Epistemic Status: Covenant Rank 8 (meaning I’ve won 8 runs, one each on each of levels 0-7) after about 20 games played.

As is my template, a spoiler-free review seeks to answer the question “should I play this game?” without giving away any unnecessary information.

A strategy discussion may happen in another post in the future. Find the game here.

Should I Play This Game?

Short Answer

In this case, my answer to that is that you should play it if you are a fan of the game’s genre.

If you are a huge fan of such games, stop reading now and go play the game.

If you are a fan of such games but not a huge fan, read on to see if the game appeals.

If you are actively not a fan of such games, this game isn’t for you.

If you’ve never played a game of this type, play Slay the Spire first. Then consider DreamQuest, which I consider a tier 2 game, and/or Monster Train.

In my tier system, Slay the Spire is a Tier 1 game. It is a Must Play.

DreamQuest and Monster Train are Tier 2 games. They are Worth It, for those who can get into them.

Dicey Dungeons is a Tier 4 game in the genre. It’s Playable. For true genre lovers only. I enjoyed watching it do its thing but have no desire to explore it in depth.

I am always happy to play Tier 3 (Good) or better games in genres I like, if there are no known fresh Must Play or Worth It games available that scratch the same itch. Tier 2 or higher games are worth playing, to me, in most genres, provided they agree with you.

Core Mechanics

Monster Train is a deckbuilding roguelike in the style of games like Slay the Spire and DreamQuest.

You start out with a deck of mostly basic cards that are mostly quite bad. Between battles, you get to do things that improve that deck. You get good cards that fit your strategy. You get rid of bad cards that don’t fit your strategy. You improve and duplicate your cards. You get artifacts that give you static advantages, similar to relics in Slay the Spire.

Each battle, you face several waves of monsters. They mostly but not always start at the bottom train level.

Each turn, you draw a hand of cards then spend our local version of energy/mana/fire to play them. Your fire replenishes each turn. You can play monsters to fight on each of three levels, provided there is space for them, and you can also play spells. Monsters that get deployed don’t go back in your deck, whereas spells stick around. When you run out of cards, your discard pile gets shuffled again to become your deck. All standard stuff.

Then on each level, they attack your stuff, your stuff attacks back if it survived, then any surviving attackers move up a level. If they survive at the top level they get to fight your pyre, which takes damage that lasts between battles. If your pyre goes to zero, you lose. At the end of each fight, the boss starts fighting. The boss will keep battling whatever is on each level continuously until one side is dead.

If you beat the Seraph at the end, you reignite the fires of Hell and win the run. You score points, unlock new cards, and the difficulty level slowly increases, as embodied by your Covenant rank. That’s similar to Ascension levels in Slay the Spire.

Like Slay the Spire, the game has its own unique art style and attitude, and music and sound effects that support its aesthetics. I don’t consider it as good at this as Slay the Spire, but it’s still effective at what it needs to do, in the sense that DreamQuest really isn’t and succeeds despite its graphics being resonant but also really really bad.

Monster Train has been growing on me. Games remain fun. It’s better and deeper than it looked at first glance. It’s possible it will end up as a Tier 2 game by the time I finish with it.

If this sounds fun to you, stop here and go in without further information. Come back once you have won a run, or at least fought the Seraph.

If you are still not sure, or you’re back after playing a bit, here’s further thoughts that are still mostly spoiler-free.

Picking a Theme

Monster Train is a game of wild swings in power level.

There are a handful of themes that can get out of hand quickly. If you don’t succeed at one of them, you won’t win. You can’t hedge your bets and you can’t play things by halves. It won’t work. Pick a side. If you start down a path and the tools don’t appear, you’ll die to one of the later battles. If you’re very lucky you can pivot, but that seems rare.

The alternative is certain death. If you don’t start down a path, you die every time.

The choice of major enhancements makes this even more clear. You can choose, after fights three and six, either to gain an extra card draw each turn, an extra fire to play cards, or an extra capacity point on floors.

I have yet to see a good case for choosing card draw, but there are definitely decks that want double capacity and decks that want at least one fire instead. This locks you in to your plans even more than before.

By the middle of a run, around floor four, you know exactly what you are planning to do and what needs to happen.

Often, you’ll know exactly which problems are the ones that might take you down. There are a handful of specific problems decks need to solve to survive the last few battles. Many strategies have one particular sub-problem that is especially likely to kill them, and will otherwise probably win.

Know Your Roots

It might not look like it at first glance, but I think the game has as much in common with DreamQuest as it does with Slay the Spire.

You’re warned from the beginning about what final boss configuration you are up against.

You are on a very tight deadline to expand your power level at times. Other times, you’re essentially done halfway through. Much of what you find will be either amazing or useless. When you get a chance to shrink your deck, you freaking pounce on it.

Each class has tight builds that can work, which you have to aim for almost from the start. In Slay the Spire archetypes are not the right way to think about the game. In DreamQuest or Monster Train, they are absolutely the right way to think about the game.

