I signed up for a game, not a lottery

The Viceroy Returns

My older brother is at something ridiculous like 0 / 40.

There's a simple fix for this problem = A guaranteed 5 star after X covers opened in a row without pulling a 5 Star cover.

There's another Marvel game out there I play called Marvel: Avengers Alliance.
They have a exclusive set of heroes that can only be claimed by collecting 8 unique comic covers (sound familar?) from lockboxes.
Each time you open one of these lockboxes, it does a random draw. It could be a new cover, or a duplicate of one you already own.
As you get closer to 8 unique, you get LOTS of duplicates, as the likelihood of getting a new one gets smaller and smaller.
However, there is a Bonus Bar that fills up each time you open a dupe. When the bar is full, you're next pull is guaranteed to be a new one.
So, eventually, with a combination of both luck AND brute force, you'll get that hero if you collect enough lockboxes.



Heavenstrike Rivals (A Final Fantasy themed tactics game kinda like Scrolls) does something similar as well when opening blind packs.
Eventually you get the best rarity if you have a really **** run of bad luck.

So for this game, do the following:

Add a "Bar to 5 Star" onto the Legendary Token page (so players can see how close they are to a guaranteed 5 star).
The bar is sub-divided into multiple divisions that eventually fill up based on pull results.
If a 4 star is pulled, fill the bar to the next division.
if a 5 star is pulled and the bar is not full, empty the bar.
If the bar is full, the pull is guaranteed to be a random 5 star cover. Then the bar empties to start again.
I'd set it so that the bar is full after 9 4 stars in a row, which would equate to the 10% pull rate they want us to have.

Stax the Foyer

DaveR4470 wrote:

simonsez wrote:

Thanks for conceding I'm correct after devising a contrivance to imply I'm wrong.

No, I pointed out that your absolute statement is only correct in limited circumstances. If you say "the sun never rises", you're absolutely wrong, unless you only meant "between 2 and 3 AM in Caracas, Venezuela".

Those "limited circumstances" encompass the whole of nearly everyone's MPQ career at current LT rates. Regression to the mean at infinite token pulls is true, but totally irrelevant.

DaveR4470

Stax the Foyer wrote:

Those "limited circumstances" encompass the whole of nearly everyone's MPQ career at current LT rates. Regression to the mean at infinite token pulls is true, but totally irrelevant.

It's not. First, they've just dumped a ton of LTs into the economy via the awards for championing 3*s and 4*s. The amount of LTs one can reasonably expect to acquire is significantly greater now than it was two weeks ago. Your point is valid in the short term -- it's statistically highly unlikely that you'd catch up to the greater-luck person (let's call them GLP) in the next 10 tokens. But if you increase the number of tokens opened, the statistical chance that you and GLP will have gotten the same distribution between 4*s and 5*s increases, approaching 100% asymptotically. And you can calculate (and graph) the exact odds for any given number of future tokens.

And there's no limit to the number of tokens available, so this isn't a zero-sum game -- the 5* tokens opened by GLP do not decrease your odds of obtaining a 5* token from any future pulls; i.e. there's not a fixed pool of 5*s available in the universe. (If that were the case, you'd be 100% correct in all circumstances -- your chances of catching up would be incredibly poor.)

So a month from now, you're right -- it's unlikely (but not impossible) that you'd catch up to GLP. But six months from now, when you've maybe opened another 30-40 tokens, it's less unlikely. When you've opened another 100, or another 150, the odds are much greater -- maybe even 50/50 -- that you'd have caught up.

If you want a guarantee that you'll catch up -- that's not possible. Odds are odds, and the only way to guarantee that would be to open an infinite amount of tokens (or so many that the odds against would be microscopic).

I see your point, though, because if there's a "fix" to RNGsus's whimsy for LTs, it won't affect your chance to catch up at all in the short run, even if you start pulling more 5*s, because both you and GLP will benefit from the "guarantee" or whatever the fix is, and it will statistically all net out to.... exactly where you started from. So you've got a legitimate complaint that the harm from landing on the wrong side of the bell curve can't be corrected.

Dwarfsteel

The Viceroy Returns wrote:

My older brother is at something ridiculous like 0 / 40.

There's a simple fix for this problem = A guaranteed 5 star after X covers opened in a row without pulling a 5 Star cover.

