Marvel Snap

Marvel Snap is on online “collectible” “card” game which was publicly released in late 2022. I signed up for it – maybe even a little before that when it was still in beta, I don’t remember – and played it for a month or two in the fall of 2022 because I was looking for a new casual game on iOS. But after a little while I got bored with it and stopped playing.

Then this past September I was feeling pretty unhappy with last year’s Magic the Gathering sets (basically, I’ve hated the limited formats for most of them, and especially loathed Wilds of Eldraine, the set which was out at the time) and decided to give it another try, and I’ve been playing it regularly since.

As a casual and simple game it has a surprising amount of depth. I don’t really care about the character skins that the cards have, so it’s the gameplay which has held my attention. Enough that I decided to write an entry about it.

Collecting

The game is a “collectible” “card” game in that the game pieces are virtual cards, and you acquire them by consistent play. You can’t trade with other players, and for the most part you have little control over how you acquire cards. The game has a “collection level” system which you move up as you play, and over the first 500 levels you acquire every card in the first two “pools” of cards. Thereafter you gradually acquire cards from the later pools (3, 4 and 5). Effectively, this means everyone gets a base set of cards after a couple of months of play, and then players randomly acquire new cards to unlock new deck archetypes.

There are a couple other ways to acquire cards: Certain cards become randomly available for purchase through certain means in the store, and every “season” (month) a new “pass” is released with a new card which you can purchase. I’ve been buying the season pass for the last few months at $10 each and have been pretty happy with it.

Each card had a ladder of rarities which you can increase as you get “boosters” and credits. These rarities just make the card look different, which has no in-game value, but increasing the rarity moves you up the collection level track. You get boosters by playing, and you get credits by achieving daily and mastery pass goals, and you also get boosters and credits by moving up the level track.

So it’s a pretty clever gamified system, and for me the motivation is to acquire new cards to unlock new decks and experience, so I keep playing. A pretty nifty feedback system. I’m sure there are people who spend real money to unlock

Gameplay

A game consists of 3 zones, and each player may have up to 4 cards in each zone. The player with the most power in each zone at the end of the game wins that zone, and you have to win 2 of the 3 zones to win the game. (If you tie for numbers of zones won, then the player with the highest total power wins.)

The game lasts 6 turns, and each turn a player gets an amount of energy equal to the turn number, losing any unspent energy each turn. Each card has an energy cost, a power number, and may have special abilities.

Each player constructs a deck of 12 unique cards from the cards they own. They start the game with 3 cards, and draw 1 card each turn. And each turn they can play as many cards they have in hand as they have energy.

One thing that took me a while to realize is that this means you’ll draw 9 of the 12 cards in your deck during the game, so well-constructed decks have a fairly consistent experience, varying mainly in the order they draw cards in, and which 3 cards they don’t draw.

I’ve also noticed that some players don’t seem to think about having to win only 2 of the 3 zones and work to win all 3 of them, whereas I’m happy to punt on a zone in order to win the other 2.

The game keeps things simple by limiting your options to basically only which cards you play. If a card has an effect which could choose from among several different things – for example, discard a card, destroy a card, affect a card in some other way – the game makes the choice randomly. There are a few combos which can result in a number of effects happening in sequence, and when this happens the game will “fast forward” to the end of the sequence. I don’t think the game has any truly infinite combos, which must be an interesting design tightrope to walk.

Oh, one more thing: Each zone has a randomly-chosen effect which are revealed over the course of the first 3 turns, from left to right. There are a lot of zone effects, so every game is a little different, and some of them are very different. Some decks can be completely hosed by a particular set of zones, while others do well with them. Those are unusual, though: Most decks are competitive with most zones, but the games play out differently. And it feels great to win a game with a set of zones that your deck handles poorly.

I haven’t yet figured the optimal strategy for playing cards while some zones are still unknown. Early on I would play in the leftmost zone – revealed on turn 1 – with my first card, but these days I tend to play in the rightmost zone unless a revealed zone makes it clear I should play something there. Sometimes this hoses me, other times it benefits me, but on balance it’s felt like the right thing. I imagine whole articles have been written about how to do this.

Other Stuff

Each season has a separate ranking track based on “cubes”, which are acquired by winning games, and lost by losing games. Each game starts at 1 cube, and either player can “snap” at any time to double the number of cubes won, up to a maximum of 8. These get you some additional rewards, but mean you have to be snapping cleverly to win more cubes than you lose. This is frankly one of the less exciting parts of the game for me.

