Your WORST MPQ moment of the day (REALITY thread)

Mr_F

@PiMacleod Propability is a.. tinnykitty. If something MAY happen once in bilion times it may happen at 1st try, as well as 168465165 in a row or never. That how much propability is a tinnykitty.

helix72

sambrookjm said:

I figured you were a math geek with your work in probability in some of the other threads I've read, helix.   Advanced Engineering degree here.

Well, hello to a fellow mathlete. It's nice being appreciated  Math has served me very well over the course of my lifetime.

One of the things I love about this game is it does give me an outlet for my math skills in various ways. I've always used "diversions" like this as ways to learn/grow/strengthen skills that I then applied in the real world. Years ago I built a Monte Carlo simulator for my fantasy baseball team, then subsequently adapted it to price corridor aggregates for large deductible insurance policies. But I digress.

It's also a bit of a mission (and related to my job) to help people understand and interpret what numbers really mean. Low odds are not the same as no odds, and over a large enough sample size, infrequent events happen. Conditional probability, sampling with replacement, etc, etc. all awesome topics. One of my favorite bar bets: if you select 23 random people, there's a 50.6% chance that at least 2 of them were born on the same month and day. Usually I can get someone to give me 5 to 1 odds and 25 people, sometimes 30. Anyway, enough ramblings for one day.

HoundofShadow

Hitting the odds needs to occur for a large sample size. If you are going to apply probability on a small sample size over many individual/separate events, of course it's going to look like a "scam".

There are streaks of players going without getting a 5* character in LT/CP store, and there are also players getting 4 5* within 10 pulls. It's just that humans are inflicted with this disease called Negative Bias. They remember negative events more clearly. Over in Reddit MPQ, more players post their good pulls frequently, compared to here. 

It's also possible to get 2 5* with a 10x or 40x (can't remember which). Once again, I saw it in Reddit MPQ. 

sambrookjm

helix72 said:

sambrookjm said:

I figured you were a math geek with your work in probability in some of the other threads I've read, helix.   Advanced Engineering degree here.

Well, hello to a fellow mathlete. It's nice being appreciated  Math has served me very well over the course of my lifetime.

One of the things I love about this game is it does give me an outlet for my math skills in various ways. I've always used "diversions" like this as ways to learn/grow/strengthen skills that I then applied in the real world. Years ago I built a Monte Carlo simulator for my fantasy baseball team, then subsequently adapted it to price corridor aggregates for large deductible insurance policies. But I digress.

It's also a bit of a mission (and related to my job) to help people understand and interpret what numbers really mean. Low odds are not the same as no odds, and over a large enough sample size, infrequent events happen. Conditional probability, sampling with replacement, etc, etc. all awesome topics. One of my favorite bar bets: if you select 23 random people, there's a 50.6% chance that at least 2 of them were born on the same month and day. Usually I can get someone to give me 5 to 1 odds and 25 people, sometimes 30. Anyway, enough ramblings for one day.

Yep...a Mathlete (MA state champ in high school!), and helped out with Mathcounts at the middle school for a while...at least, before everything went virtual.  The kids always asked how I knew that some of that stuff.  I told them it was because Euclid was my geometry teacher.

I work with Monte Carlo simulations every day at work, and have been doing probability stuff since I was in Junior High Mathcounts back...well, before a good number of the players here were born!  I have to explain to people repeatedly that each time they want me to vary a parameter, it multiplies the number of runs I need to do, so varying one parameter three times, one parameter four times, and one parameter six times means I have to do 3*4*6 trials to meet all your criteria, not 3+4+6.  "At one week per trial, I'll get back to you in a year and a half, which will cost you $$$$"  That tends to get sponsor's attention rather quickly.

Yeah, the birthday problem is always a fun one.  People think just one particular person who has the same birthday as someone else, so they're only doing 22 comparisons, rather than all of the comparisons for everyone...which is why they vastly over-estimate the number of people for 50-50.  (Note - Probability values not valid with identical twins in the room.)

sambrookjm

PiMacleod said:

sambrookjm said:

helix72 said:

This is how the math works. I've got an advanced degree in math, so I assure you, it is correct. Here is how it works:

You have a 1/166 chance to get a 5* on each of the 40 draws
So the chance of not getting a 5* is 1-1/166= 165/166 on each of the 40 draws
The chance of not getting a 5* on any of your 40 pulls is (165/166)^40 = .785
So you have a 78.5% chance of not getting any 5* on any of your pulls
Which means you have a 1 - 78.5% = 21.5% chance of getting at least one 5* in a 40 pack

I figured you were a math geek with your work in probability in some of the other threads I've read, helix.   Advanced Engineering degree here.  Yes, PiMacleod, this is how the math works over large sample sizes.

helix72 did the math correctly for no 5-stars. 
For a single 5-star, the odds are (40 pulls in the pack) * [(165/166) ^ 39] * [1/166], or about 19%
For a very lucky pack with two 5-star covers, the odds are [(40 * 39 / 2)] * [(165/166) ^ 38] * [(1/166)^2], or about 2.2%

It gets much, much worse for anything beyond that.

