Chance to Catch Up: what 5* RNG really means for players

Unknown

Randomness has always determined a lot of micro-level success or failure in MPQ. Everyone's had the game where the AI cascades 60 AP on their first turn and you lose almost no matter the team composition. This kind of random event causes you to lose a fight you should have won or win a fight you should have lost; in turn, that might cause you to unexpectedly reach or fail to reach a certain reward.

But in a day of MPQ, there are many hundreds of these micro-level opportunities. Up until the release of 5*s, large-scale success was much more meritocratic, dependent essentially on how much effort you were willing to put into the game.

With the release of 5*s, this is no longer the case. Macro-level success or failure is now largely determined by random chance. Luck causes a significant portion of the player base to fall behind or race ahead irrespective of merit. In other words, the same effort now produces significantly different outcomes at the top level of play.

To illustrate this phenomenon, I propose a new metric, Chance to Catch Up, which measures your odds of maintaining parity with a theoretical mirror player who has the opposite luck as you. If you are in the bottom X% of outcomes, find the success rate of your mirror counterpart at the top X%. Then, assuming that the next set of draws for your mirror are average, determine the likelyhood of you catching up to them. If we crunch the numbers using this metric, we can see that they're pretty dismal.

Let's take a look. For 100 LTs drawn, the bottom 20% of players will pull at most seven 5*s. The top 20% will pull 12, a difference of 5 covers. For the next 100 LTs, if the top 20% player pulls the average 10% rate, the bottom 20% player will need to pull 15 covers to catch up. He or she only has a 7.26% chance of doing so or better. Effectively, a large subset of players will be permanently behind, even though they are identical in every other respect other than RNG.

What about players with extraordinarily good or poor luck? Let's use me as an example. Since the Champions release, I have pulled a total of two 5*s for 87 LTs, putting me in the bottom 99.4% of outcomes. My mirror counterpart has pulled 16 covers, for a differential of +14. In the next 87 pulls, my CCU (rounded to the nearest hundredth of a percent) is 0%.

A high level of inequality is fundamental to the current random draw 5* system. In the few short months that this system has been in place, it has already begun to indiscriminately create distinct classes of winners and losers without any consideration of hard work or smart play. The losers will continue to be losers, exacerbated even more by a depressed ability to achieve higher placement rewards. This isn't just a concern for the small number of players with extraordinarily good or bad luck; it affects the many tens of thousands of players with merely above-average and below-average luck too. This is a system that is and will continue to be offensive to any notion of fair play, which far overshadows any of the laudable efforts the developers have made to improve the fairness and longevity of this game.

I have added the mirror-player and CCU metrics to my MPQ Luck calculator (http://dropproxy.com/f/CBF). Feel free to play around with your own data. If you've been unlucky, see how likely it will be for you to catch up. If you've been lucky, consider how easily you might not have been.

Unknown

A few additional notes:

1. This metric is a response to the many players that I've seen suggest that runs of good and bad luck will eventually even out. This assertion is not true. While it's true that increasing the sample size will make outlier permutations less likely (1 success in 20 is only bottom 40 percentile, but 5 successes in 100 is bottom 6), past draws do not affect future ones. So, while pulling 50 failures in a row is really terrible luck, you have the same chance of pulling a 5* on the 51st draw as you did on the first.

2. Yes, I know you can buy covers with CP. Let's see a show of hands of how many people have spent less than $1K and bought more than 1 5* cover? In any case, CP cover purchases require at least one extant cover; given the number of LT opportunities and the still too-fast pace of 5* releases, getting even one specific cover is potentially problematic.

3. As others have already suggested, the "Classic LT" switch exacerbates the problem of building viable 5*s. But even if the developers adopt some of the suggested remedies (awarding CP instead of LTs directly, ensuring that 5* drop rates remain at 3.3% regardless of future character releases), the essential problem with RNG that I have raised here remains.

4. As always, when measuring your own luck and CCU, try to avoid sample bias. Either use the full set of draws or a predetermined subset that isn't cherrypicked.

5. Some potential solutions for this issue are to regularly offer 5*s through some kind of regular, non-random process (the non-competitive daily rewards are too sparse to make a substantive dent), dramatically increase the opportunities for draws (while potentially decreasing the draw chance). In other words, mechanisms that already exist to distribute 1-4* covers. These would serve to reduce the inequality of random distribution without dramatically increasing general availability. Something like 5*s in high season progression and as low-probability heroic/event token rewards.

