Arena does have its own MMR implementation problem, or did a while ago and might still, but that was a whole separate problem.
TL;DR MMR in Arena is inflationary and your first #XXX games of modes using competitive MMR are just reaching the actual bell-curve/island most players are at, which created the ability for new players to reach mythic with the first random cards_I_own.dek consistently enough to make some grinders very very sad they can't even win with strong cards. This is one of the only "matchmaking is rigged" post I've ever seen backed by substantive datamined evidence.
The reason for having mmr not tied to rank is if you lose 8 cubes to a rank 100 player vs losing 8 cubes to a rank 90 player - your rank adjustment is capped, but they can adjust your mmr differently based on those two cases.
This just gives them another tool to use for matchmaking - which hopefully keeps you getting more matches that feel at your skill level.
The matchmaker for Snap seems to highly favor finding you a match quickly, which is probably why you get so many matches with wildly different rank. If you reque instantly - you get the same person all the time in Snap - so it either is barely adjusting your mmr over the last match you two had and thinks your still a good match; or doesn't care and just wants to match two ppl as fast as it can
PSN SeGaTai
+2
Options
Inquisitor772 x Penny Arcade Fight Club ChampionA fixed point in space and timeRegistered Userregular
Cube gain is analogous to chip/money gain in poker.
It is the definitive measure of skill as an outcome. It doesn't matter how many hands you win if you are losing money. Conversely, you can lose 90% of your hands and still come out on top in money, and if you do that over enough of a sample then you are a better player than someone who wins more games but doesn't earn as much money.
As I said, literally every game in existence has an MMR system as the "real" controller even with a rank system, and most of those games don't have cube management or whatever.
In any competitive game, your skill and ability to win is based on your opponent's play as well. That's what MMR is for. In theory, the system would shake out such that beating an opponent who has good cube management skills and thus higher MMR results in a relatively similar gain to beating out an opponent who has bad cube management skills and a lower MMR, the same way that me beating Kasparov would have a lot more impact on my FIDE rating than me beating my cats would.
I agree that your ability to win is based on your opponent's play, but I don't see how your skill is based on your opponent's play? What I'm saying here is exactly that cube gain provides a weird dynamic that seems to get in the way of assigning skill. Or at least, seems like (to me) a backwards way to do it. I am familiar Elo ratings and FIDE ratings, and the thing it doesn't do is make a differences based on whether your opponent concedes on turn 5 or turn 10 (the closest analogy I can think of to something like retreating). Like I could understand if they factored in when you snapped in your games (snapping early and winning would indicate to me a higher skill than snapping later and winning), but your cube gain is something different still (again, depending on whether your opponent accepts your snap or not) which is not something that you have (much) control over.
MMR should not care about what you do, it should care about outcomes. In most games, outcomes are binary. In Snap, there are 8 outcomes (discounting ties), and those should be weighted in some way. Snapping is not useful information on its own, but it is useful because it directly relates to the various levels of outcome you can have for a game.
milski on
I ate an engineer
0
Options
The JudgeThe Terwilliger CurvesRegistered Userregular
Yes, certainly, play your Galactus and Bucky and Venom and smile.
I'll just use the last reservation at that table for Hobgoblin and-- hey hey, come back!
Last pint: Turmoil CDA / Barley Brown's - Untappd: TheJudge_PDX
As I said, literally every game in existence has an MMR system as the "real" controller even with a rank system, and most of those games don't have cube management or whatever.
In any competitive game, your skill and ability to win is based on your opponent's play as well. That's what MMR is for. In theory, the system would shake out such that beating an opponent who has good cube management skills and thus higher MMR results in a relatively similar gain to beating out an opponent who has bad cube management skills and a lower MMR, the same way that me beating Kasparov would have a lot more impact on my FIDE rating than me beating my cats would.
I agree that your ability to win is based on your opponent's play, but I don't see how your skill is based on your opponent's play? What I'm saying here is exactly that cube gain provides a weird dynamic that seems to get in the way of assigning skill. Or at least, seems like (to me) a backwards way to do it. I am familiar Elo ratings and FIDE ratings, and the thing it doesn't do is make a differences based on whether your opponent concedes on turn 5 or turn 10 (the closest analogy I can think of to something like retreating). Like I could understand if they factored in when you snapped in your games (snapping early and winning would indicate to me a higher skill than snapping later and winning), but your cube gain is something different still (again, depending on whether your opponent accepts your snap or not) which is not something that you have (much) control over.
MMR should not care about what you do, it should care about outcomes. In most games, outcomes are binary. In Snap, there are 8 outcomes (discounting ties), and those should be weighted in some way. Snapping is not useful information on its own, but it is useful because it directly relates to the various levels of outcome you can have for a game.
You're being very narrow in what you're defining as an outcome. An outcome can be anything that is measurable, be it winning the game, winning all three lanes, snapping on turn 2 and losing on turn 5. All of these are outcomes, as well as "winning to a player who is 10 points higher in MMR" or "losing 8 cubes to an opponent on turn 4 to a player 100 points lower on MMR".
My point that I want to come back to is this (and I'm guessing at this point that you're going to disagree with me and that is fine, I don't want to use up more of the thread with us arguing around each other) these two examples:
1) Player 1 who realizes a winning path and snaps on turn 3, the opponent has a weak hand and retreats
vs.
2) Player 2 who realizes a winning path, and snaps on turn 3, the opponent decides to play out the game, Player 2 wins
(all other variables like MMR are the same between the two games)
In both cases, Player 1 and Player 2 have demonstrated the same amount of skill (at least how I'm defining skill, ability to recognize a favorable board and snap at the same time, and still win the game). But a system that rewards based on cube income will award Player 2 more skill than Player 1, because Player 2 won 4 cubes to Player 1's 1 cube.
Whatever system is used to measure skill, I feel as though both Players 1 and 2 should be rewarded the same amount of MMR/skill points/whatever you call it. I know that there are other games (like poker) that you can draw analogies to, and that's true, new games shouldn't need to design new systems from the ground up. But they shouldn't also be tied to the same systems that different games have, precisely because they are different. You can draw an analogy to cubes in snap and money in poker, and I agree its a pretty straight-forward analogy to make, but there are (putting it mildly) fundamental differences. Analogies are useful to draw inspiration from, but they can also ignore important differences as well.
