I take issue with randomness of pulls, please prove me wrong

Unknown

simonsez wrote:

rossmon wrote:

What I find highly improbable is that the devs spent the time and money to figure a way to code it so that you are more likely to get what you don't want, and not get what you want.

I agree, that'd be silly. But if they've skewed the odds table to favor older covers, indirectly, they've accomplished this.

I think in terms of cover-pulls-over-time that'd be true, but mainly due to NOT skewing the RNG in favor of newer covers.

I wager there's been more Nick Furys drawn from Legendarys then X-23s for the same reason there's probably more Silver Surfers out there then Phoenixes.

But that's obvious.

simonsez

scottee wrote:

That line was referring to online poker.

I guess I made a mistake in assuming you intended that to be relevant to this discussion.

scottee wrote:

But essentially you're saying that a few people being extra lucky incentivizes spending, and lots of people being extra unlucky incentivizes spending.

No, that's not what I'm saying. I'm saying that when covers get spit out in bunches the way they do, one byproduct of that is that it incentives spending. I gave you two very specific examples. Turning my specifics into a misstated generic doesn't offer anything constructive.

And stop falling back on "confirmation bias"... I'm not stating opinions based on my memory of what I pulled. I'm giving probability estimates of what people (and devs) say they pull. Yeah, I know, you think everyone (and devs) are giving fake results. Because of something.

rossmon

I'm going to create a lot of hate for this, (and I may be killing myself with murphy's law) but Invisible woman and MrF are my least covered 4*s with 3, followed by starlord thor, Rhulk, profx and x-23 at 4. On the flip side I have 9 on antman, and 7 on Iceman.

I've also never bought a 4* cover.

This is not a complaint, but just pointing out.

scottee

The Ice post referenced earlier in this thread says they've looked at the numbers from multiple angles and they're all within normal statistical ranges. I'm not saying they're lying. I'm saying they're telling the truth and people have a hard time believing them.

They have data on hundreds of thousands of pulls. And somehow posters think their experience with a couple hundred pulls is more valid.

If you think the multiple psychological effects I mentioned aren't real and aren't affecting these posts, then I'm sorry. Nothing is going to satisfy you.

TxMoose

random will naturally have some spikes and valleys over 20 characters. I have both -

spikes = rulk at 1/4/1 (only spike I have left - there were others but they've evened out over the last few weeks)
valleys = carnage at 4/-/3, px at 5/1/4, xfw at 4/5/1, IW at 4/5/1

some of those are driven by some colors never coming up at 1K in pvp since I've been hitting it - see carnage black, xf yellow.

I've also had runs on characters - starlord, hb, jean, MFer, IW. it still feels random to me, even during the anomalies/runs/droughts

simonsez

scottee wrote:

The Ice post referenced earlier in this thread says they've looked at the numbers from multiple angles and they're all within normal statistical ranges.

Ice isn't a statistician. I am. Anyone who understands probability would not be telling us that token pulls are random, and the token pulls reported by "the player who shall not be named" were accurate.

I'm not saying he's lying either. I'm saying he's a software developer, not a statistician.

scottee

simonsez wrote:

scottee wrote:

The Ice post referenced earlier in this thread says they've looked at the numbers from multiple angles and they're all within normal statistical ranges.

Ice isn't a statistician. I am. Anyone who understands probability would not be telling us that token pulls are random, and the token pulls reported by "the player who shall not be named" were accurate.

I'm not saying he's lying either. I'm saying he's a software developer, not a statistician.

You're really a statistician? And you're still arguing that people's experiences are more reliable than someone looking at the hard data? You're going to look at self-reported results on an internet forum and tell a software developer that their code is bugged? You don't care that the results here are all self-reported? That posts on an internet forum are about the most biased type of survey you could possibly have?

You should have a look at all of the internet poker data over multiple years. Top players went through bad runs and good runs that lasted tens of thousands of hands. And top mathematicians all thought it was completely normal. They must all be wrong.

Omega Blacc

I don't believe for a second token pulls are random.

I don't want to call the devs shady or anything. I respect what they do. I wouldn't, however, put it past the game to be designed to open or close the gates on rewards depending on roster size, time played, and money spent.