There's another Marvel game out there I play called Marvel: Avengers Alliance.
They have a exclusive set of heroes that can only be claimed by collecting 8 unique comic covers (sound familar?) from lockboxes.
Each time you open one of these lockboxes, it does a random draw. It could be a new cover, or a duplicate of one you already own.
As you get closer to 8 unique, you get LOTS of duplicates, as the likelihood of getting a new one gets smaller and smaller.
However, there is a Bonus Bar that fills up each time you open a dupe. When the bar is full, you're next pull is guaranteed to be a new one.
So, eventually, with a combination of both luck AND brute force, you'll get that hero if you collect enough lockboxes.



Heavenstrike Rivals (A Final Fantasy themed tactics game kinda like Scrolls) does something similar as well when opening blind packs.
Eventually you get the best rarity if you have a really **** run of bad luck.

So for this game, do the following:

Add a "Bar to 5 Star" onto the Legendary Token page (so players can see how close they are to a guaranteed 5 star).
The bar is sub-divided into multiple divisions that eventually fill up based on pull results.
If a 4 star is pulled, fill the bar to the next division.
if a 5 star is pulled and the bar is not full, empty the bar.
If the bar is full, the pull is guaranteed to be a random 5 star cover. Then the bar empties to start again.
I'd set it so that the bar is full after 9 4 stars in a row, which would equate to the 10% pull rate they want us to have.

If they implemented something like this then they would need to add an opt in/out button. Not everyone wants/can handle a five star with the game how it is and scaling. Yes, there's a chance still to pull one, but some are happier being behind the curve for the time being.

Also, then they better look into adding something similar to every other token out there. Pick your top 3-4 characters that "if you don't get a cover for one of them after 10 pulls" then you're guaranteed to. Because we all know it doesn't matter what they do to the game somebody isn't happy and only doing this to LT's will bring out the "D3 is catering to the P2W and the whales" crowd. So then everybody is "guaranteeing" more or less the same characters. More or less everyone has the same roster and the game boils down to only how much time you can spend on the game and grind. Game get's stale, people leave, game dies. That's the way I see it.

RNG, love it or hate it, is one of the significant balancing factors against those who have more time or more money to spend than others. Take that away and you lose that uncertainty factor that gives us those highs and lows we're all subconciously somewhat addicted to. With something like this we would very quickly see a lot more max covered 5 *'s than what we're seeing now.

SnowcaTT

Lightning rounds today are showing this worse than ever. Either you have gotten lucky by sacrificing to the RNG gods, or you are squashed like a bug as folks don't need to save health packs for other events (not that they'd use them anyway, since the 5*'s mostly regen) and there are no other important events going.

Saw this months ago, and knew the world was coming to this...but hoped that RNG would favor me a bit as well. Not so much! Instead, here I am incredibly frustrated that I can play harder/smarter/longer and not be able to progress over those who were a bit luckier.

simonsez

DaveR4470 wrote:

No, I pointed out that your absolute statement is only correct in limited circumstances.

It was a correct absolute statement concerning the parameters under which we're playing the game, namely, we don't get infinite token. So go find another outlet for these meaningless theoretically masturbatory exercises.

Raffoon

Thanks everyone for keeping the conversation alive while I've been on a "forum vacation to tahiti".

There are plenty of suggestions on ways to fix this. The bar idea is a good one. I spent about 5 minutes thinking of simple/easy ways to fix it in this other thread a little while back:
https://d3go.com/forums/viewtopic.php?style=1&f=7&t=37010

The problem is, the developers won't even acknowledge this as an issue.

Malenkov wrote:

Dimh wrote:

Raffoon wrote:

So, I ask the developers: How the hell am I supposed to progress my roster? I've opened almost 100 legendary tokens at this point and still have only 3 5* covers to show for it. Meanwhile, people opening half as many tokens have progressed more than 3 times as fast as me on average. On average, people that have opened the same number of tokens as me are getting to enjoy a new tier of character, while I'm stuck in the same spot with nothing to show and no recourse.

Your sample size is way too low. Let's talk when you get to 10,000 LT and remain at 3% draw rate.

edit> I apparently do not know how to quote things properly...

Sample size is irrelevant in this case. If Raffoon's argument was that his lack of success indicated that the general draw rate was extremely low, then you could make an argument based on sample size. But I don't think that anyone is actually challenging that the average 5* LT draw rate is 10%.

The problem here isn't the general problem of the *average* player, it's the *specific* problem of Raffoon drawing in the bottom 1% of LT token draw permutations.

