And the mask comes off.

[Harpospoke]

All this talk of AE farms just further reinforces my feelings of, "I didn't miss a damn thing by not touching the farms."  I generally avoided AE like the plague, but if there were worthwhile stories to experience I ran them once per character.  Never did see the draw to AE farms...was it really that damn hard to level up/earn inf/whatever that normal gameplay didn't offer?  Goodness...

Didn't bother me.   I didn't go there much after making a couple of missions to make myself laugh.   I always meant to try more of the missions people listed on the forums, but rarely got around to it.

I figure, the people in the AE building weren't complaining about the way I was playing, it was only right that I return the favor.

[Codewalker]

Its as bad as swtor having a 20% reverse engineering attempt fail 140 times in a row. for multiple people. enough that people started keeping track of it.

"X number of times in a row" doesn't matter to RNGs that are not built with a failsafe (like City's tohit failure streak breaker). Every single attempt is completely separate and not connected in any way, shape, or fashion to any other attempt. A 20% success rate can definitely fail 140 times in a row...it can technically fail 10000000 times in a row for a really, spectacularly unlucky person.

There's even a formula for it. If my math is right, starting fresh and reverse engineering 140 items in a row at a 20% success rate, the chance of all 140 failing is 0.000000000002707685%

Being generous and saying that an average character might reverse engineer 5,000 items within a certain time frame, the chance of seeing a 140 failure streak somewhere in that run of 5,000 is 0.00000000263458%

That's a very small number, so if datamining shows that significantly more than roughly 2.6 out of every 1,000,000,000 players has seen a 140 failure streak, it's a good indication to game designers (at least the ones who know math) that their RNG is broken.

[hurple]

I played from I7 to close, had 34 50's. very very rarely did AE and never got a respec drop.

I had one toon get three in one mission. 

[darkgob]

There's even a formula for it. If my math is right, starting fresh and reverse engineering 140 items in a row at a 20% success rate, the chance of all 140 failing is 0.000000000002707685%

Being generous and saying that an average character might reverse engineer 5,000 items within a certain time frame, the chance of seeing a 140 failure streak somewhere in that run of 5,000 is 0.00000000263458%

That's a very small number, so if datamining shows that significantly more than roughly 2.6 out of every 1,000,000,000 players has seen a 140 failure streak, it's a good indication to game designers (at least the ones who know math) that their RNG is broken.

Did you just use actual math to make a point as opposed to anecdotal "evidence"?  That should be considered a bannable offense IMHO.

[Relitner]

There's even a formula for it. If my math is right, starting fresh and reverse engineering 140 items in a row at a 20% success rate, the chance of all 140 failing is 0.000000000002707685%

Being generous and saying that an average character might reverse engineer 5,000 items within a certain time frame, the chance of seeing a 140 failure streak somewhere in that run of 5,000 is 0.00000000263458%

That's a very small number, so if datamining shows that significantly more than roughly 2.6 out of every 1,000,000,000 players has seen a 140 failure streak, it's a good indication to game designers (at least the ones who know math) that their RNG is broken.

The maths... they burn the eyes...

[Codewalker]

You want real eye burning... I only used the formula for the approximation of the upper bound:



Here's the formula you'd need to use to get the exact probability:



N = Total number of samples
K = Number of failures in a row
p = Probability (0-1) of desired result, in this case failure, which is 0.8

[srmalloy]

I'll say this about the city of heroes auction house.  It was a far more stable platform than what i've seen in DCUO or CO.  The prices if they were high and did start dropping, usually only dropped for legitimate reasons and it was usually gradual.  CO?  Well, often what happens is because everyone can see the prices of the items, you get what I love to call an "Underbid war", in which, due to the rather short timer of putting things up for sale(your auction would expire after only 15 days), some jerks would eventually see your prices, and put the same items up for a lower price and prevent you from selling at all.

And this is something that continues to drive me up the wall with SWTOR's Galactic Trade Network, which is not an auction house; it's a consignment house. You list an item. getting a 'default' price that you can change, and pay a listing fee based on a percentage of the default price. The item can be listed for up to two days; if it sells, you get your listing fee back and the sale price minus a service charge. If it doesn't sell, you get the item back and your listing fee is refunded. This last, combined with the fact that the 'cartel market' items (i.e., anything you have to spend cartel coins to buy, directly or through the cartel packs) has an absurdly-low default value, means that there's no drawback in listing something for an unreasonable price; you're out nothing if it doesn't sell. The rare, highly-desirable gear and mounts can be listed for tens of millions of credits, and if no one is interested, you're only out the time it takes to relist it -- unless there are other players undercutting your price, you have no incentive to lower the price until you find what people are willing to pay.

[srmalloy]

I wonder, when the game returns, who will be the first to get to 50?  You know some people are going to race to 50 as fast as they can.

Not me; the mindset behind having to get to level cap as fast as possible escapes me -- we all know where we're going, so what's the point in rushing to get there as fast as you can when you can slow down a little and enjoy the trip, instead of being fixated on the destination to the exclusion of all else?