The trouble with having an open mind, of course, is that people will insist on coming along and trying to put things in it. ~ Terry Pratchett
Like, my "off-meta" matches are seeing Thanos Lockjaw Zoo with Mojo and Kazaar instead of Thanos Lockjaw Tech Cards and seeing Zola Shuri or somebody who actually blew 6k tokens on Sauron
40s, I haven't been playing much/climbing fast this season.
Yeah, I don't want to be rude or anything and mixing up decks can definitely be successful but I do think there's a vast gulf between low ranks and like, 70-80+ in terms of how much you need to care about what decks are meta, especially right now where there are two clear top decks and both of them rely on pool 4+ cards.
Right, spool32 is talking about being stuck at 35, where it's not like that.
I mean, if you're going to walk me into saying the mean thing, if you're not at the ranks in which the meta decks are common and where you see a much wider variety of much weaker decks, you can either advance by playing a meta deck to get the free cube equity and/or you need to make a focused effort on improving gameplay and snap discipline. Shifting between non-meta decks is only going to be beneficial in the placebo sense that you might force yourself to take the game more seriously or learn the new deck instead of sticking to a non-functional script with the old one.
E: The third thing you can do is exploit bots harder, which admittedly a she hulk deck helps a lot with.
I mean you can just say Get Gud. I can take it!
0
Options
Inquisitor772 x Penny Arcade Fight Club ChampionA fixed point in space and timeRegistered Userregular
Player 1 and Player 2 both express their skill in snapping to increase their cube gain. This will hold true across a large enough sample. The more you snap when you are ahead, the more cubes you will gain, period. Thereby indicating that in both cases, these players will land at roughly the same MMR because they should face the same level of opponents with the same chance to retreat/play when appropriate over enough games.
The opponents are expressing their skill in whether they choose to retreat or play. Over a large enough sample, the opponent for Player 2 will lose more cubes than the opponent for Player 1, thereby creating the MMR difference between the two.
Your example only works if you assume an incredibly small sample size. There's a reason that statistics fundamentally require large samples - because otherwise the noise you are indicating will overwhelm the measurement.
This is going to sound condescending but I do not intend it in that way: the issues you are expressing indicate that you may want to brush up on the underlying math and theory before levying accusations of a broken system.
To put what Inquisitor said another way, you can make the same skill plays and have your MMR can go up or down depending on whether your opponent, say, plays a stock Shuri list and Aeros properly vs. attempting a points shootout into your Shang Chi. Fundamentally your MMR depends on your opponents in the short term but in the long term the skill sorting makes those individual impacts smooth out.
I ate an engineer
0
Options
ChaosHatHop, hop, hop, HA!Trick of the lightRegistered Userregular
As I said, literally every game in existence has an MMR system as the "real" controller even with a rank system, and most of those games don't have cube management or whatever.
In any competitive game, your skill and ability to win is based on your opponent's play as well. That's what MMR is for. In theory, the system would shake out such that beating an opponent who has good cube management skills and thus higher MMR results in a relatively similar gain to beating out an opponent who has bad cube management skills and a lower MMR, the same way that me beating Kasparov would have a lot more impact on my FIDE rating than me beating my cats would.
I agree that your ability to win is based on your opponent's play, but I don't see how your skill is based on your opponent's play? What I'm saying here is exactly that cube gain provides a weird dynamic that seems to get in the way of assigning skill. Or at least, seems like (to me) a backwards way to do it. I am familiar Elo ratings and FIDE ratings, and the thing it doesn't do is make a differences based on whether your opponent concedes on turn 5 or turn 10 (the closest analogy I can think of to something like retreating). Like I could understand if they factored in when you snapped in your games (snapping early and winning would indicate to me a higher skill than snapping later and winning), but your cube gain is something different still (again, depending on whether your opponent accepts your snap or not) which is not something that you have (much) control over.
MMR should not care about what you do, it should care about outcomes. In most games, outcomes are binary. In Snap, there are 8 outcomes (discounting ties), and those should be weighted in some way. Snapping is not useful information on its own, but it is useful because it directly relates to the various levels of outcome you can have for a game.
You're being very narrow in what you're defining as an outcome. An outcome can be anything that is measurable, be it winning the game, winning all three lanes, snapping on turn 2 and losing on turn 5. All of these are outcomes, as well as "winning to a player who is 10 points higher in MMR" or "losing 8 cubes to an opponent on turn 4 to a player 100 points lower on MMR".
My point that I want to come back to is this (and I'm guessing at this point that you're going to disagree with me and that is fine, I don't want to use up more of the thread with us arguing around each other) these two examples:
1) Player 1 who realizes a winning path and snaps on turn 3, the opponent has a weak hand and retreats
vs.
2) Player 2 who realizes a winning path, and snaps on turn 3, the opponent decides to play out the game, Player 2 wins
(all other variables like MMR are the same between the two games)
In both cases, Player 1 and Player 2 have demonstrated the same amount of skill (at least how I'm defining skill, ability to recognize a favorable board and snap at the same time, and still win the game). But a system that rewards based on cube income will award Player 2 more skill than Player 1, because Player 2 won 4 cubes to Player 1's 1 cube.
Whatever system is used to measure skill, I feel as though both Players 1 and 2 should be rewarded the same amount of MMR/skill points/whatever you call it. I know that there are other games (like poker) that you can draw analogies to, and that's true, new games shouldn't need to design new systems from the ground up. But they shouldn't also be tied to the same systems that different games have, precisely because they are different. You can draw an analogy to cubes in snap and money in poker, and I agree its a pretty straight-forward analogy to make, but there are (putting it mildly) fundamental differences. Analogies are useful to draw inspiration from, but they can also ignore important differences as well.
If player 1 and 2 are playing the exact same hand and the exact same opponent making the exact same decisions then yes, sucks to suck you got slightly less mmr increase because the opponent was jitterier/had a worse hand in the first example and decided to leave instead of playing it out.
The point is though, it's not meant to be "fair" on a per game scenario. If player 1 and player 2 are consistently equally as good at snapping and winning games, then they will have very very similar MMRs. It is meant to be fair in the long run.
