Biased AI cascades, and matchmaking

shteev

Time and again we are told by the dev team that the AI does not cheat when making moves and generating cascades, and that there is no bias to the random factors in the game.

However, it would seem that the ongoing nature of the long-standing matchmaking problem in events calls this claim into question. The opponents which are picked for you in events show a marked and indeed obvious bias toward non-randomness. In the current event, for example, we in GoblinPile all seem to be playing a large number of games against Killwind, a rather formidable opponent... in previous rounds tho Ive been matched many, many times with less intimidating opponents like Quebecers.

How can we possibly be expected to believe the claim that any of the random number generators in the game are unbiased? As far as matchmaking goes, there would seem to be 2 possibilities:

1) Matchmaking is intended to be completely random, but through the incompetence of the dev team a bias has crept in, and may be present in other areas of the game that they are unaware of.
2) Matchmaking is intended to show a bias, and therefore an uneven playing field is deliberately being created for events, giving some competitors an easier chance to win prizes than others.

Which is it?

And which other supposedly 'random' elements of the game are in fact biased, or even controlled? Card draw in game? The drop rate for rares and mythics? The drop rate for duplicate rares and mythics?

Morphis

To be honest I am usually against such claims.
Random is random.

Nevertheless I notice the exact same thing and this also got in my head.
What keeps them from changing it?
What keeps them from adjusting numbers to artificially make competition "more balanced" in their view.

And more than anything: what keeps them from adjusting the overall game experience create a little frustration that can push you towards buying OP stuff at launch(like Olivia and kiora) to get a (even if for like a month only)big advantage?
We have no way to know for sure and, to be honest, asking to them is pointless unless they acknowledge it being result of a bug.
I doubt that could ever happen cause probably they would never investigate the issue and claim it's the usual "confirmation bias".

jimilinho_

I've played killwind 11 times in this event, Knife(something) 10 and Merseyside 8.

Safely say it's not even close to random

jimilinho_

Placeholder for a certain someone to tell us differently

paralistalon

I don't think the AI is getting better cascades...that's just people whining (do you know how much effort would be required to rig the match like that?). We do know that the drop rates for booster packs has been horrible, but that's nothing new. There's probably around a 20% chance of getting a rare in a pack... but there may be something to the RNG pulling duplicates just based on its programming. But matchmaking is clearly skewed in an odd way. I wouldn't be surprised if there was a secret matchmaking algorithm that pairs you based on a secret winrate percentage. They did that in Marvel PQ, and people would devote entire events to "tanking" their rating by intentionally losing hundreds of games in a row so they could be competitive the next event.

Tilikum

My brother is in a random public coalition and has been casually playing this event (high school chemistry teacher) and he says he's ran into Killwind a dozen times. I don't think the matchmaking is skewed as much as it's atrocious.

Also with cascades, has anyone seen th AI match a color out of its priority to chain react another match? I can't confirm it but I thought I've seen that once.

shteev

Morphis wrote:

Random is random

Pseudorandom isn't. It's quite easy for careless or inexperienced programmers to introduce bias into an RNG, usually by tinkering with it in an attempt to make it 'more random'.

paralistalon wrote:

(do you know how much effort would be required to rig the match like that?)

Funny you should ask! With my many years of coding experience, I can tell you: almost none. Calculating which color to make a gem so that when it drops in it causes a further match would take about as much time to code as writing this sentence. And, in fact, there's an even easier way to do it... just let a random gem drop in and then change the color of it after it does. You see this happen very frequently in MPQ. If I trusted the dev team, I might be swayed by the argument that this is merely a display bug and the gems don't actually change color, but I don't, so I'm not.

Foznertep

I suspect the AI to know which gems are about to drop based on one observation: usually, once I have cast a support, the opponent tries to make a match with it to remove a shield. When a match would be possible but it makes a different match, the newly added gems will trigger a cascade that takes at least one shield of the support away.
Same thing with matches that neither are the AI's nor my colour. At first I wonder why it would make that match, just to witness a glorious cascade a split second later, not with gems previously on the board, but new ones.

Corn_Noodles

What I've noticed in the last few weeks, usually when it comes to AI matches but I have seen something similar happen to my own matches when behind, is when a match is lopsided, cascades appear to be more frequent. At times I get the feeling that the game is trying to twist itself in a way to keep things closer at least in mana generation. Since I don't see what the AI pulls for cards unless cast, I have no idea if any kind of balancing is being applied to draws as well.

