Dragon of Icespire Peak - A Review

A REGRETTABLE WASTE OF OPPORTUNITIES

Icespire GM's Screen - that almost makes up for sparse art.

This blog isn't often kind to the products of Wizards of the Coast - largely because the ‘Crawl’ (as in Dungeon Crawl) playstyle that All Dead Generations champions is very different then the one 5th edition Dungeons & Dragons seems directed towards. Yet, taking a deep look at offerings from Wizards of the Coast is one of the best ways to highlight those differences and understand them. Lost Mines of Phandelver has been the WotC sanctioned introductory adventure since the 5th edition came out, but in 2019 Wizards of the Coast published the Essentials Kit, updating the boxed set concept for 5th edition and including an introductory adventure/campaign “Dragon of Icespire Peak”.

Introductory Adventures
Introductory adventures are interesting things, doing a lot of work to define setting, and if they're part of a particularly successful system they can offer a model for adventure design to an entire generation of players. When the first edition of Basic Dungeons and Dragons was introduced in 1977 it didn’t contain an adventure, though this was remedied by 1978, when B1 (for “Basic”) “In Search of the Unknown” by Mike Carr was included. In Search of the Unknown, despite an alluring cover illustration* and amazing title has to be regarded as a somewhat experimental product, which perhaps took its role of educating the new GM too far at the cost of being fairly uninteresting and a bit unplayable. Alternatively B1 represents an insight into what early D&D looked like -- its lack of naturalism or any kind of monster ecology (it depends on random stocking) and equally unnatural map emblematic of a wargame derived early play that Gygax (for all his flaws) showed an alternative to. B1 was quickly supplemented by B2 “Keep on the Borderlands” by Gary Gygax, which is likely his best work, and still remains a read for anyone interested in adventure design or game mastering.

“Keep on the Borderlands”, the nature of the challenges in its caves of chaos and the playstyle it fairly effectively taught defined Dungeons & Dragons for TSR’s early 80’s golden age: dungeon crawls based on a stilted internal logic and ecology where scheming humanoid factions were the primary foes within a ‘points of light’ setting. That’s the power of the introductory adventure, to not only showcase an official setting (promoted or implied), but to set the tone and playstyle. “Keep on the Borderlands” was removed from D&D basic sets in 1983, which instead included a short solo adventure heavy on scripted events (as it would have to be given it’s solo nature), around the same time as the first Dragonlance module (“Dragons of Despair”) was published championing adventures of the linear, scene-based style where player characters receive immunity from harm to assure the adventure's narrative remains predictable.

Yet “Keep on the Borderland” set the basic model for the introductory adventure, one that “Dragon of Icespire Peak” even follows to some degree, it to is a regional set of adventures set in a lawless region of a world where “[e]ven farms and freeholds within a day's walk of a city can fall prey to monsters” prefaced by more general play instructions and including play aides. The adventures within “Dragon of Icespire Peak” however, and especially how they are structured, vary from the open world presentation of B2. The question for this review is thus “How well does Dragon of Icespire Peak work to introduce players and GMs to the game, and what sort of game does it introduce?” Even more specifically, and derived from looking at other 5th Edition products, “What, if any, are the contradictions between Dragon of Icespire Peak’s fiction (setting and adventure details) and the mechanical playstyle it presents?”

Dragon of Icespire Peak

A 64 (Including 15 pages of 5th editions’ indulgent monster stat blocks) page series of 13 adventure locations ranging in size from 5 (Umbrage Hill - a Manticore attack) to 30 (Axeholm - A ruined Dwarven Fortress) keyed locations and designed to be played episodically, connected by an underlying structure of regional events. I believe the adventure is written by Chris Perkins, he’s credited as the designer in the rulebook, but not the adventure itself. Likewise art in the standard 5th edition style drawn by a passel of artists specifically for the adventure, unlike early 5th edition offerings. The art and cartography has the bright colors and generic fantasy look one expects from a Wizard’s of the Coast product, though it’s pretty sparse - we have illustrations of several monsters, cover art of some adventurers confronting the titular dragon, NPC illustrations, a vista of Phalanden, a random gnome fiddling with a contraption and a skeletal horse. Given the blandness of the wotC house imagination/style and the quotidian content of Dragon of Icespire Peak this is good as far as it goes. No illustrations that are especially useful at the table (magic items, complex rooms), but the NPC cards are a nice touch and drawn in a more whimsical style than most WotC illustrations. Maps are likewise typical of a WotC product, serviceable, not especially complex but not entirely linear either -- though the small number of keyed areas in many of the locations limit the orienteering aspect, which is perhaps a hallmark of 5th edition play.

Reading through the overall introduction of “Dragon of Icespire Peak”, I’m pleasantly surprised by its openness and stated commitment to player choice. The central “job board” gimmick is something borrowed from CRPGs from before they had the budget or graphics to animate NPCs - but I guess it’s an expected trope. I sometimes wish there was a board in my town that said things like “The mayor will pay $5,000.00 to any brave souls that investigate the ruined missile silos and defeat the scabrous vermin that dwells within” but like much of WotC’s brand of D&D fantasy the job board has a logic of its own at this point. Starting quests lead to complications and new quests that all tie into an overarching region's situation of the same sort that this blog champions - factions fighting over land and power, with the adventurers theoretically holding the balance. While Dragon of Icespire Peak still assumes the party and players will be drawn to heroic motives, unlike many WotC offerings it doesn’t lay out how to act on them in a linear and entirely predetermined manner. The seemingly open basic structure is reinforced by a set of advice that might seem familiar to All Dead Generations readers, though it's fairly general in nature.