I’m not really sure how matchmaking works in Marvel Snap. I suspect they try to match you up roughly based on your collection level and maybe the season rank. I find I don’t often match up against decks which are clearly better than mine, but I also find that I tend to get up to 85 cubes and then stall out (the max reward is at 100 cubes). This probably means I’m missing out on some marginal value in my snapping, but I’ve seen people say they get up to max cubes in just a few days, so maybe not.

But the fact that I can keep moving up in collection level matters a lot more to me than getting stalled on cubes.

(It seems like optimally people would retreat when clearly behind with little chance of catching up and snap when clearly ahead with little evident chance of losing, and so most games would end up at 1-2 cubes. and advancement would be slow. In practice, this doesn’t happen, which is interesting. I often win 8-cube games where I felt ahead for a long time, and I sometimes lose 8-cube games because my opponent had stuff I hadn’t imagined which caught them up. I suspect there are built-in factors which stoke the variance, but also that there are aspects of this part of the game I haven’t figured out yet. I also suspect that some players just snap early because they like to play for higher stakes – the equivalent of a live straddle in poker – and that many people like to play out the game because you get more boosters for longer games. I dunno.)

As I said, there are also various challenges – daily ones for everyone, and extra ones for people with the season pass – which you can accomplish to get more credits. These are usually pretty easy to achieve – I’m not sure I’ve ever not achieved one with a few games of play per day.

Decks

Deck building is a pretty interesting part of the game. There are of course loads of web sites out there with decks you can build – if you have those cards. If you don’t, then you often look for a deck where you have 8-10 of the cards, and then figure out what to fill in around them. The new card with each season pass is fun because you can try to build some new decks around them.

I’m going to present a few of my decks without going into them in great depths; you can easily go online to see what the cards do if you care.

When I started playing I decided what I wanted to do was maximize the number of cards I could play, so I built a deck called “Micronauts” which looked like this:

Marvel Snap Micronauts deck

These are all tier 1 and 2 cards, and it performed really well for me. Basically these cards are all good-rate cards which put a lot of power on the board for their cost, and I could almost always play all of them by the end of the game. So if you’re new to the game, consider going this one a try. I’ve evolved the deck over time – primarily based around Squirrel Girl, Ka-Zar (which gives all 1-cost cards +1 power) and Jessica Jones – and it’s still pretty competitive. I bring it out whenever there’s a daily challenge where it wants me to win a zone with 4 cards.

I think my favorite deck came with the season pass for Ms. Marvel a few months ago. She is already an efficient card (5 power for 4 energy), but she gives adjacent cones 5 power if the cards in each zone have different costs. I looked at a few decks out there and then came up with this one:

Marvel Snap Ms. Marvel deck

I’m not sure I’ve actually changed this deck since I built it, and it wins a lot. Besides the basic synergies in the deck, I stumbled into a couple of cards – Armor and Cosmo – which absolutely hose a few popular decks in the metagame. There are few warmer feelings than dropping Armor – which prevents all cards in a zone from being destroyed – in a zone where a deck which wants to destroy its own cards has already played 2 or 3 cards.

This deck strongly prefers to play cards in the middle zone. I feel it could use a little more tuning, and perhaps someday I’ll acquire a new card which is an obvious replacement for one of the ones in it today.

My current favorite deck is this High Evolutionary deck. The gimmick of the High Evolutionary is he gives certain cards special abilities, but the nature of those abilities is that they give you a bonus for not using all your energy, and they give you a bonus if you decrease the power of your opponent’s cards.

Marvel Snap High Evolutionary deck

I don’t own a few cards commonly played in this deck, but I’m very happy with the cards I put in instead. Psylocke grants extra energy which can either ramp out cards or assist in the “not use all energy” bonuses. Scorpion and Hazmat both decrease power of opponents’ cards, making Abomination cheaper. Magik adds a 7th turn to the game, so you can play more stuff. And Iron Lad gets the ability of the top card in your deck, in addition to being very efficient, and synergies with most of the deck. It’s been a very consistent deck for me over the last week.

I have several other decks I play, but those have been my go-to decks so far.

In Conclusion

So I’ve been playing Marvel Snap almost every day for the last 4+ months, and frankly I’m kind of surprised it’s held my attention for this long. A single game takes just a few minutes, so I’m not investing a huge amount of time in it, and overall I’ve been winning at a pretty good clip, which always feels good.

So if you enjoy turn-based card strategy games like this, and either enjoy to can tolerate the play-to-acquire gamified system, then Marvel Snap may be for you. I don’t know if I’ll still be playing it at the end of the year, but if Magic sets continue to have sucky limited environments, then I just might.

