Game: HoMM3 | Campaign: Long Live the Queen | record: 100% faster with Org-mode | based walkthrough
War is hell
Table of Contents
Intro
Unreliable guides
Records set
Hero development
Declining difficulty
Mighty vs Mage
Morale and Diplomacy
Overview
Campaign
Fluff
Next campaign
Iconic handicap
Gorbag's joke
Intro
Due to chronic fatigue, I developed a personal productivity method using Emacs Org-mode. Then I had a problem: How do I show other people that it works? I needed to transparently accomplish something impossible. Fortunately, Heroes of Might and Magic 3 has an Impossible difficulty setting.
The goal of this series is to introduce strategy gamers to Emacs Org-mode so they can use their obsessiveness to win at health, wealth and worldview. Playing HoMM3 is a good way to learn Org-mode without risking your critical info on an unfamiliar app. Org-mode is a text-based personal info manager.
Below are my notes from beating the first campaign. First I'll show the records I set thanks to Org-mode. Then I'll talk strategy and fluff, hopefully tempting you into playing HoMM3 again. Ultimately though, any game that generates notes can synergize with Org-mode.
Unreliable guides
It often occurs that a community arrives at a suboptimal consensus due to hierarchical incentives. For example, HoMM3's elite multiplayers are the best at battle tactics, and they play aggressive timed games on randomized map templates. Blitzing is the meta strategy.
Even though HoMM3 is a beloved classic with an active multiplayer scene, nobody has written good strategy guides for the campaigns on Impossible. The advice basically amounts to: "Use your elite multiplayer tactical skillz to blitz the AI before it can cheat enough to crush you."
Obviously this leaves normal players struggling. Nobody has provided a conservative, reliable, user-friendly strategy for the campaigns. For example, Major Slack does a great job of achieving this balance. I consider his walkthroughs the best on YouTube.
The problem is even worse for HoMM3's first campaign, "Long Live the Queen". Its difficulty is fixed, and the first scenario is set to Easy. Changing this requires using the campaign editor.
Fortunately I found a couple of elite competitors who had done exactly that, so I was able to compare my performance to theirs.
Records set
I found only two competitors who played this campaign on Impossible: Meridian and MasterKD.
The win conditions in this campaign were not suitable for timed evaluation. The first two scenarios encourage the player to secure dominance and then develop his heroes at leisure before defeating Tan. The last scenario encourages the player to defeat Orange and then build up a large army at leisure to overcome static defenses.
Therefore I evaluated based on time to dominate the AI via town conquest, which worked in all but one game. MasterKD's Griffin Cliff playthrough was ambiguous, but he plays slower than Meridian anyway.
In Homecoming, my win was on day 26, and Meridian's win was on day 33. 7/26 = 27% faster.
In Guardian Angels, I was 150% faster than Meridian
In Griffin Cliff, I was 25% faster than Meridian.
On average, I was about 100% faster. Scenario #2 was an outlier relying on Diplomacy snowball; 26% faster is the usual improvement.
I had no intention of speedrunning, but slow is smooth, smooth is fast. My wins were more robust and tolerant of tactical errors, making the strategy demonstrated user-friendly. For example, I had a larger advantage in kingdom army size on the victory dates. I didn't invade the AI until my advantage was overwhelming.
Overall, my style is inbetween Meridian and MasterKD's. Meridian blitzes the AI. MasterKD methodically clears creep. I blitz the creep, then crush the AI.
There's a lot of unexplored depth to the HoMM3. I'll continue proving my heterodox theories in the next campaign. Developing original concepts is one of Org-mode's strengths.
Hero development
Navigation and Pathfinding are useful in Homecoming but not thereafter.
Prefer Castle heroes since starting towns are all Castle, and starting terrain is grass.
Having Dungeon or Inferno heroes is not worthwhile.
Declining difficulty
Once the enemy AI is defeated, I turn on cheats to ease the tedium of developing my heroes for the next scenario. This contributed to an observable negative difficulty curve. Scenario 1 was the hardest and scenario 3 the easiest.
Each scenario had a strategic quest that granted de-facto victory:
Homecoming had imprisoned Orrin, who was most difficult to take before Nighon.
Guardian Angels had the Angel Wings and the 2 extra pressgangable Archangels.
Griffin Cliff had the Seer's 50 griffins and Conservatory for 4 angels, which was easily completed and practically impossible for Nighon to take.
This was odd, but I guess the goal of the tutorial was to increase complexity gradually. The designers used the difficulty setting to increase difficulty rather than map design.
In all cases, the AI lost sooner than expected, usually before I could implement my full strategy.
I did not have unused operations on Homecoming, because it was difficult enough I had to script it tactically rather than operationally. Starting with only one hero also blurs the lines between tactics and operations. I felt I had blitzed a dangerous foe.
In Guardian Angels, it felt more like pouncing on helpless prey. In Griffin Cliff, it felt no different than clearing creep. That was anticlimactic.
Mighty vs Mage
You may have noticed my heterodox hero development. Might heroes do not get Wisdom or magic schools, focusing on mundane combat. Mages don't get physical combat skills, focusing on magic.
Mighties and Mages are meant to play different strategic roles. Mages cast with small high-level armies, while Mighties conquer with large low-level armies.
This was demonstrated in the first and last scenarios. The middle one was a pure snowball.
In Homecoming, Mighty Christian blitzed with low-level troops while Mages cleared creep at home.