GM Advice
The GM advice in Dragon of Icespire Peak is solid, it offers suggestions that encourage a less linear play style more focused on player agency, and while it still focuses on the idea that a story underpins a D&D adventure, it doesn’t try to convince the GM to sacrifice players’ control over their characters in service of that story. The description of adventure design and purpose on page 2 of Dragon of Icespire Peak is excellent: “A D&D adventure is a collection of locations, quests, and challenges that inspires you to tell a story. The outcome of that story is determined by the actions and decisions of the adventurers and, of course, the luck of the dice.” Other advice is similar, though there’s a subtle influence of 5th edition Dungeons & Dragons preferred vocabulary that nudges the reader away from exploration style play: adventures are still “stories” and play is still a series of “encounters”, yet the advice itself is sound. Given the authorship of “Dragon of Icespire Peak”, its openness to players having a say in events and choosing how to approach an open ended set of problems is surprising. Chris Perkin’s own game writing, an interesting series of essay on adventure design, feels dedicated to creating a cinematic play experience full of climactic moments, but without much concern for the fidelity of the setting or the centrality of player choice. Perkins of course has good ideas about how to design for this Adventure Path style, but it’s still surprising to see one WotC’s less linear adventures with his byline.

More specific advice follows the general, how to improvise ability checks, use the reference cards provided and keep location maps away from players. An additional section covers more details related to starting the adventure and the potential for running adventures for a single player. This advice is limited to the most basic ideas: a solo player will need henchmen or don’t worry about party balance; however, even this simple advice seems to push towards the direction of player choice. From the totality of the GM advice in “Dragon of Icespire Peak” it seemingly wants to offer a regional sandbox with the brewing problem of an aggressive young dragon on the rampage and the band of orc refugees who have turned to sinister gods to seek revenge. Given its stated goals it seems entirely fair, in a way that wasn’t quite fair to Descent into Avernus, to judge Dragon of Icespire Peak as a regional point crawl adventure.

Structure,  A Town, &  The Point Crawl

Structure might seem like a strange thing to begin with in a point/hex crawl style adventure, which are often critiqued as lacking story, but there’s always one there. Even the most undirected and open crawl style of adventure will have (or soon grow to have) a set of relationships between factions and the party. The town in hex A4 is raided by the bandits in hex B6, and the party’s interference will likely change that relationship. The players will wonder if the bandits in B6 were in contact with the bandits they later meet in F14 and that the party sat by while the town in A4 was sacked may have an effect on the castle in A8 which claimed rulership of it.

Still a good crawl has more than just implied or evolving faction relationships - at least more than static ones. A good crawl has something (or multiple somethings) going on in the background that the players can discover, interact with and ultimately effect: sinister forces working to pull a city into hell, a necromancer searching for power that risks raising the wrath of ancient kings, a mercenary army desperate for plunder and eyeing the region’s soft underbelly. None of these sorts of grand concerns are plot in the usual sense, the crawl doesn’t say that the the party will ultimately heroically stop Hell’s ambitions at some specific point in time under certain circumstances, but they are certainly plots. Hell plots to take the city and will succeed after some number of sessions filled with signs and portents. The necromancer will unearth a grave site each each session and the Gm rolls on a random table to see how the vengeful dead respond. The mercenaries send their agents looking for gullible wandering adventurer types who will report the size and nature of local forces (or undertake more dubious skullduggery) for coin - until they have enough knowledge to sweep in a put the region to sword and flame.

This sort of overarching situation appears in Dragon of Icespire Peak, but it’s fairly static. There’s a white dragon, newly arrived on Icespire Peak and its chased a band of orcs out of a ruined fortress. The orcs have come down from the mountains to meet up with an orc cult that worship an evil thunder god and summon its boar avatar. The dragon is also chasing and threatening other monsters and driving them into the lowlands where they endanger the oft threatened town of Phandalin. In turn the town and other more peaceable lowland groups: dwarf explorers, ranchers, hunters and a gnome commune are beset with smaller problems sometimes related to the dragon and orcs.

This seems like an entirely workable structure, with factions that interact with each other and have goals and plans -- except they don’t really. Nine of the fourteen keyed locations are directly offered as “quests” from the mayors board in town while the remaining five (including the dragon lair) are either mentioned by NPCs in the posted quests or town rumor. This isn’t that bad, though rather than a quest board I’d stick with rumor or individual hooks from NPCs in town. What’s a bit concerning is that there’s little interactivity between the quests, they’re almost entirely independent and exist to tick down a ‘quests completed counter’ to level the player characters and reveal the next set of more difficult quests.

Some of this is an artifact of 5th edition’s steep power curve and rapid leveling, though arguably it’s also an example of design that distrusts the players. A need to narrowly gauge the risk of challenges to protect characters from defeat, not trusting them to retreat from overwhelming odds. More though is a basic weakness in the structure -- the dragon Cyrovain -- doesn’t escalate the situation and force interactions between the party and inhabitants or between the various factions. Cyrovain is a random encounter, which is good, though the random encounter mechanics themselves are clunky, but he doesn’t really wreck proper draconic havoc, and the orcs don’t do more than appear as villains in a few of the locations.

