Saturday, 10 June 2023

Marvel Snap

Per my wont, I have become obsessed with another game: the 3 lane card battler Marvel Snap. I've been largely uninterested in large deck card games since 1995 because the deck building possibility space is too big - I always ran as slim and simple a Magic deck as possible using cards that were just clearly better than others like "true" dual lands and lightning bolt. Even Imbroglio's deck building is too complex for me, but the difference between 16 locations and the 12 cards in the Marvel Snap deck is apparently enough for me to dive deep into the collection and play mechanics.

I've mostly been posting bite-sized takes on Twitter which includes some microblogging about the game. A good recent example is my thoughts on Move decks at https://twitter.com/andrewdoull/status/1662591178708242433 which seems to have largely predicted the outcome of the latest card addition Ghost-Spider. If you're not familiar with the game, I summarize it at https://twitter.com/andrewdoull/status/1661002185960980482 and I highly recommend you read through that before proceeding further if you haven't played the game recently or at all.

I want to emphasize here that the game design is fundamentally solid and the designers are operating at a level of ongoing excellence that I've not seen in a live service game (Caveat: I wasn't around for the early days of Slay the Spire). A lot of Marvel Snap players are probably coughing into their tea at this statement - Second Dinner, the developers, apparently designed the game in 2 days and then took 4 years coming up with the card acquisition model, and that part of the overall experience is very rough at the moment. Undoubtedly there's a lot of internal pressures between monetization, community pressure to do the right thing, and (hopefully) pivoting to an improved card collection mechanic that they've been promising. The big error they've made is they're trying to hold back some clearly good but not great cards in order to do something with them, after creating the expectation that those cards were going to drop down to a more accessible tier in the acquisition mechanic.

That stuff doesn't especially interest me. As I've said elsewhere, you should play with the cards you have, not the cards you want, and besides I have some hot takes that the held back cards aren't that interesting (except for Zabu. And MODOK, for reasons I outline below).

What does interest me is the deck building space on the margins of the game, such as this deck which depends on you drawing 3 cards and playing them on turns 4, 5 and 6, and works extremely hard to make that happen. Part of how it works is the surprise of playing Spectrum as opposed to much more frequently played cards on a Wong -> Mystique combo, and the opponents realization that the low powered cards you played are because they have the Ongoing keyword rather than their abilities. I outlined in more detail the principles at play at https://twitter.com/andrewdoull/status/1665511669135740928.

That this deck is possible, is because of the redesign of the card Crystal, which used to work like the location Attilan but only if you played it in the center location. Because of tempo, there was almost never any benefit to playing the old Crystral, whereas the redesigned Crystal is now useful in decks looking for specific card combinations, like the one above. The devs have provided some insight into their process for Crystal specifically on Discord, an entirely transient and non-linkable medium, but luckily there are screen shots at https://twitter.com/MarvelSnapBugle/status/1667021988995399688 - in essence, the cost of redesigning a card from scratch is high enough that they're better off making a new card in almost all instances.

Which leads me to a couple of conclusions. One of which, naturally, is that I'm going to backseat design some new cards. The other of which is we haven't seen the full picture - the cards that currently look underpowered are either that way because there's new cards coming that will indirectly buff them (one new card referenced is likely Legion), because it's too hard to buff them cleanly by just adjusting their power or cost, or because they're actually better than expected.

A common complaint is Second Dinner is much happier to nerf cards than buff them. This is inevitably true, as any live service game is prone to power creep - I think the meta at the start of the current season at time of writing (Spider-Versus) was just fractionally too good to allow for much fun experimenting with decks and while the latest over the air (OTA) patch has gone some way to address this, bounce decks are likely still too strong. What interests me though is the cards that feel too weak at the moment but haven't been buffed - examples might include Strong Guy and Black Cat - because they point to data that isn't publicly available which suggests they're performing better than expected in decks which aren't in the current meta.

This is especially true if there isn't an upcoming card which will potentially buff them. The next season's cards have been announced - some of my initial thoughts on Phoenix Force, the July season card are at https://twitter.com/andrewdoull/status/1662620224410238976 which starts (hilariously in retrospect) ignoring the most obvious synergies in favour of a bunch of weird edge cases which may not even work at all. But, for instance, Strong Guy and Black Cat don't synergize with any upcoming cards  - except Black Cat with Night Nurse, and that's only going to be useful if you have The Collector out. I'm not entirely ruling this out, because +1 or -1 power is often significant enough a buff or nerf to make a difference, but I don't think its enough in this instance.