As many people have said, it's not about how many cubes you won or lost after one hand. It's based on how many cubes you're net after many many many games.
It's kind of like sometimes your sports team gets unlucky and the ref makes a shit call at the end of the game. Yes that sucks. Over the course of many many games in a season though, you are what your record says you are.
While the devs are definitely tracking all these little things like what turn you snap on and cube result; they are not weighting the results differently on the back end to how it adjusts your mmr vs connecting your mmr to your cube gain.
The only thing a player can see is how his avg cube gain, so it would be absurd to reward them based on hidden weighting based on when they snapped.
Better analogy, if two players all played the same strength of opponents and obtained the same number of cubes, they should have the same mmr.
If they didn't, it means the devs have secretly determined the "correct"way to play that is in no way observable by the player
If we take two players but one always snaps and one always doesn't snap, but they both play to the same state of game completion against the same opponents, then they should probably have the same MMR but the doubling player will have twice as many or twice as few cubes as the never snapping player.
Or both zero cube gain.
But that implies a lack of any skill differential between the two players and the general player base.
....
Although the 2x /2 MMR makes better intuitive sense if the players were consciously choosing the x2 or /2 strategy.
Like, my "off-meta" matches are seeing Thanos Lockjaw Zoo with Mojo and Kazaar instead of Thanos Lockjaw Tech Cards and seeing Zola Shuri or somebody who actually blew 6k tokens on Sauron
40s, I haven't been playing much/climbing fast this season.
Yeah, I don't want to be rude or anything and mixing up decks can definitely be successful but I do think there's a vast gulf between low ranks and like, 70-80+ in terms of how much you need to care about what decks are meta, especially right now where there are two clear top decks and both of them rely on pool 4+ cards.
Right, spool32 is talking about being stuck at 35, where it's not like that.
I mean, if you're going to walk me into saying the mean thing, if you're not at the ranks in which the meta decks are common and where you see a much wider variety of much weaker decks, you can either advance by playing a meta deck to get the free cube equity and/or you need to make a focused effort on improving gameplay and snap discipline. Shifting between non-meta decks is only going to be beneficial in the placebo sense that you might force yourself to take the game more seriously or learn the new deck instead of sticking to a non-functional script with the old one.
E: The third thing you can do is exploit bots harder, which admittedly a she hulk deck helps a lot with.
I mean you can just say Get Gud. I can take it!
To be clear, a big part of what I'm saying isn't even necessarily "getting gud" in the classical sense, it's just forcing yourself to take the game extremely seriously whenever you play it; that's kind of a skill, but when there's no meaningful unranked mode you can probably advance a good bit by just... not playing any games for fun or for dailies without forcing yourself to enter Active Serious Laddering Mode. Like, I've ranked up relatively slowly this season and it's far less because I've suddenly gotten bad or made a bunch of misplays (as far as I can tell), it's because most of my games have been idling launching a match and not aggressively snapping on Shuri hands and playing out weak games against non-Thanos decks to get some variety in, so I'm rolling in 2 cube losses and 1 cube wins to a very slow 86 so far. Disciplined laddering is a very different style of play for me and with how dogshit the meta is if I'm gonna try hard it's going to be on Theathrythm
Hee hee, go ahead and play Galactus on 6 to get 40 power worth of Nimrod on the only location…with priority. I don’t mind, I’ll just drop a Shang-Chi into this obvious play.
And my reward for hitting 60? Pixel Morph! Maybe they should fix the reward variants to have better odds for higher-tier variants…
0
Options
ChaosHatHop, hop, hop, HA!Trick of the lightRegistered Userregular
If we take two players but one always snaps and one always doesn't snap, but they both play to the same state of game completion against the same opponents, then they should probably have the same MMR but the doubling player will have twice as many or twice as few cubes as the never snapping player.
Or both zero cube gain.
But that implies a lack of any skill differential between the two players and the general player base.
....
Although the 2x /2 MMR makes better intuitive sense if the players were consciously choosing the x2 or /2 strategy.
In the absence of some sort of sorting hat you can put on that will magically assign you rank, then yes, you will always be able to poke holes in basically any method of measuring skill. This is especially so for something like Snap where you have multiple axis upon which to grade skill, both win and cube rate. That doesn't mean that the method of factoring cube rate, somehow some way, isn't the correct way. It's also not to say these things are weighted equally, my guess is that they are not.
Even if the situation existed, the always snapper might be up or down more relatively at the end of the session but in the very very long run, they should be at the exact same point (except for free inflationary cubes at rank up) but their net cubes earned through game play should be the same. Let's imagine if I paid you $100 on Monday, but you owed Milski $100 on Friday. Let's say the same scenario happens the next week with RUNAWAY INFLATION. I might pay you $200 and you'll have more in your bank account on Tuesday, but after Friday you'll be at the same spot as past you. Same decisions, increased stakes, same outcome.
Obviously it's possible the always snapper can go on some run, hit a new rank, get the bonus 50 cubes and that player will end up having a different rank outcome at the end of the season. That's by design. The system is inflationary to help players rank up, get the carrot, and feel good about themselves. For most players in most situations, you're probably better off snapping all the time than never snapping.
I'm having some trouble with the "Win a location with only 1 card" season pass mission while still actually trying to win. Does anyone have advice or a decklist that's particularly good at that (Other than shuri because I know that's fairly good at it but I dont have shuri) that I can attempt it with?
M A G I K A Z A M
0
Options
Munkus BeaverYou don't have to attend every argument you are invited to.Philosophy: Stoicism. Politics: Democratic SocialistRegistered User, ClubPAregular
I'm having some trouble with the "Win a location with only 1 card" season pass mission while still actually trying to win. Does anyone have advice or a decklist that's particularly good at that (Other than shuri because I know that's fairly good at it but I dont have shuri) that I can attempt it with?
Warpath deck
Humor can be dissected as a frog can, but dies in the process.
I'm having some trouble with the "Win a location with only 1 card" season pass mission while still actually trying to win. Does anyone have advice or a decklist that's particularly good at that (Other than shuri because I know that's fairly good at it but I dont have shuri) that I can attempt it with?
