![]() |
I'm not sure there's anywhere in the game that you encounter a dude of this description. I wonder if it's supposed to be Sheltem. |
Might and Magic III: Isles of Terra
United States
New World (developer and publisher)
Released in 1991 for DOS; 1992 for Amiga, FM Towns, and PC-98; 1993 for Macintosh, SEGA CD, and TurboGrafx CD; 1995 for SNES
Date Started: 27 August 2017
Date Ended: 12 October 2017
Date Ended: 12 October 2017
Total Hours: 68
Difficulty: Easy (2/5)
Final Rating: 52
Ranking at Time of Posting: 248/265 (94%)
Might and Magic III is a good entry in a superior lineage. It moves the game from "Wizardry done better" (which characterized the first two titles) to its own category, with an engine that is still turn-based but seems somehow action-oriented. In doing so, it preserves most of the best parts of the earlier games, including an open world, nonlinear gameplay, a hidden-but-interesting plot, and copious special encounters and side quests. There are many absurd moments in the game, but never a boring moment.
It does, unfortunately, introduce a couple of problems. The 1990s start to present an issue that is going to remain relevant all the way through the modern era: as the level of graphical detail of a game increases, we expect an equal increase in the level of content detail, and it becomes jarring when it doesn't appear.
When I say "content detail," I'm talking about the realism of the world and the way the character interacts with it. Maybe "realism" is the better term overall. I'm making this up as I go along, so forgive me if this theory isn't 100% polished, but I think it generally works, and it explains some of the problems people have with modern AAA titles like Skyrim. When developers were only capable of showing us wireframe graphics, we understood that everything was an abstraction. We didn't complain about lack of realistic layouts or an absence of obvious NPCs because we understood that we were meant to fill in those details with our minds the same way we fill in the very walls of the dungeon. But as graphics improve to show us individual bricks, we start to question the plausibility of the dungeon's very existence, how lighting and sanitation work, and so forth. This is part of the reason that Ultima Underworld will be such a breakthrough in 1992, presenting for the first time a dungeon as a (semi-) realistic ecosystem.
No one makes fun of the food systems of games like Ultima where you have an enormous feedbag with thousands of meals that deplete at a regular basis, but we do make fun of a game like Skyrim, where you can stop combat to ingest 30 cabbages. Once you get to the point that a game can graphically depict individual cabbages, you expect it to treat them like real cabbages. (In fact, let's call this whole thing my "Cabbage Theory.") Abstract hit points? No problem. But you build an engine in which you can graphically make "headshots," to the point where the arrow remains sticking out of the enemy's head? You'd better believe that I expect it to cause more damage than a torso shot. NPCs are little white icons that run around the screen? Sure, I can imagine that they only represent a fraction of who's supposed to live in this town. But when you get to the point where I can see and talk to every NPC in voiced dialogue and they all have individual homes in the city, I'm going to start to question why the capital of the entire country only has 20 people living in it.
![]() |
Walking across an ocean that feels nothing like an ocean towards mountains that feel nothing like mountains. |
Note that we don't seem to care the other way. Roguelikes often offer absurd content detail in the way that individual objects react with living things (and each other) but only the most abstract graphics. Thus, my theory is that you want your game to remain at or above the line in the graph below. If it is, players aren't drawn out of the game by its lack of "realism." If not, the resulting dissonance will probably damage the game for some players and completely break it for others.
Thus we return to Might and Magic III, where the added detail in graphics and sound raise issues that you'd rather not think about. Why, for instance, can I see all the monsters in each city but not the residents who presumably inhabit them? Terra seems hauntingly empty of people, aside from a handful of NPCs you encounter behind desks plus the people who staff the stores. When the statues in Castle Dragontooth talk about armies of thousands clashing in the northern islands, you can't help but laugh. Individual enemies are as tall as mountains; you could fit maybe six of them on the island that the stories say held 6,000.
Look at the world size. You can walk from the top row to the bottom row in 10.5 game hours and in the process pass from icy tundra to parched desert. (And this isn't an artificial world like VARN or CRON.) The 16 x 16 standard for Might and Magic's game maps is far too small for this more detailed world. The developers took pains to try to give each map its own character and backstory, but you really can't make "Serpent's Wood" or "Enchanted Meadow" all that memorable when they consist of only 8 tiles each.
![]() |
"Do you remember the Crystal Mountains of Might and Magic III?" -- No one. |
Because of these issues, it's all the more disappointing that the game doesn't take itself seriously in terms of story and quest. Most of the limited NPC dialogue and quest paths that you receive are nonsensical at best and outright comedic at worst. Take the three kings, each wanting the Ultimate Power Orbs to conquer the others. You could have made a truly compelling plot out of this, with the "good" king envisioning a land of order and harmony, the "evil" king pushing for a world in which individual strength and will are paramount, and the "neutral" king looking for balance. Instead, each one is a ranting caricature of his alignment, and you end up siding with any of them only to move the plot forward, with essentially no consequences.
This lack of seriousness was present in Might and Magic II as well, and I guess it's just something we have to live with from this developer. My more important criticism of Might and Magic III is that the game went backwards in its magic and combat systems. I missed fighting battles against dozens of enemies. I missed carefully plotting battles against tough foes that would go for a dozen rounds or more. I missed launching powerful spells against wave after wave of enemies, and adventures in which I had to cast every last healing spell to keep the party on its feet. In Might and Magic III, the average offensive spell so under-performs physical attacks that you really have no reason to cast them. The developers had to make some enemies immune to physical attacks just to justify having a mage class. Mass damage spells are hardly necessary because the screen won't accommodate more than three enemies at a time--less for larger enemies against whom you'd really need those spells. Most of the spells in the game are the same as in its predecessor, when it perhaps needed a new approach to magic to go with its new engine.
This lack of seriousness was present in Might and Magic II as well, and I guess it's just something we have to live with from this developer. My more important criticism of Might and Magic III is that the game went backwards in its magic and combat systems. I missed fighting battles against dozens of enemies. I missed carefully plotting battles against tough foes that would go for a dozen rounds or more. I missed launching powerful spells against wave after wave of enemies, and adventures in which I had to cast every last healing spell to keep the party on its feet. In Might and Magic III, the average offensive spell so under-performs physical attacks that you really have no reason to cast them. The developers had to make some enemies immune to physical attacks just to justify having a mage class. Mass damage spells are hardly necessary because the screen won't accommodate more than three enemies at a time--less for larger enemies against whom you'd really need those spells. Most of the spells in the game are the same as in its predecessor, when it perhaps needed a new approach to magic to go with its new engine.
I didn't like other aspects of combat, including the fact that all characters are in melee range and that archery--absolutely deadly in the hands of the right character in I and II--is relegated here to a "bonus" you get against approaching foes. After the first quarter of the game, it's mostly a waste of time.
Combat overall is too easy. Most foes are temporary annoyances to be brushed aside rather than true obstacles requiring just the right tactics. All the teleportation spells contribute to this lack of difficulty. If you get into a tough scrape, you just need to drop a "Lloyd's Beacon" and "Town Portal" to safety. The game isn't so big that it can easily justify these spells. Their absence, plus perhaps some nerfing of the fountains, plus perhaps an inability to save in dungeons, would have made more balanced gameplay.
As I often do, I've spent more time complaining about a good game than talking about its strengths. The paradox here is that a game must offer a certain level of complexity before you can complain about it in detail. Of course, the good points outweigh the bad. Might and Magic III kicks things to the next level in graphics, sound, and mechanics; it feels like a true 1990s game instead of a remnant from the 1980s like so many of its contemporaries. It's enormously addictive. I had to force myself to stop and write. The sense of character development is absolutely constant, the interface so intuitive that you could play in your sleep. It's the type of game for which you find yourself saying "just 10 more minutes" over and over again until suddenly it's 04:00 on a weeknight.
Nonetheless, I think the GIMLET is going to disappoint some fans. It will rank high, probably in my top 20, but I suspect it will rank a little lower than its predecessors, which in their more primitive graphics and sound offered better combat and more challenging overall games.
1. Game World. A strength that may seem like a weakness if you're not paying attention. Might and Magic III not only tells its own story but clarifies what was happening in the previous two titles. The manual's backstory and lore are well-written and complement the gameplay well, and I loved the "Corak's notes" feature that offered a little background on every map, outdoor and indoor. The party's cluelessness as to their ultimate goal is part of the charm of the series, so I won't dock any points for that. The game world could have been a little more responsive to the player's actions, but beyond that I don't have a lot of complaints. Score: 6.
2. Character Creation and Development. Someone unversed in Might and Magic could be forgiven for thinking that it draws directly from Dungeons and Dragons. During character creation, after all, you get a list of D&D-style races and classes, as well as a list of suspiciously similar attributes. But in character development, the series offers much more rapid and continual development than the typical RPG. A character might start with a might of 15 and end the game with 75. Nearly every dungeon provides a couple of character levels. The skills, while still binary (except for thievery), offer an additional means of development that most RPGs of the time didn't feature.
In fact, it's a little too much. I don't think I've ever complained about too much character growth before, but Might and Magic III skirts that edge if any game does. As I played, I routinely delayed training (after the first few hours) because it just didn't matter. If the developers had made Level 100 a distant maximum (instead of the actual Level 200) and essentially halved the game's experience point rewards, it would have resulted in better balance. Casses, races, and alignments still don't matter in any role-playing sense. Score: 4.
3. NPC Interaction. As with many first-person titles, what you get in Might and Magic is not so much "NPCs" as "encounters during which someone talks." Most of the NPCs are goofy, one-note characters who offer no role-playing options or dialogue choices. You don't even really learn much about the game world from them, with a couple of exceptions. I do like the NPCs who can join the party, but they're not really necessary and if I played the game again, I'd do without them. Score: 3.
4. Encounter and foes. Might and Magic offers a satisfying bestiary, with associated strengths, weaknesses, special attacks, and special defenses. Its frequent non-combat encounters with quasi-NPCs, statues, fountains, talking heads, and so forth are a highlight of the game. The riddles and puzzles are a little easy, but at least they're not frustrating. I would dock points for offering a "closed" system--you can kill every enemy in the game and have nothing left to fight--except that it ultimately doesn't hurt character development. Score: 5.
1. Game World. A strength that may seem like a weakness if you're not paying attention. Might and Magic III not only tells its own story but clarifies what was happening in the previous two titles. The manual's backstory and lore are well-written and complement the gameplay well, and I loved the "Corak's notes" feature that offered a little background on every map, outdoor and indoor. The party's cluelessness as to their ultimate goal is part of the charm of the series, so I won't dock any points for that. The game world could have been a little more responsive to the player's actions, but beyond that I don't have a lot of complaints. Score: 6.
2. Character Creation and Development. Someone unversed in Might and Magic could be forgiven for thinking that it draws directly from Dungeons and Dragons. During character creation, after all, you get a list of D&D-style races and classes, as well as a list of suspiciously similar attributes. But in character development, the series offers much more rapid and continual development than the typical RPG. A character might start with a might of 15 and end the game with 75. Nearly every dungeon provides a couple of character levels. The skills, while still binary (except for thievery), offer an additional means of development that most RPGs of the time didn't feature.
In fact, it's a little too much. I don't think I've ever complained about too much character growth before, but Might and Magic III skirts that edge if any game does. As I played, I routinely delayed training (after the first few hours) because it just didn't matter. If the developers had made Level 100 a distant maximum (instead of the actual Level 200) and essentially halved the game's experience point rewards, it would have resulted in better balance. Casses, races, and alignments still don't matter in any role-playing sense. Score: 4.
![]() |
My ninja's final character sheet. |
3. NPC Interaction. As with many first-person titles, what you get in Might and Magic is not so much "NPCs" as "encounters during which someone talks." Most of the NPCs are goofy, one-note characters who offer no role-playing options or dialogue choices. You don't even really learn much about the game world from them, with a couple of exceptions. I do like the NPCs who can join the party, but they're not really necessary and if I played the game again, I'd do without them. Score: 3.
4. Encounter and foes. Might and Magic offers a satisfying bestiary, with associated strengths, weaknesses, special attacks, and special defenses. Its frequent non-combat encounters with quasi-NPCs, statues, fountains, talking heads, and so forth are a highlight of the game. The riddles and puzzles are a little easy, but at least they're not frustrating. I would dock points for offering a "closed" system--you can kill every enemy in the game and have nothing left to fight--except that it ultimately doesn't hurt character development. Score: 5.
![]() |
I should have done more with the Arena. |
5. Magic and Combat. Discussed extensively above. A good system, but not a great one. Combat is a little too easy, magic a bit unbalanced. Offensive spells tend to fall into two categories: those so under-powered that it's a waste of time to use them and those that level with the caster and thus cost so much that you can only cast 5 or 6 before having to rest again. At least combat isn't tedious, though: even the toughest battle is over in well under a minute. Score: 4.
6. Equipment. Both a strength and a weakness. I loved all of the different types of equipment and potential slots; every treasure chest seemed to bring an upgrade for one or two characters. The "breakage" system is a little annoying but one of the only consequences to a character getting knocked unconscious. I wasn't as in love with the random generation of item materials and enchantments, and I would have liked to see some unique "artifact" weapons and armor.
Although they exist, I barely explored the use of magic items. You can find items that duplicate almost every spell, along with other items that recharge them, but I mostly just sold them to save inventory space. If combats had been harder, I'd probably have made more use out of them. I also didn't really explore the "Enchant" spell, which adds an effect to unenchanted items, because it doesn't work with the game's better materials. Score: 6.
7. Economy. Useful but badly balanced. After the first few hours, you don't have to worry about money except in a general sense, as you watch training costs grow and wonder where the "tipping point" will be. After your initial purchases, you really don't need money for items. You spend it on spells, training, healing, and training, and it would be nice if the whole system were both more challenging and not closed. I offer a slight bonus for the interest-earning bank accounts and the "money sink" fountain. Score: 5.
8. Quests. The game's primary strength. "Side quests" have still not crept into the average developer's lexicon in 1991, not even with Might and Magic showing how it's done since 1986. Might and Magic III excels at them. Even the main quest has a couple of choices--though more would be welcome--including that final optional area. Score: 6.
6. Equipment. Both a strength and a weakness. I loved all of the different types of equipment and potential slots; every treasure chest seemed to bring an upgrade for one or two characters. The "breakage" system is a little annoying but one of the only consequences to a character getting knocked unconscious. I wasn't as in love with the random generation of item materials and enchantments, and I would have liked to see some unique "artifact" weapons and armor.
![]() |
Things got a little formulaic by the end. |
Although they exist, I barely explored the use of magic items. You can find items that duplicate almost every spell, along with other items that recharge them, but I mostly just sold them to save inventory space. If combats had been harder, I'd probably have made more use out of them. I also didn't really explore the "Enchant" spell, which adds an effect to unenchanted items, because it doesn't work with the game's better materials. Score: 6.
7. Economy. Useful but badly balanced. After the first few hours, you don't have to worry about money except in a general sense, as you watch training costs grow and wonder where the "tipping point" will be. After your initial purchases, you really don't need money for items. You spend it on spells, training, healing, and training, and it would be nice if the whole system were both more challenging and not closed. I offer a slight bonus for the interest-earning bank accounts and the "money sink" fountain. Score: 5.
![]() |
There's not much logic to it, but it ensures that every gold piece is worth something. |
8. Quests. The game's primary strength. "Side quests" have still not crept into the average developer's lexicon in 1991, not even with Might and Magic showing how it's done since 1986. Might and Magic III excels at them. Even the main quest has a couple of choices--though more would be welcome--including that final optional area. Score: 6.
![]() |
My mage's final accomplishments. What are those two blank spots and what could I have improved on? |
9. Graphics, Sound, and Interface. I thought the monster graphics could have used a bit more realism, but they're certainly an upgrade from II. I was less enamored with the overall graphical quality than with the creative use of graphics as part of the interface. Consider how the character portraits change to match their conditions (diseased, curse, drunk, in love), the use of colored gems to depict the character's relative hit point total, the way the gargoyle waves to signal a secret or the bat opens and closes its mouth to show that enemies are near. In melee combat, you can tell how much damage you inflict from the size of the blood splatter that appears when you hit (and none appears at all when you miss). Lots of games have good graphics, but Might and Magic III is one of the few of this era to start incorporating true graphical feedback; to make graphics a key part of the interface rather than just something nice to look at.
Sound is sparse but effective and realistic where used; honestly, no game is going to do great in this category until we start hearing more ambient sounds. The interface is one of the best I've encountered, aside from casting spells, where it would have been nice to have a shortcut or a "favorites" list or something. Score: 7.
10. Gameplay. As you know, I use this category for considerations like linearity, pacing, difficulty, and replayability. It definitely gets points in the first two categories. I'm generally happy if the number of hours doesn't far exceed the final GIMLET, and here Might and Magic III does fairly well. I personally played it longer than was warranted, spending a lot of time experimenting and dithering around; it's easily winnable in 40-50 hours. And even though I didn't do much with it, the nonlinearity was welcome. On the other hand, balance issues made it a tad too easy (as did the ability to save everywhere), and it's hard to think of it as "replayable" except perhaps for a particular challenge. Score: 6.
This gives us a final score of 52, right about where I suspected it would fall. It ends up at the #19 spot and falls below both Might and Magic I (60) and II (58). Those who would give more weight to graphics and who prefer fast action combat to tactical combat will probably invert those scores across the three games. It is the fourth-highest rated game of 1991, and again I don't dispute the order. There are things I like better about the Might and Magic series than the Gold Box series, but when it comes down to the final assessment, I prefer the relatively more serious nature of Pools of Darkness and Death Knights of Krynn, the more tactical combat, and the greater challenge that they offer.
If I could play it again, I'd try something more challenging. Perhaps only four characters, or perhaps a party of nothing but knights, forcing me to make better use of special items with magic effects. I'd like to hear from someone who gave that a shot. I resisted the temptation to try a speedrun, mostly because I saw that it had already been done a few times. One guy did it in about 5 minutes, but he used cheat codes at the teleporter to get an Ultimate Power Orb and a ton of gold early in the game. A more honest one took about half an hour. He did what I would have done: got a little money early in the game, put it in the bank to earn interest for about 10 years, collected it, and donated so much to the fountain in Fountain Head that he was able to take characters to Level 150 all at once. He then bought the necessary teleport and damage spells and a ton of might potions for smashing doors and getting the pyramid key card, visited the central pyramid for the teleportation box, and used it to zip to each dungeon to collect orbs and hologram cards.
I did spend some time trying to raise my high score, fighting about 10 arena battles and donating a few million to the experience fountain before returning to the endgame. I went from 1.1 billion points to 1.2 billion.
Sound is sparse but effective and realistic where used; honestly, no game is going to do great in this category until we start hearing more ambient sounds. The interface is one of the best I've encountered, aside from casting spells, where it would have been nice to have a shortcut or a "favorites" list or something. Score: 7.
![]() |
The well-detailed and animated shop images will continue for the rest of the series. |
10. Gameplay. As you know, I use this category for considerations like linearity, pacing, difficulty, and replayability. It definitely gets points in the first two categories. I'm generally happy if the number of hours doesn't far exceed the final GIMLET, and here Might and Magic III does fairly well. I personally played it longer than was warranted, spending a lot of time experimenting and dithering around; it's easily winnable in 40-50 hours. And even though I didn't do much with it, the nonlinearity was welcome. On the other hand, balance issues made it a tad too easy (as did the ability to save everywhere), and it's hard to think of it as "replayable" except perhaps for a particular challenge. Score: 6.
This gives us a final score of 52, right about where I suspected it would fall. It ends up at the #19 spot and falls below both Might and Magic I (60) and II (58). Those who would give more weight to graphics and who prefer fast action combat to tactical combat will probably invert those scores across the three games. It is the fourth-highest rated game of 1991, and again I don't dispute the order. There are things I like better about the Might and Magic series than the Gold Box series, but when it comes down to the final assessment, I prefer the relatively more serious nature of Pools of Darkness and Death Knights of Krynn, the more tactical combat, and the greater challenge that they offer.
If I could play it again, I'd try something more challenging. Perhaps only four characters, or perhaps a party of nothing but knights, forcing me to make better use of special items with magic effects. I'd like to hear from someone who gave that a shot. I resisted the temptation to try a speedrun, mostly because I saw that it had already been done a few times. One guy did it in about 5 minutes, but he used cheat codes at the teleporter to get an Ultimate Power Orb and a ton of gold early in the game. A more honest one took about half an hour. He did what I would have done: got a little money early in the game, put it in the bank to earn interest for about 10 years, collected it, and donated so much to the fountain in Fountain Head that he was able to take characters to Level 150 all at once. He then bought the necessary teleport and damage spells and a ton of might potions for smashing doors and getting the pyramid key card, visited the central pyramid for the teleportation box, and used it to zip to each dungeon to collect orbs and hologram cards.
I did spend some time trying to raise my high score, fighting about 10 arena battles and donating a few million to the experience fountain before returning to the endgame. I went from 1.1 billion points to 1.2 billion.
![]() |
I'm sure much higher scores are possible. |
Computer Gaming World featured Might and Magic III on the cover of the May 1991 issue, and a review by Johnny L. Wilson is all positive, focusing primarily on graphical details. It was a nominee for "Game of the Year" in the magazine's November 1992 issue, which makes no sense, but lost out to Ultima Underworld, which is hard to dispute. If it had been evaluated in its actual year, it's hard to see how it wouldn't have beaten Elvira: Mistress of the Dark (which itself was being evaluated a year too late, as it's a 1990 game). Dragon gave it 5/5, which for once I understand, but again I have to comment how a magazine dedicated primarily to tabletop role-playing never seems to focus on actual role-playing mechanics in its reviews of computer games. It's always about graphics, sound, music, interface . . . anything but combat rolls and attributes.
Even Amiga magazines rated it well. Percentages range from 81% (Amiga Joker) to 93% (Amiga Action). To the extent that these reviews have complaints, it's primarily about frequency and speed of disk access on the Amiga specifically. I almost always find something that bothers me in an Amiga Action review, but here they were actually quite fair and thorough, calling it the "best role-playing adventure available on the Amiga."
In a 2012 RPG Codex interview, John Van Caneghem recalled that his team went "all out" on III, eager to meet expectations of gamers primed on two excellent predecessors. He notes that it was the "smallest seller" of all the titles, probably because many fans of the series hadn't upgraded to the 1990s platforms, but the best reviewed and highest-awarded.
To the best of my recollection, Might and Magic IV and V uses an update of the same engine, but perhaps with a better story? I honestly don't remember anything about it. It will unfortunately be the end of 1992 before I get to explore the pair, but the prospect of playing them is almost enough to get me through the rest of 1991. Before then, we'll be looking at another New World production: Planet's Edge (1992), which has a completely different interface but shows art director Louis Johnson's influence in the cut scene graphics. The title also shares several of the same programmers; I know virtually nothing about it but look forward to it.
For now, we have 13 more titles to finish in the interminable 1991. I want to do it by this blog's 8th anniversary in February.
The copywriters fell down on this one. Is the title called Power and Magic? I don't think so. |
Even Amiga magazines rated it well. Percentages range from 81% (Amiga Joker) to 93% (Amiga Action). To the extent that these reviews have complaints, it's primarily about frequency and speed of disk access on the Amiga specifically. I almost always find something that bothers me in an Amiga Action review, but here they were actually quite fair and thorough, calling it the "best role-playing adventure available on the Amiga."
In a 2012 RPG Codex interview, John Van Caneghem recalled that his team went "all out" on III, eager to meet expectations of gamers primed on two excellent predecessors. He notes that it was the "smallest seller" of all the titles, probably because many fans of the series hadn't upgraded to the 1990s platforms, but the best reviewed and highest-awarded.
To the best of my recollection, Might and Magic IV and V uses an update of the same engine, but perhaps with a better story? I honestly don't remember anything about it. It will unfortunately be the end of 1992 before I get to explore the pair, but the prospect of playing them is almost enough to get me through the rest of 1991. Before then, we'll be looking at another New World production: Planet's Edge (1992), which has a completely different interface but shows art director Louis Johnson's influence in the cut scene graphics. The title also shares several of the same programmers; I know virtually nothing about it but look forward to it.
For now, we have 13 more titles to finish in the interminable 1991. I want to do it by this blog's 8th anniversary in February.