A clock or even a set of event triggers attached to the quests could help this aimless feeling and bring immediacy to the threat of the dragon -- a threat that needs to be tracked down, killed or driven away before it destroys the entire region. Doing this though involves endangering not just the characters, but the adventure content. If the dragon destroys the outlying mine the party can’t encounter a mine infested by wererats, if it eats all the loggers there will be no one for the party to make supply runs to. For the designer this is of course upsetting, one’s ideas won’t see play and since the destruction is not entirely reactive (if it’s on a clock) a dithering party may never be able to amass the XP to confront the rampaging dragon. Acknowledging the possibility that the players can fail, that not everything in the adventure will be played and that events won’t unfurl for every table as the designer envisions them is a tricky part of classic design, but a core one. “Dragon of Icespire Peak” doesn’t quite seem to understand how to give up that control, which is rather in keeping with the 5th edition ethos, and it’s a weaker adventure because of this.

Summoning a Storm Pig/God should be on a clock.

“Dragon of Icespire Peak” though isn’t all bad, it’s a vast improvement in design over the adventure it references, borrows much of the setting from and ostensibly replaces - Lost Mines of Phandelver. This can be seen in many paces, but at the core it’s an improvement in structure. Lost Mines is a linear string of locations loosely set on a map, while Icespire Peak is a branching set of locations and scenes loosely set on a map. While both use the same cloying, predictable Forgotten Realms setting “Dragon of Icespire Peak” uses it better.

For example, both adventures use the mining town of Phadlen as a central hub and faction that they expect the party to ally with. It’s not a very interesting fantasy village, though at least Icespire Peak makes a few more mentions of the crumbling ruins it’s built atop. What “Dragon of Icespire Peak” does though is actually use the town as a location and hub for adventures, containing NPCs with some description. The town NPCs offer local rumors, a rarity in a contemporary WotC product, and have character beyond offering some sort of ‘sidequest’. The description of Phandalin may not contain much, using the WotC house style that emphasizes generalities over detail and telling unknowable facts rather then showing the character of people and places, but at least it gets across something of the rough frontier nature of Phandalin. The “job board” at the center of Phandalin is a poor mechanic for a point crawl, and it’s not the only one in Dragon of Icespire Peak. Again, Icespire Peak seems afraid to move too far from the tried structures the adventure path and the simplistic content of contemporary CRPG derived fantasy. The job board replaces connection between NPCs, it replaces player interaction in town without gamifying character time in havens.

One could offer an organic, roleplaying and relationship building style of detailed town interaction: in town factions, NPC goals, roleplaying through negotiations and rumors. Alternatively haven time can be gamified: pick a ‘downtime’ action from a menu of activities between sessions knowing how its specific mini-game functions mechanically and allow complications to appear via random table (or clock/timeline mechanics), perhaps even set player knowledge/hook engagement by faction reputation.

The job board of “Dragon of Icespire Peak” attempts to find a middle ground, and largely fails -- it limits interaction with town NPCs, uncouples most locations from rumor, and still does nothing interesting with hooks. It also fails to engage with the discovery aspect of a point crawl. While the setup exists in some larval form - a map, a series of locations with their own maps and keys, and even an overarching regional situation, but none of this exists meaningfully -- it’s not backed by mechanics (even the Rulebook in the attached essentials only says that PCs can travel 24 miles a day) or narrative import: there are no clocks to race against, no supplies to count down, no weather or random encounters to make travel a risk. Again it’s a confused failure at understanding what Dragon of Icespire Peak is trying to do -- the regional map of the Swordcoast has a hex overlay and a key, with a badly blurred notation that each hex is 5 miles, meaning that it will take a couple of days to reach Niverwinter from Phandalin but travel has no risks, consequences or play associated with it - the party simply moves from Phandalin to whichever location have been indicated by the quest board or perhaps a rumor.

This may seem simpler, a more direct way to move characters to the specific adventure locations where play is set to occur, but in doing this it eliminates exploration, which is counter productive to the goals of encouraging player engagement in the region (they won’t learn anything about it as they don’t interact with it) and empowering them to hunt out Cyrovain and his lair (There’s never an indication that locations not provided by the quest board can be accessed). The point crawl structure here is effectively a web of interconnected points, without any obstacles between them, but where some locations are hidden until either a quest is posted or the information obtained in another quest.

The most egregious example is Cyrovain’s lair in Icespire Hold, the only way to find it being a wounded orc captured by the townsfolk of Phandalin or a prayer to the luck goddess as suggested to the party by any NPC when they are the right level. It’s not just that these mechanics are clumsy, they are also unnecessary. The Dragon patrols the region, it’s driven a Manticore and a band of Orcs from their homes, both of whom have a grudge against it and who the PCs interact with. The orcs represent the major antagonists (itself a problem) in the second ½ of the adventure. It’s as if Perkins simply doesn’t understand how to create the regional sandbox or pointcrawl he hints at in the How to Play section or can’t trust the players and GMs running his adventure to unravel any mystery or engage with any location that isn’t a “quest” directly provided by an NPC.

Despite having a map, scattered locations, appearance of factions, and overall situation “Dragon of Icespire Peak” is not a hex or point crawl - it’s design is mostly linear and it’s episodic locations are only tenuously connected to each other. It certainly wouldn’t be hard to run “Dragon of Icespire Peak” with a bit more connectivity between the locations and the factions in them, to bring Cyrovain the dragon’s depredations to the fore as an actual mechanic with clock, timelines or better random encounters and give enough situational immediacy to spur player motivated investigation and adventuring on how to combat the dragon. Suddenly all sorts of opportunities, using the same set of tools become possible -- among them making common cause with the orcs (or really something other than orcs - for reasons), finding magical weapons, restoring the old dwarven fortress, and/or ultimately finding and raiding Cyrovain’s lair.

