Marvel Snap: Phase 1

Link post

Usual game approach. I’ll start off spoiler-free, then proceed to the semi-spoiled section, then to the Here Be Spoilers section.

Maximally Spoiler-Free (1 Bit of Info): Should You Play?

No.

Spoiler-Free (3 Bits of Info): What Tier Is It?

Tier 3. It is a good game.

I want to be able to put the game into Tier 2 despite its severe problems. I can’t.

I can see this becoming a Tier 2+ game over time if the outer loops are improved.

Spoiler-Minimized Overview (Significant Bits)

Marvel Snap is a beautiful mess.

There are things I and everyone else I talk about it with truly love about it.

There are other things about it I, and it seems pretty much everyone else as well, wish would die in a fire.

The thing to love about the game is its core gameplay. The inner loop.

Marvel Snap games are elegant, fun, engaging and involve interesting decisions, while also being super duper fast.

Decks are twelve unique cards. Games last at most six turns. You start with three cards and draw one each turn. Each game there are three locations, each with an ability, which get revealed one per turn for the first three turns. On turn X, you get X energy to spend on cards, which you can deploy at one of the three locations, maximum four cards per location per side. Cards do things, and provide power. Whoever has more power in two out of three sites wins.

The central twist, the ‘snap,’ is the doubling cube. You play games for cubes. At first the stakes are one cube. Each player can snap once per game, doubling that to two once the current turn ends. Stakes automatically double after turn six ends, as well, max stakes is 8 cubes. At any time, you can retreat and only lose current stakes. The decisions on when to double, including when your situation is ‘too good’ to double, and when to decline a double, get fascinating once your opponents are not rocks.

That’s the whole game. All of that is great, at the early card mix is chosen very well to ensure interesting game play. Giving everyone the same starting cards keeps it fair, at least initially.

Also, the Marvel flavor is light, but parts of it are pretty damn cool.

The thing to hate about the game is its monetization models and incentive structures designed to keep you playing. The outer loop.

This is a Gotcha Gacha Gotcha game (can we make this name change happen?). Your core outer loop experience, to the extent you pay attention to it, is a series of missions and dopamine hits and timers and multiple limiting factor currencies and the inability to get the things you want.

This stuff is highly toxic, as I explained in the Coase whitepaper that started the journey of the super cool Emergents true TCG game I designed with Brian David-Marshall (plug!) that you can play now for free.

When you play a game like this, it invades your brain. You are constantly distracted, worrying about timers. The hours of your life become pieces on a game board.

This is, in its current iteration, an unusually nice game compared to the others of this genre I have experienced. That makes it plausibly acceptable. Still highly dangerous.

One underappreciated way I judge such games is on its timers. How much felt pressure is there to time everything? Many such games are foundationally about managing timers, which is maximally toxic. Marvel Snap isn’t doing that. Yes, you do have timers to worry about, but there is substantial slack on them before you feel like you’re being substantially punished.

The two currencies that lock meaningful advancement are credits, which are the long term limiting factor if you play a lot of games, and boosters, which are the short term limiting factor or stay that way if you play few games per day.

You get boosters from playing games. Winning cubes or games ranks you up and sometimes finishes missions, but that’s all winning does. You get credits from doing missions, which mostly means playing games. You get season progress from various other missions, which unlocks various other things that loop back into boosters and credits. You spend boosters plus credits to upgrade how your cards look, which then increases ‘collection level’ which then unlocks new cards. There’s also a premium currency, gold, that you can buy and that has no great uses right now.

You get credits from completing missions, except that the way you complete missions is, with notably rare exceptions, ‘have missions to complete and play games.’ Occasionally you’ll need to put a card or two into your deck as well.

You have at present zero control over what cards you unlock in what order. So you could be facing a card or deck all the time that you do not have, with no way to get it except lots of extra grinding, even if you are willing to pay. You’re on your own, kid.

How Do We Respond To This Dilemma?

The core question of Marvel Snap is whether the juice is worth the squeeze.

To me this comes down to the question can you handle it?

If you suspect you would wind up spending money on Marvel Snap, and definitely if you would spend money not on a season pass, walk away. Gotcha games do not make your life better when you spend money. You will find yourself spending a lot and getting little, or almost nothing, in return.

If you suspect you would wind up being highly distracted by Marvel Snap, walk away. Games like this are engineered to get in your head and make you constantly think about how to optimize within the game instead of thinking about other more valuable and important things. The pull of one’s phone is already a huge problem for all of us, and this only makes it worse. If you don’t know how to get your brain to not do this, then the game is not safe for you.

