Realism in Homm:
How far to go?

by Pitsu

During my years on HOMM fan community boards, I have heard many people talk about realism and HOMM. In vast majority of cases, they made me feel that the basic idea is the more realism, the better. Yet I know that reality is not perfect. I do not live in a perfect world and I doubt that copying it would create a perfect gaming system. Of course, a completely unrealistic world would be impossible to understand and would not be enjoyable, so I have asked myself how much realism we can add to a computer game. What is the theoretical limit?  And what is the optimal amount? This article is written as a little philosophical essay about some game features. There is no way I can address everything, so I have picked out some points that I find most interesting or illustrative. This should be taken as my vision only and not as universal truth. I would be happy if you find something that you have not thought about so far, and if my article helps you to evolve your vision of how much realism is needed in HOMM, then it has served its purpose.

What is reality, what is HOMM
To compare and analyse these two subjects, the first and most important step is to define them. Reality is the material world we see around us. And let’s skip the theories of relativity, which claim that a human may be as tall as a redwood. Even phoenixes are not so fast that we need to consider such effects. Defining Heroes of Might and Magic (HOMM) is more difficult. I would do it as following: a turn-based game, where armies of heroes and creature stacks venture into a universe filled with a wide variety of objects. These objects are not only war-related, they are pretty much whatever designers have come up with. And the interactions that armies can have with the objects are also varied.  Note that I say “turn-based game”, not “turn-based strategy game”. As you will soon see, in my opinion the required amount of reality depends a lot whether HOMM aims to be more of a strategy game or a role-playing game. That problem, however, is not the subject of this article.

Adventure map
Does size matter?
What is important on the adventure map? Should it be realistic with trees higher than humans and mountains bigger than genies (Fig. 1)? Let’s look at HOMM4. It is beautiful and much more realistic than predecessors, isn’t it?  Heroes are not as big as mountains or groves. We do not see any more continents from above and adventure objects from side view. The elevation tool allows us to create more sophisticated landscapes with hills and valleys.

In HOMM3, forests and mountains are small -- or at least no bigger than adventure locations and characters. Also, there are only a few possibilities for hiding objects behind others. Thus, everything that player could interact with is underlined by size and the decorations are kept in the background. HOMM4 has some highlighting: external dwelling size depends on creature level and shrines of magic get larger for higher levels, but the main aim of designers has obviously been keeping everything (except mushrooms) natural and realistic. We can say that it is more a virtual reality than a strategic map. Only mountains would have been too big, so their size was reduced to a maximum of twice the height of trees. But dwarves are much shorter than titans or trees (see the picture again). Which style is better? It completely depends what one wants from the game. For strategy, the size highlighting is rather good because the unnecessary objects are in the background and do not disturb the overview. An overview of army placement, strategic locations, and so forth is the best way to see the weaknesses and strengths of your own and your opponent and thereby develop an appropriate strategy. For RPGs, obvious highlighting is not necessary, because it revolves around a single character. (We can discuss somewhere else whether superheroes are a result of role-playing elements or a failure of the strategy engine.) Keeping track of a single army is easier than co-ordinating movements of eight armies and needs only for the surroundings of the single army to be kept “under the player’s keen eye”.

Adventure map views of HOMM3 (top) and HOMM4 (bottom).
Pictures are taken from the editor, not the game.

I cannot resist pointing on an odd side effect of relevant sized landscape objects in HOMM4. Look at a hero travelling in forest. How far does he travel a day? He manages to pass about 25 tightly growing trees a day, a strong hundred meters if transformed to real world! It does not ruin the gameplay, and is hardly ever noticed, but can still be used as an example where HOMM3 is more realistic than HOMM4. Ship size is another thing that goes into deep contrast with overall realism of HOMM4.