I've pulled too many characters that were soon to be essential to be random. Not EARNED...pulled. This has been ever since I've started this game.

That's to my benefit, but it's not random.

There should be equal probability between all colors for all characters depending on the pool you draw from. The more characters you add, the lower the odds to pull what you need. The more slots you have, the lower the odds to pull what you need. I can accept that.

Just don't tell me "what you pull is random" is the same as having equal probability to earn whatever is available in the pool. This is why I don't trust the Vault, and this is why people hoard. I bet you anything the overall draw rate on desired toons in special packs doesn't correlate to what's listed on the token page.

And I'm sorry...you shouldn't need thousands of pulls to see how the probabilities start to pan out. You can see that much earlier.

Vinmarc43

A good evolution to this problem would be that after you have for example, 2 or 3 covers of each colors, the covers that you pull afterwards with your lucky tokens would be NEUTRAL, so this way you place them as you please and respec those colors as you win those neutral covers later.

Pylgrim

Just adding my voice here to remind people, as I always do, that the actual issue at hand is not whether there's just pure chance or coded odds. It's not about chance feeling unfair by design or by human nature. It's about it actually being unfair at this level of play. If a player can max their 5* with a few tokens, while another (me, sorry, I'm not trying to make it all about me, it's just that my case is the worst I know so far) misses an expected 80+% rate of success 16 out of 21 times, it IS an incredibly unfair and crippling difference in progression speed. Obviously not every person will be that lucky or this unlucky, but my point is that at this level of reward, you should get only people that feel good (opening useful, game-advancing stuff) and people that feel great (opening really rare and expensive stuff); at this level of reward, no one should get to feel punched in the face and spat on, repeatedly. For this, the role chance has in the prize distribution needs to be removed, or its impact diminished.


"But Pylgrim", I already can hear the inevitable disgruntled 2*-3* transitioner that for some reason is participating in an argument that doesn't concern them (in the short term), "stop being so elitist, why you want better odds of drawing what you need, while the people beneath you have to still be reliant on luck? If random tokens are good enough for us, then you shouldn't expect better. Classic rich wanting to get richer!"

Well, my dear, nameless 3* transitioner to whom I'm answering pre-emptively: a "bad luck run" of, say, 20 heroic (or similar) tokens, happens over 5-7 days, depending on your level of commitment/capacity of your roster. That means that you only need other 5-7 days of good luck to completely balance it. The sheer number of those tokens that a player can get in one month means that the statistical expected balance is much more likely to be achieved. On the other hand, a "top player" can get only about 20 LTs in one month, by spending lots of effort and game resources. A 20 tokens-long "bad luck run" will meant a month of effort wasted and will require another whole month of similar effort and a string of good luck to just return to fairness. Now, in the case of an early 4* transitioner as I, who are the ones for whom these tokens are supposedly created, it can take 2-3 months to get 20 LTs. In the same time, you'll have opened literally hundreds of Heroic-like tokens.

scottee

I think part of the problem is that someone need tokens as their main source of progression in order to play casually, while tokens and DDQ have always been supplemental to progression, not the main source. So it's definitely highly inconsistent for your own progress in the game to come that way.

The interested result of Legendary Tokens is that people seem to be going after 1st place LESS than before, because instead of targeting a specific 4* cover. It's been a lot easier for me to target 1st place covers since LT's arrived for whatever reason.

PVP 1000 and 1st place should still be the primary way people progress, not LT's.

Zen808

scottee wrote:

I think part of the problem is that someone need tokens as their main source of progression in order to play casually, while tokens and DDQ have always been supplemental to progression, not the main source. So it's definitely highly inconsistent for your own progress in the game to come that way.

This is my problem with Leg. Tokens. With normal tokens, whenever RNGeesus screws you six-ways to Holy Sunday, I always think, "well how hard was it, really, to get that token?" The answer, "not very hard at all." Results = effort, move along.

But with Leg. Tokens, you kinda have to bust tinykitty to get one. And that's what makes each and every IW pull seem like such a slap in the face.

Now that we have CP implemented, can't we at least make the Top PVE Progression the Essential 4*? There are still plenty ways to earn Leg. Tokens now for those that gotta have them 5*'s. Let the rest of us have a defined path to building our 4*'s.