Firmament

Earlier this summer I played Firmament, the new puzzle game from Cyan, the creators of Myst.

I love these sorts of games, and I wrote about what I like about them a few years ago in the context of Zed. I enjoy a mix of puzzles, setting, and story, which Cyan has historically been good at providing. I backed Firmament on Kickstarter as soon as they announced they’d be doing a Mac version, having also backed their previous game, Obduction, which I enjoyed a lot, though I thought it had a few flaws.

Firmament modifies the traditional point-and-click interface with a device the player uses called an Adjunct, which they use to connect to sockets throughout the game which provides some additional flexibility in how the player interacts with the world. It also both makes it clearer what you can interact with, but it feels somewhat limiting since everything has some small variation of the same interface. I’ve seen it theorized that the Adjunct mechanic was created to make the VR experience of the game better or more consistent or something. I guess it’s possible, I dunno. Maybe that was a concern 4 years ago when they started making the game?

A spoiler-free review first, and then some further thoughts:

The setting for Firmament is that you wake up from long-term sleep and are greeted by the Mentor, who appears to be a ghost of your predecessor. They’ve woken you up and advise you from time to time. The world consists of three settings, an ice zone, a botanical garden, and a sulphur-based power plant, along with a central structure1 called The Swan. You travel between them via conveyance pods, and have two waves of tasks to accomplish on each world to get to the conclusion.

There are basically three sorts of things in these games that I dislike: Puzzles that are too hard (this is obviously subjective), having to walk back and forth a lot to solve a puzzle, and things that are hard to see. Firmament has a few spots where there were things I just couldn’t see and I had to use a walkthrough – a video one in one case – to figure out what I was missing. This is frustrating because it feels like I just was never going to figure it out on my own. The game does pretty well on the other two points, although there was one puzzle I didn’t so much figure out as stumble into the answer for. Better lucky than good, I guess?

The game’s weakness, I think, was its story. Since Cyan’s games are solo endeavors with little capability for you to interact with anyone in the game, they all take place in environments where the people who used to be there are gone, and finding out what happened to them is part of the adventure. Firmament feels pretty thin, here, as the Mentor and one other character are the only ones you learn much about. There were clearly more people around, but we learn very little about them. I think they could have threaded more characters and more events into the game and provided a richer story to explore. As it was, it definitely felt less sophisticated than Obduction.

(I’m inclined to think that the use of AI to assist in generating parts of the game are not really at fault as this article thinks they might be. I think they just didn’t spend enough time coming up with enough story to make it satisfying.)

I think the game took me about 15 hours to complete. I did run into one bug, but it turned out not to affect me in that puzzle. Other people have run into more serious bugs, but they’ve been fixing them. If you enjoy games like these, give this one a try, but temper your expectations, especially if story is your main interest.

A few more spoilery comments after the cut:

Continue reading “Firmament”

Knotwords

Today I’m going to sing the praises of a very clever and fun word puzzle app I’ve been playing on iOS: Knotwords, by Zach Gage and Jack Schlesinger.

Knotwords is like a second cousin to crossword puzzles. You’re given a grid of rows and columns, which is broken up into sections, and for each section you’re given the letters which go into the squares in that section. So there might be a section of 5 squares, which includes pieces of 2 or 3 rows or columns, and you need to figure out which letters go where. The trick is that each row or column is a complete English word, often consisting of 2 or more sections.

Here’s a recent simple puzzle and its solution (it briefly explodes the grid when you solve it, which is what I captured for this image):

What makes Knotwords enjoyable for me is that it exploits the human mind’s ability to pattern-match, but it also allows the player to break down the puzzle into smaller pieces, building up words out of multiple sections as you go. My approach – especially with larger puzzles – is to identify the sections which appear most tractable to figure out the word, and then build out from there.

Sometimes the sections allow you to reason about the letters through knowledge of English words. For example, if there’s a section of 5 squares, with two 3-letter words which overlap, you can sometimes figure out that there are a limited number of ways to combine the letter to make valid words. Vowel placement is often key. Sometimes you can eliminate combinations, for example knowing that there’s no English word which begins with the letters ‘BK’. Knowing the 2-letter words is also helpful. The app also tells you when you’ve entered an invalid word, so you’re not just guessing. And you can ask for a hint which gives you the dictionary definition of a word you choose on the grid, but it’s rewarding to solve puzzles with a few hints as possible.