Now to the absolute size. Generally, HOMM4 has all objects bigger than HOMM3, making players feel to be closer to the ground and filling the adventure map faster (Fig 1). This is a plus for object recognition – the bigger objects are more difficult to miss or to confuse with similar objects. On the other hand, the larger objects also alter the global feeling and reduce the world size. I have heard voices on Round Table saying that HOMM4 is more a “king of the hill” than “king of the world” game, and I agree with them. Ground elevation supports this feeling, because the curvatures are more like hills than continental plateaus. Terrain elevation has so far had no strategic meaning, like bonuses when attacking downhill or such, so it can be regarded as part of the landscape along with mountains and trees. In HOMM3, one could think that small hills are too small to be seen and that if they are noticeable, they are impassable (Fig. 1), which makes strategic sense.
If the relative sizes were kept realistic, zooming out would result in the loss of smaller objects. You easily see towns and mountains but a sign or well becomes just a pixel on the screen. So if we want to recognise the objects and keep the feeling of big continents, we need to implement a smart zoom function or to make a compromise with realistic sizes. However, in principle, a continental view is not necessary.

Would it be a solution if the map-maker were allowed to define the size of decorative objects? Could RPG maps can have big trees and mountains, while strategic maps use smaller ones?

Dimensions and angles
A hot topic among fans today is whether HOMM5 should be 2D, full 3D, isometric 3D (like HOMM4) or some mixture. For the moment let’s set aside arguments about system requirements. Imagine that we are so far advanced that real-time 3D can handle tons of nano-detailed objects at once on even lowest performing computers. Try to imagine HOMM3 in 3D, while keeping object sizes, colours, details, etc. the same. Is it better than the original? Maybe. Would it change the strategy/RPG balance? Why should it? The additional dimension is like the decorative landscape objects – if too pronounced, it starts to disturb the overview. How pronounced the third dimension should be depends mostly on view angle. For 3D, it is impossible to keep the twisted “land from above, objects from side” style of HOMM3.  We are forced to build a united, real 3D world. From straight above, it would be 2D and if the weather is not cloudy, give an excellent overview. From the side we can distinguish monsters more easily. Here’s a strong plus for 2D – it allows combining different viewing angles, providing simultaneously overview and object recognition. Both of them are important for strategy, but the first is optional for RPG.

A HOMM adventure map is static and pre-defined, unlike the real world, which is always changing. Many other turn-based games--for example Civilization--have much more dynamic interfaces. Cities can be built and destroyed, forests can be chopped down, and so on. Landscape changes could be also implemented in HOMM. At certain turns the trees could change colour or snow could cover the land. Nevertheless, if we keep the turn-based principle, we have to stop somewhere. Neither HOMM nor other turn-based games can become fully realistic. And this is the point where two meaningful events (not including animations) happen simultaneously. If we allow it, we get simultaneous turns, like in Age of Wonders. The game has gotten a touch of real time strategy, and gotten closer to the dynamics of the real world, while still being classified as a turn-based game. Is that wrong? No. Several players can play at the very same moment in real time, but each of them can still control only one object at a time and only for as long as allowed by the turn. While you move your main army, your secondary army must stand still.

In conclusion, keeping the turn-based principle means that one can never make the adventure map as real as the real world with its dynamic changes. But I believe HOMM is as far from dynamic world as it can be. There is a long way to the theoretical limit of possible dynamic actions that can be implemented. The optimum has not yet been reached.
High mountains are impassable for ground troops (There is no hero named Hannibal in the game). But why can’t dragons or air elementals fly over them? The same about forests: why are they impassable? It could take longer time for big army to go through a forest than along a road, but it is possible in real world. To make Homm more real, we could implement different bonuses and penalties to movements in forests, mountains, and caves (with a penalty for flyers). Yes, native terrain bonuses are a step in this direction, but a small step. Now, however, we need a new way of creating impassable areas, or quest guards could become meaningless and storylines difficult to build. Magic barriers that even black dragons cannot ignore? Bring all the restricted areas to another plane, like underground? Could be done, but the game would be very different from what we know now. Making mountains, forests, etc. passable for at least some troops would also allow the map to become more dynamic. Passability -- or more correctly impassability -- of certain objects has so far been the main reason for keeping HOMM maps static. The possibility of adding or removing impassable objects may easily ruin the game. In HOMM4 developers took the risk and allowed mapmakers, but not players, make events that delete certain predefined objects. The feature has proved successful. We shall see whether the developers go further in HOMM5.