These numbers hold true for very large sample sizes, so you may get unlucky and get ten 40-packs without any 5-star covers.  The odds of that are 1-(165/166)^400, or about 9%.  Not very likely, but still possible...and will happen about 9% of the time, in fact!

As someone who had a 0/9/0 Nick Fury before saved covers were a thing, trust me when I say that these things happen.  The odds of nine consecutive covers of a particular color for the same character are roughly 1-in-20,000.  (1-in-3^9, or 19,683 to be precise)

See, now THAT i believe.

We have SO MANY CASES documented in these forums of people going 1 out of 50 OR MORE for LLs without a 5*.  Yet, it's just 15% chance PER PULL.  

That makes more sense to me.  Actual stories.  Math can show us all these probabilities... but we have all of these examples of where things just DON'T happen.  And sure, that's math for ya!  And yes, humanity will always report on the negative more than the positive...

...but excuse me while I save my over 10k HP, because I don't believe for a single second that I'll get a single 5* from that 40x pack.  I can't justify spending THAT much HP on such LOW chances to get something.

...which, I guess, is the basis of my posting that in the first place.  Cost versus probability.  At least the mega millions/powerball/insert your lottery here I can somewhat justify in my mind -- if you hit it, you're done!  You win!  Go ahead and retire!  ...here, if you hit, you got ONE cover out of 13 needed!  You're NOT done at all!  and you just spent HOW much HP!? 

Oh, I agree 100% that spending 10K HP to get a 1-in-5-ish chance at a new 5-star is not worth it.  I've never done that, and the only time I've bought the 10 packs is when I've been clicking too fast claiming covers to catch it.  I'm not arguing that.  I'm just arguing about the probabilities.

If you frequent this thread, you'll hear a lot of the "I went 0-for-50 in my LT pulls"  Heck, check some of my back posts, and you'll find my longest streak was with Yelena Benlova, where I was in the 0-for-70 streak for getting her.  Hence the "Worst MPQ moment" title.    But if you wander into the "Best MPQ Moment" you'll see plenty of people who have the "One pull, one five star.  I'm done!"  I've done that as well (with 5-star Deadpool) and was quite happy about it.  Assuming the 15% for an LT is spread equally among the three 5-stars, you're odds of getting the new five-star in the vault one a single pull are 1-in-20 (15% / 3, or 5%)  Average all the values for when people get their first new character from the 5-star vault, and it should come out to 20 pulls because those are the odds that we're given.

Just because something is 21.5% likely to happen doesn't mean that it will not happen to you.  It just means that it's 79.5% likely that it won't happen to you.  Think of the gambler's fallacy: "The coin's come up heads four times in a row...we're due for a tails!"  Not true - assuming a fair coin, your chances of getting the tails after four heads is the same as always: 50%.  Overall, your odds of getting five heads in a row are 1-in-32, but if you've already got the first four heads (odds of 1-in-16), the fifth flip is independent of the previous four, and has an equal shot at being either a heads or a tails.

PiMacleod

well -- please excuse me while I NEVER play these odds.   

But I guess, looking back at this entire thread, it's not like I'm saying any of you are wrong... the numbers are what they are.  It's just THAT much HP for only that SMALL of a chance at a 5*.  It's unjustifiable, IMO.

HOWEVER... this did bring one small thought to the surface, that perhaps puts my mind at ease.  This is more-than-likely something set up by D3 to allow "whales" to buy up a character they want (since HP is purchasable), while the rest of the player-base calmly waits for the CP store, maybe picking at the 100 HP discounted offer per day, if they spend at all.

What I'm getting at is if the odds were better in these HP stores, then MPQ might become more like a pay-to-win than a lot of us would like.  Sure, we could always buy stuff -- but better odds in a HP store like this would motivate some and alienate others over time, I believe.

Anon

Not gameplay related, but one of my alliance members mentioned he would be having chemo done and would come back when he can. Today marks a whole year since that post and he never came back.

Hope you managed to beat it Toney. 😥

sambrookjm

Anon said:

Not gameplay related, but one of my alliance members mentioned he would be having chemo done and would come back when he can. Today marks a whole year since that post and he never came back.