Arondite

Congrats! Your draws were exceptional!

Going by Colognoisseur's ranking of 4*s, your luck was equivalent in strength to: Iceman!

A person with the opposite luck as you succeeded 0 times, a difference of -6!
In the next 19 draws, they would have a 0.03% chance of catching up to you.

Details:
Less lucky than you: 99.14%
Exactly as lucky as you: 0.69%
More lucky than you: 0.17%

Melevorn

I have pulled exactly eight (8) 5* covers from approximately 220 Legendary Tokens. I'm at work so can't plug that into the luck calculator, but I reckon it's pretty unlucky. Note, all were cashed after the listed chance rose to 10% - I hoarded LTs while Silver Surfer was the only 5*.

As it happens, I am sitting on exactly eight (8) Star-Lord covers currently waiting to be converted into ISO (I just can't even face them right now), plus about twenty more 4* covers which are also destined for the scrap heap (not enough ISO to max level their characters). Had to sell so, so many for ISO before championing, too.

Anyway, that's my 2 cents. Anyone out there with a worse than 3.6% pull rate for 5*'s?

Unknown

Melevorn wrote:

I have pulled exactly eight (8) 5* covers from approximately 220 Legendary Tokens. I'm at work so can't plug that into the luck calculator, but I reckon it's pretty unlucky. Note, all were cashed after the listed chance rose to 10% - I hoarded LTs while Silver Surfer was the only 5*.

As it happens, I am sitting on exactly eight (8) Star-Lord covers currently waiting to be converted into ISO (I just can't even face them right now), plus about twenty more 4* covers which are also destined for the scrap heap (not enough ISO to max level their characters). Had to sell so, so many for ISO before championing, too.

Anyway, that's my 2 cents. Anyone out there with a worse than 3.6% pull rate for 5*'s?

Okay, wow. I have pulled 8 covers out of around 150, which is already horrendously bad. The calculator doesn't actually support values over 169 right now, because Javascript maxes out at around 169 factorial (!). But that is clearly pretty far into the bottom 1%. Your lucky counterpart has pulled over 30 covers.

alaeth

Hey, my luck is apparently tinykitty... 86 LT opens, ZERO 5*

I like your idea.

chaos01

Well I was Devil Dinosaur. But, after 583 days I have opened 46 token got one five star and have got 42 usable 4 stars. I have had awful luck getting 4 stars in anything but legendary tokens. With my RNG luck I'm surprised I don't find a way to get a two star out of the legendary.

spectator

Would be nice if I could work out how many legendary tokens I have opened

Unknown

Malenkov wrote:

Melevorn wrote:

I have pulled exactly eight (8) 5* covers from approximately 220 Legendary Tokens. I'm at work so can't plug that into the luck calculator, but I reckon it's pretty unlucky. Note, all were cashed after the listed chance rose to 10% - I hoarded LTs while Silver Surfer was the only 5*.

As it happens, I am sitting on exactly eight (8) Star-Lord covers currently waiting to be converted into ISO (I just can't even face them right now), plus about twenty more 4* covers which are also destined for the scrap heap (not enough ISO to max level their characters). Had to sell so, so many for ISO before championing, too.

Anyway, that's my 2 cents. Anyone out there with a worse than 3.6% pull rate for 5*'s?

Okay, wow. I have pulled 8 covers out of around 150, which is already horrendously bad. The calculator doesn't actually support values over 169 right now, because Javascript maxes out at around 169 factorial (!). But that is clearly pretty far into the bottom 1%. Your lucky counterpart has pulled over 30 covers.

I rewrote the calculator to use a free software BigNum library I found, so now it should support arbitrarily large numbers (though it is a bit slower now) Here are the numbers for your 8-out-of-250 (which is still one of the worst total runs that I've even heard of):

Tinykitty! Your draws were just awful!

Going by Colognoisseur's ranking of 4*s, your luck was equivalent in strength to: Devil Dinosaur!

A person with the opposite luck as you succeeded 45 times, a difference of +37!
In the next 250 draws, you have a 0% chance of catching up to them.

