Table of Contents
Have you ever felt the urge to pick an avatar for your deck that matches its contents? It seems natural, considering you could assign a different avatar to each of your 20 decks. At the very least, matching the avatar to the deck makes it easier to track down the deck you want to play. Looking for your Discard deck? Just pick out the one with a Morbius, MODOK, or Hela avatar!
Today, I’m looking at how often players match their avatars, titles, and card backs to the decks they’re playing. The ability to customize your cosmetic loadout on a per-deck basis was introduced in an update on May 16th, earlier this year.
How often is your opponent spilling the beans without playing a single card?
The data I’m covering comes from matches recorded by our Marvel Snap Tracker. I looked at ranked matches from all players over the past week. I limited it to matches that ended on turn 6+, to avoid skewing the data too strongly in favor of lower-cost cards. Notably, I’m looking at the cards, avatars, titles, and card backs used by the opponents in these matches.
The upside of looking at opponents is that it provides a much wider sample size. Only a small percentage of players use a tracker at all. There are many reasons for this, including being too casual of a player to care and playing the game on iOS, where a tracker isn’t available. Looking at opponents instead of players provides both a bigger sample size and one that’s more representative of the whole player base.
The downside of looking at opponents is that the tracker can’t see an opponent’s cards until they’re played. Looking at played versus in-deck deflates the usage numbers, but we still have more than enough data to draw some conclusions!
On to the fun part! Let’s look at what your opponents are telling you about their decks — for free!
Avatars have the most potential to be clues — each one besides Default and Rock represents a card that can start in your deck. Plus, there are a lot of them!
It should come as no surprise that Default is an extremely popular avatar. It’s the, uh… default. It takes first place in raw usage by a huge margin — being seen in 6.06× as many games as the second-place Loki. Rock is no slouch, either, claiming 33rd place with 0.69% of matches. That’s even more impressive when you consider the only way to get the Rock avatar was in an April Fool’s bundle that cost 400 Gold.
Here are the top 30 avatars by usage:
|16||Ant Man 🔷||1.24%|
|24||America Chavez 🔷||0.95%|
|25||Blue Marvel 🔷||0.91%|
|26||The Phoenix Force||0.88%|
The blue diamond emoji represents avatars that everybody receives by the end of the introductory Recruit Season (which finishes off with unlocking the Blue Marvel card). Misty Knight is the only early-game avatar that doesn’t make the cut, although she’s not far behind at #37. Of course, all of these avatars have popularity aided by the fact that just about every player owns them.
Recent rewards usually see a big boost (you’ll see it with card backs and titles, too). In part because they’re the new hotness and in part because they’re widely available to active players. Loki, Thor, and Enchantress from the current season’s rewards; Devil Dinosaur as a login reward; Sunspot as a Twitch Drop; and Daken, Magneto, and Wolverine from last season’s offerings.
We observed avatars for 186 characters (on top of Default and Rock). There are 233 unique cards in the game, leaving 47 without a single avatar! I wish Second Dinner would release a base-art avatar for every card as it released. I spent months wanting either Captain Marvel or Stature in avatar form. I was thrilled to finally pick up a Captain Marvel avatar through the One-Punch Women bundle (it’s replaced Enchantress as my own go-to avatar). As far as I can tell, there is no Stature avatar, despite her being in the game for almost seven months!
Across all games, the opponents played the card matching their avatar in 5.90% of games! That number climbs to 7.67% if you exclude Default and Rock.
Here’s a list of every avatar where we observed opponents playing the matching card in at least 10% of matches:
|Rank||Avatar||Play % w/|
|4||Ant Man 🔷||18.44%|
|10||Blue Marvel 🔷||16.20%|
The blue diamond emoji once again represents avatars that everybody receives early in the game.
Look at how high some of those numbers get! Nebula and Collector avatars play their corresponding card in over one-fifth of their matches!
