Playing Battle for Wesnoth

I’ve been playing Battle for Wesnoth lately. It’s a light hex-based fantasy war game that mixes influences from western war games and Japanese skirmish war games. Wesnoth is free software, and really quite impressive for that – it has nice graphics and sounds, plenty of cheesy fantasy dialogue and everything else you’d expect of a game in this genre.

For all its good sides, though, I’m quite dissatisfied by this game overall. The project workflow obviously has worked well, and the production quality is good, but the game’s fundamental design tenets seem faulty to me. Let’s see if I can figure out why.

Hexagonal War Game Positioning Example #1The Red and Blue armies are arrayed into battlelines. The numbers indicate the number of vulnerable hex sides each unit in the battleline presents towards the enemy. Given identical units and terrain, the optimal solution is a line perpendicular to one of the three hex axes, as then each unit is equally vulnerable to enemy force concentration; however, units and terrain are often not equal, and enemy positioning might encourage pushing certain units more or less forward as well. Usually your goal is to arrange your units (and enemy units) on your turn in such a manner as to leave the enemy no likely openings for fruitful attack on their turn.
Hexagonal War Game Positioning Example #2An example of a retreat movement on enemy turn. The blue unit attacks the red, forcing a retreat. This might or might not be what the red player hoped for, but now the unit that was attacked is in a more secure position, while the two units partially behind it have a new exposed flank.

Hex-based positioning

The base activity in Wesnoth and other games of its ilk is positioning your various troops against the terrain and enemy forces in such a manner as to

  1. Drop as many enemies as you can before your turn ends; every enemy killed now won’t be dealing damage on their next turn, so this is the most direct method of achieving step two.
  2. Minimize damage and risk of dying for your own troops on the enemy’s turn by controlling who the enemy can attack and with which units.
  3. Usually as distant third, you’re trying to fulfill scenario goals within a time limit. A more difficult game or scenario gives you less time, forcing you to sacrifice the above points to make haste.

I’m not a big war-gamer, but I do like the basic precept when I don’t need to study WWII era artillery history to play the game. This lack of interest in gun porn is the basic reason for why I consider Fantasy General the best game in the seminal General series of hex-based strategy games. This is relevant here, because the General games are essentially very similar to Wesnoth in terms of basic activity – you do that same optimization of troop positioning against terrain and enemy locations under a turn limit.

Now, I like the General series of games a lot. One of my favourite features in them is a small rules detail: troops under attack can retreat from under fire automatically now and then, the probability depending on the unit type and the success of the attack. This rule runs counter to the general notion that units only move on your own turn. It is sometimes beneficial to set up a retreat, as that allows a unit damaged in defence to move into a less vulnerable position. Other times the retreat becomes a hindrance, as units that do not have a venue of retreat available will disband, instead. This means that a good defensive arrangement actually is not just a big blob of units; you need to leave holes for units to retreat to, while also thinking of your defensive lines in depth – which unit ends up under attack when and if this unit retreats? Where does this unit end up if it retreats, will it be able to defend itself on the new terrain from further attacks?

This might sound picky, but much of the time when playing Wesnoth I find myself nostalgic for Fantasy General (both being fantasy games that don’t require me to distinguish between different calibers of Russian artillery). Wesnoth has almost all the rules of unit movement from the General series, such as unit zones of control, but it does not have the retreat rule. I thought at first that I could overlook that, but the more I play, the more I’m annoyed by the lack of this small rule. Lacking retreat means that battlelines in Wesnoth are completely static on enemy turns; your forces are completely vulnerable to enemy attack, leaving you much less opportunity for using multiple units in a synergistic manner. You can still benefit from having many units by forming battle-lines where units protect each other from excess concentrations of enemy force, but in my experience this tends to not be enough; forces are fragile and casualties are taken despite the best plans. Setting aside the fact that play is annoyingly simple-minded when there are no retreats to worry about, one might think that casualties are just something I need to play around and live with. However, I have a second issue with Wesnoth, one that plays into this.

Incremental Advancement

Wesnoth can be played as single scenarios against other people, but that’s not the way I swing; I play the campaign mode, in which you develop an army through many fights, managing resources on the way. There are many things I like about Wesnoth’s campaign mode; for instance, the game engine allows for delicious rules-breaking on the part of the scenario designer, so we get stuff like sea monsters suddenly attacking both sides of a fight, turning plans awry. I also like how the game’s internal reference only allows you to view unit descriptions for units that you’ve seen in the campaign; this means that you get to discover new units as you play, which is exciting.

However, the campaign mode has a dark side, one that it shares with Fantasy General. Your valuable, well-loved army improves by fighting and gaining experience, which allows the units to develop into more powerful version of themselves. All well and good and slightly addictive, except that now that I’ve played the game for several days, I’m growing sick and tired of having to shepherd the troops to prevent them from dying. Psychologically I’m completely stuck on preserving my experienced units, which means that the slightest mistake with the unforgiving, static unit positioning scheme leads to me reloading the game (which seems to take over a minute with this computer) and replaying, trying to figure out a different set of moves that’d save the valuable unit I’ve painstakingly developed from dying in the hands of capricious chance and stupid AI that seems to only care about causing casualties instead of protecting its own assets.