Details:
Less lucky than you: 0.001%
Exactly as lucky as you: 0.003%
More lucky than you: 99.996%

Pylgrim

I've been pushing this idea around for a while because it's almost the same case for 4*s. Good luck with LTs (opening a cover you need, especially one of the unarguable "top" characters) can quickly catapult some players high above those with average or bad luck (the ones opening the 12th Invisible Woman Blue cover or any repeat of the old, weak 4*s). A few lucky pulls of the likes of HB, Jean Grey or Iceman will quickly make their owners much more competitive than their unlucky counterparts, earning more tokens and so perpetuating forever the comparatively better fortune that was assigned to them by mere chance.

Unknown

Pylgrim wrote:

I've been pushing this idea around for a while because it's almost the same case for 4*s. Good luck with LTs (opening a cover you need, especially one of the unarguable "top" characters) can quickly catapult some players high above those with average or bad luck (the ones opening the 12th Invisible Woman Blue cover or any repeat of the old, weak 4*s). A few lucky pulls of the likes of HB, Jean Grey or Iceman will quickly make their owners much more competitive than their unlucky counterparts, earning more tokens and so perpetuating forever the comparatively better fortune that was assigned to them by mere chance.

The top 4*s are really great, but the power differential between Iceman/HB/JG/DPX/RH/4Cyc and a mid-tier 4* like 4hor/Fury/XFW is much less than between a 8-9 cover OML/Phoenix/SS and a maxed Iceman. Which is reasonable, 5* vs 4* and all. But it just goes to show that you can get most of the way there with more than a dozen different 4*s.

Not to mention that there's many more different ways to get 4*s - progression, placement, championing, just straight out buying them before this week. That's what I'm suggesting here (and what others have suggested in other posts). If there were all those ways to get 5*s, even we move some of the % chance from LTs to other tokens and to event rewards, it would help even out the severe imbalances that are inevitable with this kind of system.

So, not an hour after I revised the calculator, I actually pulled the OML Yellow that I was looking. So I'm up to 3 out of 89, up from bottom 0.6% to bottom 1.9%. From here, I'll actually have to pull 6 straight 5*s to claw my way up to parity. But damn if that cover isn't shiny.

avs962

My results had e in the top 59.963%. This isn't 100% accurate since I had to approximate how many LTs I've opened. The worst part is I don't feel like I'm even doing remotely well.

Simply put, I think the mere idea that end game progression is based totally on luck is appalling. Are there seriously zero people at the offices of D3 and Demiurge who think this way?

wirius

I think you all are misunderstanding the purpose of 5*'s. They are your lottery that keeps you buying into the system that you don't have a chance of profiting off of. 5*'s are a happy accident to win, and a dangle in front of you to keep playing. Look how much you want them! And it was just a long of enough dangle until championing came in to keep players playing even when those players finally realize the 5* pursuit is futile.

Bowgentle

You are in the bottom 2.13% of all draw permutations

Going by Colognoisseur's ranking of 4*s, your luck was equivalent in strength to: Devil Dinosaur!

A theoretical person with the opposite luck as you succeeded 17 times, a difference of +12!
In the next 110 draws, you have a 0.05% chance of catching up to them.

Details:
Less lucky than you: 1.17%
Exactly as lucky as you: 1.92%
More lucky than you: 96.91%

Yup, that's 5/110 for you.

Xenoberyll

Oh no! Your draws were pretty bad!

You are in the bottom 17.323% of all draw permutations

Going by Colognoisseur's ranking of 4*s, your luck was equivalent in strength to: Mr. Fantastic!

A theoretical person with the opposite luck as you succeeded 16 times, a difference of +6!
In the next 133 draws, you have a 7.158% chance of catching up to them.

Details:
Less lucky than you: 13.351%
Exactly as lucky as you: 7.943%
More lucky than you: 78.705%

10 in 133

kensterr

Don't care. If luck isn't good for new season I'll just go super casual. Playing super casual is less stressful when you don't have to rely on luck to catch up with the lucky ones. Let them play in their own battleground. Sooner or later the unlucky ones will just leave the game. Lucky vs Lucky. Fight!

Pylgrim

Malenkov wrote:

Pylgrim wrote:

I've been pushing this idea around for a while because it's almost the same case for 4*s. Good luck with LTs (opening a cover you need, especially one of the unarguable "top" characters) can quickly catapult some players high above those with average or bad luck (the ones opening the 12th Invisible Woman Blue cover or any repeat of the old, weak 4*s). A few lucky pulls of the likes of HB, Jean Grey or Iceman will quickly make their owners much more competitive than their unlucky counterparts, earning more tokens and so perpetuating forever the comparatively better fortune that was assigned to them by mere chance.