There’s not much to say here. Everybody agrees that the minimap should be like an extreme “strategic adventure map” described above: the ultimate overview showing placement of water, continents, and towns (with colours). And I bet everyone also agrees it should stay 2D, not rendered 3D or twisted-time-warped 7D.

General layout
HOMM battles so far have always been in the tactical (or strategic) style, not in the RPG style. The player controls all units perfectly (hopes that reader forgets berserkers) unlike a Diablo 2 necromancer, who can only guide his pets. Therefore, a good overview of unit placement and battlefield layout is essential. Since on the battlefield creatures are larger than on the adventure map, object recognition has so far not been a problem, although it is still very important.
In general, everything said about dimensions and angles of the adventure map also applies to the battle screen. HOMM4 removed the side/top view combo of predecessors and became more realistic. I personally felt that the far horizon (montains/hills/forest) beyond the battlefield on the top of the battle screen, gave more epic feeling for battles, but sacrificed realism. A good introduction in HOMM4 was mixed terrains. In predecessors, the entire battlefield had just one type of terrain, HOMM4 can have half-snow/half-dirt or other combined terrains. This makes them more realistic and adds to the tactical possibilities.

A battle between a hero and genies. H3 at the top, H4 at the bottom.
The battle occurs in front of the behemoth dwelling shown above.

Grid system
The HOMM4 grid system is in fact more realistic than the one of HOMM3, where troop placement was restricted to certain areas determined by large hexagons. The battlefield is not much bigger, but the variation of troop placements is significantly higher. Unfortunately, the HOMM4 system, despite its realism, completely fails to be user friendly – the player has lost a lot of control over troop movement. Sometimes it is even difficult to see whether your troops are close to opponent’s archers or whether the archers can still shoot freely. Probably it was hoped that the new system and higher variation would add strategic depth, and realism was not a goal of its own. Perhaps it would have added depth if more complex controls were provided. We can only speculate how confusing or simple it would have been in that case, but we can conclude that control and overview of unit placement and movement is more important than realism.
Heroes in or out?
Another wedge in the line of fans: Many say that the untouchable hero on the side is just so unrealistic that this is a strong reason to put them on battlefield. Others refer to Edward “Longshanks” in Braveheart movie or to Napoleon and claim that it is realistic. Not everyone’s glory is from hard battles in front lines. Since both possibilities are realistic, the main discussion is mainly about personal preferences and the fun factor. A compromise that allows heroes to be either in or out is most welcome, and brings HOMM closer to the diversity of the real world. Many innovative compromises have come from the dreams of fans, and some of them may be even possible to implement.

As I said in the HOMM definition, the game employs stacks of creatures. It is important, since unlimited stacks give endless replayability. An army consisting only of peasants can be stronger than an army of dragons even though in single-creature combats, a dragon is always stronger than a peasant. Hence, stacks are important for game fun and strategic variability. In battles they appear as single units carrying a sign “it is not only me but xxx times of me”.  In reality it would not be possible--even agent Smith could not copy himself and keep the number of persons to one, while multiplying his strength. And only the top creature takes damage in attack, which is also illogical. Cyclopes do area damage to a stack of 10 titans, and one titan takes all the damage. But the damage could be split equally among stack members, being more realistic. If 200 damage is done for 150 peasants, 100 peasants would take 1 point of damage and 50 top ones would take 2 points. In that case, the whole stack would die together.  To kill one peasant, you would first have to bring all members of the stack down to 1 point. It would make causalities in a victorious army much smaller, and make small armies completely useless for weakening superarmies. Damage distribution among only a few top creatures could be a compromise between realism and traditional HOMM, but again we should ask whether this piece of reality is a worthy addition.

