Mid-August Update

I was planning to upload another non-URR post today, but Sunday has crept up upon me with alarming speed and I haven’t actually finished the one I was working on. So, instead, here is a very short and screenshot-heavy update for this week instead with a couple of things I’ve thrown together. Firstly, I’ve done a bunch more on fortresses – they’re about 50% done now, six of the twelve planned archetypes have their appropriate generation algorithms and interact in various interesting ways with rivers, on the off-chance that a river goes across them. Here’s a triangular fortress:


I also on a whim threw together some graphics for a future release, 2 or 3 down the line, once I start putting together ships and trade routes. These are just stylized graphics to denote the different classes of ship, military and transport, when you look them up in the Encyclopedia. Alongside them you’ll see information like the ship’s schedule, its armaments, what it trades in, its captain, etc etc.

URR ships

Also did a little bit of work on making Upper Class Housing districts just that little bit snazzier (as ever, this graphic is generated, there’s over a hundred possible variations):


Next week we’ll likely have the Value of Maps entry where I’m going to talk a bit about how maps work in the Civ games (and particularly Alpha Centauri), the ways in which they try to quantify the value of map knowledge, and the importance that maps are going to play in URR. See you then!

Early August Update

This is the first development update in a couple of weeks. I’m still primarily working on my thesis in the hopes of finishing before the end of September, at which point I’ll be starting the full-time development year. Nevertheless, I’ve managed to snatch a little bit of time here and there to code and keep things ticking over.


I’ve done a few more fortress archetypes. As mentioned before, there are twelve “shapes” fortresses can generate in total; each civilization chooses one, and there can never be more than five nomadic civilizations, so it’ll be a long while until you see each archetype (not to mention, obviously, that each instance of each archetype will generate differently). I’ve thrown together the “Pentagon” and the “Double Fort” algorithms, and here are a bunch of screenshots. Fortresses are really interesting to walk around – they definitely evoke a very different feel to towns or cities or settlements. One thing I’ve found to be a very good idea is to really emphasize difference in procedural generation – sure, in the real world, a “castle” and a “fortress” may be very similar, but by placing one at the core of cities and giving the other to nomads, I’ve emphasized the differences in each one to create two totally different kinds of structure. Even though they lack NPCs yet, walking around the four different types of population centre (cities, towns, settlements, fortresses) all feel totally, totally different (which was exactly the goal) and I can’t wait to see what everyone else thinks once this release is out. Some fortresses are more or less militarized than others; some have larger or small markets, or housing, or defenses; some are more open, some are more closed and more challenging to navigate. There’s a great amount of variety, and they all also handle rivers in different ways, leading to some really interesting maps (which have been challenging to make ensure walkable generation on)…



Hunter-Gatherer Buildings

I’ve thrown together the generators for the textures of all the different building materiasl hunter-gatherer civilizations might use. Whereas feudal civilizations use variations of bricks and nomadic civilizations sometimes uses bricks, or stone, or drystone walls, hunter-gatherers are forced to be a little more resourceful. There are currently sixteen different materials, each civilization will choose a different one (a couple are tied to particular climates) and they won’t repeat, so you’ll have a wide variation every game. Here’s an example of the outcomes of each of the generators for the different materials. Going left-to-right and top-to-bottom, these are “logs”, “leaves”, “wood”, “stone”, “bones”, “mud bricks”, “snow bricks”, “interwoven sticks”, “mud”, “drystone”, “thatch”, “wattle and daub”, “rope”, “bamboo”, “leather”, “fur”:

HG Building

The materials have no particular gameplay difference, but they do lend a nice variety to the hunter-gatherers, especially as the way they are displayed in-game varies according to the colour of each texture. HG civilizations are now looking rather more complete than they were before, and the other reason for this is the introduction of standing stones.

Standing Stones

Different religions worship in different ways; some have big idols, some might have small totems in the houses of worshipers, and so forth. Some hunter-gatherer civilizations have standing stones in the middle of their settlements as the focal points for worship. Each of these is a different shape, contains a symbol of the appropriate religion, and… some other text, in an ancient language. In future versions this text will be one of many clues around the world to help you find your way through the world’s mysteries, and will be one of several reasons you might consider visiting a hunter-gatherer settlement you pass on your way to parts unknown.



The next three entries will probably be the other games criticism entries I discussed before, then we’ll have a roundup of some more code at the end of August/start of September. By then I should have more of the fortress archetypes done (maybe all of them?) and maybe some more city districts too – military districts might be next on my list, though I think I also need to make some changes to market districts; playtesting them by myself suggests that they might need to be a little more contained and a little less open to make them more enjoyable and more understandable to navigate. The plan thus remains: I’m working hard to finish my thesis before I move house in late September, and then to start the full-time year of development in October, with the intention of releasing 0.6 – admittedly the biggest URR update ever, since it is generating every town, city, fortress and settlement – within November.