Knotwords does reward having a good vocabulary, but the words are not usually rare. I’m more likely to spell a valid word and go “What the heck is that?” and then look at it differently and realize it’s a perfectly ordinary word. I’ve solved a couple hundred puzzles so far and maybe only encountered a couple of words I didn’t know.

The system is self-correcting in a way. Sometimes you build up a valid word, but it leads to other words being invalid, and then rearranging letters to fix things up untangles the other words. You quickly develop an instinct for which words are almost certainly correct (spoiler: longer words usually have only one solution).

If that sounds good to you, then you’ll be happy to learn that there’s even more to discover. In addition to the basic puzzles, there are “twist” puzzles, where each row and column has a number indicating how many vowels are in that line. For example:

Knotwords twist puzzle

And the solution. Whenever you solve a puzzle you get a happy little bunny person in the bottom right cheering you on:

Knotwords twist puzzle solved

Usually additional constraints in puzzles like these make things easier for the player (for example the “Sudoku X” variant of Sudoku), but that also means the twist puzzles can be harder because you have more information to work with.

Every day you get a new simple puzzle, a classic puzzle which gets larger and harder each day from Monday through Sunday, and a twist puzzle which also gets harder through the week.

Each month there are two puzzlebooks, one classic and one twist, each with thirty puzzles. The later puzzles in these have additional twists, such as word themes. At first I was intimidated by the puzzlebooks as I got to the midpoint, but while they get pretty large, it’s not too hard to work through them. While it takes me upwards of 25 minutes to get through the largest ones, you can always pause and come back to it later.

Knotwords has built-in gamification, such as tracking your stats on the daily puzzles, achievements, and of course tracking your streaks – which you can recover if you happen to break one!

One thing to know is that the app does not yet support syncing your progress across devices; as a result, I play it only on a single iPad. (Apparently syncing is in the works, though!)

I’ve never really made the leap to doing traditional puzzles like crosswords and Sudoku on computer, but Knotwords fills that space nicely, and in a way that wouldn’t be nearly as good on paper. It’s free to try, with an in-app purchase to unlock the full range of puzzles. I didn’t hesitate to buy it once I finished the trial, and I hope this post helps it find a few more fans.

Zed (and other Myst-like Games)

I spent a few hours on my recent vacation playing through the computer game Zed. I’d backed it on Kickstarter a while ago, and it was eventually published with the help of Cyan Ventures, an arm of the company which produced the Myst series, which I adore.

Unfortunately, I was disappointed by Zed, though it helped clarify what I enjoy about games of this sort. While the art, sound, animation, etc., are all important elements in providing a sense of being present in the setting of the game (which was the breakthrough triumph of Myst, I think), the main factors are having an engaging story, and having interesting gameplay usually puzzles or challenges to walk through the story (and usually not combat). The gameplay also allows the player a certain amount of agency, or at least the illusion thereof.

Cyan’s games do a good job of balancing both elements, particularly in Myst and Riven, as well as the sequel (not by Cyan) Myst III Exile. Cyan’s most recent game, Obduction, also hit this sweet spot for me. (I reviewed it here, though it seems I thought some of its puzzles were a bit too far on the hard side, which I’d forgotten.)

Zed is heavy on the story but very light on gameplay. The framework is that you’re playing the role of an artist with dementia walking through memories of his life to collect ideas for a final gift for his granddaughter, but it’s a heavily guided experience where you roam regions of his memory in sequence, being exposed to the narrative of his life, and collecting a small set of objects in each area, but that’s really all there is to it. There are no puzzles, nothing else really to do, and negligible agency. It looks great, the story works pretty well, but it feels like doing a walkthrough of a game rather than playing a game. I worked through it in about 4 hours (by contrast, Obduction took me about 20 hours).

Zed is the first game of this type that I’ve played which has leaned so far in this direction; I’ve seen some which leaned too far in the other direction. I vaguely recall a 90s game called Obsidian which I played and felt was all gameplay (and surreal settings) and not much story. (I didn’t finish it.) Quern: Undying Thoughts is a more recent example: It’s full of puzzles (and takes a long time to work through them), but the story is pretty thin.

Another nuance is when the puzzles are too obscure or difficult, or which are tedious because they involve too much walking around (which takes time and is no fun if you’re not discovering anything new). Myst IV: Revelation is unfortunately an example of this, with several puzzles that made very little sense to me, and I ended up using a walkthrough for a lot of it. (The story was pretty good, though.) I suspect Myst V: End of Ages was similar, but the game was buggy enough (I think it didn’t play well with the video card I had in my Mac at the time) that I didn’t get very deep into it before getting frustrated and giving up. Quern had the too-much-walking-around problem in spades.

