Thoughts: Möbius Front ’83

mobius_trucks

If you go back and read one of my earliest reviews on this blog, you might notice a slightly different tone from the one I try to take today. Back then I took bad games rather… personally (this was before I’d formulated my Bad Game Theory1) and if I played a bad one I had a tendency to lay into not only the game in question but also the developers behind it, and with quite a bit of venom, too. Now that I’m older, wiser, and a little bit calmer, I can see now that this was unreasonable and unfair. Making videogames is hard. There’s any number of reasons why a game might not come together in the way it should, many of which are wholly or partially out of a developers’ control — not enough budget, truncated schedule, interfering executives, key personnel leaving, and so on. Nobody sets out to deliberately make a bad game, and so these days while I’m perfectly happy to continue giving bad games the kicking they so richly deserve, and while I might still have some choice barbs for the corporate aspect of game development, I try and give the individual developers themselves the benefit of the doubt. It didn’t come together for them this time. Maybe it will next time. There’s no reason to get nasty about it.

Or so I thought, until I played Möbius Front ‘83.

The more I play Möbius Front ‘83, the more I find it is seriously testing my newfound equanimity.   It’s the latest game from Zachtronics, and I already have something of a checkered history with those; I started out loving the early Zachtronics puzzlers such as SpaceChem and Infinifactory, but the more of them Zach cranked out (he’s up to something like seven or eight of them now) the more their core idea of finding puzzle analogues for programming started to wear thin. This culminated in a rather ill-tempered review of Shenzhen I/O, a game which was literally about programming and which pissed me off to no end because it forced me to program badly in order to solve the puzzles. After that I swore off Zachtronics games for the foreseeable future; after all, if Zach’s schtick is to make the same game over and over again with different set dressing then I can hardly complain that I don’t know what I’m getting. I told myself that there was no point buying another one until he made something different, and which was relevant to my interests.

mobius_smoke

Then, suddenly, there was Möbius Front ‘83, a Zachtronics game that was both seemingly not a puzzler, and which was extremely relevant to my interests. It’s a turn-based Advance Wars-style strategy game set during the Cold War, with some light wargaming elements that put it about halfway between Advance Wars and Panzer General in terms of complexity. The premise has the standard Zachtronics sci-fi spin where the US is being invaded by a US from another dimension, which does at least provide a suitably non-jingoistic adversary for your 1980s-era US military forces to battle. Unlike Advance Wars there’s no in-mission economy and unlike Panzer General you don’t buy troops in between missions; instead each mission progresses through three five-turn phases (although there’s no upper limit on how long a mission can take), with both yourself and your computer opponent receiving a set number of points to buy things from a limited list of troop types at the start of every phase. In theory each mission has you accomplishing some objective, like taking all the victory points or destroying a set of SAM launchers, but in practice each one plays out exactly the same way: once you’ve purchased your tanks and infantry you rush them forward onto the map to try and take up a good defensive position because you have maybe two turns before the lead elements of the CPU force makes contact, at which point your only real option is to kill all of them before advancing to the objective over their burned and broken bodies.

There are several reasons (some might even call them “catastrophic design fuckups”, but we’ll stick to “reasons”) the battles play out the same every time. First is the AI. I have seen the AI referred to in the few reviews that exist on Steam for Möbius Front, as well as on the Steam Discussion boards, as “smart”. I vehemently disagree with this assessment, although I can see how it would at first appear to be smart as it constantly maneuvers units into positions where it can fire at you without reply. If you try and move your own forces up to deal with the threats and it thinks it’ll lose a unit, it’ll just run away. This makes it an annoying opponent to fight, but this is not the same thing as smart, especially because literally everything in the game is stacked in the AI’s favour:

  • Despite fighting against an alternate-dimension USA, the two opposing forces in Möbius Front are not symmetrical. The AI has access to anti-tank units which are as fast or faster than yours, which drastically outrange them (most of your ATGMs are range 3 whereas all of its ATGMs are range 6), and which do not need to waste a turn setting up before they can fire.
  • The AI can afford more of these super-units than you can since I’m reasonably sure at this point that every single one of its units, with the possible exception of the tanks, costs a single deployment point.  The only units you have that cost one deployment point are truck-borne infantry and soft-skinned support vehicles like jeeps and mortar carriers. Since the AI usually has static assets already deployed onto the map in addition to what it can deploy normally, it will always outnumber you.
  • The AI has perfect information and always knows exactly where your units are and what your units are. I’ve tested this by replaying the start of a couple of missions several times with different force compositions, and on the deployments where I started with a pure infantry force it always advanced its armour assets much further forward than when I mixed in some ATGM teams, where it was careful to keep them out of range of the ATGM positions despite having no legitimate way of knowing they were there.