I've been making decent progress on it with Magik/She-Hulk/Infinaut (though the current meta seems kind of hostile to it) or Venom/Zola. Warpath, as Munkus Beaver notes, is also inherently solid for this.
But if you're OK with losing a few cubes on the way, make a deck with Warpath and a few of the big bois with huge downsides like Ebony Maw/Maximus/Attuma, lead in a couple locations after 4, and retreat on 5.
looking up that Magik/She-Hulk/Infinaut deck that kinda seems fun. I have not been in a place where I actually ever see the current meta (I think I've seen like...one shuri deck period) so hopefully that isn't a problem.
Daredevil with Chuck to lock down a low power lane?
0
Options
ObiFettUse the ForceAs You WishRegistered Userregular
So.. Lets say you have Murderworld (destroy cards here at the end of turn 3) on the left. Klynta (no cards are revealed this turn) gets revealed on the right (turn 3).
Its turn 3, cards won't be revealed until the end of turn 4.
Pop Quiz hotshot, do cards played at Murderworld this turn get destroyed at the end of the turn or not?
So.. Lets say you have Murderworld (destroy cards here at the end of turn 3) on the left. Klynta (no cards are revealed this turn) gets revealed on the right (turn 3).
Its turn 3, cards won't be revealed until the end of turn 4.
Pop Quiz hotshot, do cards played at Murderworld this turn get destroyed at the end of the turn or not?
Yes, because I tried that very thing. Murderworld will flip your card and kill it.
I would download a car.
+6
Options
ObiFettUse the ForceAs You WishRegistered Userregular
So.. Lets say you have Murderworld (destroy cards here at the end of turn 3) on the left. Klynta (no cards are revealed this turn) gets revealed on the right (turn 3).
Its turn 3, cards won't be revealed until the end of turn 4.
Pop Quiz hotshot, do cards played at Murderworld this turn get destroyed at the end of the turn or not?
Yes, because I tried that very thing. Murderworld will flip your card and kill it.
I thought it wouldn't! Because literally no other destroy effect does!
I heard a comment from Cozy saying it would be neat if you could "tip" your opponent a cube for a good game. I think that's a great idea if implemented properly.
You can only tip if you win the game
You can only tip 1 cube
You can only tip if you win a 4/8 cube game
You can only tip a few times per day
Would be kind of nice to get a "Your opponent tipped you a cube!" message after a close game.
Need a voice actor? Hire me at bengrayVO.com
Legends of Runeterra: MNCdover #moc
Switch ID: MNC Dover SW-1154-3107-1051 Steam ID Twitch Page
Hey, welcome to our first development roadmap update for 2023! We’ve been having a blast exploring the Savage Land, Quantum Realm, and time-traveling to the… Past? Future? We’re not sure. We’ve had some great seasons, launched a BUNCH of new Series 5 cards, balanced some metagames, and some of us even got to Infinite Rank. Whoa. But, we’re ready to talk about what’s next for MARVEL SNAP! 👀
Coming Soon
Items listed here are in the final phase of development, and these features are almost ready to release to players! We’re cleaning up, polishing art, and squashing bugs. Our Quality Assurance teams scrape over all the little details to ensure we have as smooth an experience for players as possible. Once features are “Coming Soon,” we can start to plan release dates, timelines and share some insight with players.
The team is hard at work putting the finishing touches on a few much-anticipated features:
New Competitive Mode - Conquest
Card Acquisition Improvements
Token Shop Revamp
Ranked Mode Improvements
Infinite Rank Revamp
What is Conquest?
Conquest takes the Battle Mode format you know and love from Friendly Battles to the next level! In Conquest, players will queue up against other players to face-off in Battle Mode to earn rewards. Winning a series of battles without a loss unlocks tickets to higher leagues against tougher opponents… culminating with Infinity league at the end of each season. Players will battle their way to the top to claim ultimate victory – and the best rewards.
When you queue up for a Conquest match, you’ll start by choosing which league to compete: Proving Grounds, Bronze, Silver, Gold, Vibranium, or Infinite. Proving Grounds is always available to play, but players need the corresponding ticket to access higher tiers. Tickets are earned by winning a league series in the previous tier - for example: winning your Silver series will earn you a Gold Ticket. Players win a league series by defeating three consecutive opponents.
In addition to Tickets for higher leagues, each victory grants Medals that can be used in the Conquest Shop to unlock exclusive rewards. There’s a brand new Conquest Shop every season with new cosmetics, including a seasonal exclusive variant that can only be unlocked in the Conquest Shop.
We think that Battle Mode is a super fun way to play MARVEL SNAP and Conquest lets you play Battle Mode whenever you want, against the best competition, for the best rewards! The team is actively developing Conquest and we ran our first company-wide playtest last week. We’re hoping to release Conquest in June!
Card Acquisition Improvements
Every new card released in MARVEL SNAP should be exciting and impactful. We want players to wonder how that card could unlock new possibilities for fun or to win matches--and be able to actually play out those dreams. We’ve heard your feedback that we’re not meeting these expectations right now. It's time to change that! We're increasing the rate that all players earn Tokens, including a huge boost for Series 3 players that brings them in-line with Series 4/5 players. These changes will give all players equal opportunities to unlock brand new cards for themselves--whether you're in Series 3 or beyond.
More Tokens: We’re increasing the amount of Tokens earned on the Collection Level Track for all players who have not yet collected all Series 4 and Series 5 Cards. Starting at Collection Level 500 players will be able to earn between 200-600 tokens in Collector’s Caches and Collector’s Reserves regardless of whether they have completed Series 3 or not. This is an increase of 4X for players who have not yet completed Series 3! Once a Card Series is complete, caches that would have contained a card from that Series will now be replaced with 100 Tokens. All of these changes will result in more tokens for more people! 🎉
Choose Your Card: Series 3 cards now have their own section in the Shop! Once per season players who haven’t completed Series 3 will be able to claim one free Series 3 card to help them fill out their favorite deck and take a step closer to completing their collection. Now you can save your tokens for Series 4 and 5 cards! You don’t need to wait long for these updates, both of these changes will be in our next patch!