Perfect Dark Retrospective

A few years ago Perfect Dark was re-released on Xbox Live Arcade. Whilst it was generally met with a positive reception as an impressive modernization of a classic game, especially since the new engine was built ground-up to be as close a copy of the old engine as possible, several reviewers claimed the game’s level design didn’t hold up well compared to its modern successors. Gamespot suggested the levels are riddled with “locked doors, unused rooms, and dead ends”; Eurogamer argued that “games are now so much better at telling you where you should go next and what you should do when you get there”; Gametrailers criticized the lack of “waypoint markers or maps”; Joystiq suggests that “without a blinking arrow navigating you to your next objective, the gameplay can feel aimless”; CVG says the level design is “adrift of the clear and responsive standards we’re now used to”; while Cheat Code Central, with impressive hyperbolae, suggest that “the modern gamer will find it borderline unplayable.”

All these criticisms boil down into a core statement – that the actual navigation and traversal of the physical space of these maps is unnecessarily challenging. They posit players asking “Where do I go?” whilst fumbling around the apparently obtuse level design. This piece is going to look at this claim, consider the level design of Perfect Dark in depth, the more-modern level design of some other shooters, and then consider whether these criticisms are valid, or merely a sign of the times in an age of shooters that, perhaps, demand rather less of their players. I’ll also be looking in detail at two levels and a third set of levels that illustrate some of the things I like the most about the singleplayer campaign in Perfect Dark, and that many reviewers have tended to miss in favour of criticizing some of the weaker levels in the set. Perfect Dark has seventeen main levels of the campaign. I’m going to focus on the idea that the levels are full of pointless, unused or dead-end areas, and then also talk about one other level which displays a surprisingly effective emotional effect that, many many years later, games like Mass Effect 2 and Assassin’s Creed: Brotherhood would repeat.

Area 51: Infiltration / Rescue / Escape

Towards the middle of the game are three levels within Perfect Dark’s conception of Area 51. InfiltrationThese three levels all take place on different parts of a single huge map (as does a bonus mission unlocked when you complete the game, but I won’t cover that one here) and the first level cleverly sets up the ending of the third level. In the first level, Infiltration, you come across the door on the right before you’ve actually managed to enter the underground Infiltration2portions of Area 51. A new player might think this door is the way in; you can enter, kill a few guards, but as shown in the picture to the left, you cannot go any further (the guard who has decided to strategically insert himself into the doorway is not an essential part of the level design here). The door on the left of this picture is locked, and one might initially think this supports Gamespot’s assertion that the levels are riddled with “locked doors, unused rooms, and dead ends” which make it harder and harder to figure out where to go. Whilst this is the dead end, it is only a dead end on this level. It is in fact an important door in the overall Area 51 map, of which this is only a portion, and as we’ll see you can return to it later.

TechnicianBefore returning to this door, you’ll run into this technician manning an interceptor drone. Ordinarily you need to kill him to acquire a keycard for the lift that takes you down into Area 51 (the real entrance, not the locked one we’ve just talked about), but you can actually choose to knock him out instead. On some difficulties you have to do this quickly in the mission otherwise he’ll launch the interceptor he’s standing next to and move down into an area of the game which, later, you have to blow up with explosives. Either way – if you knock him out, you get the keycard anyway and there’s no apparent difference until the next level.

Rescue1The second level in the set is Rescue. When you start, you quickly reach this corridor; the door this guard protects is locked. However, if you left the technician alive on the previous level, you can use the X-Ray Visor you have for this mission (for an unrelated objective) to look through the door, and sure enough, the technician you kindly saved will be standing behind the Rescue2door. After a few moments he will open the door, be terrified by your presence and run away, but his work is done – he opened the door. Once through you find yourself on the upper level of the hangar you were in on the previous level, but this time the lift you arrived in doesn’t transport you between two hangars but instead takes you nearer the surface into an area you would only normally visit in the bonus mission unlocked at theRescue5 end of the campaign. On a pedestal in this room you’ll find a Phoenix, a very powerful weapon you don’t normally get until the final few levels. This is an interesting method of placing secrets that FPS games, even those with distinct levels, don’t ordinarily do – your seemingly-trivial side actions on one level have an effect on the next level to unlock a door, or keep an NPC alive, or allow you to use a vehicle (both of those also happen in the game). They are entirely unessential to completing the game. I think this may also be another example of the “dead ends” that Gamespot so disliked, but actually points towards a different issue – that the game expects the player to explore and experiment (and, given the era, to presumably discuss with other players online) to figure out how to open some of these doors. There are far fewer doors in the game that truly “go nowhere” than that article implies; many can be opened by an obscure sequence or link to another part of the map that might be visited on a related mission, or in many cases, a higher difficulty. Towards the end of the level you may come across this locked door, another apparent “dead end”, but it is actually anything but…


The third level, Escape, returns highlights the inter-connectivity of this level. Just like the first two you visit some new areas of the complex and some of the ones you’ve already been to as well, helping the player build up their mental image of the entire area. At the end of the level you are given two options – to let your ally sacrifice himself so you can escape, or you can help him escape and try and find your own way out. The default is the former; many players are unlikely to even realize the latter is an option (again perhaps contributing to the impression that the level is full of pointless doors?). However, if you offer to let your ally escape and find your own way out, the way in you took in the earlier two missions is locked, so another route must be found.