[srmalloy]

The supply of influence being so insane was the core reason I was highly opposed to AE farming, because I knew that it tended to send a very disproportionate amount of influence into the market.  Power leveling from it was one thing as one could stop at 50 but all the 50s doing it see were gaining hundreds of millions of influence in far far less time than anywhere else.

Now, this was something that never made sense to me about AE missions. Inf -- influence, infamy, or information -- was abstracted as your ability to get other people to do things for you because of your reputation. Out of respect for your service to the city, for influence; out of fear of what you might do to them, for infamy; in response to the leverage it could be used for, for information. But all of these represent how the other people viewed your character. AE missions happen entirely in a virtual world; none of the events that occur in an AE mission affect the 'real world'. Completing a mission inside AE gives you practice in the use of your powers, and would thereby gain you experience. But if a new hero went into the AE building at level 1 and did nothing but AE missions, when they came out to get enhancements, nobody would know who he was; as far as they were concerned, he'd never done anything to protect Paragon City -- in the 'real world', he wouldn't have any influence, because he hadn't done anything in the 'real world'. It doesn't matter how many times you crawled into your electronic navel and soloed Emperor Cole, or Reichsman, or U'kon G'rai; none of that was real. Tickets, sure; that's the AE system's internal reward system to encourage you to participate. But gaining any sort of reputation from defeating virtual ghosts? Not a chance. Level from 1 to 50 inside AE, and you come out and try to play on your reputation, and people are going to ask who you are; they would never have heard of you, except perhaps in a news story about the victims of AE addiction.

AE missions should give you XP and tickets; if you want to earn inf, you should have to go out in the 'real world' and do things that other people can see you doing to build your reputation.

[r00tb0ySlim]

Now, this was something that never made sense to me about AE missions. Inf -- influence, infamy, or information -- was abstracted as your ability to get other people to do things for you because of your reputation. Out of respect for your service to the city, for influence; out of fear of what you might do to them, for infamy; in response to the leverage it could be used for, for information. But all of these represent how the other people viewed your character. AE missions happen entirely in a virtual world; none of the events that occur in an AE mission affect the 'real world'. Completing a mission inside AE gives you practice in the use of your powers, and would thereby gain you experience. But if a new hero went into the AE building at level 1 and did nothing but AE missions, when they came out to get enhancements, nobody would know who he was; as far as they were concerned, he'd never done anything to protect Paragon City -- in the 'real world', he wouldn't have any influence, because he hadn't done anything in the 'real world'. It doesn't matter how many times you crawled into your electronic navel and soloed Emperor Cole, or Reichsman, or U'kon G'rai; none of that was real. Tickets, sure; that's the AE system's internal reward system to encourage you to participate. But gaining any sort of reputation from defeating virtual ghosts? Not a chance. Level from 1 to 50 inside AE, and you come out and try to play on your reputation, and people are going to ask who you are; they would never have heard of you, except perhaps in a news story about the victims of AE addiction.

AE missions should give you XP and tickets; if you want to earn inf, you should have to go out in the 'real world' and do things that other people can see you doing to build your reputation.

^^^This...makes sense to me.

[Samuraiko]

It's funny - some people saw the AE as "OMG, new content."

Others as "Hellooooooo, farm central."

Me? "BOOYAH, CUSTOMIZABLE FILMING ENVIRONMENT!"

When the game goes live again, I'll need to finish my 10-mission, 2-part story arc "The War Witch Task Force."

Thank heavens I still have all my AE files, so I won't have to recreate all my previous work.

Michelle
aka
Samuraiko/Dark_Respite

[Arcana]

Umm... are you thinking about something else? You couldn't buy purple recipes with reward merits, no matter how many you had.

There was also no way to convert between AE tickets and reward merits, or any way to earn merits from AE missions as far as I know. The merit system also predates AE.. they were introduced in Issue 13 and 14, respectively.

You could buy purples for 20 Hero/Villain merits, available only by doing morality missions (2 day minimum to get one), conversion from 50 reward merits (1 day cooldown), and the SSAs once they were added (1 week cooldown for each).

You could also get them for Empyrean merits by doing iTrials, which IIRC also had an internal cooldown on how often they could be earned.

What did make a huge difference on prices across the board was enhancement converters. In general rare stuff like PVP IOs were still expensive, but less so as it tended to average out with the cheaper ones that nobody wanted + the cost of converters.

I must be conflating AE merits and the merit system, or possibly having an episode of dementia.  And why are you all in my house?

[OzonePrime]

lol!

[OzonePrime]

You want real eye burning... I only used the formula for the approximation of the upper bound:



Here's the formula you'd need to use to get the exact probability:



N = Total number of samples
K = Number of failures in a row
p = Probability (0-1) of desired result, in this case failure, which is 0.8

Ah, memories. Stats. What fun!