Ultimately the adventure would work well presenting the situation as an open-ended problem: a dragon whose penchant for destruction is driving other monsters into settled lands and must be stopped before the beast destroys Phandalin and returns the Icespire foothills to a ruinous wasteland. Of course this sort of set-up requires two things: acceptance that the players might fail or resolve things in a non-standard/amoral manner and designing a region, rather than a series of discrete encounters.

Locations, Locations, Locations
There are several keyed locations in “Dragon of Icespire Peak”, some are small lair or raid style encounters and others more complex. They are tired 3 for levels 1-2, 3 for levels 2-4, and 3 for levels 4-5. At level 6 the party will discover Cyrovain’s Lair. This sequence is designed to have the characters gain a level each time a quest is complete. Like most milestone systems of leveling the completion system tend to inject a great deal of the author’s moralizing and plans for how the adventure should run into the design. These “Quest Goals” aren’t always incredibly precise and the adventures themselves often contemplate a few possibilities (unless orcs are involved - then it’s ‘eradicate the evil’, which is a design problem) so it’s not as bad as it might be.

The locations themselves are pleasantly varied, though they tend towards smaller maps with few keys. The first batch includes a disgruntled Manticore’s attack on the local alchemist’s windmill outside town, a ruin with some dwarven prospectors who want help with its current gelatinous inhabitant followed by an inevitable orc attack, and a colony of gnome recluses who are being hunted by a mimic. The second set are a rancher chased from his home by orcs, a logging camp cut off from supplies by ankhegs and a gold mine infested with wererats. The final set of adventures are a haunted barrow hiding a magic dragon slaying sword (and leading to a potential orc fighting location), a hunter in the woods who needs orcs cleared from a ruin, and Axeholm, a larger dwarven outpost now haunted by undead dwarves. Optionally the party may also venture to a lighthouse inhabited by an orc cultist. Finally there’s the dragon Cyrovain’s lair in a ruin mountaintop fortress.

Of these I find a few interesting for various reasons, the first three quests all try to do some interesting things, with somewhat mixed success.

The gnome burrow wants to be a strange mystery: it’s kings have gone mad as and there’s a crafty mimic slowly eating the residents. A fun enough scenario, but the actual two page adventure offers almost no tools - likely just becoming a fight with a mimic. There’s a note that maybe the mimic is a special variety that can talk (with reference to some canonical source of talking mimics -- hint if you want your mimic or whatever to talk, just make it talk, it’s okay as a designer to change monsters) which is great, but without motivations beside a love for succulent gnome meat little room for negotiation. The map here looks useful, but since it’s a friendly location (though thankfully this is not assumed), it seems unlikely that the numerous keys are helpful. A better use of the designers space would be a timeline of mimic and mad king activity that helped players unravel the mystery - instead of only offering the suggestion that the gnomes will walk around poking things with sticks and then run away, starting the fight music and leaving mimic for the players to fight. I want gnomes with schemes, traps and odd theories about the monster, not something that channels the entire situation into a set piece combat while giving me near useless location keys.

Umbrage Hill has a great name, it’s not a fake fantasy style string of syllables (or worse apostrophes), and the set up is reasonable enough. Cyrovain has driven a manticore down from the mountains and it attacks the windmill (presumably seeking a new home/lunch) where Phandalin’s alchemist/herbalist lives and works. As the only source of healing potions - a seeming necessity in 5th edition that makes the game feel even more CRPG to me - this gives the players a lot of incentive to deal with the attack. The one page encounter though is devoid of context: a tiny location, a nameless manticore -- just another set piece combat with a lone enemy. There’s the suggestion that the Manticore can be bought off for 25GP only to return with its mate and cause trouble. This is a good inclusion - but I can see it being improved on even more. The manticore and its mate (I’ve always considered manticores not to have mates - being embodiments of masculine antisocial spite and rage - baby manticores seem like an impossibility - but maybe they like to get together and hate the universe with a partner?) are powerful flying creatures. Sure they’re malicious, enjoy the taste of human flesh and get joy out of lying, but they likely want their home back and have reason to allow themselves to be hired on as scouts in an effort to kill Cryovain. A nice untrustworthy, useful and obviously evil faction is always a fine addition to a regional sandbox, but here any use or story from Umberage Hill will be the result of players and GM playing against the adventure’s goals, and extrapolating interesting tidbits from sparse offerings.

There’s nice details on the hill - a dwarven graveyard/monument (the dwarf ruins that make up so many of the adventures in Icespire Peak are exciting unwritten setting consistency even if dwarf ruins are among the most cliched of adventure locations) and the ruinous repurposed windmill both feel in keeping with the reclaimed hinterlands setting. I’d amplify the ruins though - in general amping up elements of your setting that you like is positive. If you like your dwarf ruins … decide the mountains are hollow constructs, crumbled megastructures that were once entire dwarven empires and now ruinous mile tall walls carved with incomprehensible reliefs, roofless underground maze-cities exposed to the stars where the miners of Phandalin dig for lost dwarf hoards and scrap heaps rather then ore. The dwarves of the present? They pretend this is the work of their ancestors, but they don’t know anything, they just see themselves in the severe bearded faces carved impossibly huge by hands eons dead. Push the fantasy you like to the limit, one of the weaknesses of Forgotten Realms is that it wants to make the fantastical small, a nostalgic and pastoral setting with a quasi-medieval gloss on a world that reflects our own in a non-threatening, simple way. I think that makes for worse gaming.

