Where would you find Polks, Smerdys, and Svoi pogyanes? The answer is in an Early Russian army, according to the WRG's DBM Book 3 Army Lists. The book 3 lists have recently been released, and initial impressions are favourable.

The Book 3 armies cover the period from 476 AD to 1071 AD, although as in previous army list books, some armies extend into later periods. There are 79 armies in this book. As in Book I and 2 Europe, the Near and Middle East are well represented. This book extends the Far Eastern armies to further the coverage of Central Asian steppe armies, Chinese armies and Korean armies, as well as including two earlier Japanese armies (Pre-Samurai Japanese and Early Samurai), and several South East Asian armies (Burmese, Khmer and Cham, Medieval Vietnamese). The New World is also represented in this book with Mayan, "Dog People" and Pueblo Cultures, and Toltec armies.

While I am by no stretch of the imagination an expert on the Dark and Early Middle Ages, I do have a reasonable general knowledge of the period, and I have spoken to wargaming colleagues who do have detailed knowledge of some of the areas covered by the list. Generally the Book 3 lists are favourably received. They are a vast improvement over the old WRG 6th edition lists, not only in accuracy, but also in breadth of coverage. The Chinese armies in this book, and indeed in the previous two books, have vastly improved, are more accurate and have been more finely divided into subperiods. For example, the T'ang period has now been split into two separate lists, one list representing the early period, and the other the later period.

There are, however, a few quibbles. The Byzantine armies, while much better represented in the new lists, seem to require a separate type of basing for each list, making it difficult to use the same figures across different periods. For example, the Thematic Byzantine cavalry (now termed Tagmatic kavallarioi) need to be based on double depth bases. The earlier (Maurikian) and later army (Nikephorian) kavallarioi are single depth based, with a recommendation for the Nikephorian that the kavallarioi are based 2 lancers and I archer figure on a base (I assume this particular recommendation is not compulsory, although the double depth base is). The skoutatoi evolve from Blades(l) through Pike(l) to Bow(X) plus Bow(O) on double bases over the same three lists. While these evolutions help to differentiate the different periods, and lend it a unique flavour, I am tempted to speculate that the changes might have been more theoretical than actual. It might have been better to provide a alternative categorisation for some of the troop types.

The WRG world has changed significantly since I started ancient wargaming using Sth edition. Greek hoplites are now irregular, while Avar cavalry are now classed as regular. I can live with irregular hoplites, but if Avar cavalry are regular, then why are Sassanid Persian cavalry in book 2 not regular as well? Again I would have liked to see an option here to class irregular Avar cavalry as regulars, rather than make them compulsory regulars.

The Vikings have changed, for the better in my view. They lose their ability to rush through woods with impunity, surrendering their LHI/LMI status for Irregular blades. It does mean that players will have to rebase at least some their Viking army since some elements are classed as Blades(O). Most figures can be retained on existing LMI bases as Blades(F) if the player wishes, representing the earlier Viking armies, but they tend to be classed as Blades(l) or (0) in the later period. The beserks are still there, but only up to two elements are allowed. The good news is that beserks do not need rebasing!

Many armies are new and there is certainly plenty of variety to choose firom. As in the previous two books, numerous existing lists have been split up from one all encompassing list into two or more lists. The Japanese and T'ang have already been mentioned. Other examples include a Tribal (ie Early) Mongol, Early Slav, and several Frankish lists. The lists seem well researched and reflect recent discussions and findings. Although Phil Barker and Richard Bodley Scott "compiled" the lists, they received material and opinions from the likes of Duncan Head and Chris Peers, amongst others.

For the player that likes to take the side of the underdog, there are armies such as the Emishi (aboriginal Japanese) or the aforementioned Dog Peoples - armies that appear to be basically irregular Bow(I) or irregular Psiloi(O).

The composition of an army list is subjective to an extent: what should the minimum number of infantry be in a T'ang or Viking army; how many kavallarioi should a Byzantine army have? It is no easy task to get the balance right. The DBM books on the whole achieve this balance satisfactorily. The composition of most armies in DBM book I and 2 feels much better than the 6th edition lists. Most armies have a quarter to half of their points taken up with compulsory troop types, and the quantity of ancillary troops is reduced. Book 3 continues this trend, forcing players to take a reasonable quantity of troops types characteristic of the army being used. The disappearance of super troop types - the double or triple armed troops characteristic of earlier WRG rules, and the restoration of a balance between the weapons/troop characteristics also helps, making the majority of armies viable on the table. The key factor in choice of armies, especially for competition, will probably not be weapon type (the old pike, pilum or wedge cavalry favourites), but control characteristics. Impetuous armies may well prove too hard to control.

The new lists, probably as much if not more that the rules themselves, will help to establish DBM as one of the mainstream rule sets. They now cover the ancient and medieval period far better than any previous set of lists. It looks as if rebasing of some elements may be unavoidable for certain armies, but if you like DBM then you will love the new lists. Seventh and Sixth Edition players now have the choice of using existing lists that are demonstrably out of date, rewriting their lists, or reverse engineering the DBM lists.