mobius_helicopters

Strangely enough it’s quite easy for an omniscient AI whose anti-tank missiles have twice the range of yours to always be able to find a position where it can safely shoot you. Compounding this is the massive RNG inherent in each anti-tank shot, where a tank shell or an ATGM will do anywhere between two and six points of damage. Armoured vehicles have three HP and between one and three points of armour that absorb damage before it bleeds through to the vehicle’s health. What this means is that, if you’re shooting a tank which has three health and two points of armour, the outcome can be anywhere between “it does two points of damage and bounces off the armour with no effect” and “it does five or six points of damage and one-shots the tank”. The super ATGMs and fast LAVs the AI can field only have a single point of armour, but that doesn’t matter because the only way you can get into range with a tank is to move and fire, and firing on the move reduces your damage to between one and four for the same range of outcomes. If you attack with one tank you’re hoping you roll a four, otherwise the return fire will likely kill you. If you attack with two tanks you need to roll at least a two and a three, or at least one four. These are still not great odds, especially since the AI outnumbers you and so your target will likely have friends move in to blast the hell out of your exposed tanks the moment they move out of a safe position — and since more units means they get more damage rolls they have much more chance of rolling a lucky 6 and annihilating your poor tanks.  The variance of anti-tank damage is so high that you cannot ever predict how an attack will go, and since you cannot afford many armoured assets to begin with (and losing them often spells doom for a mission as you do need them to support the actually useful elements of your force) it’s absolutely not worth the risk of exposing them.

So attempting to control Möbius Front’s open spaces with tanks and armoured vehicles is impossible; you will be swarmed and destroyed with no possibility of success until all of the enemy anti-tank vehicles are dead. This is a task that falls to your infantry, and the infantry war at least has some interesting design to it. All infantry come as a multibuy package complete with their own transport vehicle — so one point buys a truck and two Rifle Squads, two points buys an M113 APC, two Rifle Squads and an ATGM team — and infantry squads are also surprisingly resilient thanks to the low damage of anti-infantry weapons in Möbius Front actually making the outcome of infantry vs infantry combat predictable. Most anti-infantry attacks do just one damage, with larger calibre weapons doing between one and two if the shooting platform didn’t move beforehand. As a Rifle Squad has four HP this lets them weather up to four attacks before dying — a marked contrast to your cardboard tanks, which will be instantly deleted from the map as soon as they move out of cover.  The catch is that Rifle Squads only have a range of 1 (i.e. they can only shoot things that are adjacent to them) and vehicle weapons have a range of 3, allowing them to shred your poor infantry if they’re caught in the open. This is why each Möbius Front battleground is covered in a copious number of forest tiles, which both block enemy line-of-sight past the outermost tile — critical for hiding your armoured vehicles from those ATGM swarms — and which infantry can conceal themselves inside; vehicles cannot move into forest tiles and they cannot attack infantry inside forest tiles unless they’re adjacent to it, which brings them into range of a Rifle Squad’s LAW launchers.

mobius_writing

This is why battles play out the same way every time, then; the enemy forces are so fast and powerful and numerous that you have to get your infantry into the nearest forest and create a chokepoint before they arrive. There’s next to no scope for tactical movement here as isolated forces will be hunted down and destroyed (remember, the AI always knows where you are) and you need all of your troops in one place to successfully fend off their attack and destroy them. If you do manage to claim a nice patch of forest, however, you’ll notice the last quirk of Möbius Front’s AI: it is pathologically terrified of actually going on the attack. Its sheer speed allows it to make contact quickly and it’ll happily massacre your vehicles if it can do so without being shot in return, but if there’s any move that would actually require it to move within range of your weapons without one-shotting the target it’ll instead choose to hang back one hex. It will continue to do this indefinitely — it won’t try to outflank if you have the obvious approaches covered, or to punch through your line with massed forces, it’ll just sit there and wait for you to attack it.

As your tanks are much more useful as a deterrent (a single tank “hidden” behind your lines will prevent the psychic AI from rushing past your infantry with its fast units as it’s too scared of losing one to the tank) than they are as actual breakthrough tools, each of Möbius Front’s battles has much more in common with World War 1-era trench warfare than it does a Cold War-style battle of maneuver. After the initial frantic rush to the forest, most of your turns are spent slowly and painstakingly maneuvering a huge clump of infantry around to try and get them into a favourable position for a massed assault. Coming out of the forest and crossing open spaces is risky — although not impossible if you have the numbers or favourable terrain — but pretty soon you learn that the AI does have a couple of quirks that you can abuse to bait its armoured vehicles into LAW range. In particular it hates ATGM teams, which only have two health; these are the one unit type that it will drive right up to the forest edge in order to kill before they can get set up. I can count on the fingers of one hand the number of times an ATGM team has survived to actually fire their missiles once they move into a hex the AI can attack, but I have a much, much larger set of examples where the sacrifice of the ATGM team has enabled me to kill one or two armoured vehicles in reply, as they had to drive into LAW range in order to do the deed. Another trick is to move away from a forest edge for a turn, baiting the AI into thinking that hex is safe and moving there, and then moving back; the AI normally won’t move into range of your infantry, but if you move your infantry into an attacking position it’ll elect to attack you first instead of running away, enabling a kill.