Speaking of Dwarf ruins, the last of the starting quest is simply titled “Dwarven Ruins” and of the three it’s the most dungeon crawl like. The party finds a pair of dwarven explorers/treasure hunters who ask them to help deal with an ochre jelly they’ve encountered in the ruins. Of the three, it’s also the best. The NPCs have personalities and goals - terse and simplistic, but that’s perfect for dwarf prospectors/explorers. Dalyzn and Norbus (Fantasy names, but at least simple ones) have the makings of interesting recurring NPCs - almost functional as henchmen, with a body of knowledge (nearby dwarf ruins) useful to the PCs and it’s a shame that they aren’t used as such - establishing themselves in Phandalin to buy dwarven artifacts or offer the possibility that if the party gets on their good side (by fulfilling their demands for valuable dwarven antiquities) they might be able to call in Dwarven aid or if they aren’t happy entice a pack of surly dwarven adventurers that will impede and double cross the PCs.

The ruins themselves have a couple of interesting elements - a magical trap, and a deviation form the most cliched fantasy dwarf stereotypes. Arguably this location interrogates the cliche of the ruined dwarfhold, and that’s better than a lot of alternatives. The ruins are a temple to an evil dwarven god of greed, and it feels fairly naturalistic, with Norbus’ and Dalzyn’s explorations leaving signs, breaking through a secret door, and spiking things open until they reach the lair of the jelly. Secret doors are a recurring theme to the Dwarven Ruins and they’re used decently. Their presence indicated by prior discovery and their positions mirrored to make their presence and location something PC’s can deduce rather than guess or discover only through excessive, time consuming caution. Most importantly the secret doors don’t act as obstacles to continuing the adventure - instead they conceal treasures. This map in general seems to do what it wants to - create a feeling of archaeological exploration while raising some scary implications about the nature of the dwarven culture whose ruins dot the region. It’s unfortunate that it ends in a seemingly gratuitous orc attack - one that’s both unavoidable and results in an automatic combat.

“Dragon of Icespire Peak” uses its orcs badly. Beyond any larger consideration of barbaric, dark skinned, bestial humanoids in fantasy fiction and game design Icespire Peak’s orcs are simply not used well from the perspective of a regional point crawl. They’re included multiple times, and always as foes without any possibility of negotiation or understanding, and yet the orcs are Cyrovain’s first victims. Leader slain and defeated by the dragon the orcs of Icespire Peak retreat to the lowlands where they fall in with an evil storm god cult. This seems like a compelling narrative, a meaningful part of the regional situation, and there’s nothing wrong with the set-up. The set-up of a dispossessed faction that makes trouble is a decent one as it has multiple solutions - the status of foe, refugee or ally entirely depends on the players and this scenario offers fertile ground for moral play and long term effect on the region. All squandered by “Dragon of Icespire Peak”

Personally I’d make the orcs the men-at-arms of some wretched robber knight, human mostly, maybe dwarven if I kept the rest of the region the same, but this is a personal choice. Orcs, like brigands, represent a distinctly comprehensible evil - they’re people (or monster-people) who despite the fundamental ability to reason and a shared set of needs don’t have the same values as more friendly factions. Unlike a dragon however, they aren’t a singular reptilian intelligence pathologically dirven to accumulate wealth. Still players can negotiate with dragons, and so should be able to negotiate with the far more understandable orcs, especially when there’s a shared goal of driving off a murderous flying magic lizard. From a design perspective the orcs are simply wasted a faction that could provide a counterbalance to the dragon, but whose goals (their own comfort and power, summoning a terrible deity) are generally opposed to Phandalin’s, except where Cyrovain is concerned.

Two last locations are worth noting, the larger dwarven ruin of Axeholm and the dragon’s lair.

Axeholm is a dwarven fortress with 29 keyed locations, haunted by a banshee and a bunch of dwarven ghouls and few stirges. Covering only four pages the keying is unusually terse for a Wizard’s of the Coast Product (as is all the keying in “Dragon of Icespire Peak”) and most of the fortress is empty. One of the things I’ve talked about a fair bit in looking at keying and design in Descent Into Avernus, was the ways that the mechanics of 5th edition don’t support dungeon crawling. Here again we have a map seemingly designed for dungeon crawling (two levels with a basically hub-like structure where corridors or series of rooms radiate from a center), but the map is largely empty of obstacles, encounters and description. It has a meaningful spatial element - and even some interesting spatial puzzles such as the difficulty of entering the sealed fortress, but there doesn’t seem to be the necessary mechanics to support dungeon crawl play here. Less even then in 5th edition’s Player’s Handbook and Dungeon Master’s guide. The Essentials kit has less then a page on movement, most of which is about combat and special movements (surprise! You make another DC check!), and no reference to timekeeping. Supply exists only as a list of items characters may carry, with no mention of how they are exhausted or how much characters can carry. There are literally no systems for encumbrance, timekeeping or non-combat movement in the D&D Essentials kit - something that is understandable given its priorities, but which makes running an adventure like Axeholm a chore. What does exploration of a dungeon look like without rules for exploration?