[Arcana]

There's even a formula for it. If my math is right, starting fresh and reverse engineering 140 items in a row at a 20% success rate, the chance of all 140 failing is 0.000000000002707685%

Being generous and saying that an average character might reverse engineer 5,000 items within a certain time frame, the chance of seeing a 140 failure streak somewhere in that run of 5,000 is 0.00000000263458%

That's a very small number, so if datamining shows that significantly more than roughly 2.6 out of every 1,000,000,000 players has seen a 140 failure streak, it's a good indication to game designers (at least the ones who know math) that their RNG is broken.

Math I can still do (probably).  To put it in easier to understand terms, the odds of a 140 failure in a row run of something with an independent chance to fail of 80% is 36,931,914,471,142 to 1 - 37 trillion to 1.  The odds of a single player seeing such a run after 5000 tries is about one in 38 billion.  Incidentally, the recursive exact computation for the second calculation is *just* within the bounds of Wolfram Alpha, unless you're a pro subscriber.

And yes, I don't think there's any way the MMO in question has sufficient subscribers to make this observation credible.  However, there are three things that can cause this and bad RND is only one of them.  Misuse of the RND in the code is the second, and improper observation is the third.

And now that I've double checked my notes, I did conflate AE tickets with alignment merits.  That's another thing that's fallen off the edge of FIFO memory.

[r00tb0ySlim]

Math I can still do (probably).  To put it in easier to understand terms, the odds of a 140 failure in a row run of something with an independent chance to fail of 80% is 36,931,914,471,142 to 1 - 37 trillion to 1.  The odds of a single player seeing such a run after 5000 tries is about one in 38 billion.  Incidentally, the recursive exact computation for the second calculation is *just* within the bounds of Wolfram Alpha, unless you're a pro subscriber.

And yes, I don't think there's any way the MMO in question has sufficient subscribers to make this observation credible.  However, there are three things that can cause this and bad RND is only one of them.  Misuse of the RND in the code is the second, and improper observation is the third.

And now that I've double checked my notes, I did conflate AE tickets with alignment merits.  That's another thing that's fallen off the edge of FIFO memory.

[Minotaur]

Math I can still do (probably).  To put it in easier to understand terms, the odds of a 140 failure in a row run of something with an independent chance to fail of 80% is 36,931,914,471,142 to 1 - 37 trillion to 1.  The odds of a single player seeing such a run after 5000 tries is about one in 38 billion.  Incidentally, the recursive exact computation for the second calculation is *just* within the bounds of Wolfram Alpha, unless you're a pro subscriber.

And yes, I don't think there's any way the MMO in question has sufficient subscribers to make this observation credible.  However, there are three things that can cause this and bad RND is only one of them.  Misuse of the RND in the code is the second, and improper observation is the third.

And now that I've double checked my notes, I did conflate AE tickets with alignment merits.  That's another thing that's fallen off the edge of FIFO memory.

Even in CoH there were lots of stupidly rare events going on, my favourite was one of the veteran attacks with a 75% to-hit chance and no modifiers either way (I had the last to-hit displayed at the time so I know this) missing 24 times in a row in among other attacks that were hitting normally so no streakbreaker, then later in the same day it missing 21 times in a row. I suspect some circumstances cause the RNG seed not to regenerate properly and it not actually be random.

[Vee]

I actually thought I had just missed some way to get purples and merits from AE given that Arcanaposts tend to spout metaphysical certainties.

[LaughingAlex]

Even in CoH there were lots of stupidly rare events going on, my favourite was one of the veteran attacks with a 75% to-hit chance and no modifiers either way (I had the last to-hit displayed at the time so I know this) missing 24 times in a row in among other attacks that were hitting normally so no streakbreaker, then later in the same day it missing 21 times in a row. I suspect some circumstances cause the RNG seed not to regenerate properly and it not actually be random.

I could tell the RNG seed wasn't always wholly random, as I'd frequently see the same attack in an attack chain miss(and often one that does high damage in a single hit).  I also had a peculiar moment when a mob rolled near identical numbers 3 times in a row and hit me three times in a row when he only had a 33% chance I think to hit me.  Now I know I didn't have phenomenal defenses in that fight but, seeing the numbers role so close together as like 26, 28, 29 wouldn't be expected.

I also always, always, always tried to get my end to-hit up to 95% and over that, because of how the rng often worked.  Personally I have never trusted random number generators since that moment and I still find RNG bullcrap to this day in other games.  I'll never forget the numerous "lose five in a row in blackjack in new vegas with TEN LUCK when the win odds are supposed to be a staggering 80% win ratio with TEN LUCK" events, total random number god moment there.  OR 80% chance to hit enemy in vats, fires 6 slugs, all of them narrowly miss.  On hardcore, against a fricken super nightstalker.  Or 6 of my ten precious 10mm rounds in the early game missing that convict leader narrowly in vats when I seriously needed them to hit and had also over 80% chance.

Random number generator my #$%.

[JanessaVR]

AE missions should give you XP and tickets; if you want to earn inf, you should have to go out in the 'real world' and do things that other people can see you doing to build your reputation.

This makes a huge amount of sense to me.  Perhaps the CoT or APR people will take note of it for their efforts.