This is the story of my attempt to beat the game on Insane. I haven't seen any posts like this recently, so I hope people find it interesting.

I love strategy games and I've been playing them for about twenty years. I really enjoy Fallen Enchantress/Legendary Heroes. I've been steadily increasing the difficulty level since I started playing, although the AI has got a lot tougher since the 1.2 and 1.3 patches. I've won the game using most of the standard factions, and I was consistently beating the AI on Ridiculous even with 1.3 AI, as long as I didn't get an awful start. So I decided now was the time to take the Insane plunge. I was expecting it to be hard, but I thought if I chose one of my favourite factions, it should be achievable. I underestimated how hard it would be!

Beating Insane was seriously hard work. It took three attempts, but I got there in the end. This is my account of how it went.

The rules:

1: No custom sovereign; default settings, four Insane AIs, Insane world, all victory methods enabled.

That's just the way I play. Other settings will make other strategies viable.

With the benefit of hindsight, I now think Ridiculous monsters (rather than Insane) might have been harder. As someone has noted in a different thread, Insane doubles starting monster levels, Ridiculous adds four to them. This means that monsters up to level four are actually easier on Insane than on Ridiculous. As it's those low level monsters who give you experience and loot at the all important start of the game, the lower level basic monsters on Insane probably made it easier; by the time you're taking on monsters > level 5 you're on the way to winning anyway. But I didn't realise this when I started my challenge, so I make no apologies for "only" having monsters on Insane.

2: No cheesy exploits.

No queueing pioneers. No gaining unlimited mana from Cull the Weak on Skeleton Horde. No reloading just because I'd left my capital unprotected for five turns and I realised that I needed to get my main army back one turn earlier. On the other hand very powerful tactics that no-one has banned yet, like Maul on prone units, or Titan's Breath in general- if they ain't banned it, it's legal. On Insane, the AI doesn't play fair, so if I can see a legal tactic which is always successful, I will use it.

3: I reload sometimes. I'm not ashamed of that.

On one particular occasion, I reloaded the same battle about twenty times before I won. This was in the interests of experimentation as much as anything; in a really tough battle, what is the best strategy? Clearly it was a battle that could be won, because in the end I did win it; what was interesting was to see what order I should do things in to win that battle.

If anyone's curious, I did not reload twenty times in the game I won. I probably reloaded two or three times after I made some stupid mistake which potentially threw away fifteen hours of gameplay. You don't really get second chances on Insane.

4:  I reloaded any start that didn't have at least one Essence.

My strategy revolves getting a powerful Mage sovereign who can win battles as quickly as possible. Having zero mana at the start of an Insane game? I'm not that much of a masochist.

So, those were the rules. How did it go?

Attempt 1

I played as Resoln. This was the first ever faction I played, and I still have a soft spot for them. Sure they can't have metal armour, but I generally think the game is won or lost by the time you can think about getting metal armour. High dodge troops (Wraith + monk robes) can be great, especially before the AI gets researching crazy Accuracy.

What went well: free troops from shards is good. I think I had a Death shard from the start; Wither causing loss of 3 attack without any chance of resisting is devastating against Darklings. I had a lot of Darklings around my starting position. The strategic mana cost of Wither (without any tactical casting reductions) is significant, but the fact that it can't be resisted and can be cast outside of your borders makes it very powerful. I basically used Wither single-handedly to expand my empire. Eventually I met the AIs though, who already had a massive lead in terms of Research and expansion. The game revolved around a struggle over a couple of very good city sites. I settled an Essence 3 site with a Fire Shard attached, which made a big difference to my mana economy. I settled another good site, but ended up defending against Magnar. I was not really ready for a war. My best stack was Strong, Magnar turned up with a Deadly stack and another Strong stack. After reloading many times I finally worked out a way to kill Magnar's Deadly stack. Blizzard followed by Raise Skeleton Horde followed by casting Curse on Magnar (he didn't have great resistance) followed by casting Infection on Magnar followed by casting every other negative spell I had on Magnar and any other summons I had was what finally did the trick.