Let’s look at a set of avatars: #2 The Collector, #6 Loki, #9 Agent Coulson, and #14 Quinjet — all likely representing the same Loki-Collector deck type that has been both strong and popular since Loki’s debut. Loki and Coulson have had avatars made recently available (Loki from hitting rank 80 or Season Pass level 46, Agent Coulson through the Days of Summer Past bundle). The Collector and Quinjet haven’t, and yet players are bringing those out from their collections to match their decks.
It’s interesting to see which avatars are the biggest tattletales. The stats back up my own anecdotal experience that Destroy and Discard avatars often mean a matching deck. There are already four Discard cards in the top 30 (including Wolverine). If I extended the table a bit further, you’d see Blade, Swarm, Hela, and Dracula all holding a rank between #31 and #37 with their card being played in at least 9.8% of their matches.
It’s possible that the avatars topping this list are doing so just because their corresponding card is popular. Like, it doesn’t mean much for Nebula avatars to be playing Nebula in 22.71% of matches if all the other avatars are playing her that often, too. I can account for that!
Relative Card Matching
Here are the top 30 avatars that play their corresponding card disproportionately the most often:
|Rank||Card||Play % w/|
|Play % w/o|
21 of these top 30 appeared on the previous list (all those with Play % w/ Matching Avatar over 11%).
As an illustrative example, let’s compare Zabu to Armor. They hold very similar ranks looking strictly at how often their matching avatar plays them: #20 and #19, respectively. If we look instead at how much more often their matching avatar plays them than non-matching avatars do, it’s a completely different story. Zabu rises to rank #12 and Armor plummets to rank #79. Zabu was played in 13.35% of matches by Zabu avatars, but only in 3.51% of matches by non-Zabu avatars — a dropoff of 9.84%! Meanwhile, Armor was played by Armor avatars in 13.42% of matches (similar to the Zabu number) and by everybody else in 10.07% of their matches — a much smaller dropoff of 3.35%.
Alternatively, we could express the differences in play rate between matching and non-matching avatars using division instead of subtraction. Like, Zabu was played 3.80× as often by opponents rockin’ the corresponding avatar. Meanwhile, Armor avatars only played armor 1.33× as often as everybody else.
Here are the top 30 cards by the ratio of how often their matching avatar played them versus non-matching avatars:
|Rank||Card||Play % w/|
|Play % w/o|
|11||Ronan the Accuser||2.15%||0.37%||5.81×|
Does this list finally feel right to you? This one feels the best to me. I think it’s the closest to matching the way our brains process how patterns of frequency feel. It’s actually been theorized that humans naturally think logarithmically. That is, if you ask a young child what number is halfway between 1 and 9, they are more likely to answer 3 than 5.
The rarely played cards that top this list might be better examples of what I’m trying to describe.
If you asked me to list off the avatars that give away their decks the most, the list I’d invent would look the most like this one.
I find it interesting that MODOK can top all three lists about matching, but of the Loki-Collector bunch, none make this final list (Coulson is the closest, at #49). In other words, MODOK is played a lot more by MODOK avatars than others, while pretty much everybody is playing lots of Loki, Collector, Coulson, and Quinjet.
Before moving on, let’s look at the bottom of the pile. Fifteen avatars played their matching card less often than non-matching avatars. Here they are:
|Rank||Card||Play % w/|
|Play % w/o|
Some of them are kind of weird, right? We have three Discard cards on this list, despite so many others of their archetype in the top 40. Professor X avatars play him less than half as often as others? Some of them I get. The Punisher avatar (I think the one from the Heroes for Hire season is the only Punisher avatar) is really cool but the matching card stinks. Similar stories with Gambit and Hulk — I can imagine people who really enjoy the character but aren’t interested in using the card.
Admittedly, this far down the list, the problem might just be one of sample size. Across almost 300,000 games that went to turn 6+, we only saw 137 Professor X avatars.
Rock is actually on this list — as the only avatar that represents a token. It’s in 155th place, between Shang-Chi and Cosmo. That’s right: Rock avatars play Rock at least once in a match more often than non-Rock avatars do (1.28× as often).