In Griffin Cliff, the Mages used their starting MP to rip the scenario apart. They cleared roadblocks, Diplomacied, and teleported to strategic points. The Mighties scouted and harvested. Orange lost before the Mighties could do much fighting, but Diplomacy Mighties did snowball conquer two of Orange's three towns.
Morale and Diplomacy
Overview
Morale is a central Castle theme. Castle gets high morale via angels and Leadership.
A point of negative morale is much more harmful than a point of positive morale is helpful. This incentivizes adding irregulars to an army until its morale is neutral.
Diplomacy therefore synergizes with Castle's Morale theme, providing many irregulars to absorb surplus morale. (This is very different than Inferno's method of handling irregulars, which is to sacrifice them.)
Most skills are suitable for either Might or Magic heroes. However, some skills are suitable for either support or combat heroes, e.g. Estates. Diplomacy is a combat hero skill. It is wasted on a support hero who will rarely have a large army, and then only late in the game.
This is a controversial point. The multiplayer consensus is to use a single main hero for combat. However, in campaigns I use at least two combat heroes: a Mage and a Mighty.
Campaign
The Long Live the Queen campaign emphasizes morale and Diplomacy:
The Library of Enlightenment in Homecoming requires Advanced Diplomacy to enter.
One should level at least one Diplomacy hero in Homecoming, ideally two: a Mighty and Mage. This is easy after securing the surface.
This allows snowballing in the next scenario, Guardian Angels.
In Griffin Cliff, it allows Castle recruitment in the starting zone, and underdark diplomacy to vanquish Orange.
Fluff
Peacetime militaries generally devolve into ineffective obsolescence. However, a few months in the meatgrinder works wonders — one of which is the transformation of civilized men into ruthless penetrators.
Historically, whores and war go together like brothels and ports. Either soldiers are provided whores or they make them on the spot. See the film The Last Valley for details. Note the priest absolves the girls for joining the new whorehouse.
PIC: queen on throne. CAPTION: Separation of Church and State is like cleavage.
Are you beginning to suspect that war isn't as noble and glorious as you've been told by state-sanctioned media? Quelle surprise.
Today the line between combatant and logistics (REMF) is blurred, but traditionally the latter were called camp followers. These included merchants, wives, whores, and everything inbetween.
Why the high morale for human armies? My theory is that their camp followers offer the most bang for your buck. Yes elves are prettier, but only the officers can afford them.
Humans are unspecialized templates who hybridize easily with other species, with a high libido and a roughly 1:1 sex ratio. There is more than one kind of silver tongue.
Doubtless human empathy towards orc and elf has led to spectacular diplomatic triumphs at court. However, when a pack of centaurs is considering an offer they can't refuse, the sergeant probably just gives them free booze and a tour of the whorehouse, which earns a commission per stallion tamed. That is why the Tavern gives +1 morale.
Next campaign
In the last campaign, I won too quickly, preventing me from fully testing my strategies. The next campaign is Dungeons and Devils. So I'll make it devilishly hard.
Iconic handicap
The first scenario is Devilish Plan. The orthodox strategy is to select the 100 imp starting bonus and blitz Red's main town. Meridian won on m1w3d1, faster than MasterKD's win on m1w4d7. This doesn't help the non-elite players who will have to deal with a strong AI who resists the blitz and survives to month 2. Instead, I'll show how to turtle and counterattack.
Meridian's method was fastest, but it mostly skips the map. It's a boring generic strategy with high risk of failure. For example, it's easy to get lost in the roadless unscouted grassland — which he does and has to reload! A strategy guide should do better. I promise to scout the prairie like Comanche before plunging in like Custer.
Here's my handicap. The orthodox starting bonus is 100 imps. Instead I'll pick the Slayer scroll, because it is iconic for the dragon-centric win condition. The spell helps kill huge beasts such as dragons.
Presumably Slayer is the worst choice. It's easier to just defeat the AI and then build up at leisure. The AI can't even build Dragon Cliffs.
However, the AI does have two Refugee Camps which could randomly offer behemoths, hydras, dragons, etc at any time. This is a major threat to a blitz. Thus turtling is the more reliable strategy, despite the conventional wisdom that one must blitz the AI on Impossible.
In the previous campaign, I picked the consensus bonus because it was also iconic:
14 Pikemen for Homecoming's Halberdier blitz to Orrin
an Angel for Guardian Angels
Lion's Shield of Courage for Griffin Cliff
Devilish Plan sets a precedent: I'll always pick iconic over orthodox. Picking the worst bonus gives the AI a chance. Strategy is written for struggling players, so I must suffer with them.
Below is a bit of hellish fluff about how the Inferno player ended up with the Slayer scroll. Perhaps you've had a boss from hell?
Spawn vs Clown (Violator) | Spawn (Director's Cut) | YouTube
Gorbag's joke
The problem with working for devils is the ubiquitous treachery. Gorbag, your boss, knows you're gunning for his job. That's a given. The prick also has it in for your clan. Infernal politics. That's why your backpack is empty save for this useless scroll, and your retinue consists of nothing but your personal guard.
By all accounts, you're on your way to a freshly-sprouted volcano about to be stomped by AvLee's rangers. The Dragon Queen was tipped off (surprise, surprise) and is buttoned up tighter than a baby's butthole.
Speaking of assholes, you pull out the scroll and examine Gorbag's tracking seal. You can't even sell the "strategically invaluable artifact" to fund your campaign; he'll have you crucified in a heartbeat. If you somehow survive this, you promise to introduce Gorbag to your favorite sexecutioner. Lady Pain may not be a lady, or even conventionally female, but no one is better equipped to give Gorbag a lesson in professional courtesy.