Escape2At this point much of the level unlocks (but not the way “backwards”) and you can make your way to the other side of the complex. Whilst guards continue to spawn you will find your way to the door shown above, but this time, it’s unlocked. You pass over a bridge and come to the room on the left, which you won’t have visited before. The door just shown in the right of the Escape3picture leads back to where you met your ally the mission before, whilst the blood-splattered door ahead (blood splatter optional) leads into this room. Seem familiar? This is the other side of the locked door from all the way back in Infiltration! You open the door and you’re back outside, and the moment you step over the threshold, the mission ends, and you’ve gone full circle through the entire Area 51 map, even though it took three different missions to do it.

Whilst some FPS games have you circle back around in order to explore a new area, few have you do this between levels. Indeed, most modern FPS prefer to have a linear path not just between levels but within levels themselves; games like the Legend of Zelda or Dark Souls may emphasize the opening up of new shortcuts and the like, but it’s rare for an FPS to do the same. Whilst many players will not even figure out that this ending exists, once you do it the entire map comes together rather suddenly in one’s head – you haven’t just been going deeper and deeper into the complex, but rather moving around the complex, and coming back out through the locked door you might have visited in Infiltration is a great moment the first time it happens.

In the picture below you can see the parts of the overall Area 51 map (thanks to memoryhacking.com for the background image; the three coloured outlines are my additions) that are used in each part of the level. Rescue overlaps with Infiltration, Escape overlaps with Rescue, and Infiltration in turn has a little overlap with Rescue. When you choose the “hidden” ending in Escape, much of the Rescue part also unlocks, but that’s only for the “final run” to escape Area 51 on the hoverbike.


Thus, in the entire three levels – despite their complexity – there is not a single door that is not used at some point. Once the player understands that locked doors are not just those doors you get in other games that are textures rather than objects (a la Half-Life 2) but do always lead onto some other part of the level (though I confess there are one or two minor exceptions, but only one or two in the entire game), it becomes apparent that these doors are actually things the player is meant to remember and keep track of. Locked door this time? Well, if the next level takes place on the same map (or parts of it), perhaps it will be open this time? Or maybe there’s some secret sequence that will unlock it and give me access to a secret? Once you realize this, finding a locked door is actually quite an interesting moment as you try to figure out how best it may be opened. I think the criticism over dead-ends shows a lack of appreciation of both the complexity of some of the maps, and (as I’ll talk about next) the fact that some parts of some maps are only unlocked on higher difficulties, or if you perform a particular (sometimes quite obscure) action within the mission.

Deep Sea: Nullify Threat

On higher difficulty levels these days we rarely see FPS games that put forward more objectives. You still go to the same places on the same levels and do the same things, but it’s just “harder” – enemies have more health, more accuracy, better weapons, fire faster, smarter AI, etc. However, Perfect Dark often has areas of its levels which either require a particular difficulty setting or a particular trigger to access. An amazing example of this is the Deep Sea level. During this mission you have to “disable a megaweapon” on a sunken alien vessel, but depending on which difficulty you play on, you will visit three totally different maps and disable the weapon in three totally different ways. If you never do all three difficulty levels you will never even see some parts of this level (and even if you DO play it on the highest difficulty, some parts of the level can only be sighted using the see-through-walls scope on one of the weapons, and cannot be physically accessed). This takes both the non-linearity and the emphasis on new areas and new objectives for higher difficulties to a higher point than any other mission.

Deep2On Agent difficulty you find your way through a series of corridors into the room on the right, illuminated by a pulsing green light. In this first example you wait until your ally catches up with you (which can take a while as he loves wasting time on the path there) and wait until he disables the weapon by using the console. Although you only see this area on Agent, it’s a surprisingly atmospheric room (and hints towards just how many rooms in the later Timesplitters series would have pulsing lights).

Deep4On Secret Agent the same teleporter takes you instead to a balcony overlooking this room which you quickly descend towards down a spiraling corridor. Whereas the first appears to be a control room, this appears to be perhaps the firing mechanism of the weapon itself? The second beam feeding into the middle has to me always evoked the little ignition fire on a flamethrower; as something small that triggers the far greater weapon. Although you’re now clearly near the core or the firing mechanism of the weapon itself, your ally is still required to disable it.

Deep5Having seen the control room and the firing mechanism of the weapon on the other two difficulties, on the Perfect Agent difficulty you pass through a series of rooms that seem to form the power supply of the weapon. In these you have to destroy these glowing pillars, before then doing something very interesting. You’ve been equipped with a weapon that can see Deep6through walls and you have to use it in order to shoot out the final few pillars which cannot be accessed or seen ordinarily. That these final components cannot be easily reached – combined with the fact that stopping the megaweapon on this difficulty needs destruction instead of a technical fix, reinforces the sense that you’re crawling around the least accessible parts of the ship; its engineering decks, engines, fuel reserves, power supplies, and so forth. The fact you visit a different part of the ship on each difficulty makes this one of my favorite levels in the game – the “higher difficulty areas” are larger than any other level, more varied, and although nothing explicit is said the aesthetics of the three areas seem to suggest the three different “components” of the weapon I’ve described here. I also (this will be a common theme for regular readers) also really appreciate the effort put in to what could easily be termed “optional content” – a lot of players will never see some of these areas, but they still took the time to put them in, and it really rewards you for attempting the higher difficulty levels.