Note that I still say Class here rather than Class Pair. In my experience, you lean into one or the other theme, and use only a little from the other. There are good units and a few good spells everywhere, but you can’t mix multiple card-intensive themes if you want to win. Trying to do both at once can work if you’re being handed the world, but mostly not otherwise.

Some of the cards, even without upgrading, are stupidly powerful. With upgrades, those cards become absurd. Getting and supporting those cards is usually key to victory. At a minimum, you need to put a strong upgrade set on multiple creatures, or you’re toast.

There’s no question the game incorporates a lot of innovations from Slay the Spire too. The branching paths, the card choices, the upgrading and so on are clearly refinements of those ideas, for better or for worse.

Pros and Cons

These are the big ones that stand out for me, including stuff that didn’t fit into the above.

While there are a lot of cons listed, that’s because I am holding the game to a very high standard.


Genre tends to be insanely great by default, and it’s a new twist on that genre.

Many problems feel importantly different than past games in genre.

Decisions on how to configure your team across three levels can be very tricky and complex in interesting ways.

Lot of the stuff feels good to do. Good fun factor.

Remarkably deep planning is possible, and becomes necessary on harder levels.

Consistent aesthetic is well executed.

Short journey makes each game a contained unit.


Runs and battles can both take a while after they are mostly or entirely over.

Once you know your path you often don’t make many interesting decisions.

No undo button for clear mistakes even more frustrating than in Slay the Spire.

Absurdly unbalanced in many places. Some crazy good relics and events and cards.

And there’s the snowball effect of getting an amazing thing and getting to copy it.

Short journey forces focus on endgame from the start. Things that don’t work at the end are non-starters at the start. There’s only really one ‘transition’ in deck construction, which is after the first 2-3 battles, then that’s it.

First few battles have low stakes, can feel like wasting time once you’re used to them.

Pyre health doesn’t trade off with other things in interesting ways often enough. When things get through and actually kill you, usually they massively overkill you.

Game stats point to stupid stuff and leave out stuff that I’d be interested in, please fix.

Unlocks are really, really slow. Game may look a lot less interesting than it is because of this.

Runs start to blur together. Not enough different cards.







New Comment
4 comments, sorted by Click to highlight new comments since: Today at 11:06 PM
Note that I still say Class here rather than Class Pair. In my experience, you lean into one or the other theme, and use only a little from the other. There are good units and a few good spells everywhere, but you can’t mix multiple card-intensive themes if you want to win.

This feels often true, but I just noticed that Hellhorned/Remnant can work really well together. Spam imps for their effects, let them die / kill them off with Inferno or Imp-portant Work, then reform them and repeat.

Had a particularly enjoyable Covenant 9 run ( monstertrain://runresult/21069f33-c51b-4b3f-b12a-a22a6c4c7c52 ) where I also lucked into having Flicker's Liquor and it was just insane, all of my heavy hitters on the top floor and the lower floors swarming with imps and dregs. I was mostly playing them for their effects and didn't even care if they killed enemies (because the top floor would have taken care of anyone anyway), but they had been reformed so many times that they actually dealt significant damage on their own. Infernos helped too, and many enemy waves never even made it to the top. Seraph was consuming spells but my Impish Scholars meant that that didn't really even matter: often the consumed spell would already be back in my hand before the turn was over. At best I was playing something like four copies of Onehorn's Tome (with a total of two copies in my deck) a turn, and the battle was over before it ever even got to the final wave.

Alas, it isn't out on Mac, so I can't play it, but I will give DreamQuest a shot, which I haven't heard about until now.

25 hours played, this game has been growing on me as well. Though I too would like it if there was more room for creative improvisation, as in Slay the Spire, rather than strict optimization. A lot of the units have just outright fun concepts, and it would be nice if mixing and matching them more freely would make gameplay sense. Often I just do so anyway, even if it means that I'm less likely to win. (This might explain why I still mostly lose even on Covenant Rank 1.)

An interesting thing that I've noticed, and people seem to agree with, is that the game feels shorter than Slay the Spire despite one run taking about the same time in real-world minutes. (My successful StS runs are about 1-2 hours; Monster Train, about an hour and a half.) When I finished my first run, I assumed that it was just the first world, since it felt roughly like the end of the first act in StS. That's an interesting psychological observation by itself. Seems to be a result of the upgrade-battle-upgrade loop having been reduced to a much smaller number of battles and upgrade points, but with those points being correspondingly larger.

25 hours played, this game has been growing on me as well. Though I too would like it if there was more room for creative improvisation, as in Slay the Spire, rather than strict optimization. A lot of the units have just outright fun concepts, and it would be nice if mixing and matching them more freely would make gameplay sense. 

At 220 hours played and with Covenant 25 wins on 14 of the base game's 20 possible class pairings, I don't feel this way anymore. Improvisation and relentless optimization aren't opposed; the need for relentless optimization just requires me to improvise better. It just required getting familiar enough with the mechanics and cards to be able to do that.

New to LessWrong?