In July 1984, DL2: Dragons of Flame was published. Written by Douglas Niles, it was the second adventure in the Dragonlance Saga, an ambitious fourteen-part series of adventures that told a story about dragons returning to the world.

Dragons of Despair, the first adventure in the series, had been a spectacular success. This wasn’t the case with this adventure. It took a great story idea and showed exactly what problems could be caused by putting it into adventure form. The term “railroading”? It pretty much describes the opening section of this adventure. You could argue that DL1 also had railroading, but it wasn’t executed as poorly as in this adventure.

Of course, in 1984, the typical experience of most players of D&D was with sandbox campaigns or megadungeons: campaigns where player decisions were the drivers of the campaign. The Dragonlance adventures were introducing a new form of telling D&D stories. These days, we call this style of adventures an “adventure path”, but it was brand new back then.

Even so, Dragons of Flame pretty much deserves its poor reputation. It takes a heavy-handed approach to the adventure, forcing a lot of player actions. The first half has almost no player decisions at all. Instead, they’re at the mercy of the story. Get captured by the draconians. Get put on carts. Meet Gilthanis. Get rescued by the elves. Meet elves. Meet Laurana. Laurana captured. Party given mission. Adventure begins… why is it page 19 of 32 already?

The problems caused by the first half of the adventure almost obscure what an interesting scenario the second half presents. The Dragonarmies have taken a number of captives and brought them to the old fortress of Pax Tharkas. The heroes are sent to the keep to free the captives – and to provide a distraction to allow the Qualinesti elves to escape. The general selfishness and unlikeability of the elves is quite a surprise to those who had traditionally seen elves as being good-aligned in D&D. The task of rescuing the prisoners is complicated by the chief villain, Verminaard, being smart: he’s separated the men, women and children, so that the men won’t rebel without knowing their families are alright, and the children are being cared for by a senile red dragon, who thinks they are her own children. It’s an unusual scenario – really, a puzzle scenario – which must be solved by wits and role-playing instead of brute force-of-arms. An optional ending provides closure for if the adventure runs as planned (and the adventurers sneak the families out instead of fighting). Despite Douglas Niles’ name on the cover, this feels very much like something Tracy Hickman has developed.

Actually running this section is quite challenging. When you get down to it, it’s Verminaard and his dragons that are the problem, as there aren’t that many hobgoblins and draconians in the fortress, and those that are present can be isolated and killed – a little more difficult with the pregenerated characters, given that Raistlin is only a 4th-level magic-user at the time. However, fighting a red dragon with a lot of women and children about? Not that simple a task for good-aligned heroes!

The strength of this adventure is in its descriptions: plenty of text describing the fortress, the elves, and the main non-player characters. It’s very easy to properly describe the world to the players as a result.

There’s an additional wrinkle in this tale. You see, Dragonlance was the first time TSR had tried creating a multi-media property. (At that time, that meant adventures, novels and the odd calendar. AD&D computer games were still a few years away). The first Dragonlance novel, Dragons of Autumn Twilight came out in November 1984, only a few months after this adventure debuted – and before most people had run it. And so, a lot of potential players read the novel.

Dragons of Flame is the adventure of the series that is most like the novel. The later novels diverged greatly from the events of the adventures, but that wasn’t the case in the first book. Dragons of Despair is open enough that it doesn’t feel like you’re forced down the book’s path. Thanks to the railroading of the first half of the adventure, it really felt like you were forced to play the novel. And, unfortunately, a traitor gets introduced in Dragons of Flame… and the identity of the traitor is spoiled in the novel. The idea of the traitor is fantastic. Unfortunately, it doesn’t quite work if everyone knows who he is!

There must have been a moment when the designers realised what had happened. This led to two things: first, DL5, which acts as a sourcebook for the first four adventures, offers suggestions to change the identity of the traitor. And second, several of the later adventures gave multiple goals and endings, much like Tracy Hickman had done in Ravenloft, to ensure that players wouldn’t know exactly what was going on. It’s quite possible that Hickman would have done that anyway, but it did stop a repeat of the problem caused by the revelation of the traitor in the novel!

TSR put a lot of effort into these modules. The artwork is Dragons of Flame is very good; the maps likewise. The long poem, “The Canticle of the Dragon” adds two new verses to describe the heroics of the first adventure – something that, alas, never happens again – and a four-part elven hymn gives a good insight into the mindset of the elves.

Ultimately, despite a lot of good material, Dragons of Flame forces the players’ path too much. It’s quite close to being the weakest of the Dragonlance series. However, the next adventure would see the series move into uncharted terrain…

(The D&D 3E revision of this adventure has just been released on DNDclassics.com as part of the adventure Dragons of Autumn)