Carrington Institute: Defence

This is the last level I want to look at. It takes place directly after Deep Sea and I think the setting of this level is interesting (the other two levels having illustrated my thoughts on the level design of Perfect Dark). Until you reach this point in the game, the “Carrington Institute” is your home-base or your hub, akin to Firelink in Dark Souls, the Normandy in Mass Effect or Monteriggioni in Assassin’s Creed 2. You can leave the campaign at any time to come back here and wander around, explore the institute, talk to any of the staff, do a bunch of tests (which range from the challenging and enjoyable like the firing range to the… less challenging and less enjoyable like a lot of the equipment tasks), read up on information on the plot, weapons, locations, vehicles and whatever else, and generally take a break. Back when I played on the N64 I found it surprisingly easy to just wander around here and take a break from the missions themselves; it becomes a place of easy relaxation and safety.

Car1bWhich is why (much like Mass Effect & Assassin’s Creed: Brotherhood would do a decade later) it’s so very, very clever when one of the later missions has you defending the institute and rescuing the hostages who are these people you’ve been chatting with and helping out up to this point. In these two screenshots, the folks in green are your friends, the blue/black guards the invaders holding them Car2bhostage; you have to leap into the offices and gun down both the hostage takers within a second or two before they take out your friends. It’s a very effective technique; whilst I enjoy a lot of other missions, this one has always felt to me the most urgent and the most “real”, somehow, as your enemy invades the place you’ve been comfortable wandering around and calling home between missions. It’s a surprisingly effective technique, and whilst I don’t know if Perfect Dark was the first game to ever give you a hub which is later explicitly attacked, it is certainly chronologically the first I know of. Whilst this isn’t strictly a level design observation, when one plays the Defence mission one then realizes that, actually, the Institute has been designed to offer an interesting singleplayer mission as well as being a believable hub for the rest of the game.

Changing FPS Standards?

Now, let’s look at a comment from Eurogamer.

Perfect Dark’s not afraid to throw dead ends at you seemingly for the hell of it, or repeat textures so much in its huge maps that you can get a little dizzy.

I think we have covered the first half of that comment – that these locked doors are either for other levels using the same map, or for higher difficulties on the same map – but let’s consider the repeating of textures. This, alas, is somewhere Perfect Dark does fall down, but at the same time it tries (sometimes) to negate the most negative effects of the copy-paste level construction that raises its head in a few places. Several levels can be confusing until you come to know them. For all I praised the interesting design and the secrets of the Area 51 levels above, when you first play them (especially during Rescue) it can be hard to orient yourself inside in the endless sequence of similar pristine corridors. However, the game does offer something to help – there’s a range of different door shapes only used in a small number of places. Whether these were specifically placed to help people orient themselves is, perhaps, besides the point – once you come to recognize the various types of doors, the similar-looking corridors stop blurring into one another. Whilst I cannot confirm this, why have so many different types of door otherwise? Why go to the effort of having different sizes, different textures, different animations, some with/without glass, if this variation didn’t serve some purpose to help the player orient themselves?


Since (as we’ve covered) the levels are rarely straight lines, this makes the need for the player to get their orientation even more important. If the levels were just straight lines like The Library in Halo CE, for example, dull or repeating textures are never an issue since you never get turned around. The Perfect Dark levels are rarely straight lines, make no mistake, but as above they do at least try a lot of other things to signpost your way which just require a little more attention. For example, DataDyne : Research (the second mission) could be seen as a maze of similar corridors and identical doors, and as a young teenager playing the game on the N64 I found it unclear at times how to progress. However, the game subtly points you in the right direction; the mission briefing mentions you need to head towards Sector 4, the “highest” sector, and sure enough you start seeing doors with Sector X markings on them, helping you mark your progress through the level. If you didn’t read the briefings, though, perhaps you wouldn’t assign much significance to these visual markers?


Equally, consider the level Pelagic II. Even now I sometimes get confused on this level, but they at least tried to help the player find their way around by changing the colours of the lighting on each floor; the lowest floor is blue, the first floor (where you start) is red, then yellow and green at the top. Once you figure this out it aids you greatly in figuring out which part of the ship you’re in.


Every level that suffers from this problem does try to do something to alleviate it and help the player get their bearings, but it’s debatable how much use the Pelagic II’s different colours are when, even though there are four colours, there are a dozen identical corridors in each colour. In this regard, I believe the second half of the criticism many reviewers isolated – the reuse of textures – is a legitimate one on some levels of the game, even if some of these tried their best to cancel it out. To alleviate this problem a designer has two possible solutions – either make levels much more distinct throughout (like many PD levels are) so that you can easily distinguish between the different regions of the level, or make the levels so linear that you cannot possibly go wrong. 