As I’ve already mentioned, I’ll be looking at similar card-matching statistics for card backs and titles. I just wanted to set the expectation that these tend to be much less informative than avatars. I think there are a few reasons for that.
Compared to avatars, there are much fewer card backs, and they have a lower rate of corresponding to a specific card (although plenty do).
Here are the top 30 card backs by usage:
|Rank||Card Back||% Matches|
|23||Infinite Spider-Man 2099||1.17%|
|29||Game of the Year||0.83%|
For context, we observed 53 card backs over the last week. You can check ‘em out at our Card Back database page.
The default card back is quite a bit more popular than the default avatar. About 4% of matches were against an opponent who changed their avatar but not their card back.
All six of the default Snap card backs make the top 30. They certainly have an edge because they’re owned by every player in the game. The Snap Cube (rewarded at rank 5) and the Marvel (given at Recruit Season level 9) card backs have similar popularity advantages.
All three of last season’s card backs remain popular (Infinite Daken, Daken, and X-23). This season, we’re already seeing lots of Infinite Loki but not nearly as much of the regular Loki. That’s, of course, because the golden version is acquired by reaching rank 100 and the base version by reaching season pass level 48. Roughly half of players will unlock the Infinite Loki back before the regular Loki back.
It’s cool that at least one card back from each of the past 10 seasons make the cut for top 30!
March: X-Men and Sentinel
April: Team Dog (Cosmo) and Team Cat (Goose)
May: Mixtape and Infinite Mixtape
June: Ghost-Spider, Infinite Spider-Man 2099, Spider-Man 2099
July: Echo, Infinite Echo, and Phoenix Force
August: Daken, Infinite Daken, and X-23
September: TVA, Infinite Loki, and Loki
As stated, we observed 53 card backs. Card backs that are older or were only available in real-money bundles are owned by the fewest number of players, of course. The two rarest are Shield and Hydra, which were available in the last beta-only Season Pass. You might have expected them in the Sword & Shield Beta Rewind bundle, I sure did! Those card backs were rewarded based on your level in ranked mode instead of from the Season Pass like in previous seasons. We don’t know how the Shield and Hydra card backs will be re-released, but thanks to an answer on the official Marvel Snap Discord (screenshot below), we know they will be.
Relatively Popular Pairings
It’s not really worth looking into the most popular pairings of card backs and cards, it’s a repetitive list of the cards that are most popular right now. Nearly every one of the top 50 is a pairing with Nebula, The Collector, Jeff, or Loki.
Instead, I’m going to skip right ahead to relatively high popularity, similar to the relatively high matching rates we ultimately looked at for avatars.
For this section, I’m requiring a card back to have been seen in at least 400 games. That ends up excluding nine backs, but less than 1% of matches. Similarly, I’m requiring a card to have been played 100 times. This only excludes Kang. Without these restrictions, small sample sizes cause these card backs and cards to dominate the top list. One additional restriction — that the combo was observed at least 25 times.
Let’s stick with top 30 here:
|Rank||Card||Card Back||Play % w/|
|Play % w/o|
|1||Spider-Man 2099||Spider-Man 2099||1.69%||0.27%||6.26×|
|7||Strong Guy||Snap (Black)||1.28%||0.37%||3.46×|
|8||Black Panther||Black Panther||5.70%||1.65%||3.45×|
|11||Strong Guy||Snap (Blue)||1.22%||0.37%||3.30×|
|17||Jessica Jones||Snap Cube||4.25%||1.36%||3.13×|
|21||Groot||Space Age Love||2.29%||0.75%||3.05×|
|22||Strong Guy||Snap (Green)||1.13%||0.37%||3.05×|
|29||Sword Master||Snap (Blue)||4.77%||1.68%||2.84×|
|30||Strong Guy||Snap (Yellow)||1.07%||0.38%||2.82×|
I’m actually surprised it’s not Surfer on top, but Spider-Man 2099 ekes out the win.