Token Shop Revamp
Token Shop is an awesome way to target the cards you want most, but we think we can do it even better. We’re dividing the Token Shop into a few different sections that will allow players to target new and desirable cards faster. We’re also revamping the look and feel of the Token Shop so you can swipe through to see what’s new.
Weekly Spotlight: When a new Series 5 card is released, it will be immediately featured in its own section for its first week in MARVEL SNAP. Once you buy the Weekly Spotlight card, this section will be hidden until the next new card is released. If you want the newest cards at release, here’s what you’ve been waiting for!
Series 4 & Series 5: This section functions nearly the same as previously, but only features Series 4 and Series 5 cards. This section will feature a card you don’t own from all currently available Series 4/5 cards and rotates every 8 hours. Series 5 cards will be added to this section once they leave the Weekly Spotlight.
Ultimate Variants: Ultimate Variants are some of the most epic variants in the game. We felt they deserved their own section too! We’re moving Ultimate Variants out of the way of seeing Series 4 and 5 cards and into their own rotation for those of you who want to make your decks that much cooler.
We’re aiming for these changes to make our April patch!
Ranked Mode Improvements
We’ve been hearing your feedback about Ranked Mode and we agree it’s time to make some changes. Ranked Mode is a tough feature to get right and we want to find the right balance through some additional iteration. We’re making a few updates before our next season that will address issues with some players feeling unable to progress or players that feel they are not facing worthy opponents.
Matchmaking Algorithm: We’re improving our matchmaking algorithm to create higher quality matches for all players–including an update that will prevent players from matching against an opponent more than 30 ranks away.
Infinite MMR Floor: When someone reaches Infinite Rank, we’ll take a snapshot of their MMR (Matchmaking Rating). For the rest of the season, their MMR cannot fall below that value. This change is to address players that could intentionally lose many games at the Infinite Rank floor to drop their MMR for the next season. You’ll still be able to increase your MMR while playing in Infinite Rank.
Infinite vs Infinite Matchmaking: Players that reach Infinite Rank will only match against other Infinite Rank players.
We know there’s more work to be done. We think the current journey to re-climb to a previous season’s rank is too difficult. To that end, we are exploring changes to how many cubes are required per rank, how many bonus cubes are granted when a player achieves a new Tier, and potentially how much of a reset is incurred each season. These numbers are actively being crunched right now, so we don’t have more details to share, but rest assured, our goal is to deliver a satisfying seasonal experience for players who wish to get better at the game, and see their rank increase as a result.
We’ve also seen a lot of feedback around matching by MMR and why we don’t simply match by Rank. The team has been weighing the positives and negatives of both these options. In a pure match-by-rank situation, it can lead to an awful start-of-season experience, as players are matching up against opponents who are far superior to them. However, we are exploring alternate ways of matchmaking to try to deliver a more satisfying overall experience, as there are some upsides to matching-by-rank. We have some ideas that we think can capture the best of both worlds, and will be trying out updates to the matchmaker in future seasons. We’ll absolutely be sharing more as we lock in details–you’ll hear more from us soon!
Infinite Rank Revamp
For most of us, the climb to Infinite Rank has been our peak achievement… and we have great news! We’re completely revamping Infinite Rank so your experience doesn’t end at Infinite - it’s just the beginning! We want to create an exciting and fair battleground for our most skilled players to compete on the Infinite Rank Leaderboard. Once players reach Infinite Rank, they are separated into their own ecosystem that will highlight two important numbers:
Skill Rating Score: This score is your personal skill rating score; it’s similar to your matchmaking rating. You’ll be able to see how your skill rating score changes after each match. If you’re more interested in increasing your personal skill from season to season, this is the number for you.
Leaderboard Rank: Your friend has always claimed that they’re the Best MARVEL SNAP Player - well, now we’ll find out. This is the leaderboard number based on the ranking order of each Infinite player’s skill rating score. We’re fine-tuning the experience and visuals for the new Infinite Rank ecosystem and currently scoping out the work, but we’re hoping to release it as soon as possible! More details on timing soon!
In Development
Game features in this section are actively being developed, iterated, and improved upon. Our production team dedicates team resources, generates tasks/timelines, and ensures we’re moving forward. Features and systems in this bucket can change dramatically – we might find our initial idea wasn’t fun OR we might discover we can achieve our goals in an unexpected way!
Here’s some features that are currently in development:
PC Widescreen UI
Smart Decks
Avatars & Titles by Deck
Personalized Shop
Global Matchmaking
In Concept
We have BIG dreams about what we’d like to bring to the future of MARVEL SNAP! The first step towards these dreams is turning ideas into concepts through discussion, documentation, and planning. Features currently in the “In Concept” phase have been initially scoped out and given shape but have not yet had any engineering or development work done in the game itself.
This is where we’d love your feedback the most! Concepts listed in this section may be elevated in priority, rescoped, or scrapped entirely… or we might hear something completely new from the community!
Here are some of the future concepts we’re thinking about for MARVEL SNAP:
Guilds (Social Systems)
Collectible Emotes & Card Emojis
Mythic Variants
PC Controller Support
Seasonal Audio
Test Deck Mode
Test Deck Mode
Where’s Unranked Mode? What’s this Test Deck Mode?! We know that players want a quick-and-easy way to try out new decks or strategies. Since before Global Launch, we’ve been talking about the idea of Unranked Mode as a solution to this problem. But, as we explored this feature, we found that Unranked Mode comes with some not so great additional challenges. So, we set out to reimagine what this feature could be and Test Deck Mode evolved from there! We’re considering adding a button to the deck editing screen that will let you instantly hop into a low-stakes game against the computer. We’re still in early stages of concepting here, but we’d love to hear your thoughts on this new path.
Let us know what you think, and please continue to share your ideas for the future of MARVEL SNAP!
MNC Dover on
Need a voice actor? Hire me at bengrayVO.com
Legends of Runeterra: MNCdover #moc
Switch ID: MNC Dover SW-1154-3107-1051 Steam ID Twitch Page
looking up that Magik/She-Hulk/Infinaut deck that kinda seems fun. I have not been in a place where I actually ever see the current meta (I think I've seen like...one shuri deck period) so hopefully that isn't a problem.
Well, I'm about +40 cubes since I started trying it so here's hoping it lasts! Back to rank 34.6 in short order.