It's a system that produces extreme divergences of winners and losers and has nothing to do with merit or commitment. If Raffoon's luck is worse than 99.2% of players, there exists a theoretical player on the other side of the probability spectrum. That player has drawn 19 5*s to Raffoon's 3 and has done EXACTLY the same amount of work.

Malenkov GETS IT!!

And also the Skrull Empire guys too. Thanks for fighting the good fight while I was temp banned.

simonsez

DaveR4470 wrote:

So a month from now, you're right -- it's unlikely (but not impossible) that you'd catch up to GLP. But six months from now, when you've maybe opened another 30-40 tokens, it's less unlikely. When you've opened another 100, or another 150, the odds are much greater -- maybe even 50/50 -- that you'd have caught up.

Instead of just pulling numbers out of your ****, why not work them out and you'll see how wildly wrong "50/50" is. Even without doing any math, if you're talking about a fixed number of tokens, it can NEVER be 50/50, because this implies that on the second run of pulls, the expected number of 5s for the people who had bad luck will now be greater than the expected number who previously had good luck.

I'll repeat my sim results from the previous page. For those people who pulled less than 10% on 150 pulls, their chances, over the next 150 pulls, of catching up to someone who pulled more than 10% on their first 150 pulls, is only 16%. Not 50/50. Not even close.

morph3us

Alrighty, a question, then.

Firstly, for what it's worth, I think random isn't a good way of determining end game progression, given the overall sample size (ie number of LTs opened per player) you're going to need in order to determine a reasonably fair distribution amongst the playerbase, even if it is a good way of keeping overall 5* cover numbers within a target range for a population from the devs' perspective.

Now, assuming they implement some sort of "streakbuster" mechanic (I have visions of people piling on top of a naked person on a sporting field, but moving on), be that a progression bar or otherwise, presumably that streakbuster mechanism is going to increase the distribution of 5* covers above that 10% target. Do they adjust the odds of a 5* LT pull down slightly to account for that?

FierceKiwi

I was going to log in to type out a long reply about how dumb it is to play the idiotic infinite token argument but thankfully Simon already pointed out how stupid it is so I'll just quote his brilliance.

simonsez wrote:

DaveR4470 wrote:

No, I pointed out that your absolute statement is only correct in limited circumstances.

It was a correct absolute statement concerning the parameters under which we're playing the game, namely, we don't get infinite token. So go find another outlet for these meaningless theoretically masturbatory exercises.

XandorXerxes

simonsez wrote:

DaveR4470 wrote:

So a month from now, you're right -- it's unlikely (but not impossible) that you'd catch up to GLP. But six months from now, when you've maybe opened another 30-40 tokens, it's less unlikely. When you've opened another 100, or another 150, the odds are much greater -- maybe even 50/50 -- that you'd have caught up.

Instead of just pulling numbers out of your ****, why not work them out and you'll see how wildly wrong "50/50" is. Even without doing any math, if you're talking about a fixed number of tokens, it can NEVER be 50/50, because this implies that on the second run of pulls, the expected number of 5s for the people who had bad luck will now be greater than the expected number who previously had good luck.

I'll repeat my sim results from the previous page. For those people who pulled less than 10% on 150 pulls, their chances, over the next 150 pulls, of catching up to someone who pulled more than 10% on their first 150 pulls, is only 16%. Not 50/50. Not even close.

By my math, it's a 43% chance of pulling 16+ 5*s in 150 tokens.

That said, if there's a 10% chance of drawing a 5*, if player A starts from 10 5*s and player B starts from 20 5*s, for the next however many pulls they're going to have a 10% chance of pulling a 5*. It is by definition less likely that player B "catches up," but how severe that "catch-up" requirement is depends entirely on how far the starting distance between player A and player B is (assuming an equal number of pulls). Even in 1,000,000 pulls, there's going to be variation on how many get pulled - but the odds will be close to 10%.

The biggest problem is that it's the initial dispersement. Those who invested heavily / got really lucky will have an innate advantage, but that's fairly true at any level. Those who put in more time and effort will still do better than those who don't, as they will get many more chances to pull the 5*s. It'd be more annoying to me that it's not skill based if I didn't feel this game was already incredibly luck-driven.

Stax the Foyer

DaveR4470 wrote:

Stax the Foyer wrote:

Those "limited circumstances" encompass the whole of nearly everyone's MPQ career at current LT rates. Regression to the mean at infinite token pulls is true, but totally irrelevant.

