Breaking down XCOM's Tactical Combat

I have a long relationship with the Firaxis XCOM reboots. One of my very early efforts with Unity involved cloning a simplified set of XCOM’s core game mechanics. To my surprise, my copy of the strategy layer felt incredibly poor. This was one of my first, very harsh, lessons in ‘narrative’. XCOM’s strategy layer is intimately tied to the games loose narrative.

The choices you make as you assume the role of ‘Commander’ are deeply soaked narrative, even if there’s very little traditional ‘plot’ - the game mechanics and the ‘role’ you assume work together in tandem to create a rich experience. The whole of the result is greater than the sum of its parts: the mark of a superb game.

That said, let’s break down some of XCOM’s core game play.

TACTICAL COMBAT
More than any other game I’ve ever played, XCOM taught me that there’s simply a massive gap between good strategy for a game, and the player’s fantasy of what the strategy is.

  • On it’s surface, XCOM is a game about tactics, daring flanking maneuvers and tense firefights.
  • In truth, XCOM’s actual strategy has very little to do with this. The most important element in XCOM is not moving, and the core to a winning strategy revolves around uncovering the fewest possible squares.

XCOM’s tactical maps: Enemy ‘Pods’
A map in XCOM has a bunch of walls, objects and assorted stuff for your guys to take cover behind. The enemy deployment is built around ‘pods’ - these are usually groupings of 3 alien bad guys. These pods are scattered randomly around the map.

This image is a breakdown of one of the XCOM:EU maps, and where some of the enemy pods might be deployed. Your men need to move through the area, and clean out each of the bad guy pods.

The Core Strategic Element: Pods won’t actively engage you
There are corner cases and rare exceptions, but in the vast majority of cases enemy pods are immobile and will not actively engage the player.

Instead, once the player moves into line of sight, the pod becomes ‘active’, the aliens in the pod take cover, and at the end of the player’s turn they start shooting.

As a player, activating an enemy pod at the end of your turn is the absolute, hands down, worst thing you can do. The result will often be that the game will kill one or more of your little men for making this mistake.

The Core Strategic Tool: Flanking
The game presents the player with one major combat bonus: when you ‘flank’ an enemy (move a guy perpendicular to, or behind an enemy’s cover), you get large bonuses. Flanking is presented to the player as their main goal in any combat situation.

This is emphasized by giving ‘flanking’ its own color in the largely 3 color HUD.

The color ‘yellow’ is largely reserved for to either indicate flanking or just a couple special moves. This would not be so meaningful, except the color scheme is largely limited to teal (general info on friendlies), red (bad guys) and yellow (flanked guys).

A huge part of the combat system revolves around flanking the enemy.

Where things get interesting: Punishing the player for flanking

So we have the two core elements to XCOM’s tactical design:

  • The worst thing the player can do is move into a new square (you might activate a new enemy pod)
  • The near term goal for the player seemingly involves moving behind the enemy, this almost always involves moving into new squares.

I do not believe that any other game will punish the player with anything like the severity that XCOM does (potentially one shot perma death) for trying to achieve game’s major micro goals.

It’s a fascinating design. XCOM’s design, literally goads you into flanking the enemy, then kills you for it.

Almost every recent turn based game have copied the shallow elements of XCOM’s tactical design. Everything from cover icons, to the movement system, but none have really taken this conflicting player incentive.

There are more really interesting parts to XCOM, and it’s design in the large and small - but this core element is so philosophically different from most games, in an era where so much game design revolves around gratifying the player, rewarding him for tiny step forward, XCOM stands in harsh contrast by doing the exact opposite.

5 Likes

Interesting stuff. I knew a lot of this on a base level from playing it, but had never broken it down like this. I found this part the most interesting:

I think what you said is true for the most part (although a good portion of the missions also put you under some time pressure to encourage you to go as fast as you can without losing people).It sounds like pretty poor game design to encourage the player to crawl through the game very slowly, but it doesn’t feel that bad when you actually play it. I can’t help but wonder if the game would be more fun if the pods did actively engage you. It would certainly be more realistic since super advanced aliens would obviously have a way to communicate with their allies that are 100 feet away.

I pretty quickly recognized XEU’s AI didn’t “activate” until you exposed it and personally I found it highly susceptible to run-of-the-mill kiting tactics. Combined with the very limited number of maps and the small size of the maps, I feel they severely crippled something that had much greater potential.

That aspect of XCOM actually bugged me a lot. I’m beta-testing a game right now (Depth of Extinction) that largely clones the XCOM tactical combat system (they hate it when people say that, but I see it as a good thing – the general system is a lot of fun to me). Their first public release had similar pacing and at the time I didn’t quite realize how much the plodding nature of XCOM had bothered me. They’ve made a lot of subtle changes recently (not only revising the mechanics but also good UI choices) that significantly increase the tempo and I’m enjoying the new flow quite a lot.