At this point I had built up a stack which I thought could beat any AI stack. The trouble was that I couldn't necessarily beat multiple AI stacks. So I had to fight a scorched earth campaign, razing my cities so the AI would have to build pioneers to get those sites. This worked up to a point, but eventually one of the AIs started casting the Spell of Making. I had found the Master Quest site by this point, and I think I might have been able to win by Master Quest given enough time. But the AI completed the Spell of Making a long time before I was ready, so I had to concede defeat.

This might have been winnable if I'd played it better, especially if I could have completed the Master Quest. I made a few decisions I'm not sure about, e.g. making my capital a Conclave rather than Fortress. The thought was I needed more mana, but I'm not sure if it made sense in a 1 Essence city. However if I was playing it again now I might have gone all-out to get Blizzard scrolls. Playing as Resoln the way to win is supposed to be Corrupting Shards and then casting Wither; I never had enough mana to Corrupt Shards, and I never had that many shards. Spiders are cool but I never had enough spare mana. Money was always low, however Blizzard scrolls are relatively cheap, so maybe that would have been a way to win. One thing about this game was that it made me appreciate Wither, it made me appreciate Infection, and I actually found a use for a Familiar; having a Familiar meant I could cast strategic spells such as Freeze and Wither twice as often, even if the Familiar did nothing except sit in my capital out of harm's way.

Attempt 2

I played as Magnar. This was interesting; I still think in many ways Magnar are overpowered by virtue of the easy mana they get (either from Consume Shard or from Cull the Weak, even if you don't use Cull the Weak on Raise Skeleton Horde), however I really struggled on Insane. The problem I had was that even though I potentially had lots of mana, I didn't have a way to beat scary AI stacks. To beat Insane AI stacks the Oracle has Blizzard and Infection; Magnar has... what? Fireball ain't good enough. Slaves are cheap but they aren't good quality troops, they tend to die (don't gain levels) and they just aren't good enough for taking on Insane AI stacks. As Resoln it might have taken me twenty attempts but I was eventually able to work out a way to beat one Deadly AI stack; I then used a scorched earth policy and Razed my own city rather than let it fall to the remaining Strong AI stack. As Magnar I was arguably slightly unlucky in that the first AI to declare war on me turned up with four Strong stacks, but after a few attempts where I couldn't even get close to beating one of those Strong stacks, I decided to call it a day and try again.

Attempt 3

I played as Umber. Ah yes, Kulan, I'd forgotten how much I liked you. Beastlord = free and powerful army from start to end of game. Air and Death are probably my two favourite schools of magic. Titan's Breath with -1 turn casting time will win you the game. It won my Insane challenge virtually single-handed. On top of that is how all Kulan's magic works together; to get -1 casting time you need to get lots of spell mastery, which is going to increase your chance of casting Titan's Breath and Mass Curse. My entire strategy revolved around getting a decent Beast stack, then heading for Titan's Breath. The fourth level of Air spells requires level seven, so if I was too low in level I took a step towards Savant (-1 casting time) on the mage path. Every step towards Savant gave me more Spell Mastery, meaning Curse and Blind were more effective and Titan's Breath would be more effective when I could cast it, so every level was a big win, there were no wasted traits. Some of the most obvious Beasts to Tame as a Beastlord are Bears; Titan's Breath knocks units prone; Kulan has Air Magic and Beastlord; Mauling a prone unit is a certain kill. Did I mention how complementary all Kulan's traits are? I personally think it's overpowered how Maul on a prone unit is a certain kill, however I can see arguments why it should be a certain kill, so I did it, even if I think it's over the top how a Bear can kill a prone 300 hit point unit, 2 hit points at a time. The main problem is tedium, even at 5x tactical combat.

At the point where I could cast Titan's Breath and Cloudwalk (with Cloudwalk you never need more than one stack again as long as you have at least 80 mana), and had a Cave Bear, a Silt Kath and maybe an Umberdroth or two, I thought I should win. However mana was still quite tight, sometimes I didn't have the 80 mana I needed to cast Cloudwalk; three more levels until I could get to the fifth level in Death Magic and could cast Sacrifice, which solved that problem.

