Install Steam
login
|
language
简体中文 (Simplified Chinese)
繁體中文 (Traditional Chinese)
日本語 (Japanese)
한국어 (Korean)
ไทย (Thai)
Български (Bulgarian)
Čeština (Czech)
Dansk (Danish)
Deutsch (German)
Español - España (Spanish - Spain)
Español - Latinoamérica (Spanish - Latin America)
Ελληνικά (Greek)
Français (French)
Italiano (Italian)
Bahasa Indonesia (Indonesian)
Magyar (Hungarian)
Nederlands (Dutch)
Norsk (Norwegian)
Polski (Polish)
Português (Portuguese - Portugal)
Português - Brasil (Portuguese - Brazil)
Română (Romanian)
Русский (Russian)
Suomi (Finnish)
Svenska (Swedish)
Türkçe (Turkish)
Tiếng Việt (Vietnamese)
Українська (Ukrainian)
Report a translation problem
Your ego is saved but you won't be growing as a player.
How long have you been playing the game? What decks do you use?
Please educate yourself...
- Here is the Ruleset Schema Second Dinner uses as part of their AWS gamelift service, Flexmatch...
https://docs.aws.amazon.com/gamelift/latest/flexmatchguide/match-ruleset-schema.html
- Here are the property definitions...
https://docs.aws.amazon.com/gamelift/latest/flexmatchguide/match-ruleset-property-definitions.html
- Here is the Ruleset restriction, denoting they can't be edited whilst in production, and the region limit of rulsets...
https://docs.aws.amazon.com/gamelift/latest/flexmatchguide/match-ruleset-schema.html
From the above, it's clear deck based matchmaking, whilst been entirely plausible in its potential to be implemented, is impossible to coexist alongside actual player attributes of CL, MMR and Rank. And also find you an opponent within the current queue time frames as consistently as people describe it to be. To argue that deck based matchmaking exists is to also argue that MMR, CL and Rank play little no factor in determining your opponent. As in order to find you an opponent with the added scrutiny of comparing the cards in each other's decks through specific strings or values: They are entirely redundant metrics to compare based on wait distance limitations put in place before you are matched with bots after a search for an adequate player fails.
Also with the inability to edit a Ruleset whilst in use, SD must have a near omniscient magical ability to predict every card that they will ever create, their future OTA's and patch changes, and their eventual counters to keep this "deck based matchmaking" working flawlessly for almost 2 Years. Suffice to say, this isn't the case.
- Here is an example tutorial if you want to check out the basics...
https://www.youtube.com/watch?v=LaR1CWictb4
- Here is an incident in 2023 explaining a perfect oppurtunity to implement a prevention in matchup that never occured (despite the presumed ability to do so)...
https://www.dropbox.com/scl/fo/e9ys40dix4gmqp0ha2dqt/AD_VIqMGjto1gAESKKfNjwg?rlkey=g7pjmavx7b1gtmwyq9qhtsjw4&st=6monhyom&dl=0
- Here are two tests of small sample sizes, but attempts none the less at actually compiling data...
https://www.reddit.com/r/MarvelSnap/comments/10l08lb/mythbusters_does_matchmaking_care_what_deck_you/
https://www.reddit.com/r/MarvelSnap/comments/15un9rs/ive_begun_documenting_my_matches_to_see_if/
- Here are dev responses to questions on matchmaking...
https://www.dropbox.com/scl/fo/ls2lvgjpgq66plj37mnka/AH3ww1jxYSohZe9zrQgm80A?rlkey=oot8bmfmyqohn3mfg2a6fxzfd&st=wcfjhkul&dl=0
- Here is Ben Brodes response on the notion of Deck Based Matchmaking and the challenges of even attempting it...
https://www.youtube.com/watch?v=coDDgWaVlKY&t=5327s
Pity really, because otherwise this would be a fun game to waste some money on, but as it is I wouldn't even contribute my spare changes to this.
Funny, how people don't read. And this post will be believed, while the above post, with references and detailed explanations of why this cannot be true, will be ignored.
It's not that deep man.
if SD truly wanted to adjust the play experience through matchmaking, they wouldn't have purchased an off the shelf solution and publicity advertise the AWS services they use.
I have never had this behaviour in other games. But then I don't usually play predatory mobile games.
For me its just been every deck having the perfect curve and locations. I can play the same deck and never get it, I.e. I'm seeing a lot of scream and agent venom and in both cases its those cards on 2 every turn. For me that's a rare event. Would be nice if some opponents missed and had bad draws like me but that never happens. Then it'd the locations, if I play junk I get locations where cards have + energy which hurts my deck. If my opponents have s Wong deck it'd almost certainly going to be onslaughts citadel or kamar-taj. There is no way through pure chance that 20 opponents in a row could have locations favourable to them but its legitimately happening.
Why is your account been singled out but your opponents seemingly have benefits?
What is the differential here?
I was playing the war machine/storm/legion deck.
Turn 1 I realize I'm against discard. I already know for a fact, they will play Gambit and it will take out my war machine.
Turn 4, I play War Machine. Without any doubt, I know what's coming next turn.
Turn 5 I play Legion, they play....Gambit, and out of all the cards on the board, Gambit hits....War machine.
Amazing. Except I knew that was going to happen 4 turns earlier.
That's just one example out of thousands now. The game is basically on rails, it essentially runs on a script.
Recently I had one where I had 6 cards in hand and Sakaar is going to put one down and I knew it would be Kang, and it was...that was a 1 in 6. Then the turn restarts and now its a 1 in 5 and I know its putting down Mystique this time and it does. So added together those odds are low yet there was never a single doubt in my mind.
Answer me if the game is "random" why Is every outcome entirely predictable?
But from listening to players here and elsewhere, people have similar experiences to me. Nobody comes forward and says "yes I'm that guy where every location goes my way"
So basically, its probably just bots.
I mean... doesn't that resonate any meaningful info to you?
You're discrediting evidence solely based on the fact that it doesn't exist.
That's a bias friend. You've assured yourself "it's on rails" based off of social media conjecture, and dismissed the notion it isn't because there aren't claims to state the opposite in a majority.
Then you call us dense for not seeing it, but back up your claims with "it's probably just bots"...
Comeon now.
For someone who claims it's rigged, you sure expect the results to be rigged in your favour. Your offered example clearly suggests you would have preferred a more favourable target for gambit to select from an array other than war machine... So by this logic, it sounds like your an advocate for entirely predictable outcomes with modified arrays that benefit one player over another.
This begs the question, why argue against it if you are wanting a more expected outcome that favours you over your opponent? It sounds like you actively want match ups that are rigged in favour.
I was referring to the post by Doovies. Let me repeat, there is no personal experience involved, just a disussion of known facts about the game, and what they mean for the possibility of deck-based matchmaking.
Also, no verdict on developer statements. I do think lying to customers is a no-no in any business, unless you are a scammer after a quick profit. You may withhold info, you may issue vague statements, but if you are caught in a lie, your credibility is gone, and so are the customers.