I might have to check that game out. XCOMs tactical combat is one of my favorite things in gaming. If anything, using that system is a selling point.

https://hofstudios.itch.io/doe-story

1 Like

You probably didn’t notice the plodding pace of XCOM so much until you were playing an inferior version, one where not all of the elements that went into XCOM are present. This is what I also found with my attempts to clone parts of the game. The bits and pieces I cloned really sucked.

In really, really well made games, the whole is much greater than the sum of it’s parts. Different pieces and elements click together. In less well made games, the weaknesses and problems are more apparent.

I don’t think the goal of XCOM’s combat was to be ‘fun’ - at least not in the way that we tend to talk about fun (generally ego gratification). I think their goals for the combat were:

  • Tons of Tension
  • Unknown and Uncertainty
  • Harsh Consequences

I think the most important thing here isn’t ‘fun’ - I think the most important thing is ‘fair’.

The most interesting thing about XCOM’s system is that it so directly punishes specific actions. The connection between a given click and a disastrous outcome is super clear.

“Oh god, I shouldn’t have run up to shotgun that alien in the face… now my whole squad is gonna get killed!”

By making the connection between action and consequence so clear, the player is more likely to walk away feeling like they were in control. “I messed up, but I can do better next time” - is far more likely if the consequences are direct. When the consequences are more subtle or disconnected, you’re more likely to get “I PLAYED PERFECTLY, THIS GAME IS RIGGED, F THIS”.

I think “I can do better next time” is probably the most important ingredient for true replayability. (Maybe it comes second to “I wonder what would have happened if I did … instead”.)

Having the game’s main cause of failure be in direct opposition with the near term goal that the game presents to you (flank for a bonus!) - this cannot have been a design goal up front. It’s just too weird of a design.

It had to be the result of iteration. The thing that’s interesting is that this is where they stopped iterating. So even if it wasn’t a fully conscious design goal, they recognized that that conflict created the game play they wanted.

2 Likes

I don’t agree that the “worst” thing you can do is activate enemies at the end of a turn. The worst thing you can do is run straight through the level like a lunatic and activate enemies while you’re standing in the middle of an open area. But if you move from cover to cover, it’s not a problem at all. If an enemy “pod” appears at the end of your turn, with you in cover, then they’ll spend their first turn either taking shots from too far away or maneuvering closer to you.

Nor can I agree that the game “punishes” you for flanking. The game punishes you for running out in the open. It punishes you for charging straight at a group of enemies. But I would frequently discover a “pod” (in the more natural environments like your picture above you kind of learned where to expect them to be) then slowly creep up on them until I got to a flanking or otherwise sufficiently lethal location where I could take them out freely.

As for other games which work in a similar manner–Wasteland 2, the Shadowrun games by Harebrained Schemes, and I imagine most ranged combat “tactical RPGs.” Probably Jagged Alliance 2, though I couldn’t make it 10 minutes into that game.

In the early missions (where the difficulty level is by far highest in the game) a single shot at a dude in high cover will kill him about 1 in 15 times. So you should lose a guy pretty reliably about 20% of the time on the next turn per 3 active enemy. If they can flank you on the turn you activate then chances are probably closer to about 50-60% of losing a dude outright.

The odds and stuff change depending on how far into the game you are, the difficulty level, the exact pod, etc. This roughly around like gatecrasher (mission one) on commander.

The difficulty curve in XCOM 2 is also really interesting - EU had a similar curve - but in 2 I think they were much more specific about the story and difficulty. I think, in general, the hardest mission in the game is gatecrasher (the first mission) on commander or legendary. Every now an then you’ll generate an exceptionally hard map, but it’s really rare, the proc gen is shockingly consistent (they must have done an insane amount of work here).

The difficulty overall also tends to peak as the story begins in full then drops off rapidly. So, the first real story mission (give or take) is the ‘make or break’ point in the game.

So XCOM 2 difficulty is organized like this (roughly - at commander, legendary is slower and I never beat legendary ironman, so i’m not sure about the end game):
Hour 0: VERY HIGH
Hour 5: Low/Medium
Hour 10: HIGH (first story mission)
Hour 15-20+: Low (story missions come fast and furious here, and you roll through the rest of the game)

They very intentionally timed the difficulty to drop to near zero once the story missions start in full. I am fairly certain that the most frequent point people give up a failed game is either at mission 1 or right around the first story mission.

They front load the difficulty, cuz like, people would get really mad if you wiped their team on the last mission 30+ hours into a game. Most of the mid-late game upgrades are so absurdly broken that it’s almost impossible to lose a guy. This design is like 100% intentional.

Note: Wasteland 2, Shadowrun all use very very different and more conventional level/encounter design. In JA enemies are all active at all times, XCOM will specifically deactivate AI so you don’t get overwhelmed unless you activate them via movement (remember, fairness).