Pylgrim

scottee wrote:

PVP 1000 and 1st place should still be the primary way people progress, not LT's.

Ideally, yes, at least on the 1k reward. I don't know what scores do you have to reach for top 1, but it's usually 1500 for me. Taking in account I cannot even reach 1300 reliably (or without spending $7 worth of shields per event) that's an unrealistic goal for me, and people in my position. So, the 1k reward. That's a certain cover every 2.5 days, except every 10 or so days, when a cover that I already have maxed is rewarded, in average. So let's say, 8 covers in a season. I just counted, I'm missing 103 4* covers, many of which I have 0. If they stopped releasing 4*s as of Totally Awesome Hulk (hint, they won't), in order to get to 5 copies of one cover—which would be coming back every 69 PVPs—it would take almost two years of rotation. If they don't stop releasing 4*s, in that time 144 new covers would be added to the game, which means not only 144 more covers to get but also a greater dilution of the rate at which old covers come back around. 1 person among every 1000 (you, for example), gets to halve those numbers. Yoohoo.

So no, the "primary way" for 4* progression you speak of is nowhere near sufficient. The fact that the supplemental way (LTs) is based on pure chance means it's not much of a help.

scottee

Pylgrim wrote:

scottee wrote:

PVP 1000 and 1st place should still be the primary way people progress, not LT's.

Ideally, yes, at least on the 1k reward. I don't know what scores do you have to reach for top 1, but it's usually 1500 for me. Taking in account I cannot even reach 1300 reliably (or without spending $7 worth of shields per event) that's an unrealistic goal for me, and people in my position. So, the 1k reward. That's a certain cover every 2.5 days, except every 10 or so days, when a cover that I already have maxed is rewarded, in average. So let's say, 8 covers in a season. I just counted, I'm missing 103 4* covers, many of which I have 0. If they stopped releasing 4*s as of Totally Awesome Hulk (hint, they won't), in order to get to 5 copies of one cover—which would be coming back every 69 PVPs—it would take almost two years of rotation. If they don't stop releasing 4*s, in that time 144 new covers would be added to the game, which means not only 144 more covers to get but also a greater dilution of the rate at which old covers come back around. 1 person among every 1000 (you, for example), gets to halve those numbers. Yoohoo.

So no, the "primary way" for 4* progression you speak of is nowhere near sufficient. The fact that the supplemental way (LTs) is based on pure chance means it's not much of a help.

Oh, I'm not saying it's sufficient. I'm just saying people should see it as the primary way. It's the way you can determine what cover you get upon achievement. The 1000 progression cover and PVE Legendary Token are the two ways that a 3* roster can start the 4* transition. That eventually makes it possible to do some of the DDQ's. That's the current best path available, even if it's not as fast as most would want.

Pylgrim

scottee wrote:

Pylgrim wrote:

scottee wrote:

PVP 1000 and 1st place should still be the primary way people progress, not LT's.

Ideally, yes, at least on the 1k reward. I don't know what scores do you have to reach for top 1, but it's usually 1500 for me. Taking in account I cannot even reach 1300 reliably (or without spending $7 worth of shields per event) that's an unrealistic goal for me, and people in my position. So, the 1k reward. That's a certain cover every 2.5 days, except every 10 or so days, when a cover that I already have maxed is rewarded, in average. So let's say, 8 covers in a season. I just counted, I'm missing 103 4* covers, many of which I have 0. If they stopped releasing 4*s as of Totally Awesome Hulk (hint, they won't), in order to get to 5 copies of one cover—which would be coming back every 69 PVPs—it would take almost two years of rotation. If they don't stop releasing 4*s, in that time 144 new covers would be added to the game, which means not only 144 more covers to get but also a greater dilution of the rate at which old covers come back around. 1 person among every 1000 (you, for example), gets to halve those numbers. Yoohoo.

So no, the "primary way" for 4* progression you speak of is nowhere near sufficient. The fact that the supplemental way (LTs) is based on pure chance means it's not much of a help.

Oh, I'm not saying it's sufficient. I'm just saying people should see it as the primary way. It's the way you can determine what cover you get upon achievement. The 1000 progression cover and PVE Legendary Token are the two ways that a 3* roster can start the 4* transition. That eventually makes it possible to do some of the DDQ's. That's the current best path available, even if it's not as fast as most would want.