It's not. First, they've just dumped a ton of LTs into the economy via the awards for championing 3*s and 4*s. The amount of LTs one can reasonably expect to acquire is significantly greater now than it was two weeks ago. Your point is valid in the short term -- it's statistically highly unlikely that you'd catch up to the greater-luck person (let's call them GLP) in the next 10 tokens. But if you increase the number of tokens opened, the statistical chance that you and GLP will have gotten the same distribution between 4*s and 5*s increases, approaching 100% asymptotically. And you can calculate (and graph) the exact odds for any given number of future tokens.

And there's no limit to the number of tokens available, so this isn't a zero-sum game -- the 5* tokens opened by GLP do not decrease your odds of obtaining a 5* token from any future pulls; i.e. there's not a fixed pool of 5*s available in the universe. (If that were the case, you'd be 100% correct in all circumstances -- your chances of catching up would be incredibly poor.)

So a month from now, you're right -- it's unlikely (but not impossible) that you'd catch up to GLP. But six months from now, when you've maybe opened another 30-40 tokens, it's less unlikely. When you've opened another 100, or another 150, the odds are much greater -- maybe even 50/50 -- that you'd have caught up.

If you want a guarantee that you'll catch up -- that's not possible. Odds are odds, and the only way to guarantee that would be to open an infinite amount of tokens (or so many that the odds against would be microscopic).

I see your point, though, because if there's a "fix" to RNGsus's whimsy for LTs, it won't affect your chance to catch up at all in the short run, even if you start pulling more 5*s, because both you and GLP will benefit from the "guarantee" or whatever the fix is, and it will statistically all net out to.... exactly where you started from. So you've got a legitimate complaint that the harm from landing on the wrong side of the bell curve can't be corrected.

I do understand the math, don't worry. The disagreement is just about where and when that convergence will occur. I think you're overestimating the effect of the champing LTs (and underestimating the dilution effect of additional 5*s).

The burst of LTs from champing is a one-time burst, for the most part. The trickle of LTs at every additional 10th 4* cover is negligible. I'm on the other side of that burst, having champed almost all of the 3*s and 4*s that I intend to, and still don't have a viable 5*, lacking both OML yellow or SS blue.

There are positive and negative feedback factors on either side of the distribution curve, as well. Having viable 5*s makes reaching progression in PvE much easier, as healthpack usage isn't a factor. Not having viable 5*s will make PvP progression somewhat trickier, although that effect won't be as pronounced. Both of these factors will be accelerated by the championing LT wave, as some people will benefit, while others will be left further behind.

Plus, there are psychological factors at play as well. LT pulls are the result of a fairly substantial level of effort, DDQ4 aside. Someone playing hard for a year will eventually catch up, but how many people will sustain that level of effort when they've been falling further and further behind?

There's a hypothetical convergence out there, but I think it's well beyond 13 covers/5*, and at least several months out, probably a few years out. That's the issue. Realistically, I'm looking at maybe 30 LTs/month, probably less. Under the current setup, I've probably got at least 6 months minimum before I can expect to catch up to the GLPs, more likely a couple years.

Plus, if additional 5*s are released, further dilution of the 5* pool will likely push that window out, unless there's a drastic increase to the overall drop percentage.

The Viceroy Returns

Dwarfsteel wrote:

If they implemented something like this then they would need to add an opt in/out button. Not everyone wants/can handle a five star with the game how it is and scaling. Yes, there's a chance still to pull one, but some are happier being behind the curve for the time being.

Also, then they better look into adding something similar to every other token out there. Pick your top 3-4 characters that "if you don't get a cover for one of them after 10 pulls" then you're guaranteed to. Because we all know it doesn't matter what they do to the game somebody isn't happy and only doing this to LT's will bring out the "D3 is catering to the P2W and the whales" crowd. So then everybody is "guaranteeing" more or less the same characters. More or less everyone has the same roster and the game boils down to only how much time you can spend on the game and grind. Game get's stale, people leave, game dies. That's the way I see it.

RNG, love it or hate it, is one of the significant balancing factors against those who have more time or more money to spend than others. Take that away and you lose that uncertainty factor that gives us those highs and lows we're all subconsciously somewhat addicted to. With something like this we would very quickly see a lot more max covered 5 *'s than what we're seeing now.

I realize this suggestion isn't perfect for sure, but it's something at least.