Posts
TL;DR MMR in Arena is inflationary and your first #XXX games of modes using competitive MMR are just reaching the actual bell-curve/island most players are at, which created the ability for new players to reach mythic with the first random cards_I_own.dek consistently enough to make some grinders very very sad they can't even win with strong cards. This is one of the only "matchmaking is rigged" post I've ever seen backed by substantive datamined evidence.
Kinda funny how many wins I'm racking up with Wave > Agatha > Taskmaster > Aero.
Simple, easy to remember.
Twitch: akThera
Steam: Thera
This just gives them another tool to use for matchmaking - which hopefully keeps you getting more matches that feel at your skill level.
The matchmaker for Snap seems to highly favor finding you a match quickly, which is probably why you get so many matches with wildly different rank. If you reque instantly - you get the same person all the time in Snap - so it either is barely adjusting your mmr over the last match you two had and thinks your still a good match; or doesn't care and just wants to match two ppl as fast as it can
It is the definitive measure of skill as an outcome. It doesn't matter how many hands you win if you are losing money. Conversely, you can lose 90% of your hands and still come out on top in money, and if you do that over enough of a sample then you are a better player than someone who wins more games but doesn't earn as much money.
MMR should not care about what you do, it should care about outcomes. In most games, outcomes are binary. In Snap, there are 8 outcomes (discounting ties), and those should be weighted in some way. Snapping is not useful information on its own, but it is useful because it directly relates to the various levels of outcome you can have for a game.
I'll just use the last reservation at that table for Hobgoblin and-- hey hey, come back!
You're being very narrow in what you're defining as an outcome. An outcome can be anything that is measurable, be it winning the game, winning all three lanes, snapping on turn 2 and losing on turn 5. All of these are outcomes, as well as "winning to a player who is 10 points higher in MMR" or "losing 8 cubes to an opponent on turn 4 to a player 100 points lower on MMR".
My point that I want to come back to is this (and I'm guessing at this point that you're going to disagree with me and that is fine, I don't want to use up more of the thread with us arguing around each other) these two examples:
1) Player 1 who realizes a winning path and snaps on turn 3, the opponent has a weak hand and retreats
vs.
2) Player 2 who realizes a winning path, and snaps on turn 3, the opponent decides to play out the game, Player 2 wins
(all other variables like MMR are the same between the two games)
In both cases, Player 1 and Player 2 have demonstrated the same amount of skill (at least how I'm defining skill, ability to recognize a favorable board and snap at the same time, and still win the game). But a system that rewards based on cube income will award Player 2 more skill than Player 1, because Player 2 won 4 cubes to Player 1's 1 cube.
Whatever system is used to measure skill, I feel as though both Players 1 and 2 should be rewarded the same amount of MMR/skill points/whatever you call it. I know that there are other games (like poker) that you can draw analogies to, and that's true, new games shouldn't need to design new systems from the ground up. But they shouldn't also be tied to the same systems that different games have, precisely because they are different. You can draw an analogy to cubes in snap and money in poker, and I agree its a pretty straight-forward analogy to make, but there are (putting it mildly) fundamental differences. Analogies are useful to draw inspiration from, but they can also ignore important differences as well.
I mean you can just say Get Gud. I can take it!
The opponents are expressing their skill in whether they choose to retreat or play. Over a large enough sample, the opponent for Player 2 will lose more cubes than the opponent for Player 1, thereby creating the MMR difference between the two.
Your example only works if you assume an incredibly small sample size. There's a reason that statistics fundamentally require large samples - because otherwise the noise you are indicating will overwhelm the measurement.
This is going to sound condescending but I do not intend it in that way: the issues you are expressing indicate that you may want to brush up on the underlying math and theory before levying accusations of a broken system.
If player 1 and 2 are playing the exact same hand and the exact same opponent making the exact same decisions then yes, sucks to suck you got slightly less mmr increase because the opponent was jitterier/had a worse hand in the first example and decided to leave instead of playing it out.
The point is though, it's not meant to be "fair" on a per game scenario. If player 1 and player 2 are consistently equally as good at snapping and winning games, then they will have very very similar MMRs. It is meant to be fair in the long run.
As many people have said, it's not about how many cubes you won or lost after one hand. It's based on how many cubes you're net after many many many games.
It's kind of like sometimes your sports team gets unlucky and the ref makes a shit call at the end of the game. Yes that sucks. Over the course of many many games in a season though, you are what your record says you are.
The only thing a player can see is how his avg cube gain, so it would be absurd to reward them based on hidden weighting based on when they snapped.
Better analogy, if two players all played the same strength of opponents and obtained the same number of cubes, they should have the same mmr.
If they didn't, it means the devs have secretly determined the "correct"way to play that is in no way observable by the player
If we take two players but one always snaps and one always doesn't snap, but they both play to the same state of game completion against the same opponents, then they should probably have the same MMR but the doubling player will have twice as many or twice as few cubes as the never snapping player.
Or both zero cube gain.
But that implies a lack of any skill differential between the two players and the general player base.
....
Although the 2x /2 MMR makes better intuitive sense if the players were consciously choosing the x2 or /2 strategy.
To be clear, a big part of what I'm saying isn't even necessarily "getting gud" in the classical sense, it's just forcing yourself to take the game extremely seriously whenever you play it; that's kind of a skill, but when there's no meaningful unranked mode you can probably advance a good bit by just... not playing any games for fun or for dailies without forcing yourself to enter Active Serious Laddering Mode. Like, I've ranked up relatively slowly this season and it's far less because I've suddenly gotten bad or made a bunch of misplays (as far as I can tell), it's because most of my games have been idling launching a match and not aggressively snapping on Shuri hands and playing out weak games against non-Thanos decks to get some variety in, so I'm rolling in 2 cube losses and 1 cube wins to a very slow 86 so far. Disciplined laddering is a very different style of play for me and with how dogshit the meta is if I'm gonna try hard it's going to be on Theathrythm
Twitch: akThera
Steam: Thera
I am serious. And don’t call me Shuri.