The real question here is, why am I so absorbed in saving individual units that I actually reload turns to find an optimal play? Specifically, why am I doing this when it slows the game down into a crawl, and ultimately ends with me quitting in frustration? My current theory is that this is because the campaign model in this genre of war game is flawed: the only kind of genuine advancement you can have as the player is development in your army of playing pieces, while the thrust of the game itself is nothing of the sort. Consequently my goal in a given scenario is not really to win it, but to win it without losing any of the elite units I’ve myself decided I wouldn’t lose. This is frustrating because the game does not care about my internal state and about which units I want or don’t want to sacrifice. Play degenerates into masochistic optimization when inevitable casualties trigger reloads. Meanwhile the campaign drags on as a linear narrative that I’m not really invested in, being that I’m lost in this army management reward loop. From my viewpoint as a player the only achievement there is is to improve my army; the scenarios themselves come and go, and what you do in them doesn’t matter that much as long as you manage to finish successfully.


I have nothing bad to say about Wesnoth in general terms; the game looks good, the price point is excellent (the game being free), and the Finnish-language translation is just zany, considering that usually only children’s games get translated around here. Despite this I don’t think that I can stand playing the game a lot more – it’s just too frustrating to try to improve and preserve a growing number of units in an army that is doomed to die because there are too few tools available for protecting and preserving units. Something in the game is working at cross-purposes.

Thinking of similar situations in other games, I notice that players do not get stuck in frustrating reload loops when playing games that feature only simple pass/fail conditions. In other words: losing is fine, but unofficially losing because you managed to win the scenario but failed to preserve your real measure of progress, your elite army, is annoying. I don’t think that the game would be feasible if losing any units at all were an official loss condition, but at least then the designer would have to account for this psychological drive he creates by making army improvement such a central part of the game.

Thinking further, how would I approach this genre, then? I like these hex-based games, and I certainly like the idea of arranging game content into campaigns – if for no other reason, then because that allows me to not choose the scenario I play next myself. It also helps that in a campaign game I grow familiar with my army, and therefore don’t need to spend two hours at the beginning of every scenario familiarizing myself with the 50 units at my disposal. So campaigns are good for these games in this regard.

One interesting approach here would be to take a page from the roguelike playbook – that’s another genre I’m very fond of, and I never get this sort of frustration in roguelikes. One basic design precept of these games is that there is no saving the game aside from time-transfer of play; if you fail, you begin again. I love this strict design sensibility, it enforces good design habits much better than almost any other conceit you’ll encounter in video games. One of those habits is the fact that in almost any roguelike you’ll care to name the player is never disadvantaged by surviving: you’re almost always better off continuing to play after the most awful set-backs. That’s probably a principle I’d pick for this sort of hex-based wargame campaign mode as well: you don’t want to make the game less fun for the player when he’s doing badly, which is exactly what happens when the player loses his fancy special units and has to regrow them from scratch. It’s a death by thousand cuts, something you never see in a roguelike: in those games the play-state only ever grows more interesting as play progresses, whether you do well or not.

As for how to get that sort of behavior within the framework of a hex-based war game campaign… I’d probably start by having dead units trigger interesting extra stuff. Survivors from the lost battles come back as scarred commanders, to pick a simple example; this way you actually benefit in a backwards manner from bloody losses, which is cool. I’d probably also try for a more ambitious campaign model than these games usually have; I don’t care for the plot that much, I’d rather have macro-level strategic choices than a plot railroad. Or, if plot railroad is mandatory, perhaps it could become more interesting when you lose, rather than just becoming impossible to finish. All sorts of options when you realize where the problem lies.


10 Responses to “Playing Battle for Wesnoth”

  1. Nick Novitski Says:

    All well said. I think there’s a place in the “market” for turn-based strategy in a randomly generated fantasy world full of rapid and interesting choices which cannot be undone.

    I suppose I’m going to have to try out Fantasy General sometime.

    I believe that Wesnoth suffers most from having a nearly impenetrable map/campaign editor. A good viewing and testing environment for custom maps allows rapid prototyping of divergent playstyles and genres. Then the successful ones are suddenly seen as the fully featured games they are, and in the case of open source development, can be quickly ported out to a standalone application, or rolled back into the base platform. One of the best things about Nethack is the number of times its development has been branched into new projects. If Wesnoth ever reaches the point where there are a half-dozen “Wesnoth-likes” kicking around, it’ll have truly arrived.

  2. Ben Lehman Says:

    With me, and Wesnoth, I got to the point where the game started punishing me for *not* shepherding my experienced units (basically by presenting me with challenges that only a “leveled up” unit could deal with) and then I quit.

    Your critique of the movement system is pretty good, I think.


  3. Eero Tuovinen Says:

    Interesting blog, Nick; I don’t think I’ve seen it before.

