers team either. It barely enters their minds when it comes to balancing nerfs, much less balancing buffs. But, I'm not the thought-police. I don't tell people how to spend their time.

Alternate animations won't make Siphon Life execute faster, and the complaint seemed focused on the fact the attack is slower than the tier1/2 attacks, not that it visually looked odd. The complaint about SL was compared to the relatively long animation time of Midnight Grasp.

Not a problem.

Do we now. I was specifically responding to this:

If you want a fast-acting attack chain, it still has one. If you want the optimal chain, there's that also, which is not terribly far off from the fast one. But if you want to use the optimal chain, you have to use the powers in the optimal chain. Complaining that one of the powers isn't to your conceptual liking is one thing, just don't use it. But claiming that all of the powers in the optimal chain should match your personal preferences is now connecting min/maxing to power choice options. This game does not, and as far as I can tell will never honor such a request. So you link the two at your own peril.


When it comes to balancing philosophy, being called silly is the least I've been called. It really no longer bothers me. In fact, if a balancing rule or perspective isn't trivialized or ridiculed at least a couple of times, I start to wonder if its completely wrong.

24 times higher than the numbers quoted in the thread (x4 for the resistance floor, and x6 for the maximum combat modifier difference of -49 between player and critter).

I don't think your impression of how movies get made is realistic.

Also, whether the SL chain is "clunky" is more a matter of taste than objective fact. So its really more like both paint jobs are fine, and Bill bought the car with the slightly higher horsepower, then began complaining that he hates that particular color.

That is exactly what I'm stating. If you have a choice between two cars, one of which has a great paint job and 300 horsepower, and another car with a crappy paint job and 308 horsepower, you can complain all you want about the lousy paint job of the latter car after you buy it, but who are you going to complain to? The dealer that you bought the car from, after *knowingly* buying the car with the lousy paint job when another option was available?

Don't try complaining to me, 'cause I'll be half a mile away driving off in the other car.

I'm not commenting on whether the movie might "succeed" for a single viewer. If it succeeds for you, great. I'm talking about the fact that the movie failed to reach the audience it intended to reach, and there are consequences for failing at that level that will make it harder for the next guy.

And while I can't blame the director for what the marketing people do, I can still blame the director (especially when he also worked on the screenplay) for very obviously failing to synchronize scope with reach. He made a $90m movie that needed to reach triple the audience just to break even, before a single marketing dollar was spent. If you make a $90m movie that you actually *know* only a subset of a subset of the gamer-geek population will go see, you're either the best Hollywood con man who managed to steal a bunch of money and donate it to the Scott Pilgrim fan club, or you're an idiot.

Since I don't think Edgar Wright is an idiot or Robin Hood, I think its more likely that he simply failed to consider that his movie was unrelatable, and that this was more than a trivial defect in an otherwise niche-genre movie.

I'm not telling you what to like, or how to choose how to spend your entertainment dollar. And as I said, I don't think Scott Pilgrim is a bad movie as such - I liked it for what it was. But I can still take a step back and say it failed in at least part of what it tried to do, which was project Scott Pilgrim to a wider audience.

And there has to be a balance between artistic freedom and commercial success. Who is supposed to subsidize your personal taste in movies otherwise? Hollywood is not going to spend a hundred million dollars on Scott Pilgrim movies that only make back forty as a form of community service.

The devs usually don't buy the argument that adding an option invalidates a previously good one. If DM has a nice attack chain that doesn't include SL, buffing SL so that it just so happens a chain that includes SL at a particular (and high) level of recharge edges out the chain with SP doesn't mean DM no longer has a nice attack chain without SL. That attack chain still exists. If a player chooses not to use it for strictly min/max reasons, they lose the right to complain about the conceptual clunkiness of it.

In other words, never say you do something strictly for numerical reasons and then complain about its appearance. Of all the people that believe this is valid criticism, none of them appear to be CoX game designers.