Sacrifice is a great spell. Population are meaningless in the game unless you're about go up a population level. Sacrifice can be cast as many times in a turn as you like, and will turn half a city's population into mana. If you choose a level three or four city that has got +1 growth which you decide is taking too long to go the next level, in fact you don't care if it never goes to the next level... potentially that's several hundred mana you can spend right now. Playing on very high difficulty levels the AI frequently gets cities to level five. When you capture those cities that's hundreds of mana to cast right here and now, which is plenty to cast Titan's Breath, Mass Curse, Haste, until you win your battles.

I didn't actually have a great start for Beastlord. One late Cavebear, no Great Wolf. A few Spiders. A Silt Skath. I did get an Umberdroth, which with my maxed out Spell Mastery I was able to Tame. I was able to get another Umberdroth from a quest; ok, I'm not complaining. But there were not actually that many monster Beasts around my start. I only had one city till season 60-70. To repeat, I only had one city till > season 60. When I first met my first AI I was about power 14, they were about power 60, mainly because I'd built myself a Beast stack. I didn't build my second city till about season 70. After getting the best stack I thought I would get, I attacked the AI city with my Beast stack and Titan's Breath, and annihilated it. At that time I was power 33 and they were power 254.

Having got to Titan's Breath, Cloud Walk and Sacrifice, it was pretty easy. A high level Commander in my main stack to give high initiative, Command and Battle Cry didn't hurt, however the win was built round Beastlord, Titan's Breath, Cloud Walk, Sacrifice, in that order. I did however have one scare where the Insane AI which had started in the far corner of the map started the Spell of Making. What I should have done is completed the Master Quest in double quick time. In fact I've proved by reloading the save game that I can win this game before the AI casts the Spell of Making. However at the time I wanted to beat the entire map, and I avoided completing the Master Quest. This found a possible bug in that I continued playing past the point where the AI should have completed the Spell of Making, bearing in mind it was in the opposite corner of the map and I couldn't attack it fast enough. Possibly another AI attacked it, but it seemed like a bug to me that it failed to win when it apparently stopped casting the Spell of Making when it should have won, only to start again a couple of turns later.

Anyway, I eventually beat the map. I won using Master Quest because I thought it might be more fun. (I could easily have Conquered the last few AI cities in another few turns if I'd wanted to.)

As in previous games, the Dragon Eyes quest was a lot harder than the Master Quest. In the Dragon Eyes quest I lost half my army until I could cast Shadowbolt and Kill. If I hadn't successfully cast Kill on a dragon the first time I cast it at 34% failure I would probably have lost.

The Master Quest on the other hand... there's only one thing to Kill. Provided I have enough units to take the damage I will cast Shadowbolt till I can cast Kill with 90%+ chance of success. Provided I have the army units and mana I will win. Doing it with Master Quest asap (as I should have done) gave me a victory at season 135 with a score of 27,585. Completing the map, even though the AI on the other side of the map was casting the Spell of Making, gave me a victory twenty hours of gameplay later at season 169 with a score of 34,210. However I don't actually understand why the AI on the other side of the map failed to cast the Spell of Making. The Master Quest at season 135 was the way I "should have" won if the AI had been playing correctly. If both victories had been disabled then a Conquest victory would have been fairly straightforward.

Conclusions

Beastlord is great. Titan's Breath with -1 casting time is great. Cloud Walk and Sacrifice are great. If you have these things, you too can beat Insane.

Factors which make you more likely to beat Insane: other than Beastlord, Air magic, Death magic, I would imagine increasing the number of shards and increasing the amount of time before you meet an Insane AI (i.e. fewer AIs, bigger map) would help. On Insane, you really don't want to meet the AI until you've had a chance to build your Deadly/Epic stack; rushing the AI is not normally an option. Custom sovereigns obviously help. Based on my experience, having Spell of Making enabled favours the Insane AI, whereas having the Master Quest enabled favours the player.

Hope that's helpful. Have fun. I would say I didn't particularly enjoy beating Insane difficulty, because I had to find a way of playing that the AI couldn't cope with, and then exploit it ruthlessly. It was an interesting experience, but I may revert to completing all factions on Ridiculous and maybe playing with some mods. On Ridiculous you have a bit more room to experiment.