Ythalacuatha

And this frustrates the mess out of me.
I play claustrophobia, and it lands in a place like this:


Blue Blue White
White Red Blue

In this example claustrophobia lands on any of the blue, and is immediately popped by opponent.

but if the opponent plays claustrophobia or Hix prison guard or suppression bonds
it lands on the bottom row somewhere, and I actively try every strategy to pop it for ten turns.

I really REALLY wish ALL supports could land on a chosen spot.

They have that ability, because some cards have that ability. And in Marvel PQ Captain america had
that ability.

The unfair randomness of support drops have affected my play. I don't much like playing blue anymore,
and I rarely play with claustrophobia or suppression bonds.

Ythalacuatha

Cragger

The Killwind thing is kinda aggravating.

I think the "guest star" that we keep seeing shows up because they have purchased and maxed out the newest planeswalker. Maybe it makes everyone jealous?

Theros

Killwind and badger. I dont recall facing some1 different

Tilikum

MADAFAKA wrote:

Killwind and badger. I dont recall facing some1 different

Grats on Platinum! No wonder I haven't seen you in a while.

BigMao

You guys should calculate a p-value or something. Like, what are the chances that your matches are a result of random chance?

glggwp

unless they show me the code, i don't believe their "unbiased random" claim.

killwind

Hello guys,

I have also been trying to sort how I get paired up. I see Mersicide, SellerSloan, Quebecers, Largatha, Majicob, Shteev and BaDJer, regularly.

My coalition members are constantly complaining about running into me in events.

So lets try and baseline this. Let me give you my stats. I am an avid Quick Battle player, I constantly either outright win the QB or I place top 10 if im not feeling the grind.

On the quick battle side of the house I have 7400+ wins 940+ lost. Win percentage is 88.7%. I am Platinum tiered in all colors as of 2 weeks ago. I have mastered 333 of 679 cards.

All this data is so that we can find a common denominator as to how I am constantly being paired up with you and my coalition members.

killwind

Last piece of relevant information I have been playing since first release.

Ohboy

The matchmaking is obviously not random and everyone is getting the same opponents.

While that is still somewhat fair, it's lacking. I wish we could go back to more random opponents.

As for pseudo random Vs real random... Let me assure everyone here that it's not rocket science to ensure that for a small set of numbers generated and an unknown seed, there is no difference between the two.

EDHdad

There have been plenty of times when the AI had a great cascade just waiting for them in plain sight, and then they make some other match instead. I've seen this happen thousands of times. I mean, the AI playing Koth won't even match 3 red gems when there's a match-4 of some other color available. So I'm really skeptical that the AI has some magical intuition to make perfect matches with gems that aren't even on the board yet.

A more plausible explanation is that, when new gems are introduced to the board, they are statistically more likely than the existing gems to cause a match.

As for constantly getting paired with the same person, my belief is that you're probably getting paired with people who are on a similar level as you with respect to the event. So if you're grinding the event 1 minute after it starts and go on a 35 game win streak, you're probably getting paired with some other person who has also been grinding since the first minute and who is also on a 35 game win streak. Or some other factor related to the way your are playing the game (for example, a top 10 player might only be paired with other top 10 players). If so, then a person in the #1 coalition is probably destined to frequently be facing the same opponent. I doubt that you have an equal chance of facing the #500 player in an event that you have of facing the #5 player.

nexus13

These are probably two very different issues, only tied together by if they are supposed to be random processes or not. I definitely saw killiwind and badjer (pardon misspellings) multiple times in the last event. I didn't see shteev at all (thankfully), and we ended up next to each other on the final ladder. It doesn't seem to be power matched because I often saw people with wildly different ranks. I was suspecting killiwind might be a dev test account or something since so many people encountered it. Maybe it's an issue of how active a player is. All the QB keeps putting killiwind back to the front of the opponent pile or such?

bk1234

The Killwind thing is frustrating because people play him in both Platnium and Gold.

Another is Gunbunny who I played 38 times in the last ****.

Also I agree with the cascades. This is seen especially with Koth. AI cascades every turn. I have taken to exiting out of my match when any Koth comes up and adjusting my deck for manna denial.

When I play Koth I'm lucky to get a red match.

Like Killwind (and many others who have commented on this thread) I've been playing since the first release so I can see trends and this one has gotten progressively worse as the game has gotten more competitive.