Bytten

Playing turn-based strategy games on the PC is one of my favourite ways to relax. I love Sid Meier's Civilization series, but on the indie front, I'm also an avid player of Sillysoft's Castle Vox. I was looking forward to clearing some free time to review Blood, Grain and Steel for a while. I downloaded and installed the game and was immediately disappointed when I booted the game up and couldn't play because of a graphical glitch. However, after shooting off an email to developer David Walters, and receiving a very speedy reply with a hot-fix, I was up and playing soon enough.

I don't have a lot of space to work with here so I won't go into the back-stories of the two rival forces in the main campaign game. To begin, the player pledges allegiance to either the Kaitan or the Tirus forces, and enters battle for control of the island of Gisaria. The war is fought in both a turn based strategic overview mode, and a turn based tactical battle mode. Regardless of which mode is active, a player may act with just one unit, army or territory per turn, and all turns are of equal length.

The goal of the game is to take the opponents capital city tile. The island of Gisaria is quite large, consisting of 140 territories. Capital cities are placed on the board in what seems to be a random position each time a new campaign is started. Armies are moved around the map, and when a conflict needs to be resolved, a new territory sized tactical battle map opens up. Forces are deployed at each end of the map and the battle is fought. Once the winner has been decided, play resumes on the strategic map once again. Apart from the main objective of taking the opponent's capital, the secondary objective of the players is to take control of as many territories as possible. This facilitates the production of grain; the game's only resource. Each territory produces 1000 grain per turn which accumulates until the stockpile has reached 10,000 grain. Armies can be created by using 10,000 grain. Armies can also transport grain around in seemingly unlimited quantities, and deposit it in friendly territories. Once a threshold has been reached (from memory – 15,000 grain) the territory is fortified and will be harder to take by the enemy.

When a battle is triggered, the defender uses the amount of grain in their territory plus the amount of grain invested in any armies garrisoned there to buy units. Any remaining grain after this is used as supply for the remainder of the battle. Battles are won by destroying the opponents grain supply depot (at least one must be bought and placed), killing all their commanders (again at least one must be bought and placed) or by natural attrition when the opponents supply runs out. Each active unit placed costs supply in grain at the end of each turn when they are in battle. A special siege unit also costs grain to fire.

The units that can be deployed in battle are footsoldiers, cavalry, paladins, infiltrators, catapults and supply depots. Each have differing attributes and each becomes more or less vulnerable depending on the terrain that they occupy as well as the opposition unit faced. The details of each unit are clearly documented in the excellent pdf manual, and you'll probably want to print out a copy of the battle results tables and have them handy for you first few skirmishes.

The most significant feature of the game is that the outcome of battle (including every single combat result) is based entirely on a set of pre-determined rules, and has no random component whatsoever. It makes the game play out much like a game of chess, with each move being deliberate and meaningful. It also exposes the biggest flaw in the game: a terribly bad AI. I don't like picking on AI because I'm sure that it's the most difficult part of the whole game to code. Still the AI in Blood, Grain and Steel refuses to guard its supply depots against attack in the battles, and moves troops around seemingly at random in the strategic game. It does not protect itself from infiltrator attack especially, and to a lesser extent cavalry. It will target the players supply depots on occasion but rarely with more than one unit. It is sadly no match for any player with any degree of experience in turn based strategy games. On my first playthrough, I took the conservative route and won after about 1000 turns. On the second campaign I was more aggressive and won by turn 750. On the third experimental playthrough I managed to defeat the game in 180 turns, and I'm sure I could do it in even less. There are no difficulty settings.

Three multiplayer modes are available. Hotseat, play by email and LAN can be used to hook up a game between two human opponents, though my experience with the only choice for remote games (PBEM) is that both players need discipline and dedication in order to see a game to completion. This is all well and good, but simply finding opponents in these indie games communities can be tricky. It's a real shame that the single player campaign is such a pushover.

The game looks fantastic. The action is presented on neat, functional yet attractive boards. Units are designated by pieces that look like chessmen in a fully 3-D environment that can be zoomed, panned and rotated at will. The sounds (including some voice work) are good, and the music is suitably epic for a game such as this.

Although the player can choose to fight skirmishes on randomly generated boards, or practice their tactics on any board from the campaign in a one-off match, the real attraction of this game for me was the campaign mode. The current level of the AI is simply only good enough for people who have never had experience with turn based strategy games or for players on their first playthrough to come to terms with the rules. Either buy this one for the multiplayer alone (make sure you have an opponent in mind first though – there's no matchmaking service here) or save your cash for a more challenging strategy game. The presentation is top notch, but unfortunately, Blood, Grain and Steel needs an AI overhaul to be a viable single player game.