Threats and obstacles are separated from each other - still discrete encounters. The players reach Axeholm’s gates, bypass them, and begin to explore empty stone halls or if they were noisy getting in, face the fortresses’ entire ghoul population. The halls have little description beyond noting that pillars are made of stone, defenses are everywhere, and the floor is quite dusty. The party will wander about, their movements unworried by risk and incapable of discovering nonexistent clues, effectively meaningless until they stumble on a ghouls, number determined by party size, in nests of bones and dispatch them. There’s little treasure, except that carried by ghouls, no puzzles or traps and nothing that allows the players to learn many of the details of former room use supplied, or even the story of how the fortress became a ghouls fouled ruin. All the monsters either attack or attack and lack significant intelligence to parley with. Nothing moves from its lair in the fortress except as triggered reinforcements. There’s no writing within the fortress, or decoration, secrets are limited to a single secret door with nothing to indicate its presence or possibility. As much as Axeholm wants to provide the dead feel of a ruin abandoned to tragedy, it can’t manage that with only a string of combat encounters. Even the carefully thought up defensive structures, nice details all, are meaningless here -- simply filler that will be ignored because the former use of the rooms is unimportant and the encounters are lair set-pieces. A fortress of drops, giant cauldrons and artillery becomes interactive and useful if the location has some dynamism, if its residents attempt to use the environment or move about allowing the party to.

Axeholm represents the larger possibility of a regional complication as well - squandered again. The quest suggests that in the event of dragon attack (one that never happens) the town should flee to Axeholm - create a defensible fortress to protect from the dragon. This is an interesting plan, more interesting given how easily the place could be turned into a trap for a dragon with it’s defensive structures, hallways of ballista and murderholes. Likewise, it seems like a nicer dragon lair than the ruined Icespire Fortress (or perhaps a place manticores or orcs would love to move into) with easy access to the sea and prey filled roads around Neverwinter. Finally it’s another dwarven ruin! It’s filled with undead dwarves - it’s a shame not to combine it with the evil greed dwarf town of the Dwarven Ruins and the Dwarven explorers there. Even the cairns on Umbrage Hill could feed into this. None of this has to be a serious, but a thread of implied secrets (the ruins of a greed worshiping civilization of Dwarves that used to live around Icespire Peak) is just prime hook material and setting detail left to rot. Axeholm could be such a useful place on the map whose presence generates story. Unless it’s just a location to kill ghouls in because there’s no real factions, no spatial element to any maps, and little connection between adventures.

Finally, through some badly mangled illusionism the party will find itself hiking up Icespire Peak to confront Cyrovain the White Dragon. I enjoy that Cyrovain has a somewhat silly name which speaks to its character - it’s an ice breathing dragon of the greediest, dumbest sort. “Cold Vanity” is a good name for it, in fact I’d rename it something more along those lines, “Vanity of Ice” maybe. Cyrovain’s a fine dragon name though, excellent because it’s also a characterization for a regional troublemaker. Cyrovain likes killing things, sometimes eating them, destroying stuff, and stealing shiny objects. Cryovain isn’t too smart, but has the ingrown and absolute belief that everything else in the world weaker then it is it’s prey or plaything. Susceptible to flattery and displays of threatening magic, open perhaps to bribery, but unwilling to keep a deal a moment longer then it’s little greedy mind finds it useful.

That’s my Cyrovain, and indeed the adventure describes him as “slow witted and easily baited”, which is good enough characterization. There’s even a paragraph about how and when the dragon will react to intruders in its lair, which includes basic tactics. There’s a force of mercenary brigands in the fortress, waiting to rob the dragon - but they don’t provide dubious allies or even offer a race to find treasure, they’re simply a warm up fight before the dragon - another squandered opportunity, made worse because there’s a sidebar that gives the four brigands more character then any other NPCs in the entire adventure.

The fortress itself, like Axeholm is the site of a tragic disaster -- a trapped garrison starving through a long winter. It’s also minimally, but poorly, keyed and of course without any sort of exploration mechanics. There a few good bit of dungeon dressing in Icespire Fortress, a taxidermy winter wolf head, skeletons arrayed in vignettes that tell of their demise and a feeling that the dungeon has been lived in and used by multiple waves of inhabitants. Unfortunately the designer spends far too much of the limited descriptive space of undiscoverable, unimportant history - a skeleton is found in an empty room where the description tells us this was once the supply room and the dead man was killed by his fellows for trying to steal while the fortress was starving (and just left there, not cannibalized - but that’s perhaps to grim a disaster for a WotC product). It’s poor keying, poor design - another way that “Dragon of Icespire Peak” wastes potential and fails short of its proclaimed intent.

A larger, frustrating problem of “Dragon of Icespire Peak” and one almost ubiquitous in 5th Edition Dungeons & Dragons products from WotC is a lack of usability compounded by poor organization. There are plenty of examples, but the one that kept confusing me was the arrangement of the locations within the book, arranged alphabetically rather then in chronological order, finding the right adventure at a given moment and contextualizing it in the larger adventure can be hard, especially when adventures that do refer to each other or act as a sequence aren’t nearby in the text. This is just one small thing, but like most WotC products “Dragon of Icespire Peak” makes little effort to help the GM run it.