(The AI does have its own infantry forces, but it’s noticeably much worse at using these than it is armour because they’re pretty much identical to yours and it’ll never ever make attacking moves with them, even when you’re shelling it to death with indirect fire weapons and artillery. Mostly it’s just an annoyance that slows you down rather than a genuine obstacle.)

mobius_air

When I’ve reached the point of describing, in detail, the AI glitches that enable me to actually win Möbius Front’s missions, it’s quite apparent that what we’re dealing with here isn’t really a functioning wargame. A proper wargame would give me multiple potential strategies and interesting risk-reward decisions, whereas I just play Möbius Front the same way every time, not because I want to, but because I have to. The AI simply has too many advantages to win conventionally. Trust me, I’ve tried many times, only to lose expensive tanks and APCs and helicopters to those RNG anti-tank damage rolls again and again. When I’m facing that level of bullshit, the only thing that I can do is fall back on spamming the most cost-effective units  available to me — infantry — and tricking the AI into making dumb moves so that I can kill it. The numbers are all so massively unbalanced in its favour that I do wonder what on earth happened during Möbius Front’s development to make it so. Was it an attempt to make an actually decent AI that fell apart, and so inflating the stats and numbers and giving it panopticon vision was the only way Zach could make it a challenge? Was he operating in a tiny bubble with zero outside feedback for the duration of the game’s development, and so thinks that having to exploit an AI that he wrote and knows inside out is acceptable? I can’t for one second believe that this is the sort of thing that external playtesters look at and go “Yep, this is A-OK!”

In fact I can’t believe that Möbius Front was playtested at all, and this is because of the thing that finally broke me: Chapter 2, Mission 7. This is a perfect storm of everything that’s bad about Möbius Front, and which serves as a fantastic summary of why the game fucking sucks.

  • It’s the first mission where the AI deploys significant numbers of helicopters on Turn 1; helicopters have a move range of 10 and can move across the entire map in two turns, and these particular helicopters have anti-armour rockets that they can fire after moving.
  • The AI gets the first turn, during which it moves the helicopters halfway towards your position. Your deployment phase counts as your first turn, during which all you can do is drive your troops onto the map – you cannot set up your SAM units or unload infantry from their transports. If you do not have a unit that can immediately shoot the helicopters, on the AI’s turn 2 it will fly the helicopters the rest of the way to your deployment zone and start shooting up your vulnerable tanks and troop transports. Usually it will roll 4s and one-shot everything.
  • You have access to precisely two units which can shoot the helicopters. One of these is the Chaparral SAM launcher, which requires a turn to set up and which you cannot set up on the same turn you deploy it onto the map. The other is the M1 Bradley, which can fire immediately and which does deter the helicopters for a turn, but which costs a whopping 3 deployment points out of an overwhelmingly generous total of 5 that you get for Phase 1. (The CPU gets 10 points, which is why it also has a shitload of armour that makes contact the turn after the helicopters.)
  • If you just buy Chaparrals they’ll be killed before they set up. If you buy one Chaparral and one Bradley you do get the time to set up the Chaparral, but then they’ll just fly in from outside range (it has a range of 6) and kill it (the Chaparral has three HP and no armour) before it can ever fire. The Bradley on its own cannot do enough damage to kill the helicopters before it is killed in turn. If you buy two Chaparrals and one Bradley you run into yet another problem, which is that the terrain is extremely unfavourable and doesn’t let you deploy the Chaparrals so that they can cover one another; the helicopters will always be able to kill at least one of them from a safe position.

mobius_aaargh

As far as I can tell there is no way to win this mission on Standard difficulty aside from hoping the helicopters roll low on the RNG and do not one-shot my Chaparrals (this has not happened on the last fifteen attempts, so I’m not holding my breath). There is nothing that you can deploy — nothing — that will not be slaughtered by the helicopters. This isn’t a matter of “oh, you’re just playing it wrong, maybe you just need to change your strategy” because those helicopters are coming at me before I’ve had a chance to implement any strategy. I can’t exploit any quirks in the AI because it doesn’t give me the time or resources to do so. And I’m tired of playing the first two turns of this mission over and over again, and watching my troops die over and over again with nothing I can do to stop it. I’m tired of constantly struggling uphill, of having to contend with the superior numbers and superior firepower and psychic powers that are propping up this otherwise braindead AI. I’m tired of every mission taking 30 or 40 turns because that’s how long it takes to kill the legions of AI trucks and tanks and troops, one unit at a time. I’m just absolutely staggered that this is a mission that exists in a wargame in 2020; it’s the epitome of Möbius Front’s miserable design ethos, and it’s what tipped me over into posting a rather foul-mouthed rant on the gaming Slack I’m part of, as for the first time in years I took a bad game personally and I was feeling spectacularly aggrieved towards the people who had inflicted this monstrous failure of design and balancing upon me.