The sad truth, I think, is that contemporary FPS games have done both. They’ve sacrificed the complexity and many-layered levels in order to make things clearer, but they’ve also sacrificed the subtle things that point you in the right direction. The path forward is obvious, but that no longer matters when there’s only one path you could take anyway! You could preserve the complex levels whilst adding more signs and symbols where you need them, or lose the complex levels and not have to do any more effort in laying out the path, but many modern FPS games seem to have done both. In their Perfect Dark review, Joystiq acknowledged that:

“Love it or hate it, modern games have us weaned on hour-long tutorials and overzealous narrators who explain every detail to us ad nauseam. So, when you jump into an older game like Perfect Dark, you’ll be surprised at how self-sufficient you’re asked to be. “

And I think this says it all. It came with some downsides – getting lost on the Pelagic II research ship is an infuriating experience that everyone should experience at least once in their lives – but at the same time we’ve sacrificed the completeness of Area 51 and the many “hidden” areas of the sunken alien vessel in Deep Sea. Modern FPS games have perhaps gone too far by taking both solutions to the issue when only one solution was really needed, along with other structural questions – what counts as a “mission”, for example? – that are questions for another day. Ultimately, if you want an FPS world that has more of an exploration focus than most these days do, give that XBLA version a download – you won’t be patronized by the game, you can explore some of the surprisingly “complete” and convincing levels the game has to offer which often really do feel like part of a cohesive world, and you certainly won’t be disappointed.

Dark Souls 2 Design Ravings

Dark Souls 1 may be my favourite game. It has stern competition from the original Command & Conquer, but there are no other games in the last few years I’ve poured as many hours into. As you might expect, therefore, when Dark Souls 2 came along I quickly preordered one of the super-special editions, happy in the knowledge I would be sending some more money to a company who had made an absolutely stunning game the last time around, and a game I only paid £10 or so for after it had been out a year. Having played it through once – and then having gone back to the original for my first playthrough in around half a year – I found myself with so much to write about I just had to put some of these thoughts to paper. If you’re not interested in some detailed thoughts about the two Souls games I’ve played (I have yet to play Demon’s Souls), feel free to skip this one. Otherwise, let’s begin. I’ll start off talking about the things the game does better, then I’ll talk about the things the game does worse (alas, a list that is both much longer and deals with much more significant parts of the game), and then conclude with some final thoughts on the whole business.

For Better:

User Interface:

The UI is clearer – you can see more items at once, check out a larger selection of items more easily, and so forth.

PvP Lag:

Although I have been hearing issues about the “soul memory” system in the game – designed to help balance PvP encounters, but apparently it can be exploited in order to do the exact opposite – the loss of lag is a hugely important point here. Even if the PvP mechanics may be weaker this time around I still prefer to play this in DS2, not 1, simply because the lag is pleasingly minimal. It doesn’t have the minimal latency of something like CS:S, but it’s still eminently playable now, and doesn’t boil down – quite as much – to farming backstabs by exploiting the lag.

NPC Invasions:

There seem to be many more NPC invasions now which come at a variety of points in the game. These NPCs have an interesting range of builds, appear whether you are human or hollowed, and are much more challenging than the ones in the original. The fact they appear regardless of your status is also a huge step in the right direction for the concept.

Resting at Bonfires:

You now only need to touch a bonfire to activate it, not sit at it. Entirely minor, but nice.

Multiple Soul Uses:

You can now choose how many in a pile of souls to use at once instead of having to pop each at the same time. Truly wonderful.

Humanity/Human Effigy Improvement:

Humanity is no longer an item you only use if you actively want to PvP, or if you want to kindle a bonfire. Instead, the new “Human Effigy” item gives you a real reason to be human – each time you die your maximum health is reduced slightly, so the Effigy becomes an item that you possibly want to use before particularly tough sections, maybe even between bonfires (though as I’ll go onto later, the bonfire placement is far too kind this time around), and balancing how often to use them vs their scarcity is a good move in the right direction. Equally, the removal of some specifications around invasions and being human/hollowed combines with this change very well.

New Items:

There are many more usable items (like firebombs, resins, etc), and the extra variety is appreciated. I couldn’t find more than one or two of them on my first playthrough however, nor anyone who sold most of them, but I shall assume I missed it.

For Worse:

Level design.

This is one of the big ones. The level design is, unfortunately, scarily simplistic compared to the two dozen labyrinths we traversed back in Dark Souls 1. This was a feeling I got to an extent upon playing DS2 first, but after I went back to DS1 and explored areas like the Painted World anew, this realization simply could not be ignored. Here is a map of one early area in Dark Souls 2, Heide’s Tower of Flame…


…and here is a map of a mid-game area from Dark Souls, Sen’s Fortress:


Just flick your eyes between those maps (thanks to the Dark Souls wikis for both). Yes, the first is a sketch and the second is detailed and neat; yes, Heide’s Tower of Flame is simply a smaller area; yes, Dark Souls’ Ash Lake and Lost Izalith, for example, are very linear (though even Ash Lake has more alternate routes than HToF!); but the problem is, the Tower of Flame map is indicative of almost every area in Dark Souls 2 – a straight line with one or two tiny diversions – whilst the Sen’s Fortress map is equally indicative of almost every area in Dark Souls 1. The only large, densely-packed area is the Lost Bastille, which has five bonfires. In Dark Souls, an area of that size – the Burg, say, or the Painted World, or the Depths – would have one or possibly two, but the level would be designed in such a way that there are multiple paths to take, shortcuts to unlock, and the level design would be incredibly dense as a result. This pattern unfortunately repeats across the game. Most areas are fundamentally a straight line with a few minor tangents that basically lead back to the same original direction; almost no-where is that multi-layered maze-like level design that worked so incredibly well in the original DS.