Time for a career in politics. "As fast as most would want" is an understatement of governmental PR magnitude that subtly places the blame on people for wanting too much. As I outlined in my previous post, it would take /me/ around 2 years to finish progression if they stopped releasing new 4*s now, which won't happen. And I already have over a hundred 4* covers. For the person in your post, just starting the transition, it would take much longer. The fact that "eventually" they may do some 4* DDQs, just to earn a random drawn is not a great encouragement, is it? And the fact that—to go back to the crux of the argument—for people not blessed with great luck, that rare, hard-to-acquire random draw will spit junk on their faces every so often, makes that gruelling progress to an ever-increasingly unreachable goal more frustrating and pointless.

scottee

...To complete every single 4*. You don't need to complete every single 4* to make the 4* transition. You really only need two right now. I'm not sure why everyone keeps counting the max number of covers. You only need the key characters covered. This is actually why targeting specific characters you know you need is important, and why dilution is much bigger issue in the rotation of prizes than it is in token odds. Because progression through tokens was always supplemental and progression through targeting rewards was always and still is the much more efficient way.

Pylgrim

scottee wrote:

...To complete every single 4*. You don't need to complete every single 4* to make the 4* transition. You really only need two right now. I'm not sure why everyone keeps counting the max number of covers. You only need the key characters covered. This is actually why targeting specific characters you know you need is important, and why dilution is much bigger issue in the rotation of prizes than it is in token odds. Because progression through tokens was always supplemental and progression through targeting rewards was always and still is the much more efficient way.

Fair enough, I am a completionist and you are not. And yet, the problem remains the same. I got my second red Hulkbuster cover around 2 months ago if memory serves well. That means there will be around 3-4 months before he's rewarded again. At that rate (not counting dilution caused by newer covers), I'll get the 5th red 1 year after that. Theoretically, I can get around 80+ LTs in the meantime but with chances of 1 in 63 (again without dilution), I won't get my hopes up. So I could be maxing my HB around 3 years and a half after I started playing the game... but by then, the meta would have moved into the 5*s and HB may not even be the best 4* anymore. So things don't look that much better for someone focusing only in 2 or so characters over a completionist.

evildead666

Every LT I have pulled has been a 1 of so I kinda feel your pain

XandorXerxes

scottee wrote:

...To complete every single 4*. You don't need to complete every single 4* to make the 4* transition. You really only need two right now. I'm not sure why everyone keeps counting the max number of covers. You only need the key characters covered. This is actually why targeting specific characters you know you need is important, and why dilution is much bigger issue in the rotation of prizes than it is in token odds. Because progression through tokens was always supplemental and progression through targeting rewards was always and still is the much more efficient way.

Two major flaws in your arguments:

1) If the initial cover being random isn't the problem, but the subsequently generated covers that aren't sufficiently random from the first cover because the RNG isn't properly randomized, the large number spread of covers won't reflect it. For every person that pulled 10 2*s, there's another person that pulled 7 golds. Both of which should be (relatively) rare, but even if those were common occurrences it would look statistically normal over a large sample size.

2) It doesn't matter if you don't need to complete every 4*. You can't complete any 4* because they rotate. It takes 2 years to fully rotate through, which means it takes just under 2 years to get the first initial 4* completed - assuming the math is right and assuming no other 4*s are released.

I don't gamble because I have really bad luck. Do I need to stop playing this game? Because finishing 4* and / or starting 5* in any reasonable amount of time seems to be really reliant on it.

OneLastGambit

I think the main problem with the argument that people have on this issue is the stubborn refus to review all the data. They see one person post (unconfirmed with evidence - ICES word is not evidence) an unlikely (not impossible) result and that's the only thing they pay attention to because it happens to be the only thing supporting their argument, while the vast majority of the evidence (the 1000s of pulls which DON'T provide unlikely results) is continually ignored as it doesn't support their hyptothesis.

They simply say "ice confirmed this guy's results as ridiculously long so it must be true " ice also confirmed that the stats support the RNG is working perfectly fine so even using the shaky evidence of anecdeotal reporting from ice you're still only paying attention to the part that supports your argument and ignoring that which doesn't