Hope you managed to beat it Toney. 😥

Oof.  That's way beyond worst moment in MPQ...  That's something that actually matters in the grand scheme of things.

sambrookjm

Used Maria Hill and Agent Coulson in the Agents of SHIELD match in 12 Days of Marvel.  Coulson ticked her Best of Intentions counter down to zero, and killed her off.

shardwick

Anon said:

Not gameplay related, but one of my alliance members mentioned he would be having chemo done and would come back when he can. Today marks a whole year since that post and he never came back.

Hope you managed to beat it Toney. 😥

Here's hoping that he did beat it and then maybe life got in the way so he opted to focus on other things instead of MPQ. 

@PiMacleod I think it's interesting what we put value in. I would scoff at spending that much hp on a 40 pack for a shot at a new 5* cover yet don't have a problem spending around 4k in hp for a fifth of a 5* cover. Still waiting for the game to offer Apoc in the shard store. Haven't seen him in probably three or four weeks now and I didn't get him back then because I was focusing on iHulk.

DeNappa

40 legendary pulls, 1 5*. Sigh, I hate this game. Merry X-mas?

PiMacleod

Anon said:

Not gameplay related, but one of my alliance members mentioned he would be having chemo done and would come back when he can. Today marks a whole year since that post and he never came back.

Hope you managed to beat it Toney. 😥

I don't know Toney.  But I'm now more concerned with this post than any meta, new character release, etc.  I hope he beat it too... and maybe, just maybe his values changed to the point where a mobile game just isn't on that list anymore.  We can only hope.

Akoni

Anon said:

Not gameplay related, but one of my alliance members mentioned he would be having chemo done and would come back when he can. Today marks a whole year since that post and he never came back.

Hope you managed to beat it Toney. 😥

I hope Toney beat it, too. Cancer scares the *bleep* out of me. I know two people who died of cancer and many more who won their battle. Eventually cancer wins the war, but every extra day we can get is priceless.

Mr_F

Wasp Lover trying to use her 1 vs 1 PVP.

After 5 initial win, I started to loose.  To everyone. Because Doom healed every round. Or prof X match 4 every other turn. Or best: if lv 290 Chavez can beat lv 390 Wasp with easy then something is not right.

Ya may ask how? Match 4 spamming giving Chavez to many AP. Not the 1st time: maybe its just confirmation bias, but it happens to me every time I face her.

So, I hapilly join @OJSP with "Conspiracy Group". My hypothesis: Chavez has encoded increased match 4 rate. Now I have 5 characters to skip as match as possible - because its freaking flipcoin match.

Prof X, Polaris, Karnak and Chavez. And Legion. Yes, worst 4* on so many categories that I will use him AFTER all other options are downed and I have 0 healthpacks. He is SO BAD for me that I will Talos and Emma before him.

Just to clarify: Legion being worst 4* for me is not based on his usefullness. Its literally everything else that made me to dispise him.

Dogface

Also under confirmation bias: I swear that everytime I face Pr5f X, he gets all the match-4's and match-5's, but whenever I play him nothing. 

helix72

Dogface said:

Also under confirmation bias: I swear that everytime I face Pr5f X, he gets all the match-4's and match-5's, but whenever I play him nothing. 

I confirm I experience the same

sambrookjm

helix72 said:

Dogface said:

Also under confirmation bias: I swear that everytime I face Pr5f X, he gets all the match-4's and match-5's, but whenever I play him nothing. 

I confirm I experience the same

Exhibit A was the 1-v-1 PVP event.  Prof X was a popular choice, especially since there are no team-ups so that increases the likelihood of a match-4.  It was definitely borne out when I was playing against him in several matches in a row where they got multiple match-4 cascades on the same turn.  The record was six...which took out Apoc in no time flat.

I try my luck using him...I got a single match-4 in four matches, losing two of them.

sambrookjm

Hitting the 300 roster spot barrier, so now my additional characters cost 2000 HP instead of 1000.  Lovely.

HoundofShadow

Where did those additional slots go to?

1-star: 1
2-star: 14
3-star: 47 x 2 = 94
4-star: 107
5-star: 50
Total: 266

Dupe 4*?

helix72

For me:

1-star: 8x1 = 8 (1 max of each)
2-star: 14x2 = 28 (1 max champ of each plus 1 farm copy)
3-star: 47x3 = 141 (2 max champs of each plus 1 farm copy. Double max champs save on healthpacks in PvP)
4-star: 106x1 + 1 extra copy of Polaris =107 (needed to keep her from tanking Blue over Okoye and didn't want to throw away the covers). Once I start max champing 4*s I'll be adding dupes.
5-star: 51x1 =  (1 of each)

That's 335 with no dupe 4*s and D3 adding a new character roughly every 2 weeks.