How I'd Change Icespire Peak
Mechanically Icespire Peak would be better as an actual regional point crawl. Define the major factions: Phandalin and friends, the dragon and the orcs. Flesh the factions out with dysfunctional internal dynamics and create clocks, tables or timelines for them - what triggers their attacks on eachother, when they take over certain areas etc. Review the minor factions: manticores, undead evil dwarfs, gnomes and add actual goals, knowledge they can share and efforts that can be triggered by player intereference. Finally there need to be tools to answer questions like: 

“How well can gnome engineers repair several decaying dwarven ballista, what do they want in return and are they competent artillerists?"

"Who else in the region can arm, aim, and fire a ballista at a dragon"

Finally supporting mechanics on supply, travel and finding locations would help answer more technical concerns:

"Who knows where adventure locations and factions are?"

"How hard it is to travel up a snow covered mountain without a mule?"

Dungeon locations could use mechanical support as well: timekeeping rules, encumbrance, random encounters and more dungeon dressing/secrets/non-combat obstacles.

That’s really all one would need mechanically to make "Dragon of Icespire Peak" a more compelling sandbox adventure.  This is why I keep mentioning missed opportunity.

Thematically and fictionally I think Forgotten Realms could use an entire overhaul to inject even a tiny bit of wonder into the setting, but here a start would be replacing the orcs/half-orc cult with human brigands and robber knights. If it must be part of the adventure, bring the dwarven ruins theme to the fore and re-frame Icespire Fortress as one - on the massive, monumental scale I mentioned before.  Finally rework the dragon as a monstrous embodiment of destruction drawn to a specific variety of human/dwarven suffering (the sudden tragic destruction by avalanche, loss of their heating and earthquake of the last bastion of the Icespire’s dying, greed worshiping dwarven culture).

Many elements of “Dragon of Icespire Peak” are painfully cliched, and the reason I call it painful isn't just personal taste. Designing adventures that draw only from the cultural currency of post-Tolkien mainstream fantasy (hugely influenced by AD&D’s implied setting), and what I'm calling vernacular fantasy here, is an onerous limit on design. Obviously a fairly shallow pool for inspiration, but also one that gives the designer few tools to hold player attention or create wonder.

Dealing in Cliche
Fans of Forgotten Realms style world building will argue that cliches allow for immediacy and clarity because everyone playing has a shared frame of reference. If dwarves are gruff miners obsessed with axes, brew and beards every player with even a passing familiarity can visualize their hirsute faces, barrel chested and mail-clad bodies. The GM needs only string cliched concepts together to form easily read referential chimeras rather than describe creature or locations.  It’s awfully convenient: instead of describing seamlessly carved stone halls wider and shorter then the golden ratio (say 7 feet wide and 7 feet tall) decorated with angular inlay in darker stone the GM simply says “Dwarf Fortress” and rather then describe an actual monster just says “Dwarf Ghoul”.

This is boring, and worse than boring it makes it harder to engage player interest. Not in the sense that a droning noise is boring because it’s annoyingly meaningless or aggressively ugly, but in the way that performing a rote task is boring - allowing for distraction and inattention. The players’ minds drift off from the excitement of the adventure, away from the GM’s description and most likely to a mathematical sort of evaluation of a ghoul’s mechanical strengths and weaknesses - no longer engaged with the fantastical story, but instead with the rules.

Worse, cliched content is harder to write. Everyone may know what an abandoned dwarf hold looks like, but after checking off the basic expected elements of forge, drinking hall, living quarters, mine and barracks (there might be a few more bits) it’s hard to add things without them seeming out of place highlighting the triteness of the expectations by comparison. It won’t seem creative to add a series of botanical domes to a cliched dwarf fortress, it’s more likely to create a jarring oddness that at best pulls the lack of detail and specificity in the rest of the location to the fore.

One of the most important elements of writing useful location keys is including evocative detail while picking cliched locations and themes makes it harder or actively works against it. Cliche act against players to feelings of wonder and any sense of discovery (even if it’s new to the campaign players have already imagined or seen a ruined dwarf hold a hundred times in other media.) Cliche also activates a sense of cynicism. Literally world-wear, the players recognize the setting content and are unlikely to feel any need to interrogate the world because expectations are set: the tomb is always full of undead, the wizard’s tower always home to a summoned demon run amok, and every monumental cave winds down to a dragon etc. Reliance on cliches risks the expectation that everything in the location or game is following the well known path of contemporary vernacular fantasy which makes departing from it feel unfair, “unrealistic”, or punitive because it destroys the players’ assurance that they know the basic ways of the world - ways that weren’t learned through play but laid down by other media. GM, player and designer become increasingly trapped in the expectations of well known cliches, making for a dull, exhausted setting that's much harder to breath new life into.

One can of course invert cliches or overindulge them in interesting ways to attempt a cure, and “Dragon of Icespire Peak” is best when it makes time gestures in that direction. The suggestion in Dwarven Ruins that the inhabitants of the fallen hold weren’t stoic miners but instead skull collecting greed cultists is such an inversion. Cruel, dangerously religious dwarves aren’t a fantasy trope (or at least not much of one outside maybe Norse Myth), and it's almost a direct inversion of the classic honorable, hedonistic dwarf cliche.

Above I suggested that because the Icespire Peaks contain so many dwarven ruins the geography should be reworked into a ruined dwarven megastructure. This is an example of overindulgence of a cliche working to transform it. For example, vernacular fantasy dwarves always build monumental statues of their bearded ancestors but this might become interesting again when an entire mountain range is bearded ancestor statute (weather worn, partially collapsed to reveal the galleries of a lost city within) and it will  hopeful reawaken any initial wonder that the player might have had thinking about Tolkien’s dwarves and their culture on first exposure. Dwarven ancestor worship is staid, but carving an entire mountain to look like an ancestor suggests something so bizarre and inhumanly excessive - a singular, centuries-long, obsession or collective undertaking with a profound mastery of the environment - that it might again be wondrous.