One of the many problems with the current system is that the tokens themselves haven't been upgraded to match the game's overall progression. The game has very much out-paced the usefulness of Heroic & Event tokens for several of the top tiers of players at least. They are still being given out as though the 2 Star tier is the rarity base the Player should be using for a long time, which hasn't been true for over a year. 2 Star covers dropping from PVP fights and DDQ have made the 2 Star range a drive-by on the way to 3 Star progression.

Heroics have 3 different rarities crammed into them, so it would be hard to emulate. There needs to be a token in between Heroic and even the new one they are supposedly making.
A token that is 90% 3 Star and 10% 4 Star. Then there is at least SOME consistency with their tokens. We don't even know what the new 20 CP token odds are.

The currently small pool of 5 Star is something this system would have trouble with as you point out. However, they claim more "are coming" in that Game of Thrones way. After most players burn through their haul of LT"s from the first rank of 3 Star Champions, it will take some time to open 10 total. Unless in a buyer's club alliance, which can't be designed around anyway unless CP's are removed from all store purchases. Something like this would have to wait until here is a solid enough stable of 5 star characters to have a difference.
Opponent diversity is pretty lacking right now in top end PVP as it is; you only ever see the same 4-6 4 Star characters in various combinations after about 900 points.

The biggest problem with the entire 5 Star system is that they initially said that 5 stars would not be purchasable, but then they went and added CP to the store entries, so they effectively are.
The Super Whales are buying Stark Salaries for the CP alone. They could give a rat's about the HC, especially now that it's only more or less used for Roster Slots and the occasional shield since they won't be able to buy direct covers anymore.
And even worse, they made them purchasable AND luck based. So even after one spends heaps of money, they STILL aren't guaranteed to get one. That's the crux of this post completely, and it sounds an awful lot like gambling to me...

So "D3 catering to the P2W crowd" is already here.
I've seen several games balance RNG & progression without completely **** over the player via luck. RNG has to have checks and balances in their implementation or else both ends of the spectrum become pronounced and noticeable.
I just wish they'd acknowledge that it's a problem is all and do something to address it.

slidecage

i pulled a surfer and a jean right after one another few months back... since then

17 with not a single 5 star.

Raffoon

They say that they want to improve the long-term health of the game.

So what do they do?

DOUBLE DOWN on the Legendary token idea.

Unbelievable.

Why not make a system that lets you NOT completely rely on luck to get the 5* covers that you need? Some luck in the system is fine, but ALL luck is absurd.

tanis3303

Yea, I now share your rage Raff. This latest change with the tokens is them, reading the months worth of **** about the absurd luck based distribution, and saying:

"Hey guys...sorry you didn't pay the entry fee of $XX,XXX to play with the new tier of characters...if you didn't tho, and/or you had **** luck...well, Kitty you! Lol, cry moar brah!"

This change means people that didn't already finish Silver Surfer (and very soon Logan as well) will NEVER finish him. You'll get roughly 2 shots per month at him via earned CP. Two, possibly three chances...Per. Month. Not guaranteed pulls, a tiny kitty slot pull. Smh...I'm already waaayyyy behind the power curve, and now I can never even hope to catch up.

Bowgentle

I just did some math.

28 LTs from championing.
21 from DDQ
15 from PVE (give or take 1 or 2)
40 from PVP (including season progression and alliance rewards)
8 from CP (probably a bit more)

So let's say, a nice round 110 LTs.

I have pulled 5 covers spread across 3 5* characters. Five.
Plus the one SS from day 730.

I used to have a top roster. Until they introduced pure luck-based progression.
Thank you, devs.

Monged4life

You wont like my comparison then, as someone who doesn't want 5*s I have now pulled 6 out of my ~30 Legendary tokens (I don't have the time to grind for the PvE one or the roster to reach 1300 in PvP).

TheRealJRad

I've pulled 2 5* since LTs came out, the second was a SS from the champion LT surge. I've had an "it is what it is" attitude about it. It would take me so long to get a 5* viable, it's not worth stressing over. What is grinding my gears is my 4* pulls outright refuse to give me Ice/Hulk/JG tier characters. I'm not an expert in numberology, but it seems less than random when people are consistently pulling Star-Lords, but can't sniff a Hulkbuster (just as an example).

Unknown

Another 17 LTs and zero 5*s, putting me at 87 total LTs drawn since I blew my stash after the release of championing and only 2 5*s, putting me at the bottom 99.4% of draw permutations.

Not only is that all my accumulated CP, it's also all the "easy" LTs from championing maxed 3*s. With classic LTs and the continued release of new 5*s, I guess I really am just boned.