The top 4*s are really great, but the power differential between Iceman/HB/JG/DPX/RH/4Cyc and a mid-tier 4* like 4hor/Fury/XFW is much less than between a 8-9 cover OML/Phoenix/SS and a maxed Iceman. Which is reasonable, 5* vs 4* and all. But it just goes to show that you can get most of the way there with more than a dozen different 4*s.

Not to mention that there's many more different ways to get 4*s - progression, placement, championing, just straight out buying them before this week. That's what I'm suggesting here (and what others have suggested in other posts). If there were all those ways to get 5*s, even we move some of the % chance from LTs to other tokens and to event rewards, it would help even out the severe imbalances that are inevitable with this kind of system.

So, not an hour after I revised the calculator, I actually pulled the OML Yellow that I was looking. So I'm up to 3 out of 89, up from bottom 0.6% to bottom 1.9%. From here, I'll actually have to pull 6 straight 5*s to claw my way up to parity. But damn if that cover isn't shiny.

Oh yeah I was not trying to compare the magnitude of the issue, just pointing out that the "progress-or-fail-by-chance" is an issue that is present at all levels, especially at the two highest tiers. Luck should generate variance, not dictate progression speed.

simonsez

This is excellent, thanks for doing this. I'm glad that someone else is pointing out that once you fall behind, your chances of catching up are negligible.

kalex716

wirius wrote:

I think you all are misunderstanding the purpose of 5*'s. They are your lottery that keeps you buying into the system that you don't have a chance of profiting off of. 5*'s are a happy accident to win, and a dangle in front of you to keep playing. Look how much you want them! And it was just a long of enough dangle until championing came in to keep players playing even when those players finally realize the 5* pursuit is futile.

I would agree with you if nobody was in fact getting them. But people are lucking out, and its giving them a huge edge.

I'm in the middle of the 4 star transition, with about 120 4 star covers total, but unfortunately, i don't have any of the top 5 4 stars even close to being covered. I'm way behind. Getting 1000 was hard enough before because I'd have to use boosts and stuff at the 800-1000 mark just to out compete hulkbuster/jg combo's to push me over the 1k mark, but now i'm seeing decently covered 5 stars into that mix as well. This is directly impacting my achievable goals in a negative way, for no good reason other than math being those 5 star players benefactor.

PeterGibbons316

Malenkov wrote:

Randomness has always determined a lot of micro-level success or failure in MPQ. Everyone's had the game where the AI cascades 60 AP on their first turn and you lose almost no matter the team composition. This kind of random event causes you to lose a fight you should have won or win a fight you should have lost; in turn, that might cause you to unexpectedly reach or fail to reach a certain reward.

But in a day of MPQ, there are many hundreds of these micro-level opportunities. Up until the release of 5*s, large-scale success was much more meritocratic, dependent essentially on how much effort you were willing to put into the game.

With the release of 5*s, this is no longer the case. Macro-level success or failure is now largely determined by random chance. Luck causes a significant portion of the player base to fall behind or race ahead irrespective of merit. In other words, the same effort now produces significantly different outcomes at the top level of play.

This is spot on. One of the biggest problems with this game that I have always complained about is the limited availability of rewards. If you need ISO for something and want to grind it out by playing the game, the best you can do is 20 ISO rewards over and over. Need HP? Sorry, the amount you can earn is fixed. CP? Fixed. Tokens? Fixed.

Having RNG based progression is not a bad thing if you are able to continuously earn rewards with random draws. In an RPG you can keep raiding the same dungeon and killing the same boss over and over for that sweet drop you are looking for. And while random, you can eventually get what you want with enough attempts. That mechanism isn't present here. So with every event, everyone gets one shot at a sweet drop, but only 10% of players actually get it. The other 90% fall behind, for no other reason than blind luck. And they are not given a chance to catch up, the only shot they have at catching up is with another draw in the next event. It's a terrible mechanic for progression.

XandorXerxes

There was already a big luck element involved in the 3 and 4* transition, it's just that the people I usually see on here (long-timers) largely missed it after their initial climbs. This new 5* tier has put the older players in the same spots as the newer players trying to hit the next tier level - good luck getting lucky enough to get the covers you want/need. No surprise, the older players like the new reward structure about as much as the newer players do - not at all. To make matters worse, the 5* tokens are more scarce and some take dedication and effort.

This isn't to say that this isn't a bad thing - it's to say that it's not really anything new.