2 Likes

It’s been a while since I played XCOM (I didn’t play the second one), so I may have been conflating some stuff with Wasteland 2, which I’ve played more recently.

I put a ton of hours into EU/EW/2. I also tried to copy a bunch of stuff (and failed!), so I’m pretty confident in my analysis (probably a small mistake here and there).

Trying to clone bits and pieces and failing though, that was a weird experience. What seems simple on the surface - it really isn’t - there’s a crazy amount of work that went into making those game mechanics click together. It’s nowhere near as easy to copy as it seems, many games did a better job copying stuff than I did, but it’s not a coincidence that none of them deliver the same level of experience that XCOM does (imo).

Again:

I think both these comments are really telling, and show what good a job Firaxis did.

2 Likes

Are the XCOM 2 maps procedurally generated? Not the same handful over and over like the earlier ones? Maybe I need to pick that one up after all…

Yeah - they did a kick ass job on the random generation. They have mod tools to extend out the map generation, so info on how exactly they organized it is public and it’s modular enough that modders can add stuff (I think it’s like a hrybrid half tile approach). I haven’t really looked at it in detail, but the results are top notch. The maps feel about as hand crafted as they did in EU/EW but each map feels surprisingly unique.

If I were doing proc gen, I would really dig into the modding tools and take a look at how they did it. Whatever it was, it worked. Probably one of the best, if not the best procedurally generated map systems I ever saw. Miles ahead of diablo - which felt super repetitive.

2 Likes

Didn’t xcom 2 mostly solve the “crawling forward” problem by a) starting you off in stealth mode on most missions where you can see the tiles that will activate the enemy AI, and you can get quickly into positions to lay an ambush, and by b) having timers that force you to move towards the objectives on most missions?

Bumping a semi-old thread, but XCOM 2 is on sale on Steam right now, 60% off.

3 Likes

@frosted : about the RNG to hit chances we talked about in another thread: I’ve started a second playthrough on commander difficulty and I’m pretty sure the RNG is biased in favor of the player in the easy mode. I don’t remember raging much at things like missing three 70%+ shots in a row on my first playthrough, and now I’m basically ragequitting even when I’m still winning the fight, because I find the amount of missed shots so aggravating. I’m still very sure I would enjoy the game more if there was a way to get the to hit chances to 100%, like in JA2 where you have more fine grained control on how much AP you want to spend aiming your shots.
Maybe the idea of “xcom chess” should really be tried some day. Imagine equipment and abilities give your characters different kinds of chess moves to perform, the base class dictates the kind of move they always can perform, and the enemies are locked in to 1 move pattern per enemy type, no healthpoint system needed, all attacks kill instantly.

1 Like

It’s only fun if you don’t play against a Wookie.

3004339--224012--1.png

1 Like

Wait until you play on legend and the dice are fair :smile:

It’s pretty amazing how hard the game cheats in your favor, and how little we actually notice! If you have any RNG at all in a single player game, you gotta cheat in favor of the player.

The really amazing thing is that if you go on the XCOM forum, people will insist the game is cheating against them when it’s still actually cheating in their favor!

I also kinda wondered if we just misunderstand % chance and maybe ratios might make us assess the chances better. Like 70% is close to 2 out of 3 - so 1 in 3 times it wont work. I wonder if we show chance in ratio if people would be less likely to rage

1 Like

My gut feeling is that the general public would be even worse at interpreting a ratio. And you still have to think about presentation. I bet that stating 1-in-3 to miss would produce a completely different reaction than 2-in-3 to hit.

I bet there’s a tipping-point where that’s true of percentages too. For example, telling them 40% to miss might make a miss easier to swallow than getting their hopes up by telling them 60% to hit. UX sleight of hand, more or less (no pun intended)…

This is super true. I actually tested it! I presented chance to miss instead of chance to hit. It really transformed how you thought about the odds. Only problem is it felt too weird, we’ve all played waaaay too many games that use “chance to hit” - reversing it is too weird and felt bad :cry:

1 Like

I think the opposite is the case. I already convert those into ratios in my head, and that’s exactly what leads to expectations like “2 out of 3 shots should statistically hit on a 70% shot, I’ve already missed once, the next one should be a hit”. When that one misses again I start to get angry and think “ok, the next one must be a hit now”, and when that one misses again, I might ragequit in a game that I can just save and contine later. Imho true randomness (like the chance of missing three 90% chance rolls in a row) for to-hit chances is inherently un-fun and the numbers should be fudged in a way to make deductive reasoning about results work to a degree. Like pregenerate 1…100 and sort the numbers so that every 10 numbers you’ve gotten one of the possible 1x, 2x, 3x,…9x results at least once, and make sure that no sub-50% results come more than twice in a row. I think that would be a lot less frustrating to play with, because it eliminates the extreme bad luck in a row situations.

1 Like