Anyway, I do love this style of game, and will play most games of this type that I come across as long as they’re on platforms I own (Mac and iPad, basically). Here are some others I’ve played:

  • Alida (2004) This one was pretty well balanced, with maybe a couple of puzzles that were too obscure.
  • The Talos Principle (2014): I’ve been playing on my iPad. The puzzles are pretty good, but the story is nearly nonexistent.
  • The Witness (2016): All puzzles – often very frustrating puzzles – and no story at all. I ha-a-ated this game and gave up after about 5 hours. The graphics are pretty mediocre for a modern game, too.
  • Tipping Point (2007?): Another game I’ve played on my iPad. It’s okay but I lost interest about halfway through and haven’t gotten back to it. I’m not quite sure why I haven’t found it satisfying.
  • Grim Fandango: Originally released in 1998, I bought the remastered iPad version when it came out a few years ago. I’m not sure this game really belongs in this category, though it seems adjacent at least. I didn’t get very far in it because it involved a lot of walking around from place to place, and frankly I got bored. It sure is stylish, though!

Are there are others currently available that I should try?

I’ll also try games that are clearly not really intending to quite be this sort of game but are similar in some key ways. Some games by Simogo feel adjacent to Myst but not quite the same thing. Device 6 is more of a story with a few small puzzles, as is The Sailor’s Dream. I enjoyed both, although Sailor left me feeling a bit empty at the end. I also tried Year Walk, but it felt like it was all walking around and not much progress.

Anyway, my disappointment with Zed isn’t going to dissuade me from playing more games of this sort. In fact, I backed Cyan’s next game on Kickstarter, Firmament, once they committed to a Mac version. And heck, I kinda feel like playing through Obduction again.

Tower Defense

In case I didn’t need another way to waste time, I recently discovered the tower defense genre of computer strategy games. Specifically, I discovered them for my iPhone. I think this puts me, what, about 3 years behind the curve for the genre, and a year behind for the platform?

Anyway, Tower defense games involve placing towers on a map in order to fend off invading hordes of creatures. The towers are statically placed, but they can be upgraded or torn down. You have a certain number of resources with which to build towers, but you can more resources as you fight off each wave of attackers.

I was initially intrigued when I saw the demo during this year’s WWDC of the game Star Defense (links to individual games herein will take you to the App Store in iTunes). Of course, that was months ago, and I just this weekend got around to downloading some tower defense games. I actually decided not to start with Star Defense since it seemed like a relatively advanced example of the genre, with 3-D maps where many others have 2-D maps.

A cow-orker of mine pointed me at TapDefense, in which the hordes of hell are trying to storm the gates of heaven, and your towers all have medieval or magical themes. TapDefense has the cardinal advantage of being free. It also has the advantage for a newbie of having good built-in help, as well as a tutorial.

But one of the nifty things about the App Store is that so many good products are quite cheap. So I bought two more which seemed to have good reviews: geoDefense, and Sentinel: Mars Defense, which were both only 99 cents. I ended up going right to Sentinel mainly based on this review of its sequel, Sentinel 2:Earth Defense (which itself is only $2.99).

Sentinel has great graphics and sound, but I’m glad I didn’t make it my first-ever tower defense game, since its help is pretty minimal. On the other hand, having had that first experience, it was pretty easy to figure out what to do. The bad guys come in five varieties (fast-and-wimpy, slow-and-tough, flying, teleporting, and big-slow-and-really-really-tough) and each wave consists of one type of baddies which are tougher and more numerous than the last batch you saw of that type. So you need to diversify your towers to deal with all the different types, but you get a bonus if you spend minimal resources in doing so. The Easy setting is really, really easy, while the Hard setting is pretty challenging.

The tower defense genre seems to be a comparatively passive game, where you place a tower or two, do a few upgrades, and then see if your changes deal with the attackers. If they don’t, then you may need to quickly place a few extra towers to deal with any who got by, but for the most part you’re watching the results of your handiwork, which is fun, but also a bit monotonous – in a hypnotic way. I found that a half an hour slipped by in my first game of Sentinel before I knew it – it didn’t feel that long.

As a mix of combat game and puzzle, the genre appeals to me, although the monotony makes me wonder if it will have any staying power with me. Though I’m not going to judge the whole genre on just a couple of examples, as it’s easy to envision variations on the theme. But it’s something new and different to me, and it runs on my phone – a feature of the iPhone I’ve underutilized, this game-playing thing – so I’m going to give it a whirl.