Surfer nearly topping the list doesn’t surprise anybody, right? I don’t think I’ve ever seen the back used for a non-surfer deck (although our stats say the Surfboard card back played Silver Surfer less than 10% of the time!).
I can explain, though, the presence of so many early-game cards. Notice how they’re all cards you don’t use much by the time you’re halfway through Series 3? Elektra, Angel, Domino, Punisher, Sword Master, Jessica Jones, and Strong Guy are all up there matched to early-game card backs. That’s because early-game players are (mostly) the only ones using these cards… and they only have access to the early-game card backs. Why is Strong Guy holding four of the top 30 positions? It’s more about how little he is used by other card backs than about how strong his presence is among early-game card backs. Related — notice that the purple Snap card back isn’t up there. Since it’s the default, it gets way more usage by end-game players than the other Snap card backs (and thus Strong Guy’s presence on purple isn’t nearly as strong).
There are some funky things, for sure. Daken’s most unique card back is X-23, with 3.33× the usage of all other card backs combined, versus 2.01× for the Daken back. Infinite Daken was strangely pushed out of third place by Space Age Love (the back featuring Star-Lord and Gamora’s faces).
I looked at all the characters who have card backs that correspond to them. I even played a little loose, like counting the Mixtape back for any of the Guardians of the Galaxy or counting Asgard for any Asgardians. I also removed sample size restrictions for cards whose match had low usage (such as Stegron, Namor, and Venom).
Here is the list of cards that has a matching card back that is also the back where the card sees the highest relative usage:
- Black Panther
- Gamora (Space Age Love)
- Jane Foster Mighty Thor (Mjolnir)
- Loki (Infinite Loki)
- Master Mold (Sentinel)
- Silver Surfer (Surfboard)
- Spider-Man 2099
- Thor (Mjolnir)
If you allowed some further stretching, you could include these. It seemed a little wrong to me since they all have a better match:
Granted, many of the cards who have possible matches are hurt by being cards from the early game. For instance, nearly every player has Ant Man. The Ant-Man card back is much rarer, being a reward from the free season pass in February 2023, and therefore inaccessible to anybody who started the game since.
There are more titles than card backs or even avatars, but titles have the lowest rate of corresponding to specific cards.
Here are the top 30 titles by usage:
|1||Lowkey Loki Lookalike||3.81%|
|3||No Ring Toss On The Headgear||2.19%|
|4||Knockoff Tesseract Dealer||2.10%|
|5||Loosely Related To Wolverine||1.88%|
|7||Thanos Was Right||1.49%|
|8||What Is This? Wizard Poker?||1.47%|
|9||I Am On The Toilet||1.37%|
|12||Guess Who’s Back?||1.26%|
|13||The One and Only||1.18%|
|14||I’m Outside Your House||1.16%|
|15||Cool Cool Cool||1.15%|
|16||HEY. Calm Down||1.08%|
|17||I’m Sorry It Has Come to This||1.07%|
|18||Don’t Worry I Don’t Bite||0.97%|
|19||Your Average Blazing Space Bird||0.95%|
|20||Got a Bad Feeling About This||0.90%|
|21||On Reveal: You Lose||0.90%|
|22||Not Potty Trained||0.89%|
|23||AKA “The Boss”||0.88%|
|24||If I Lose, Do Not Take Credit||0.87%|
|26||I Don’t Have A Title I Like Yet||0.84%|
|28||Embarrassment To My Family||0.81%|
|29||90s was the Best Decade||0.81%|
We observed 235 unique titles over the past week. While the most popular titles don’t do nearly the numbers that the most popular avatars do, the advantage of recency is still on display. Among the top 5, two are from last season, and the other three are from this season.
You’ll notice that variety among titles is much higher than with either avatars or card backs. Even the most popular title was observed in less than 4% of matches.
While you have to use an avatar and card back, the use of a title is optional (and the default is to not have one). Across nearly 300,000 matches, the opponent was equipped with a title only 59.94% of the time!
Relatively Popular Pairings
Just like with card backs, looking at the most popular title-card pairings is pretty much a gallery of the current most-popular cards. Instead, we’ll skip to relative high popularity. That is, we’re looking at how often a card is played with each specific title versus all other titles combined. We then look at which pairings have the highest gaps between the with-title rate and the without-title rate.
The only limitation I’m applying is that a card-title pairing must have been observed at least 25 times.
Here are the top 30:
|Rank||Card||Title||Play % w/|
|Play % w/o|
|1||Agatha Harkness||Playing Agatha||10.26%||0.63%||16.29×|
|2||Agatha Harkness||Definitely Not Ultron||7.23%||0.64%||11.30×|
|3||Silver Surfer||The Silver Couch Surfer||6.83%||1.57%||4.35×|
|4||Daredevil||Future Mrs. Man-Thing||22.76%||5.67%||4.01×|
|5||Nimrod||Wolverine Is My Spirit Animal||7.64%||1.94%||3.94×|
|6||X-23||At Some Boring Funeral||15.24%||3.87%||3.94×|
|7||Ghost-Spider||Thwip! Get Webbed||3.11%||0.80%||3.89×|
|8||Miles Morales||Stuck In Spidey-Landing Pose||8.39%||2.20%||3.81×|
|9||Daken||Trigger Scent: Nachos||5.40%||1.44%||3.75×|
|10||MODOK||The Apocalypse Sucks||6.85%||1.86%||3.68×|
|11||Deadpool||Wolverine Is My Spirit Animal||9.11%||2.69%||3.39×|
|12||Rocket Raccoon||Playing Agatha||7.39%||2.21%||3.34×|
|13||Deadpool||Merc With A Mouth||8.93%||2.70%||3.31×|
|14||Dracula||The Apocalypse Sucks||6.60%||2.00%||3.30×|
|15||Daredevil||Here Comes A Fastball Special||18.24%||5.67%||3.22×|
|16||X-23||Merc With A Mouth||12.20%||3.87%||3.15×|
|17||Daredevil||Bruno Makes Me Swoono||17.53%||5.67%||3.09×|
|18||Venom||Wolverine Is My Spirit Animal||17.24%||5.61%||3.07×|
|19||Brood||The Silver Couch Surfer||8.00%||2.66%||3.01×|
|20||X-23||Wolverine Is My Spirit Animal||11.58%||3.87%||2.99×|
|21||X-23||Do Not Fist-Bump Wolverine||11.45%||3.85%||2.97×|
|22||Polaris||Thwip! Get Webbed||4.31%||1.46%||2.95×|
|23||Daken||Lady Deathstrike’s Nail Filer||4.27%||1.45%||2.94×|
|24||Knull||Merc With A Mouth||9.82%||3.36%||2.92×|
|25||Arnim Zola||Wolverine Is My Spirit Animal||6.65%||2.30%||2.89×|
|26||Knull||Wolverine Is My Spirit Animal||9.61%||3.36%||2.86×|
|27||Silk||Thwip! Get Webbed||7.18%||2.52%||2.85×|
|28||Nightcrawler||Shampoo Scent: Terrigen Mist||17.96%||6.33%||2.84×|
|29||Sabretooth||Loosely Related To Wolverine||5.26%||1.86%||2.83×|
|30||Korg||Definitely Not Ultron||7.23%||2.59%||2.79×|
There’s some fun stuff here. People who declare they’re “Playing Agatha” actually do so 16.29× as often as people who don’t rock that title. A few spider-people matched with spidery titles. Destroy cards love referencing Wolverine (Wolverine himself doesn’t appear until rank #69, along with “Wolverine Is My Spirit Animal”).
If we counted being without a title as a title, title-less Punisher would hold 26th and title-less Strong Guy, 29th. This for the same reason they showed up with relatively popular card backs: hardly anybody is playing these weak early-game cards with a title. Like, Punisher is played almost 3× as often by titleless players than players with a title.