And my reward for hitting 60? Pixel Morph! Maybe they should fix the reward variants to have better odds for higher-tier variants…
In the absence of some sort of sorting hat you can put on that will magically assign you rank, then yes, you will always be able to poke holes in basically any method of measuring skill. This is especially so for something like Snap where you have multiple axis upon which to grade skill, both win and cube rate. That doesn't mean that the method of factoring cube rate, somehow some way, isn't the correct way. It's also not to say these things are weighted equally, my guess is that they are not.
Even if the situation existed, the always snapper might be up or down more relatively at the end of the session but in the very very long run, they should be at the exact same point (except for free inflationary cubes at rank up) but their net cubes earned through game play should be the same. Let's imagine if I paid you $100 on Monday, but you owed Milski $100 on Friday. Let's say the same scenario happens the next week with RUNAWAY INFLATION. I might pay you $200 and you'll have more in your bank account on Tuesday, but after Friday you'll be at the same spot as past you. Same decisions, increased stakes, same outcome.
Obviously it's possible the always snapper can go on some run, hit a new rank, get the bonus 50 cubes and that player will end up having a different rank outcome at the end of the season. That's by design. The system is inflationary to help players rank up, get the carrot, and feel good about themselves. For most players in most situations, you're probably better off snapping all the time than never snapping.
Warpath deck
I've been making decent progress on it with Magik/She-Hulk/Infinaut (though the current meta seems kind of hostile to it) or Venom/Zola. Warpath, as Munkus Beaver notes, is also inherently solid for this.
But if you're OK with losing a few cubes on the way, make a deck with Warpath and a few of the big bois with huge downsides like Ebony Maw/Maximus/Attuma, lead in a couple locations after 4, and retreat on 5.
Twitch: akThera
Steam: Thera
Its turn 3, cards won't be revealed until the end of turn 4.
Pop Quiz hotshot, do cards played at Murderworld this turn get destroyed at the end of the turn or not?
Yes, because I tried that very thing. Murderworld will flip your card and kill it.
I thought it wouldn't! Because literally no other destroy effect does!
I hit him with Enchantress the turn after Hazmat went off and all their cards dropped in power.
Wasn’t sure it worked that way.
ok, so one other destroy effect works like that
No card destroy effect does!
Twitch: akThera
Steam: Thera
And like Fisk's Tower, Kingpin will destroy unrevealed cards that move to his location on turn 6.
It’s me, I’m a recipient of this cube welfare.
3478 with Absorbing Man and Viper left. And Attuma. 1800 away from Thanos(?).
Would be kind of nice to get a "Your opponent tipped you a cube!" message after a close game.
Legends of Runeterra: MNCdover #moc
Switch ID: MNC Dover SW-1154-3107-1051
Steam ID
Twitch Page
Edit: Massive text dump from Reddit...
Hey, welcome to our first development roadmap update for 2023! We’ve been having a blast exploring the Savage Land, Quantum Realm, and time-traveling to the… Past? Future? We’re not sure. We’ve had some great seasons, launched a BUNCH of new Series 5 cards, balanced some metagames, and some of us even got to Infinite Rank. Whoa. But, we’re ready to talk about what’s next for MARVEL SNAP! 👀
Coming Soon
Items listed here are in the final phase of development, and these features are almost ready to release to players! We’re cleaning up, polishing art, and squashing bugs. Our Quality Assurance teams scrape over all the little details to ensure we have as smooth an experience for players as possible. Once features are “Coming Soon,” we can start to plan release dates, timelines and share some insight with players.
The team is hard at work putting the finishing touches on a few much-anticipated features:
New Competitive Mode - Conquest
Card Acquisition Improvements
Token Shop Revamp
Ranked Mode Improvements
Infinite Rank Revamp
What is Conquest?
Conquest takes the Battle Mode format you know and love from Friendly Battles to the next level! In Conquest, players will queue up against other players to face-off in Battle Mode to earn rewards. Winning a series of battles without a loss unlocks tickets to higher leagues against tougher opponents… culminating with Infinity league at the end of each season. Players will battle their way to the top to claim ultimate victory – and the best rewards.
When you queue up for a Conquest match, you’ll start by choosing which league to compete: Proving Grounds, Bronze, Silver, Gold, Vibranium, or Infinite. Proving Grounds is always available to play, but players need the corresponding ticket to access higher tiers. Tickets are earned by winning a league series in the previous tier - for example: winning your Silver series will earn you a Gold Ticket. Players win a league series by defeating three consecutive opponents.
In addition to Tickets for higher leagues, each victory grants Medals that can be used in the Conquest Shop to unlock exclusive rewards. There’s a brand new Conquest Shop every season with new cosmetics, including a seasonal exclusive variant that can only be unlocked in the Conquest Shop.
We think that Battle Mode is a super fun way to play MARVEL SNAP and Conquest lets you play Battle Mode whenever you want, against the best competition, for the best rewards! The team is actively developing Conquest and we ran our first company-wide playtest last week. We’re hoping to release Conquest in June!
Card Acquisition Improvements
Every new card released in MARVEL SNAP should be exciting and impactful. We want players to wonder how that card could unlock new possibilities for fun or to win matches--and be able to actually play out those dreams. We’ve heard your feedback that we’re not meeting these expectations right now. It's time to change that! We're increasing the rate that all players earn Tokens, including a huge boost for Series 3 players that brings them in-line with Series 4/5 players. These changes will give all players equal opportunities to unlock brand new cards for themselves--whether you're in Series 3 or beyond.
More Tokens: We’re increasing the amount of Tokens earned on the Collection Level Track for all players who have not yet collected all Series 4 and Series 5 Cards. Starting at Collection Level 500 players will be able to earn between 200-600 tokens in Collector’s Caches and Collector’s Reserves regardless of whether they have completed Series 3 or not. This is an increase of 4X for players who have not yet completed Series 3! Once a Card Series is complete, caches that would have contained a card from that Series will now be replaced with 100 Tokens. All of these changes will result in more tokens for more people! 🎉
Choose Your Card: Series 3 cards now have their own section in the Shop! Once per season players who haven’t completed Series 3 will be able to claim one free Series 3 card to help them fill out their favorite deck and take a step closer to completing their collection. Now you can save your tokens for Series 4 and 5 cards! You don’t need to wait long for these updates, both of these changes will be in our next patch!