The second issue with the level design is more aesthetic and thematic. For any game with an “exploration” factor the ability to surprise the player is surely paramount. This may seem obvious, yet many games fail this simple test – the only place in Skyrim that might awe the player is Blackreach, for example. In the rest of the game, even though you don’t know specifically what lurks in each cave, you do know what lurks there in more general terms. You know what a Dwarven ruin will look like, what a cave will look like, what a draugr-infested place looks like, etc. I really value the moment where the game confounds your expectations, or shows you something you didn’t even know was there, or gets you totally rethinking what you thought you knew.

One of the best things for me about the Souls series is their ability to constantly surprise you. Maybe for a boss you’ll fight another player, or maybe you’ll uncover a secret in the most obscure of places, or find a region whose architecture makes you utterly reevaluate the fictional world in which you play (a la Ash Lake). Unfortunately I found the first half lacking in this regard – one boss surprised me and sent me spiraling into a little pondering of the in-game lore, but otherwise nothing really struck me and amazed me. The gameplay was fine, the storytelling decent (though too many items boil down to “This sword was used by an ancient hero – or was it?”), the aesthetics of the world (mostly) as great as ever, but nothing made me rethink what I thought I knew. However, the later half of Dark Souls 2 really picks up in this regard. In the first half there was probably only one area which had the same impact as being snatched by the crow out of the Undead Asylum, seeing the Moonlight Butterfly perched upon a distant ledge, descending into Blighttown and realizing an entire civilization squats amid the foundations of Lordran, being cursed for the first time, or winding up in the jail in the Duke’s Archives. All those moments in Dark Souls both impressed me with the creativity and the audacity of the game’s design, and made me feel the game could continue to surprise me both aesthetically and mechanically.


Passing through Darkroot Garden, you see this perhaps minutes, perhaps hours, before you ever encounter it close up…

Although the first half lacks any real surprises and often feels like a rehash of DS1 – indeed, see the plot spoilers below, there is a reason for this – the second half of DS2, in fairness, has many such moments. At one point I was sure I was approaching the penultimate boss which I thought would just be followed by a 5-minute ride to the final boss, and then discovered the boss I expected was actually an NPC and three entirely new areas had just opened up to me; at another, rather than descending down into the depths of the planet (a common Souls theme) I found myself climbing up higher than one has ever been in a Souls game (this is debatable based on your interpretation of Drangleic geography, but I think this region is subjectively significantly higher than Anor Londo was). Another point saw a visually arresting area unlike anything before in a Souls game whilst elsewhere assorted secrets referencing the original game point towards all manner of lore speculation. Whenever I thought I knew what the was going on and what the game “was”, on some fundamental level, it changed things up and surprised me, and that is something an exploration-focused game should always try to do.

The last aspect of level design lacking is that foreshadowing the previous game did so well. Often you could make out something in the distance, either a structure or a boss, and later would find your way towards it. This made the entire world feel incredibly connected (see the next section) and quickly raised the interest and excitement to get to this location on the player’s part. The “here’s something you’ll be fighting!” is less common this time around, or rather, is almost entirely lacking. In DS1 Ceaseless, Kalameet, the Moonlight Butterfly and the Iron Golem are all enemies you likely see long before you ever fight them, and you wonder – are these bosses? NPCs? When will I fight them? To its credit, Dark Souls 2 undermined this expectation nicely by turning what I was sure would be a boss into a very cryptic (and very large) NPC towards the end, and similarly I was very surprised when encountering a character I’d heard a lot about (and knowing the Souls’ series perspective on “King” characters), but these were really the only two examples. You cannot see outside the little corridors the game channels you down at any point, and the corridors themselves are, for the most part, sadly very dull.

World design.

Roughly speaking, the world structures of each game are as follows (in both cases you begin in the middle).


Dark Souls 1 is split into two halves – the first half of the game where you navigate the central interconnected wheel, and then you gain access to a fast-travel system and the four spokes open up. Once you’ve reached this point you know your way around the core well, and fast travel is only available to pursue the paths that branch out. It is worth noting that the levels on this path – Duke’s Archives, Tomb of the Giants, Painted World, etc – are every big as multi-layered as the others, but they just don’t connect with each other. When you’re in these regions you can still see other parts of the map, and the world still feels huge because you’ve explored most of it before gaining access to the fast travel system. By contrast, the latter game is a very simple structure. You have four main paths, and later you gain access to the branching path in the bottom corner. No regions connect with each other. This, combined with the fact you can now warp from the start, combine to make the world in Dark Souls 2 seem flat, uninteresting, unconnected (because it is!) and simply a far less interesting place to explore.

In Dark Souls, on one’s first playthrough, one is overwhelmingly likely to pass through the “expected” route – through the Burg, the Parish, the Depths, Blighttown, and through to Quelaag. However, if you really look around (and fight through some Drakes), you can sequence-break that and move around a lot of Blighttown. Alternatively, if you take the Master Key on a later playthrough, you can totally change the order you do everything by opening the tower Havel is locked in, or even more importantly, the door between New Londo and Blighttown allows you fight Quelaag first if you’re so inclined. You can also fight the Four Kings as anywhere between the second, and the penultimate boss, of the entire game. Sadly, nowhere in DS2 is there a comparitive level of freedom to decide on your own route even once you know the game well. Even though the hub area has many routes, most are closed off when you start, and it only really offers you five linear paths, and the second half of the game – although far stronger – also sticks to the generally linear regions. There is no sense of mastery over the game world on repeated playthroughs, nor a reinforcement of the idea that you are exploring one area.