    I recommend Fantasy General, it’s one of my long-time favourites. It’s “computer game high fantasy” from before the Japanese flavour came and conquered; I quite like the sketched artworks and quirky setting that’s obviously been written from the viewpoint of translating WWII game mechanics into a fantasy game. The game is quite difficult, though; I tend to play it every few years, but have yet to finish it. (Probably something to do with me playing it on the highest difficulty level.)

    Concerning the Wesnoth modding community, it seems to be quite fertile. I like the in-game mod download facility, but found that it didn’t quite entice me to download anything, at least not yet: while the system lists various mods that are available, it doesn’t tell me anything about their contents, so I don’t know if I’m downloading new campaigns, scenarios, or what. I have no reason to expect that additional content that is “more of the same” is going to change my judgment on the game, though.

  4. Nick Novitski Says:

    You’re not missing much. It still doesn’t have its Defense of the Ancients or Tower Defense.

  5. Vince Says:

    I bogged down on Wesnoth as well. I enjoyed the LIGHTness aspect of the wargame (having long-ago had a similar reaction as yours to serious wargaming).

    Decent game though – even if you never finish it, worth some time.

  6. Kaj Sotala Says:

    I’ve had a bit of the same experience with Wesnoth: it’s one of the most polished and completed open-source games I’ve seen and technically good, and I’d want to like it, but the gameplay seems too simple-minded. I don’t really get the feeling that there would be multiple viable tactics from which I could choose. Or even that there’d be only one tactic, but optimizing that one in painstaking detail would be fun. It feels like there’s just one viable tactic, and it’s not even all that fun to pursue. Of course, I’ve had this experience before with other games and later on realized I was just too blind to see the alternative tactics, so perhaps more vistas would open themselves to me if I studied the game more. Your description makes it sound like I won’t, though.

    As for the issue with safeguarding your units: the first solution that comes to mind would be to treat killed units as simply temporarily knocked out. Once you finished the scenario they would recover, though they obviously couldn’t gather experience while knocked out, so you’d still have an incentive to keep them alive for as long as possible. Perhaps you could add an interesting game mechanic if the surviving units had to somehow safeguard the knocked-out remains of the defeated units to the end if you wished to recover them, possibly earning experience for doing so. (Though you’d want to avoid setting up a situation where you’d exploit the mechanic by purposely killing part of your units in order to get safeguard XP for the rest.)

    It might also work to give you some number of points for buying new units and purchasing extra XP between scenarios. A dead unit could be repurchased at a discount – or, if being brought back from the dead gave powerful “scarred veteran” bonuses, at an extra cost.

  7. Eero Tuovinen Says:

    Yeah, my brother suggested that one as well – simply making death of units non-permanent is a fine way of confessing (as designer) that the game is really about the character/unit development and not goal completition per se; most computer rpgs work that way, for example.

    I’d like to be more ambitious if I were making this kind of game, though – I like my games unforgiving like salted liquorice, bitter and logical. It seems too easy to ignore the fact that you’re basically just doling out a string of scenarios connected with plot as a distraction, while the real game is just a grind for experience points. Many extremely popular games work in this way, but I don’t like that – in fact, I think I’ll write a new post to explain what I mean in more length.

  8. Games and Toys and Defeat Horizons « Game Design is about Structure Says:

    […] me of follow-up comments via email. « Playing Battle for Wesnoth Blog at • Theme: Garland by Steven Wittens and Stefan […]

  9. Toni Ruottu Says:

    I’ve been playing Wesnoth for quite some time. I started by playing “Heir to Throne”, which I still regard as one of the most successful campaigns. It is quite linear indeed.

    In the old days there was no autosave, and the players were given advice to never fight chances with reloading the game. Later, by popular demand the autosave feature was added, probably because of the misaligned goals as you present them.

    I’d have to try the retreat tactics to know, how they feel to me. However in general it seem interesting. I guess the easiest way to try out something like that initially without adding ways of setting retreat paths, would be adding a “yellow” treat to some unit. Such a unit would then automatically try escaping to the closest non-hostile hex once it looses some specific amount of hp.

    This probably wouldn’t help too much in laying out tactics the way you described, but would provide current players a way of getting used to the idea, as well as the developers a way of exercising with implementation of such a feature.

    I guess single player games in Wesnoth are currently just a neat way of learning the tactics for online games. When you play against other people, there is no reload. The people who reloaded in single-player because their strategy failed, learned more than the folks who reloaded because chances failed them. Thus too tight reloading would in the end work against you.

  10. Nick Novitski Says:

    I guess single player games in Wesnoth are currently just a neat way of learning the tactics for online games.

    There’s an amusing site,, whose reviews often mention a idealistic design principle which can be re-stated: “your action game should be interesting enough to control that it could be a fighting game instead.” That is, don’t depend on the level accoutrement, the context of your little avatar’s actions, to inspire fun; those very actions (running, jumping, backflips, shooting) should be fun, every time you do them, instantly and forever.

    So perhaps the equally iconoclastic bizarro-twin of that claim would be: “your multiplayer strategy game should be interesting enough to interact with that it could be a singleplayer strategy game instead.” In other words, don’t depend on the human context to make the game fun…but I have no idea where to make those kind of changes.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

%d bloggers like this: