Behind the Curtain: Mapmaker as Entertainer

by Charles Watkins


Much has already been written about map ratings—what goes into them and what they mean. As a player I rarely enter ratings and as mapmaker I really pay them no attention. This is because map styles vary so much that there is little base for a fair comparison. And players gravitate to the kind of maps they like, so there’s never a representative sample. It’s like asking people to rate Mother’s cooking. It seems to me that the best way to rate a map is according to how much you enjoy it. If you have a good time, it’s a good map. If you have a great time, it’s a great map. What this means is the mapmaker is really performing as an entertainer, and in this series I’m looking at how mapmakers can create maps that entertain their players. What makes a map enjoyable? Last time I started the discussion by saying that the main sources of enjoyment are discovery, success, and aesthetics. I went through several ways to appeal to the player through discovery—foreshadowing, surprises, and so on. Today, the discussion turns to success, the feeling of triumph over worthy challenges and the satisfaction of a game well played.

Part 2: Success

Everyone likes to win, but if victory is too easy, it loses its luster. On the other hand, if it seems too hard to win, players become frustrated and give up. Like so many other aspects of life, getting there is half the fun, so we are looking for ways to make the pursuit of victory as enjoyable as possible.

Since not everyone will win, the mapmaker should include multiple challenges, so that each player will have a chance to savor some measure of success. This means planning out a game in a succession of milestones, so that players get a sense of attainment throughout play. Ideally these build to a grand finale, culminating in some additional reward.

What sort of goals can you set? On the way to the game ending victory condition (complete takeover or something else), players seek to explore the map, build up towns, amass armies, and level up heroes. The means to these usually involves combat, travel, magic, and treasure. These present the mapmaker with enough choices that players should be able to enjoy the way to victory as much as victory itself.

Sweet victory

Everyone likes to win, but some victories seem sweeter than others. Experienced players often complain that the AI is a pushover, so they don’t enjoy playing single player maps.# To some extent, that’s true for just about every computer game this side of Deep Blue. But I believe that the Heroes4 AI can put up enough fight to challenge experienced players—and make average players hesitate before choosing the “Champion” difficulty setting. However, it is up to the mapmaker to arrange for this, and I don’t mean simply handing out extra resources to computer players.

For multiplayer maps, players need to know that they all start off on equal footing and that it was best play that won the victory, not just who got lucky. Most makers of multiplayer maps take pains to see that resources, artifacts, and adventure objects are distributed equally. In fact, I think some go too far when each player is given the exact same set of artifacts. This results in peculiar endings where the winner ends up with four of everything.

Challenging computer players

My first advice to mapmakers who want to have challenging maps with satisfying endings is to forget about having player positions that can be played by either a human or computer. The AI is simply not up to the task of successfully emulating a competent human player. So stop complaining about that and use the AI for what it does well.

Why can’t the AI perform as well as a human? There are too many reasons to go into here, and soon I’ll start another series on getting the most out of the AI, but for now it’s enough to say that it is far easier to treat the computer positions as relatively fixed and waiting for the human players to advance to the point they are ready to take them on. Instead of starting off the computer player with a minimal town, a handful of low level creatures, and level one hero, give it a built up town, a substantial garrison, and several tough heroes. When human players eventually reach it, the computer’s town will look more like a town that a human player would have developed at that point in the game.

A second reason for designing computer-only positions is that they allow a game to reach a satisfying climax and then come to an end. Heroes3 was particularly bad about endings in that when the AI had been thoroughly beaten, it would persist in sending out lame little armies that players had to hunt down and exterminate. The game ended in a whimper instead of a bang and this was something of a letdown to players.

Heroes4 introduced the three day waiting period, which was intended to bring the game to close soon after the last enemy’s towns and heroes had been defeated. This might work for multiplayer games, but I don’t see the use in single player maps. Without a hero how is the computer player supposed to capture a town? (A human might send a creature to Tavern or Prison to recruit one, and if there’s a weak town nearby, possibly retake a town.)


Players like options for how they attain the map’s objectives. If there is just one way to reach a goal, they begin to feel manipulated. Some like to bludgeon their way to victory. Others like to sneak. Still others like to be clever. A good map offers something for everyone—indirect routes to objectives as well as direct ones.

Some story maps make it seem as though “the game is playing you,” as the mapmaker delivers a narrative that directs the player’s actions. This effect can be blatant, as where the mapmaker out and out tells the player to do something, or more subtle as the mapmaker telegraphs the instructions in prose. Personally, I like to have my story unfold in response to actions that the player would logically take in the course of the game. You can be pretty sure that a player will notice a lightly guarded Gold Mine and visit it at some point, so you can also be pretty sure players will move to where you can deliver some text to advance the story. The difference from the narrative approach is that players get a sense of being in control – and are occasionally delighted to learn that their choices have been anticipated and have consequences in the story.


Like many character-based fantasy games, much of the attraction of Heroes stems from the way heroes advance, becoming more powerful and thus able to take on new challenges. To a lesser extent, towns and armies also improve during the course of the game, which is good because, unlike most other games, opposing armies also get better as time passes.

I first experienced the rush of ‘leveling up’ as a player of Chainmail, the precursor to Dungeons and Dragons. Chainmail was designed as a combat system for miniatures and included a system for upgrading armies as a reward for victory. You could carefully nurture a squad from novice to elite status and see them become more and more powerful.

The computer game Rogue was very popular in the early days of Unix, and lives on today in the form of Nethack. It is also apparent that Diablo’s roots go back to Rogue. In all of these, players start out as level one characters with minimal abilities and rise in level as they fight their way down into the dungeon. With each new level come additional hit points, improved abilities, and new powers. Of course this is nothing new to players of Might & Magic and other popular RPGs. But since there is at least one well-known member of the Heroes community who has never played an RPG, I thought some background information might be appropriate.

The thrill of leveling up comes from deep in the psyche. We all have feelings of powerlessness and aspirations to become more capable and important than we are now. Leveling up allows us to satisfy this inner need. Even if we are trapped in dead-end jobs, overburdened by debt, and tied down by past mistakes, we can still experience the sensation of leveling up. And this makes the ‘real world’ a bit more bearable. Playing Heroes is like having a Graduation every day.

So what does this mean for mapmakers?

If players are drawn to the game by the prospect of advancement, the mapmaker can use advancement as an incentive that players can strive for. And advancement also provides the most tangible reward that is possible in a computer game: gold lets you buy more armies and armies let you fight bigger battles, but leveling up gives you new and improved skills, access to better spells, and sometimes even entry to new locations. With more gold or armies, you are still just able to buy things and send more armies into battle. But leveling up lets you do new things, which keeps the game interesting. (Yes, some people like to see nice cutscreens, but I doubt the ones in Heroes are good enough to attract much of an audience by themselves.)

Earning it

One reason HEROES has difficulty settings is to provide a suitable challenge for all skill levels. As players get better, they can ratchet up the difficulty so that the game always provides a good contest, nether too easy or too hard. For the mapmaker, this means there is a balance to be struck between allowing players to advance too quickly and holding them back too long. As a player, I don’t much care for maps that are so resource-poor that it takes many weeks to build up heroes, armies, or towns. At the same time, I don’t want everything handed to me on a silver platter.

There are quite a few maps where the player starts out with multiple towns, high level heroes, and formidable armies. Though these can be enjoyable, they also make me feel somehow cheated. I prefer to start out from scratch, not only for the fun of developing something out of nothing, but also because I like to make my own decisions.

I have my favorite advanced classes (like Cardinal) and I can’t use them at first when the mapmaker starts me off with a level 5 Ranger. One reason mapmakers do this, I imagine, is to encourage the player to use the starting hero, who is likely intended to be the protagonist in whatever story is being told. I know that when I’m faced with a “Lose Hero” loss condition, I’m inclined to park that hero in my starting town and do my adventuring with new heroes that are not bound by the condition. If the starting hero already has some levels, I’m less likely to waste them by shifting to another main hero.

Of course, there are other ways to avoid this besides awarding extra levels at the start—easiest is withholding the Tavern for a while—but storytellers should realize that Heroes players do not normally identify with a single hero. Because most stories revolve around a single protagonist, mapmakers may find it difficult to tell them in the context of Heroes.

In maps I’ve made, I’ve tried several ways to deal with this problem. One approach (Agent of Heaven) was to deny the player access to a Tavern until the intended main hero had time to develop and the player had become attached to her. Another time (Grandmaster), I made the protagonist the mayor of a town who hired mercenary heroes and sent them out on missions. In another map (Kid Heretic) I reduced the pool of available heroes to a collection of characters who were part of the story. In each case I was able to give the player the fun of seeing main characters advance within the context of both the game and the story.

Players will vary, but I’d say the right pace of development would be for heroes to level up about once a week. Likewise, I’d like to see towns add new creature dwellings once a week on average—figure on levels 1 and 2 in the first week, level 3 in second, and level 4 in the fourth. At the end of the first month, the player is probably ready to develop a second town.

Character progression

This same principle of scaled advancement applies to power-ups. Players are more gratified by working their heroes up the hard way, through combat and treasure hunting, rather than coming across Altars, Learning Stones, and so on. This is not to say that maps should not include these power-ups, but that they should be carefully placed and well enough guarded that players get the sense that heroes are advancing mainly through their own efforts.

There is some scalability built into the Heroes level progression in that more and more EPs are needed to reach higher levels. However, the progression is somewhat flatter than in some RPGs, where EP requirements double for the next level. It seems to me that a reasonable balance is to allow the player to gain one level by power-up for each level gained the hard way. If the main hero gains a level the hard way about every week, then after six months, the hero will have about 24 levels. Add to that another 24 from power-ups and the hero is on pace to max out in about a year. In the sixth month, a hero with 48 levels will be near the top rank for both sides of an advanced class. Smaller maps may need to advance heroes faster, but the same principle applies.


So far the discussion of advancement has focused on the heroes. Now let’s look at advancement of the player’s empire, where many of the same principles apply.

As with heroes, there is a balance between making expansion too easy and too hard. A mapmaker can start players with a minimal starting town—a ‘fixer upper’—and still strike the balance by providing the right amount of resources for development. This can be very tricky when you take into account that starting resources vary with game difficulty and the needed resources vary with the town type.

Personally, I don’t care for maps using timed events to take away or add resources. Even if they are cleverly written, these events are completely beyond my control and I start to get the feeling that the game is playing me. I especially don’t like it when I choose to play at a lower level (I usually play on Expert) only to have the mapmaker take away all the resources on day 1. If it is necessary to take the resources, I’d prefer to be involved, perhaps by paying a toll to get where I’m going. At least then, I feel somewhat in control. The design principle here is that players like to think their own decisions are what determine how their heroes and towns advance.

Building Armies

In addition to building up towns and leveling up heroes, players also seem to enjoy amassing formidable armies. I think Heroes is missing something with its relatively limited town development. Towns often top out in less than a month and after that the Grail is the only chance for further development. So the mapmaker may as well concentrate on the heroes and the armies.

Maybe not all mapmakers realize it, but it is possible to arrive at quite good estimates of player strength as the game progresses. If you assume that players will follow a typical build progression—levels 1 and 2 in week 1, level 3 in week 2, and level 4 in week 4—you can add up the numbers of armies available and get close enough for design purposes.

When I plan a map, I think in terms of how many weeks I expect it will take for players to reach various objectives and set opposing monster levels accordingly. This works for players that tend to use one large army. To allow for smaller armies or slower players, I try to give players a 3:2 advantage. This gives the maxed out army a chance to get through a good fight unscathed, while still holding out a reasonable chance for smaller forces. For town defense, I give computer players some additional help to compensate for the weakness of the AI there, reversing the balance to 2:3. Incidentally, the same approach works with hero development, since EPs awarded are proportional to hit points defeated.

Pumping up Armies

An interesting part of the game comes when players begin to strip the lower level creatures out of their armies to replace them with top troops from other towns. I’m wild about shooters, so I like to put together a powershooter force with Titans, Monks, Cyclops, etc. Others may enjoy playing an all-Dragon combination or something even more exotic. In any case, by setting up the map to make such combinations available, the mapmaker can offer players a novel experience at about the point that accumulating more-of-the-same is becoming less fun.

A Sorceress has assembled an army of Fire-based creatures. Such theme-based armies can fit into a storyline and provide some variety for players used to one-town armies.

Scaling up Treasure

Another way to heighten players’ enjoyment of advancement is to award progressively better treasures as the game goes on. Heroes4 includes some potentially unbalancing artifacts in the Minor category (e.g. Crown of Enchantment) and some real spoilers in the Major category (e.g. the staves and tomes). The Relic category, especially with the WOW expansion, contains some real doozies (like Archmage’s Hat). Tossing these out as random treasures not only risks giving players unintended advantages, it also dulls their appreciation of quest objects and other prizes. Mapmakers should consider how such artifacts fit into the theme of the map, and whether their value is proportional to the effort required to obtain them.

I’d say Heroes4 has done a good job of calibrating the rewards for capturing a treasure house to the difficulty required to defeat the guards. By the time I’m able to defeat the Black Dragons in the Dragon cave, the major goodies awarded are not unbalancing. Still, at the end of most games, I end up with an incredible number of unused magic items. After accumulating five or six magic swords, finding another one is not such a thrill. This tells me that maps are often too generous with lower level items.

It would be better for the mapmaker to consider what main items players should have at various points in the game and distribute them accordingly. And as I said before, there’s really no point in laying out treasure houses for computer players. They will not use the items they find there and so these end up in the hands of players. Better to outfit computer heroes with whatever items they need to put up a good fight, and maybe use a Defeated script to take away items not intended for players. Better still to minimize the number of powerful items to begin with and rely on spells and skills to create strong opponents.

Clever scripters can set up completion scripts that vary the value of rewards according to variables related to the player’s current level, days elapsed, or even army composition. These allow treasures to scale up as the player advances.

I should also put in a reminder about Prisons. While freeing a level 1 Priest in the first week would be a prize for most players, getting that same hero in Month 6 will probably be a disappointment, especially after a fight to reach the Prison. Ideally, a skilled Prisoner can be the ultimate reward – for unknown reasons, Heroes jailers allow prisoners to retain their artifacts and even their armies while doing time

Difficulty settings

Part of the art of giving players a rewarding sense of accomplishment at victory is setting the difficulty level appropriately, not just at the end but throughout the course of the game. Players may choose to play on the easier difficultly settings or on the higher ones. There is frequently confusion between the difficult setting of the map, the difficulty of play selected by players—and the actual difficulty of a particular map resulting from how it is designed.

There are some Heroes4 maps that try to make the player’s selection determine more of the course of play, but I haven’t gotten around to replaying them on different settings, so I haven’t had a chance to see the difference.

I tend to take the player’s difficulty selection more as an indication of the level of their play—new players tend to be unsure about their skills and choose easier settings, while jaded masters just sneer and go right to “Champion”. (Personally, I prefer to play on “Expert” because I like larger maps and when I crank up to “Champion” I find that later in a long game the level 4 neutrals are just too tough for me.)

The resource differential is felt mainly in the first couple of weeks of the game, so some care should be taken to see that players on all levels have a fair chance of getting off to a good start. There is enough variation in the power of level 3 monsters that random level 3’s should not be used as guards for critical resources. For instance if Life runs into Cyclops guarding the source of Crystal, the town’s development can be stymied for some time, while if the guards are simple walkers like Mummies, they will be much easier for shooter-heavy Life to deal with.

In ‘Conquest of Four Lakes’ we see a Monster-3 guard for a Pillar of Eyes, a level 2 dwelling that has its own guards. Clearly, this is overkill. The mapmaker must have been thinking about Trolls rather than Genies or Goblin Knights.


A good reason to avoid making a map too hard at the start is cheating. When players get the idea that the game is too hard, they start going for the Load Game option, hoping that they may be able to succeed in a replay. Not only does such an experience fail to add anything to the enjoyment of the game, it cheapens the sense of accomplishment when players finally manage to succeed. On a poor designed map—for instance one infested with unexpected ambushes—players begin to feel entitled to cheat, and once they start, the behavior may continue throughout the game.

I once made a map* that had designated “save” positions where players were given permission to use save/load to avoid restarts. The idea was that once they had shown they could accomplish some major objective, it was pointless to make them do it all over again to make up for a mistake later on. Since permission was explicitly given at the “save” points, players may be less likely to cheat later on. And when they eventually reach victory, their elation is not offset by guilt.

In conclusion, the fun of the game is not so much in winning as in the way you win. The most entertaining maps sustain players’ sense of accomplishment throughout the game by offering a succession of intermediate goals leading up to a big finish.

Want to talk about this some more? Meet me in the Mapmaker Forum!

* On the other hand, I’ve also heard complaints about the AI cheating, though I have never seen it myself.

Next time: Aesthetics...