However, if you are confident you can handle the game – and this can be as simple as ‘the moment I am not properly handling it I am confident I will then quit’ then there is value here. That is especially true if taking short breaks from other things to play a game is refreshing to you rather than denying you the time to breathe.

Marvel Snap’s core gameplay is the best I have ever seen for something that plays this quickly. It is a remarkable achievement. The question is whether you can safely access that without Getting Got by what is wrapped around it. Where Getting Got here means either tricked into spending money, or being distracted by the pull of playing games when you need to focus on other things or simply missing out on richer, more in-depth experiences – if you find yourself playing this game for hours at a time, either you are a professional streamer, or something has gone wrong.

If I did not have any professional reason to play, I would consider my own continuing to play the game to likely be net unwise. If the game continues to occupy substantial brain space for me a week or two from now, in ways that on reflection I do not approve of, I should definitely uninstall it. Continuing to play will depend on finding a way to make it available when it is a good idea, and not coming to mind when it isn’t.

Economic Analysis /​ Outer Loop

The way you progress through the game is kind of weird.

You can think of yourself as having two core numbers you are trying to make go up.

One of these is cubes, this game’s version of the ladder. That is purely a measure of your skill and success. Technically it gives you some very small prizes along the way. I found those to be insulting, net worse than offering me nothing at all. It doesn’t otherwise factor into the economy.

Instead you have one clear measure of your economic success in the game, which is your collection level. Quite the He Admit It moment. Make number go up.

How to make number go up? Spend credits and boosters on chromatic card upgrades.

Boosters are a proxy for games (technically turns) played. Play a game, get six boosters. At first, or if you don’t play many games per day on days you play, this is the limiting factor. You get tiny bonus amounts on reward tracks, as well, I guess.

Boosters are for specific characters, but this can safely be ignored, they’ll give you things to upgrade periodically and you upgrade them, shrug.

Credits then seem to be the limiting factor if you play a lot of games. Once you run out, you can get them from the store for free (on a timer), you can get them from missions (on a timer), so mostly you get them on timers.

You can also, in theory, buy them. Don’t do this. To extent you have gold given to you for free, you should almost certainly save it until game gives you a better opportunity.

You will complete most your missions by accident via playing games. Occasional ones you’ll need to put a card into your deck to do it, or something.

Thus, the game is sending you a very simple message. Play a lot of games. Play continuously. We will slowly give you cards for it. That’s it.

As toxicity levels of Gotcha games go, that’s not actually that bad. There are (so far) no weekly events full of constant grinding. There are no super precious currencies free players need to earn in dregs. And so on. Things could be so much worse.

Think of it kind of like Magic Arena, except you can’t draft or play sealed deck, you can’t enter tournaments, and the store won’t sell you booster packs.

Mysterious Chromatic Upgrade

In Emergents, we want to ensure the cosmetic upgrades are both cool and unique, and in linking progress on each card to its own unique quest lines. Players wanting a particular card to be cool can work towards that goal, either to have it be cool or to sell it to someone else. I think we are doing a great job.

Of course, I am biased.

Marvel Snap does a fantastic job of making the basic copies of their cards cool and unique. Given the amount of space available on a phone, it’s an A+ job.

The upgrades are a different story. First you get a ‘frame break’ meaning part of the picture is slightly outside the frame. Then you get ‘3-D’ which is, technically, slightly 3-D. Then you get ‘animated’ and finally ‘shiny logo.’

If I squint during a close-up, I can sometimes see the differences on the upgrade screen. The frame breaks are sometimes a little cool. The ‘shiny logo’ is the tiniest, least interesting chromatic boost I’ve ever seen.

During a game? Nothing.

What makes this super confusing to me is that chromatic upgrades are something some people care about a lot. Others are indifferent. When we tie card access, which almost everyone cares about, directly to chromatic upgrades, we centralize something many players (including me, with this implementation at least) care about not at all, or very little. Yet the game treats this as a super big deal, wastes my time with it, intermediates its stuff through it.

Quest Progress

When you first unlock the campaign and all its chapters, the sudden barrage of missions, and the progress you get towards them, is hilarious. Suddenly here are dozens of numbers ticking up, with progress reports scrolling in and out of the screen, almost none of which require any conscious attention on your part. Then eventually you have to go click through all the completions. Again, it’s so weird, teaching me not to care about such things.

The weirdest part is that a thing will pop up that says ‘hey look you did the thing’ and flash. Then if I tap it, it will take me to another screen, where I then need to find the thing I did and tap it again, in order to get another screen that gives me a tiny bit of currency. Finding all these things can be trickier than it sounds or has any right to be.

I have chosen not to study the Modern Art of Gotcha Game Outer Loop Design in overly detailed fashion. Things like this still baffle me.

Then there’s the issue of the three card pools.

The Three Card Pools

The game does not explain this, but you can find out online.

Beyond the cards everyone starts with, the cards are divided into three pools.

Everyone always unlocks all Pool 1 cards first. You do so in a completely random order that you do not control, which is frustrating, but you do unlock all Pool 1 cards first.

While you are ‘in pool 1’ you do not in practice face anyone in Pool 2+. The game pairs people in large part by collection level, so it is more fine grained than this, but what is clear is that zero of my opponents have played a Pool 2+ card that wasn’t randomly generated during the game.

At least one friend is trying to avoid Pool 2 for now, because progression there is slow and it means facing lots of cards you don’t own. Whereas at the top of Pool 1 everyone has full access. Plus progression slows a lot as you move into Pool 2 and again into Pool 3.

On reflection, I think the pools are a mistake. I get what they are trying to do, but it means that the budget deck players all have access to the same cards. Which is nice in your first few dozen games, until it isn’t, because everyone is playing mostly from the same effective pool.

A number of those cards are good enough that essentially everyone plays them. I’d say about half your deck at this level is usually more or less ‘forced.’ Blue Marvel in particular is given to everyone and is far and away the strongest card, which is highly annoying and warping, and almost everyone uses that and Iron Man as their two 5-cost cards, at least for now.

The game has recently decided to pair me against players making some strange card choices. I don’t know what is going on inside the pairing algorithm. I do know that I noticed I played enough of the same cards as my opponents I wanted to try Cable, which draws from the enemy deck, at which point opponents started often having Cable. Then I added another similar card, The White Queen, and opponents in the next two games suddenly had The White Queen. Huh.

As things progress (see spoiler here), decks become less ‘here are all the good cards available’ and more ‘this is a deck that does a thing’ with a dial of how much to flood the zone with cheap stuff. There’s clearly a movement deck, a self-destroy deck, an ongoing deck, an on-arrives deck, a discard deck, and a few explicit build-around cards. Not that I know any of them are any good. Hard to tell.

Games look a lot less diverse when you intentionally pair similar decks together.

You May Wish To Employ Strategy

All right, that stuff is out of the way. What’s the strategy, at least at the lower levels, right now?

One thing that was twisting my first two days was Bar Sinister. It was a ‘feature location’ that came up a lot, where you fill it with anything you place there. This makes the game a lot about who can and can’t play four copies of Blue Marvel there. If you have Onslaught, which I don’t, then it is worth noting that Onslaught/​Blue Marvel/​Nightcrawler is an automatic win on Bar Sinister (so there was an obvious ‘best deck’ and if it snaps you kind of have to fold) and there are substitutes for Blue Marvel there that work fine (how to do it left as exercise to reader).

Once that ended, things felt a lot more sane. The rotation of feature locations seems like a good thing in general, this was simply a really strange example.

There are three subcategories of strategy. There’s playing the cards, there’s building a deck, and then there’s snapping and retreating.

Playing the cards is basic ‘play the game’ or more precisely ‘do the thing your deck does,’ while trying to know what they are doing, figure out what cards they are likely to play where, and do something that wins against as many plausible paths as possible. It should mostly be obvious what cards they have – either ‘same as yours’ when you start out, or later what makes sense for them. There isn’t that much to say.

That said, here are some simple pointers that might not be obvious to some people.

  1. Assume they ‘have it’ and ask how you can win anyway.

  2. Most people will do the obvious thing, few will respond to what beats it.

  3. Deck construction is a tell to what level a player is on.

  4. Telegraphing your move is more expensive than it looks.

  5. Ask if you have a way to disguise what you are up to for another turn.

  6. Filling up all four slots in a location makes you predictable and cuts off options.

  7. When in doubt, don’t play to the face down locations.

  8. Knowing what your opponent has is more valuable than you think.

  9. Turn three is often when you can most confidently know where they’ll play.

  10. You can’t bluff a rock. People are bad at, well, basically everything.

Deck building is about choosing a theme if any, then choosing the best cards you have, and finding the right mana energy curve for what you are doing.

For basic collections, there are essentially two choices. You can either pursue a strategy with more 1-drops and Kazan, or you can play a balanced curve, which is what I’ve been doing. You can then either theme out, or not.