Token Shop Revamp
Token Shop is an awesome way to target the cards you want most, but we think we can do it even better. We’re dividing the Token Shop into a few different sections that will allow players to target new and desirable cards faster. We’re also revamping the look and feel of the Token Shop so you can swipe through to see what’s new.
Weekly Spotlight: When a new Series 5 card is released, it will be immediately featured in its own section for its first week in MARVEL SNAP. Once you buy the Weekly Spotlight card, this section will be hidden until the next new card is released. If you want the newest cards at release, here’s what you’ve been waiting for!
Series 4 & Series 5: This section functions nearly the same as previously, but only features Series 4 and Series 5 cards. This section will feature a card you don’t own from all currently available Series 4/5 cards and rotates every 8 hours. Series 5 cards will be added to this section once they leave the Weekly Spotlight.
Ultimate Variants: Ultimate Variants are some of the most epic variants in the game. We felt they deserved their own section too! We’re moving Ultimate Variants out of the way of seeing Series 4 and 5 cards and into their own rotation for those of you who want to make your decks that much cooler.
We’re aiming for these changes to make our April patch!
Ranked Mode Improvements
We’ve been hearing your feedback about Ranked Mode and we agree it’s time to make some changes. Ranked Mode is a tough feature to get right and we want to find the right balance through some additional iteration. We’re making a few updates before our next season that will address issues with some players feeling unable to progress or players that feel they are not facing worthy opponents.
Matchmaking Algorithm: We’re improving our matchmaking algorithm to create higher quality matches for all players–including an update that will prevent players from matching against an opponent more than 30 ranks away.
Infinite MMR Floor: When someone reaches Infinite Rank, we’ll take a snapshot of their MMR (Matchmaking Rating). For the rest of the season, their MMR cannot fall below that value. This change is to address players that could intentionally lose many games at the Infinite Rank floor to drop their MMR for the next season. You’ll still be able to increase your MMR while playing in Infinite Rank.
Infinite vs Infinite Matchmaking: Players that reach Infinite Rank will only match against other Infinite Rank players.
We know there’s more work to be done. We think the current journey to re-climb to a previous season’s rank is too difficult. To that end, we are exploring changes to how many cubes are required per rank, how many bonus cubes are granted when a player achieves a new Tier, and potentially how much of a reset is incurred each season. These numbers are actively being crunched right now, so we don’t have more details to share, but rest assured, our goal is to deliver a satisfying seasonal experience for players who wish to get better at the game, and see their rank increase as a result.
We’ve also seen a lot of feedback around matching by MMR and why we don’t simply match by Rank. The team has been weighing the positives and negatives of both these options. In a pure match-by-rank situation, it can lead to an awful start-of-season experience, as players are matching up against opponents who are far superior to them. However, we are exploring alternate ways of matchmaking to try to deliver a more satisfying overall experience, as there are some upsides to matching-by-rank. We have some ideas that we think can capture the best of both worlds, and will be trying out updates to the matchmaker in future seasons. We’ll absolutely be sharing more as we lock in details–you’ll hear more from us soon!
Infinite Rank Revamp
For most of us, the climb to Infinite Rank has been our peak achievement… and we have great news! We’re completely revamping Infinite Rank so your experience doesn’t end at Infinite - it’s just the beginning! We want to create an exciting and fair battleground for our most skilled players to compete on the Infinite Rank Leaderboard. Once players reach Infinite Rank, they are separated into their own ecosystem that will highlight two important numbers:
Skill Rating Score: This score is your personal skill rating score; it’s similar to your matchmaking rating. You’ll be able to see how your skill rating score changes after each match. If you’re more interested in increasing your personal skill from season to season, this is the number for you.
Leaderboard Rank: Your friend has always claimed that they’re the Best MARVEL SNAP Player - well, now we’ll find out. This is the leaderboard number based on the ranking order of each Infinite player’s skill rating score. We’re fine-tuning the experience and visuals for the new Infinite Rank ecosystem and currently scoping out the work, but we’re hoping to release it as soon as possible! More details on timing soon!
In Development
Game features in this section are actively being developed, iterated, and improved upon. Our production team dedicates team resources, generates tasks/timelines, and ensures we’re moving forward. Features and systems in this bucket can change dramatically – we might find our initial idea wasn’t fun OR we might discover we can achieve our goals in an unexpected way!
Here’s some features that are currently in development:
PC Widescreen UI
Smart Decks
Avatars & Titles by Deck
Personalized Shop
Global Matchmaking
In Concept
We have BIG dreams about what we’d like to bring to the future of MARVEL SNAP! The first step towards these dreams is turning ideas into concepts through discussion, documentation, and planning. Features currently in the “In Concept” phase have been initially scoped out and given shape but have not yet had any engineering or development work done in the game itself.
This is where we’d love your feedback the most! Concepts listed in this section may be elevated in priority, rescoped, or scrapped entirely… or we might hear something completely new from the community!
Here are some of the future concepts we’re thinking about for MARVEL SNAP:
Guilds (Social Systems)
Collectible Emotes & Card Emojis
Mythic Variants
PC Controller Support
Seasonal Audio
Test Deck Mode
Test Deck Mode
Where’s Unranked Mode? What’s this Test Deck Mode?! We know that players want a quick-and-easy way to try out new decks or strategies. Since before Global Launch, we’ve been talking about the idea of Unranked Mode as a solution to this problem. But, as we explored this feature, we found that Unranked Mode comes with some not so great additional challenges. So, we set out to reimagine what this feature could be and Test Deck Mode evolved from there! We’re considering adding a button to the deck editing screen that will let you instantly hop into a low-stakes game against the computer. We’re still in early stages of concepting here, but we’d love to hear your thoughts on this new path.
Let us know what you think, and please continue to share your ideas for the future of MARVEL SNAP!
Legends of Runeterra: MNCdover #moc
Switch ID: MNC Dover SW-1154-3107-1051
Steam ID
Twitch Page
Well, I'm about +40 cubes since I started trying it so here's hoping it lasts! Back to rank 34.6 in short order.
So, it’s NOT doing that now. Confirming what I said above, that Infinite players can and are getting matched with anyone.
Twitch: akThera
Steam: Thera