Another glaring weakness is, as hinted at above, is that even the levels that lead to each other are not connected! You almost never see things in the distance you’ll later find your way to, and the rare times you do – for example, you can see the Tower of Flame from Majula – the perspective and location are all wrong. However, the most shameful example of this involves the middle of one of the early paths the game sends you down, when you find the structure shown below.

Earthen peak

You will note, I am sure, that there is nothing atop this windmill-castle-thing. No lift, no path, and certainly no volcano. Indeed, there is no volcano anywhere even remotely in sight. As you progress through this region (Earthen Peak), you reach the top of the structure and defeat the boss. After that, you get in a lift and go upwards… into the air… and emerge into this area, Iron Keep, a castle within a volcano with lava and flame for miles around.

Iron Keep

Which, basically, means this area of the world is laid out like this…


…and the upper half is entirely invisible from the ground (Credit for first two pictures go to Matthewmatosis (Youtube), last picture to maplejoker-anno.tumblr). This hammered home to me more than anything else that these were just levels in a game, not segments of an actual world I was exploring. Anything like this would not have been considered for a heartbeat in the beautifully pieced-together map of DS1 – where you can stand atop Sen’s Fortress and look down upon the Parish, or walk through the Tomb of Giants and catch a glimpse of Ash Lake – and, whilst this example is more extreme than others, this repeats throughout the game. A quick look at the map viewer for the game shows a huge number of both logical and thematic discrepancies. No-Man’s Wharf is way below the sea level seen from the Bastille, Sinner’s Rise is entirely invisible from up above, going up the lift from Aldia’s Keep launches you into a sky area you hadn’t seen anything of before (though a visually amazing region), whilst the Shaded Woods path goes from a forest into an area akin to the Giant’s Causeway (again, aesthetically impressive, but entirely incongruous) and then into a desert all without any real warning.

Ultimately, Lordran felt like a single location that had stood for thousands of years and finally fallen into ruin, through-out which you might run across a few other souls on their own personal quests or missions, picking through the ruins of this astonishing megastructure/city/fortress that stretched far up into the sky and deep underground. Drangleic, alas, feels like a sequence of tunnels and people specifically laid out to greet the player at every turn. It’s more like a sequence of “levels” akin to, say, early FPS games where you basically “teleport” between each level (Goldeneye style), rather than anything approaching a connected world. In fact, although the areas in Demon’s Souls are not connected, I believe they do not overlap either in as horrifying or jarring a way as some of the DS2 ones do.


I recently bought the Dark Souls design works book, which is fascinating. At the back is an interview with Hidetaka Miyazaki, the design lead on the first two Souls games, and in the interview he mentions that one of his inspirations for the lore of the Souls games was reading Western fantasy books/RPGs – his English wasn’t amazing, so he wasn’t entirely clear on what was going on, and had to form his own connections. This in turn inspired the way the story/lore are told in Souls games, and I for one love it (and it is something I aim to emulate in URR). I think the storytelling in DS2 remains strong and encourages you to piece things together, but those pieces are far less interesting than the pieces in the original game.


Seriously, major ones. I know some people – like me – read at a pace where just seeing “spoiler warning” is not enough to stop you reading before you read on a few words, so let this sentence be the final warning. If you’re like me and adore the Souls story, stop reading at once. Otherwise, let me state up front – Dark Souls 2 introduces the idea of a repeating or cyclic history. I have no objection to the “cyclic nature of history” plot ideas – Japanese games seem more willing to use this than their Western counterparts, but this is far from an exclusive rule (e.g. Mass Effect). However, in this case I think From Software made a colossal misstep by assigning a cyclic history to the Dark Souls lore. Now, I should state, to an extent the existing lore suggests a cyclic history, on a very small scale – the bonfires keep those cursed endlessly resurrecting, yes, but on the grander scale of history Dark Souls 1 has no mention of repeating patterns. The world began with the dragons; the Lord Souls were discovered; Gwyn forged “civilization” atop the archtrees; but now even his soul is beginning to fade, and must either be replaced or allowed to die. Nothing repeating or replaying is mentioned. By contrast, a little delving into DS2 lore makes it quite clear this is the case – “many civilizations” have risen and fallen on the spot where Drangleic stands (which is where Lordran once stood); the four Lord Souls have been taken by other creatures, who have been defeated by other “chosen undeads” (though DS1 does imply this very concept is a fictional creation if you talk to Kaathe or consider the role played by Gwyndolin), who have each had to make a choice about keeping the First Flame going or allowing Dark to spread. You begin the game tasked with seeking four Great Souls, and each of these corresponds to one of the four from the original game – Seath, Gwyn, the Witch of Izalith and Nito (though Seath had a fragment of Gwyn’s soul, the nature of the Dark Soul is different in DS2). They have taken new forms, but their origins are apparent for those who care to look. The new location of Seath’s soul features a Duke and an entombed Dragon; the new owner of Gwyn was also a king who overreached; the new owner of the Witch’s soul tried to relight the first flame and appears to have been cursed by Chaos; while Nito’s new owner is also a foul amalgamation of corpses dwelling far below the Earth. This isn’t “cyclic history” – this is just repeating the same ideas, which were wonderful the first time, but we’ve seen them before.