Yet another problem: 4 stacks of 1000 wolves can surround a stack of 10 pikemen, and the pikemen can still escape the circle, just by walking away. That does not happen in reality or a in non-stack oriented game. Moreover, 6 stacks of 1 sprite can completely block the movement of a behemoth, which is also not logical. Thus, the idea of stacks and how they have been implemented are completely illogical. However, very few people even think that eliminating stack would make the more realistic or entertaining. I personally think that stacks make the game “simple to learn and difficult to master” and gives enough strategic variation to justify the loss of realism. It is close to the optimal mixture of simplicity, realism, strategy and fun. More realism could make it more difficult to learn and I am not sure that the fun and strategy factors will not suffer under it.

What about towns? It seems odd when a HOMM3 tower, placed in desert, is covered with snow and nobody is walking on the streets of the world’s most glorious capitol.  HOMM4 showed a reflection of the surrounding terrain, but despite that fans are not happy with towns. The reason is that the dwellings are placed with no meaning. There is plain land and all of a sudden a huge cliff appears. Moreover, the cliff is in different colour than the surrounding. In HOMM1-3 the buildings melted naturally into the background, which gave more natural (realistic) look.

Lets look at the Rampart and imagine how different terrains could be implemented (Fig 3). Snow would be easy – just add a layer that covers ground and trees with snow. However, all buildings could also have a different, snow-covered version. Desert would need new ground and possibly less forest on the background, but might look beautiful. Lava and cursed ground: mainly dead trees and dark ground, and perhaps moonlight instead of sunlight. Note that in many cases there is even no real need to make a completely new set of structures.


HOMM3 Rampart town panorama.
Some dwellings and mage guild not fully upgraded.
The scene shown above is not animated, but in the game the only living beings in the town are the dragons where a decent number of them can be seen flying around the dragon cliff. Didn’t the player have enough crystal to buy them out? Why have they left home if the others have gone to war? It is not necessary that the inhabitants act according to the player’s mouse click or that the number reflects creatures in dwellings, though it would be realistic to have more life in towns (even in necropolis).

Something where realism is important is game sales. Screenshots on the website or the CD box are the first ones that potential new customers look at. The more photogenic the screenshots, the more likely customers are to buy it. Photogenaity and realism are not synonyms, and I think both of them are important to make a game appealing. A photo by an amateur can be very realistic, but makes the viewer think only about a common workday. He has seen the object thousands of times before with his own eyes. A professional can photograph the same object under slightly different angle and the viewer will see the picture as art, as he sees the object in a new light. Not all optical effects caused by the camera have a negative influence on the beauty of the photo. What I want to say is that it is not the realistic look itself that makes people like a game like HOMM. It is the artistic view of reality.  

You want to hear some conclusions? Lazy as I am, I would like to say “do it yourself!” and actually my aim is not to make you believe what I do. I am happy if this little article helps everyone to think more clearly about connections between realism and HOMM.  You are welcome to could draw your own conclusions, but first I will give you mine.
Even theoretically, we cannot make HOMM as dynamic as the real world. Neither can we keep the sizes of adventure and landscape objects in line with common sense -- nor can we add too many decorations -- without disturbing the strategic balance of the game. For RPGs, these limits are less restricted.
But where is the line across which realism would not add more to the game? I personally think that a lot more realism can be added with dynamic changes of adventure maps and towns. Battles and 3D view, on the other hand, do not need additional realism.

There are plenty of illogical things, but many of them may be necessary for the game to play well. It is not realism that makes the game good -- it is something else.

Finally, I would like to thank Charles Watkins for correcting the most odd grammatical constructions that I wrote in the preliminary manuscript.