Ooh, can I please tell you about something fun that showed up when I ran these numbers last week?! (Quick aside, I ran the numbers last week because I meant to have this article out sooner. That got disrupted, and the delay was so long I decided to re-run all the numbers today.) Anyway— Mjolnir!
I’ve removed tokens (as in cards like Demon and Squirrel that you can’t put directly into your deck, but are created by other cards) from most of these lists for a variety of reasons. Last week, the pairing of Mjolnir with the “Can Lift Mjolnir” title held 24th place (before I removed the tokens). Thor’s top title was also “Can Lift Mjolnir”, but that pair was all the way down at 148th place. So Mjolnir was played 3.5× as often with that title than with all others combined, while Thor’s relative play rate with that title was 2.38×. In other words, players bearing the “Can Lift Mjolnir” title were backing up their claim: they played Mjolnir about 60% of the time they played Thor, while other (less worthy) Thor players only put his hammer on the board about 40% of the time!
Does it Matter?
Does it even matter if you can tell what your opponent is playing early? Often not. Some decks give away their identity through play a lot earlier than others. Part of the reason that Discard and Destroy decks don’t mind revealing themselves from the get-go is because they usually do so by the end of turn 2 anyway; Bucky Barnes and Morbius are dead giveaways!
Over the last couple weeks, I paid attention to when an opponent’s avatar helped me make a useful in-game decision. In all that time, I collected three examples.
- Valkyrie Lockdown vs. Morbius Avatar: I played like I was trying to stop Morbius on 2. I usually play Echo early, but I decided to try sniping Morbius instead. I played Ant-Man on 1 in a weak attempt to take priority. My opponent played nothing, so it was enough! The middle location was safe for Morbius, so I snapped and played Echo there. Sure enough, Morbius revealed in mid and promptly got his ability slapped into non-existence.
- Jane Foster/Hit Monkey vs. Leader Avatar: My opponent played Wave on 3. At that point, I typically expect Galactus. I had priority. I would normally play Jane Foster in that position, hoping for her 8 power to stop Galactus. I noticed the Leader avatar and prepared for that instead. I played Hulkbuster, who merged with an existing card. My opponent played Leader, who had nothing to copy.
- Valkyrie Lockdown vs. Nimrod Avatar: It’s turn 2. Neither of us have played a thing. I had added Armor to the deck to deal with Alioth. The plan with Armor is to play her later, after I know which lane will be the Valkyrie lane. However, the Nimrod avatar was my opponent’s way of telling me they were playing Destroy. I snapped and played Armor that turn. My opponent played X-23 and Deadpool into the same lane. They retreated.
Whether you’re climbing the Ranked ladder or competing in Conquest, every cube counts. In these three cases, paying attention to my opponent’s avatar helped me win some cubes. I might’ve won anyway, but I’ll gladly take the edge! It’s a small edge, mind you — I probably played over a hundred matches and only have three stories about turning my opponent’s avatar against them. Granted, I’ve been playing Conquest most of that time, and I think being able to predict your opponent’s deck within a single match is a lot more useful in Ranked.
It seems plenty of people are eager to tell you about their decks before a single card is revealed. Their avatars, card backs, and titles sometimes give hints that prove genuinely useful. Use that to your advantage! Similarly, be careful about giving your opponent that same edge.
I challenge you to watch out for opportunities to use the info your opponent is sharing. Who knows, maybe it’ll earn you a few extra cubes, too!
I hope you all have a great week! If you’ve got an idea for a cool data- and/or math-heavy analysis, I’d love to hear it! Please share it with me in the comments below or else on Twitter/𝕏! 💜
This analysis was powered by the data from our Marvel Snap Tracker. If you play on PC or Android, consider giving it a download! It has many features including a live overlay, the ability to sync your collection to our site (letting you review your personal stats), and powering useful tools like the Meta Snapshot.