As much as I wish to say this is just a part of the story… it just strikes me as lazy. By the game’s own admission the first half is a “repeat” as you gather the same four Souls. For those Lore-buffs, you will note that the Lord Souls are what gave Gwyn et al their power, not something inherent to those individuals/creatures themselves. That may be, but that’s no excuse to just say “other creatures now have these souls and are rather similar to their original owners!”. The second half of the game when you break away from the four Souls from the first game is far more interesting and varied and had far more “surprise” moments than the first half, which (sadly) only contained one or two. In general the original lore – about Navlaan, Aldia, the Dragon, the Giants – is genuinely great, and I’ve enjoyed keeping track of the internet’s collective deductions about all this new lore, but the rehashed lore is very uninteresting.



In Dark Souls 1 each of the few NPC characters has their own story which might not necessarily link up with your tale; they have their own objectives, and you can only trade/talk with them for certain periods. They move through the world and may die or live on their own quests. Completing these quests was an interesting extra challenge in the first game and one lacking in Dark Souls 2. NPCs don’t generally have their own quests, and once they get to the game’s hub, they then just don’t go anywhere! They just sit still until the end of the game and never do a blasted thing. Many of DS1’s characters were far more tragic – Solaire and Seigmeyer spring to mind, not to mention Sif and Artorias – whilst DS2 lacks a single NPC I felt a damned thing for, though I guess Lucatiel wasn’t half-bad. And I am convinced that “Laddersmith” is not a real profession.



There is a fine line between “homage” and “reusing the same art and AI assets we used in the first time”, and alas, Dark Souls 2 falls squarely on the wrong side of that divider. Two entire bosses from DS1 are simply lifted wholesale into DS2 whilst others (Najka, Royal Rat Authority) are lifted from DS1 but reskinned and with a few new moves, and others (Royal Rat Vanguard, Dragonrider, and then… Twin Dragonriders!) are, as much as it hurts me to say it, just phoned in. I am not rose-tinting DS1’s bosses – Ceaseless has hitboxes large enough to strike you in the next country, the Bed of Chaos is a travesty – but the selection is far more limited, many have far less impact (as they have less lore behind them) and some are simply direct copies, even with the exact same name as in DS1. DS2 does have some great bosses – the Smelter Demon has an interesting core mechanic (proximity does damage), the Executioner’s Chariot is very different and original, the Demon of Song is a magnificent concept (though not the most fascinating fight), but… the list is short.

Repeated Chest Ahead!

Repeated Chest Ahead!

Other, much smaller issues:

Four K-, er, Rings:

You can now put on four rings instead of two. This is a totally pointless change; any meaningful decision-making between your rings is gone. In Dark Souls 1 the Abyss forces you to use one ring for the Covenant of Artorias and thereby restricts you to a single ring; Lost Izalith all but requires the Orange Charred Ring to survive on lava; many rings are very strong in Dark Souls 1 and – unless you’re a good enough player to have the Red Tearstone Ring on 24/7 – offers a lot of tricky choices. Even the amazing Ring of Favor and Protection, boosting three key stats, breaks permanently when removed; its cousin in DS2 can simply be repaired. Four rings is far too much, and there was no point where I felt uncertain about which to wear. Also, the Ring of Binding is vastly overpowered (for singleplayer, anyway).

Randomized Crow-trading rewards instead of fixed rewards for specific trades:

This is inherently stupid.

NPC Graves:

If you kill an NPC, they return as a grave, and you can still interact with a ghost. So there is now no penalty for killing NPCs, and the weight of your decisions in DS1 are lost. Souls games are meant to eschew the weak hand-holding of so many modern games, and yet they’ve removed a minor but excellent aspect of it here.


Cursing – which used to be a permanent reduction of your max health by 50% until it is cured – is now an almost meaningless ailment, thereby constituting another step back from the amazingly bold/confident (ballsy?) game design of DS1.


So what went wrong? For the most part I enjoyed my first playthrough, even though some bits left me cold the first time I was playing them. Other parts impressed with their aesthetics or some of their enemies, but upon further consideration their linearity compared to the first game became apparent. The fundamental problems are two-fold – the development team tried too hard to create Dark Souls 2 instead of a new Souls game, and – for whatever reasons – they moved away from the maze-like structure of the first game (at both the individual level scale and the entire world scale) towards a world of straight lines with minimal deviation. Too much of the first half draws from the first game; too few bosses, NPCs or areas feel at all original; and too many regions just boil down to straight lines which try to cover up wildly inconsistent geography. There are many small improvements, but the few major steps backwards are just too significant. After finishing DS2 I returned to DS1 for the first time in six months and instantly fell back into its incredible world. I thought I was certain to want to 100% DS2, and yet nothing from that world actually calls me back. There’s so much to go through to get to the more interesting segments, and even those segments are still just walking from one point to another. I’m sure I’ll revisit it some day, but Lordran, not Drangleic, remains my Dark Souls home.