There’s plenty of other things I could take Möbius Front to task for, such as the garbage UI which expects you to measure ranges manually instead of giving you any kind of in-game overlay telling you which hexes your troops can hit from a given position, but this is my biggest problem with it: it’s a vicious and unforgiving travesty of a game that is intentionally designed to punish aggression and constrain the player’s freedom of action so that there’s only one “valid” way to complete a mission2, but in order to find out what that is you need to try and fail, again and again, until you stumble on the right solution. It doesn’t have the safety valves that exist in the games that it’s modelled after; it doesn’t have the economy of Advance Wars or the wide-open spaces of Panzer General, and I think that this is quite intentional because that would give the player too many ways to complete a level. If I really wanted to go back to the bad old days of being nasty, I might even voice my suspicion that after making the same game seven times in a row Zach’s brain is so puzzle-poisoned he’s incapable of designing games outside of this paradigm now. Because make no mistake, Möbius Front might look like Advance Wars but it actually has far more in common with one of his more straitjacketed puzzle games like TIS-100 or Shenzhen I/O, except this time whenever your puzzle solution doesn’t match the required outputs the entire thing is deleted and you have to start all over again3. And that’s possibly the most upsetting thing of all: I swore I’d never play another Zachtronics puzzle game and I’m very annoyed that the Cold War theming and cute graphical style of Möbius Front have tricked me into buying the worst one he’s made yet.

  1. Which is that playing bad games is sometimes necessary in order to understand what a good one looks like.
  2. That doesn’t involve abusing the AI, anyway.
  3. There’s actually three very simple fixes that would make Möbius Front a halfway decent puzzler, if not a good wargame. First, remove the fog of war and give me vision over the entire map. Second, remove weapon RNG and have things do fixed amounts of damage so that I can predict and plan properly. And third, add an “Undo Move” button back to the start of the current turn.
Tagged , ,

8 thoughts on “Thoughts: Möbius Front ’83

  1. Nikola says:

    I had a much earlier and more prosaic negative reaction – it was *boring*:-/. I kept waiting for something interesting to happen,some quirk or variation or unique mechanic,a twist on the genre… But couldn’t find it. It felt like a mediocre low budget game that implements the same basics I’ve played a million times before,with nothing new it added to genre.

    Eventually I asked myself: “would I keep playing it if it weren’t Zachtronic and I’m expecting something interesting to come?”, and realized I’m not actually missing anything. It’s just not there :-/

    • Hentzau says:

      See, I was actually enjoying the opening few missions of Mobius Front because I thought there was a promising set of systems there, like Advance Wars crossed with Wargame. I can remember quite clearly the point at which I started to spit nails because it was at 2.2 hours played — just past the point where Steam will give an automatic refund.

  2. ilitarist says:

    That’s savage.

    Zachtronix game puzzle me cause they’re quickly becoming more boring than my programmer day job. Guess that wasn’t the worst thing about them.

    • Hentzau says:

      I just can’t understand why they’re all so similar. Well, I *can* — recycling the interface (for example) between Shenzhen I/O and Opus Magnum drastically lowers dev time, it’s why he’s put out eight games in nine years. But at this point a new Zachtronics game is more like a new Football Manager than anything else.

  3. Carlisle Dave says:

    Well this has saved me a purchase.

  4. I just got this and from every angle it just looks like “strategy wargame skin with a puzzle game under the hood”, so I imagine it’s like Wargroove’s “puzzle” mode but with longer missions and the whole game is like that.

  5. Interesting, perhaps there have been some updates since then because the AI seems to be a lot less static. Been having a lot more success with AT teams actually killing things especially set up in depth, and the player 1 cost 6-range ATGMs seem a lot more cost effective than Parallel Earth ones especially in multiplayer. Game actually works quite well with Talon Voice Commands too. Anyway, thanks for taking the time to write out a review – always interested in experiences within the wargame sphere.

  6. Pad Linek says:

    Chapter 2, Mission 7 is certainly doable now; it’s C3M10 that really sucks donkey balls ATM :)

    But yes, besides that one little nitpick, the review is spot-on.

Leave a Reply