For the balanced curve, it seems clear to me you want exactly one 4, two 5s and one 6. The two 5s cover you in case you miss on 6, so you can likely go 5+1, especially since the top two 5 drops are both crazy good. Having two 4s will often strand one if you draw both, and playing 2+2 or 3+1 on turn 4 is not so bad anyway.

That leaves 8 slots for the 1-3 drops. Going 3-3-2 is the obvious choice, and seems mostly to be right. You could also go 4-2-2 if you want, at which point one could shift a 3 to a 4 to get Kazan, and you’d then play Squirrel Girl – that’s about the extent of the flexibility here at low levels. Decks seem like they’d need to be pretty weird before something different was a better idea.

The other option is to use Domino, which means you are locked into having a 2-drop, so you only need one. I think it’s worse because several of the two drops are very good, but that might mostly reflect my card access, in particular not having Squirrel Girl yet.

More advanced collections get to do more interesting things.

Card effectiveness is often about some cards having slightly higher rates on average than other similar cards, so in context there is usually a ‘right’ answer, although not always.

Snap!

The most interesting decisions in the game are when to snap and when to retreat.

How you snap and retreat depends heavily on how opponents react to your snaps, and how they snap and retreat both when you do and don’t snap. You can’t bluff a rock. I am ranking up as I play, and play is improving, but I am not yet at the point where opponents seem to be snapping and retreating anything close to optimally.

For a while the answer is ‘you always win and no one ever retreats even when you snap on 6, so snap on 6,’ and while that is 90%+ effective you should totally do that.

Assuming you want to rank up (i.e. win more cubes). This section assumes you do indeed want more cubes.

Over time, you will start losing more and they will start retreating more and snapping more, and more strategically, and start reacting to your snaps. For a while, they learn how to occasionally win much faster than they learn how to handle snapping and retreating.

That means, yep, keep snapping late when you know you are going to win, or failing that on six when you are favored.

There is then a phase transition, currently in the high 30s, when your opponents are now smart enough that they will see ‘my opponent wants to play for 4 cubes’ as a sign that giving up 1 cube might be a good idea, especially on turn 6, often enough to make you play… well, not game theory optimal or anything, that comes later, but you will need to play more reasonably. Then, over time, more reasonably than that.

What does that look like?

There are three big changes versus the Backgammon double.

  1. Hidden information. Your double represents a hand the way it would in poker.

  2. Automatic double at the end. You don’t want to force your opponent into a wise retreat, and it is much tougher to stick it out for +/​- 4 versus a loss of 1.

  3. There is no third double. If I double, you can double back, but it ends there.

This makes it more a cross between a poker bet and a double.

(To explain the third point: It is hard to profitably redouble in Backgammon, because you are giving the other player back the cube. In Marvel Snap, both you can know what they don’t know, and also you don’t hand them the cube back. Thus, redoubling is much stronger.)

When should you snap first? It’s a good question. One can think of three distinct types of snaps.

  1. You are bluffing, trying to get them to retreat.

  2. You have the edge and are putting them to a decision with no good answers.

  3. You are winning and a retreat in response would make you sad you snapped.

Perhaps people start folding enough to allow bluffing when you get high enough. Perhaps there are then successful players all over the place going around bluffing.

I doubt it, on both counts, unless you are very very high up. There are going to be too many people playing honest games, and too many who want to see what happens. Poker players ‘get curious’ often enough even with real money on the line. The chance of an opponent who cannot beat any plausible hand and strategy I could have not withdrawing on turn six remains very high. As in, they turn over their cards, and you think ‘wait, that wasn’t enough even if I did almost nothing, huh?’

Thus, I will continue, for now at least, to never bluff. If I snap, it is because I have a clear advantage and I worry waiting would lose my customer. If you snap, I assume you are doing the same. I will withdraw a lot. What I won’t do is assume you know what you are doing or have the best possible hand.

Note that under these circumstances, you should be exploitable by someone bluffing. As in, you should be willing to net lose cubes if they snap on turn 3 with a bad hand, versus if they hadn’t snapped. Who cares? It’s not like they are going to do it that often.

Conclusions /​ Where Do We Go From Here?

I do expect this game to have staying power. It has a lot going for it – its quick and fun inner loop, its guessing and bluffing games, the Marvel flavor and so on. The outer loop mechanics are proven to be part of many successful games over the years.

However, when I thought about this question for myself, the answer was obvious, at least as far as my phone is concerned. This is too distracting, it’s a net negative.

I uninstalled as I finish writing this. I feel good about this decision.