So that is a strong hint that there's a deck using those discard cards that wins a lot of games. I've seen something similar with Shanna - there's some viable Shanna decks out there that won't really hit the meta until Shanna drops to tier 3 this month which effectively means it's in the main card acquisition pool. Shanna works because you can get to around 20 points in each lane if locations permit, and that's the typical threshold for winning 2 lanes. If the locations permit is the big caveat which is why this deck isn't played that often (and why Legion is likely to be a big buff for it).

So let's talk MODOK. MODOK discards all your cards, and should theoretically synergize with Strong Guy who gets +6 power, but in reality MODOK is only used in Hela discard decks, which normally play Apocalypse and Apocalypse cannot be discarded (he gets buffed instead). Some Hela decks instead run without Apocalypse, but they are frustratingly unreliable because they keep discarding Hela, unless MODOK is hidden behind Invisible woman turn 5 along with Hela on turn 6. For a so-called invisibility power, this is incredibly telegraphed and takes a lot of set up and luck.

But I ran into a deck about a month ago that absolutely wrecked me using a combination of cards I haven't seen before: it ran MODOK without Hela or Apocalypse, and instead relied on Collector and Morbius. This is also a natural synergy, but it normally uses a lot more discard cards, whereas this deck relied on card acquisition cards like Sentinal to have a 7 card hand on the final turn and dropping Agent 13 and Morbius. That's a +16 power with two cards, and an empty hand, which makes me think that Strong Guy would work well in the mix (for another +6 in another lane). So my theory is that there's a deck, that is just wrecking house, but no one is picking up on it whereas some game designer at Second Dinner is just smiling and waiting for people to figure it out.

The problem is I don't have MODOK to test it. Which goes back to the card acquisition dilemma I mentioned earlier.

Sixty Years in Space: Devlog 3

 

A lot of my concerns about the length and complexity of 60 Years in Space could simply be addressed by looking at the 5th edition Dungeons & Dragons Players Handbook and seeing how many rules players would be willing to put up with just for character generation. But there's two clear differences: there is an incentive to learn about what your character can do, which is cleverly partitioned by class so that - for instance - a barbarian never has to read about mage spells; and that a lot of the invented or borrowed Dungeons & Dragons lexicon has entered general language usage over the 49 years it has been played.

Science fiction has a similar challenge, except each science fiction world invents its own terms and adopts its own definitions for more general terms such as robot. In a fictional world, the author has control over the pacing and scope of the new language, using this to teach the reader at a pace they are comfortable with. By procedurally generating the world I could do something similar: have a massive list of all possible science fiction futures, but only have a limited number come up with each possible play through. I could even let the players win the game or have the world or the universe end - whereas most science fiction TTRPGs consist of a string of subgenre sand boxes which let players play in their favourite without spilling over into another.

This still doesn't solve the language problem. If I wrote the supplements using invisible ink and used a magic marker pen to reveal it as you play, I might be able to avoid you worrying about the names of every possible type of human-robot-animal-alien hybrid mix. But I have to be consistent in defining what a theriomorph, for instance, is so that when you see that word in the rules you know it means a human who has changed themselves to have an animal-inspired form. It helps, of course, if you are a classical scholar or speak Greek, but in the end I ended up using a lot of words that fall outside normal vocabulary. And as a result, much of the five books consists of the definitions of things.

I also coined some new words while doing this - apologies in advance if one of these is your band name or user handle. This is not an exhaustive list, but the most interesting.

  • amorcracy - rule by a group of people who sleep with each other
  • beginling - the first member of a species (from endling)
  • bisentient - intelligent being consisting of one body and two minds
  • fandemic - a behavioural modifying pathogen which makes infected both moderately addicted to a pleasurable activity and driven to  infect people around them to get them addicted to the same activity
  • posterhuman - human who has completely replaced their body with a robotic body, but retains a humanoid form for aesthetic reasons
  • skototropolis - a city that hangs underneath an aerostat
  • simgularity - a singularity that has occurred in a simulated environment and is therefore not transferable to the real world
  • storgearchy - rule by people who have children

  • Theseuvian - either a robot who has replaced themselves with human parts or a human who has replaced themselves with robot parts

  • tunafication - increasing proportion of fast twitch muscle fibers in meat
  • xenonaut - someone who explores digital environments looking for naturally evolved life