Using this technique I’d recast Icespire Castle as a titanic geothermal forge - the last redoubt of the ancient dwarven society. Then tragedy -- an earthquake closed the steam vents, burying many of the already diminished people in an avalanche while the survivors, lost in their constant scheming and greedy accumulation at each other's expense, couldnt rebuild and were driven away from the broken shards of ancient glory.  Sites of such sudden disasters of course attract wyrms, who feed mostly on the psychic reverberations of terror and tragedy.

Not entirely uncliched, but there’s still another useful technique for building compelling adventures using cliched starting material. Research and real world analogue. Not the clumsy sort of decision to repaint a fantasy race with the basic cliches of a human culture (avoid idiocy like “These orcs are vikings, they wear furs and have horned helmets”) which simply substitutes one set of unthinking cultural dross for another, often insulting and offensive, one -- but considering the use of a location and real world analogues.

Now that's a mountain bastion...

What is a dwarven fortress most like? Not a castle or a mine, but instead the bomb proof bunker fortresses of modern warfare - especially those designed to combat land attack. Look at some maps of the 20th century fortifications and begin there, thinking how they might change with the introduction of fireballs and dragons with considerations for gas attack (pits are a common measure against heavier than air gases), vent systems, ambush galleries, remote or concealed watch posts, reinforced areas -- there’s an entire language and art of underground fortification. It’s possible to find maps (though actual underground fortification design rarely offers the complexity that makes for a good dungeon crawl. Likewise the geothermal forge, it will be more interesting if it’s based on something like Iceland’s geothermal power stations, using steam to create power and heat nearby lakes for recreation.

The point is to be able to offer a convincing setting that diverges from the cliche, even the inverted or overindulged cliche, and forces the players to interrogate the cliched content for in-game advantage. If the players have to question how the giant pumps and sulphur bricks are meant to be a defensive measure in the abandoned fortress they’re going to be engaged with puzzling out the material culture and other implications of a people willing to use poisonous gas in underground warfare. If the great hot spring pools filled with dead floating orchards are frozen over the fate of the geothermal dwarf hold’s inhabitants is both clear and more interesting. For the designer new and unique elements are easier to come by, and since they have a clear internal logic they can resist the power of cliche that insists dwarf holds must be a certain way.

Icelandic Geothermal Plant

FINAL THOUGHTS - A COMPARISON WITH LOST MINES OF PHANDELVER
I’ve mentioned “Lost Mines of Phandelver” a few times in this review, the 2014 introductory adventure that “Dragon of Icespire Peak” seems designed to replace and which is set in the same region of the Sword Coast. Many of the problems in “Dragon of Icespire Peak” and Lost Mines are similar and this review isn’t all that different then the one I wrote for Lost Mines back then. The fundamental issue for me with Forgotten Realms content is the glib reliance on bland vernacular fantasy, vanilla fantasy or simplified Tolkien by way of 1980’s D&D’s implied setting. In comparing the two products I want to set aside that problem - it’s almost universal in WotC products and it’s a huge part of the brand identity (though I do think some never products like Descent into Avernus and Tomb of Annihilation push against it gently).

Lost Mines and Icespire Peak share structural problems, a seeming breakdown between the sandbox and player choice focused techniques they espouse in their GM advice and the nature of the adventures they offer. I’d call "Dragon of Icespire Peak" better at living up to its goals, it offers the bones of a point crawl and while it still lacks the mechanical underpinnings of one it doesn't rapidly revert to being a linear forced narrative to the same degree as Lost Mines of Phandelver.  The problems of constantly forcing combat even where it doesn't make sense are worse in Lost Mines as well, at least some of the opponents (I'm thinking of the manticores) will accept parley. I'd argue that as a narrative device the Orcs in Icespire Peak are far worse however, breaking the most interesting and obvious source of tension and roleplaying in the adventure for seemingly no reason.

Fantasy names are less obtuse in "Dragon of Icespire Peak" which is a small thing that pleases me far more then it should and it makes some half-hearted gestures to break away from 5E's cliches. Plus the minimal art is at least designed for the adventure. The dragon in "Dragon of Icespire Peak" receives better treatment then the one in Lost Mines, where it serves as a sort of confusing side trek.  In Icespire Peak, Cyrovain is overshadowed as a threat by the orc marauders in a way that sidelines the dragon and makes the final assault on his lair anticlimactic, but at least the titular dragon is part of the adventure from the beginning rather then an afterthought.

In all I find "Dragon of Icespire Peak" an interesting introduction to 5th Edition, mired in many of the problems I see in other 5th edition products - mostly an unwillingness or inability to conceive that it might require different design principles then the Adventure Path to play an exploration game vs. a tactical combat one.  Perhaps this is an unfamiliarity with the design principles of older editions, but often it comes across as a terror at giving control to the GM and players, a counter productive insistence that the designer should set the course of the adventure and tell its story.

*B1 was the first object I spent my own money on - $3 or some paltry amount off a sale rack at Waldenbooks. I was entirely drawn in by the stilted illustration of adventurers surrounded by a weird array of fungal growths -- still a powerful and enduring image of fantasy adventure -- that B1 largely fails to deliver.