- What are the drawbacks of Heroes 3? (What didn't you like about it?)

Ave, Heroes! I stumbled upon this topic and decided to write a "longpost" dedicated to the shortcomings of Heroes 3, which—alas—in my view, are quite numerous. Let me say right away that Heroes 3 is a good game; otherwise, it wouldn't have hooked me in my same teenage years when I was wiping out hundreds of dragons and so on "without losses" on King.) Later, after watching streams and studying the technical side, I realized I played poorly, but that's another story. The main thing is that I am writing this with some PvP experience (non-tournament), so I know what I want from Multiplayer.
I will write point by point; if I compare, it will be with the fifth part, as it is my favorite and, in my opinion, the most successful in the series (I think one of the reasons for its creation was, besides the development of the series, the shortcomings of the third part). I am speaking primarily about "pure" 3, say, Shades of Darkness (in the HotA mod the balance is tweaked, but that's a mod, it has its own nuances).

1. General Balance.
I think one can notice a trend—the more rules and restrictions there are, the worse the balance. The balance of Starcraft, if I'm not mistaken, was polished for years to achieve a state where any opponent's strategy could be answered with a counter-strategy. The balance of the third part is well known to all—there are seas of restrictions, up to banning playing as Conflux and Necropolis, or using control magic (TP, Fly, and DD). In the fifth part, not a single castle is banned; Necromancers also have an advantage in troops, but it is offset by the hero's weak attack and lack of morale (there is a good chance against them for both mages and warriors).

2. Map Balance.
In Heroes 3, external dwellings (so-called "external dwelling") are of huge importance, and they are scattered unevenly—one person gets a centaur dwelling, while their opponent gets 2 manticore "external dwelling", and even with a summoning portal. It's clear how this affects gameplay. Plus—a level 2 dwelling doesn't always appear in the Castle. Plus—hives and conservatories are scattered unevenly (there is even a topic on this forum suggesting banning visits to these objects). In the fifth, respawn is uniform, i.e., either level 1, 2, 3 dwellings are available to everyone—or to no one. Objects that give "free" troops do not exist at all. Accordingly, map balance is much better.

3. Dependence of hero movement range on the speed of the slowest unit.
One might think—what's wrong with that? IMHO, extra micromanagement—the hassle of "speeding up" heroes, which beginners are taught from "birth"—is simply unnecessary. Why can't I take a hundred zombies or dwarves in my army? I doubt the developers intentionally introduced this parameter into the game; rather, it seems to me that no one suspected people would start using this in an attempt to win an extra cell of movement. However, I am writing this as IMHO, since I'm not sure—in principle, it can be viewed as an element of strategy.

4. Standard Heroes.
There are 128 heroes in the game (144 with Conflux). Of these, in the overwhelming majority of cases—excluding diplomats, sometimes logisticians, Solmyr-Jeddit-Solmyr, etc.—Tazar, Crag Hack, Mephala, Neela, Gundula turn out to be the main ones; that is, specialists whose bonus actually affects the course of the final battle (and thus, the entire game). The defenders' bonus is indeed too strong—it's no joke to have an additional 15 percent defense in melee and against arrows at level 20! That is, there are few truly playable heroes. For example, in the fifth there are only 64 heroes; of them, 6-8 are considered "OP", plus say 15-20 "weak", while no less than 40 are quite playable—5 heroes per faction. However, there are always people playing other heroes—I just wanted to say that the problem of their "inequality" is more acute in the third than in the fifth.

5. General Similarity of Castles.
Despite the presence of 2 classes—mages and warriors, 8 heroes each—and city specialization (Tower—magic-shooters, Stronghold—attack-melee), the specialization is quite conventional, because after level 10 parameters grow roughly equally; as a result, we have similar heroes (30-30-20-20 and vice versa). There is no racial ability except necromancy, so you can successfully appoint some Tazar to command your army of Knight, or even some Orrin to lead a demonic army (which is completely wild from a roleplay perspective). In the fifth, for example, each faction corresponds to a certain direction—besides the racial ability, there are also stats (Necromancers get spellcasting and defense, but their attack is weak, while Knights get defense-attack and have problems with magic), so confrontations became more vivid; you feel the difference between factions.

6. Stupidity of Neutral AI.
I am focusing specifically on neutrals, as in the fifth, the computer AI is also no match. There are many videos here and on YouTube showing how Dragon Utopia, crypts, and cons are taken with a minimal army using the "leading" method for neutrals. Essentially, this is a banal exploit, not brilliant tactical skill—it's enough to model and remember the computer's behavior. In the fifth, a neutral won't be distracted by a single centaur; he will go straight for the archers, making battles more intense. There is no such term as "eating" (attrition)—battles in Dragon Utopia happen "as usual", troops on the left, dragons on the right. You also need to cover your troops and remove retaliation, but not simply "stand and endure" by distracting dragons with a low-value stack. Because all stacks are native, all are valuable (I'll dedicate the next point to this). And note that battles are also tactically difficult, and Dragon Utopia are taken by masters on the 2nd or even 1st week with minimal forces.

7. Value of Troops.
It has already been written here about the decreasing role of level 1 in battles, when troglodytes, for example, are replaced by angels. To which the writer received a caustic reply—saying that the inability to use high levels speaks to the player's skill, or something similar. Indeed, battles in Heroes 3 are battles of heroes. Battles in the fifth are battles of factions (races, castles). There are three reasons for this. First, there are fewer troops in external dwellings and no "free" wyverns and angels. Second, level one troops became more valuable. With battle frenzy (an ability that increases a unit's min-max damage by 1), my imps can hit harder than archdevils. In the third, first levels eventually "sag" in damage and survivability (exceptions like skeletons are rare) and are replaced by troops of other factions (I often saw screenshots of Crag or Tazar with gold dragons-gorgons-archangels-black ones etc.). Third, every unit plays its own important role—zombies are a good "anchor", imps are damage dealers, etc.; having 50 zombies is more profitable than 3 dragons. This is also subjective—personally, I prefer the struggle of factions over a Rampart hero with Dungeon and Inferno troops running to a Fortress to pick up gorgons.

8. Skill System.
In Heroes 3 there are 27 skills, not counting necromancy. Of these, two-thirds are actually useful. One could certainly name a map where mysticism or own sight would be indispensable, but in the overwhelming majority of cases, there is a certain template set—offense, defense, logistics, earth magic, wisdom. And that's already 5 slots. If you add a second magic (often air), and variable skills like leadership, pathfinding, or resistance, for example—only two will remain. I think this is a weak point of the third part. In Heroes 5 there are only 12 development branches, but not one is useless (for example, education, it's training, giving an additional stat every 2 levels at expert), and the choice of branches often makes you think—what's better: blessing the army en masse, or reducing flying enemies' speed by 2 and initiative by 20 percent? There are many alternatives.

9. Magic System.
In Heroes 3 there are 70 spells, in the fifth—48. However, even on this forum there are quite a few topics where people ask—in what cases is misfortune useful, for example? A good example and answer—none, because in the original, damage under misfortune is not halved. However, joy-sorrow, quicksand, etc., also have a limited area of application. Moreover, schools are not equal—a vivid example is the magic popularity poll, where earth magic leads by a huge margin (I think for many people the priority is earth-air, sometimes earth-water). It got to the point that in some mods the town portal was moved to the school of fire for balance. In the fifth, schools are logically divided "by effect"—buffs (Light), debuffs (Dark), direct damage (Chaos), summoning creatures and other things (Summoning). Battles became more interesting—duels between knight-buffers and mage-summoners, debuffers and casters, etc., became the norm. Moreover, there are no longer, or almost no longer, useless spells (perhaps 1-2 claim to be—like Sorrow or Magic Fist). Plus, heroes in the third cast during a creature's turn—there is still a difference in "first move" depending on the castle and an excessive weight of slow-haste spells. The introduction of the "initiative" parameter fixed this situation.

10. Absence of the "initiative" parameter or dependence of creature movement on speed.
There are two points here—first, with initiative and speed being indivisible, slow creatures are the biggest "bottlenecks" (like dendroids). Second, mostly, though not always, the 7th level has the highest speed—a unit with huge attack and damage. A situation where an archangel or gold dragon moves and kills a stack in one hit is, in my opinion, not as good as if, say, pegasi moved first, then centaurs, then dwarves, and only then the 7th level monster-damage dealers. In the fifth it's implemented IMHO better; the same ents (dendroids) have low initiative 7, but speed 6, meaning with tactics and a speed aura (speed + 1) they will arrive immediately as soon as they move. Moreover, small levels move first, then "heavyweights"; there is time for the 7th level to cover itself or be neutralized.

That's probably not all, I forgot much, but considering humanity's tradition of thinking "decimally", that should suffice.)

Summarizing, I repeat—the third part is a good game; perhaps in strategic terms it surpasses the fifth part, but the tactical side belongs to the fifth—a flexible skill system, the introduction of initiative, interesting and logical magic, caster units, buffing-debuffing troops—all this makes battle more complex and exciting.

The strategic level, however, can also be disputed—the reduction in the number of important objects (1 Dragon Utopia - maximum 2), gold, and resources on the map make the task of map control extremely important. On the other hand, it has now truly become possible to play with 2-3 heroes; "chains" are used, but not as they were before. I think everything is relative.

The post turned out to be a kind of "praise" for the fifth part unintentionally; the third is a good part, comparison with the fifth suggests itself, the above—my personal opinion. I didn't want to offend anyone.

Separately, I want to note that fans of Heroes 5 generally love and respect the Third, because a significant part (if not most) grew up on this game. At the same time, fans of Heroes 3 often speak disparagingly of the fifth part, sometimes even resorting to the "wrong graphics" argument (!). Indeed, "baby bottle syndrome" and a certain conservatism are inherent in people; I also had to relearn, and the fifth initially repelled me with its unusual brightness. However, it's a boomerang—now many who are used to 3D find 2D graphics outdated.

I consider the question of graphics in a good game to be the last thing. I consider Heroes 5 to be the last successful part of "Heroes".