Banishment rules change needed?

[Seravy]

See http://www.realmsbeyond.net/forums/showt...#pid635178 , I believe the current "banished wizard" mechanics don't work.

The problem is, the penalty for being banished is extremely harsh (not outright losing the game but close to it - no combat spellcasting, and no income!), while protecting the capital is often impossible - you can launch multiple attacks a turn while the other player has only a single stack to fight off your potentially 20+ attacks. As long as you can kill at least one defending unit, you can kill all of them, you just need more forces. This means that if all players have equal (or even slightly lower) tier units, whoever can execute such an attack wins. And that's always the human player, the AI can't plan for such a thing. Which means the player has to be stuck on lower tier units and that's neither fun nor good for game balance.
I believe the AI has the best possible garrisons they can possibly have now and it definitely isn't enough. So they inevitably will be banished. On the other hand the AI can't use the same tactic effectively so the human player won't.

This is not good. The only way to fix this is if we change the penalty for being banished to something the AI can deal with. Something that still provides a strong enough advantage to make the capital important, but not enough to outright decide a game. This thread is meant to discuss this.

Some ideas :

-Banished wizards can still cast spells in combat, but the distance modifier is 5x for them anywhere. (Limbo is further than Myrror)
-Banished wizards still gain power from all sources, but they can't produce RP or SP. All their power is automatically spent on producing MP.
-If needed to balance out other changes, the cost of Spell of Return could be higher - lower penalty for being banished but for a longer time. This is also more AI friendly as the human is way better at timing all their attacks to happen in the same 2-3 turns. (the banished wizard still can't cast other overland spells - they are casting the Spell of Return, which is a fairly important penalty by itself)
-If it is sounds too drastic, it can be optional (though it should default to being on)

[zitro1987]

Moved my idea, which is more based on difficulty in taking an intact AI's capital, rather than penalties of a lost capital.

I hate the idea of #1 (combat spells) - it's not very intuitive.

One suggestion of mine to slightly alter these strategies was not taken, but I don't mind throwing it out there again in a modified format and in a different context:

Complex and Most Logical - Fortress Defense alteration - A formula of weaker later-game bolts 15-35 dmg without armor piercing (or 10-25 with armor piercing) depending on "Power" and higher skill results in more lightning bolts a turn, therefore being much more powerful late-game (200 skill -> 2 bolts, 400 skill ->3 bolts cap)
-Now this idea (especially using non-armor piercing) encourages reduction of power as the bolts are weaker but in higher counts later in game. If late-game wizard heavily weakened, the bolts, even if up to 3, are not very threatening to advanced units.
-A very advanced and intact wizard with dozens of cities would have 3 powerful bolts, dissuading attempts to take a capital from an intact opponent. Someone may prefer to steal nodes and take non-capital cities.

Simple - Fortress Defense Alteration: A formula from 15-55 dmg without armor piercing (or 10-45 with armor piercing), depending on caster's "power" output instead of skill.
-It still encourages strategies of weakening an empire before the capital, but not as much as 'complex' as bolts are based on a single powerful strike
-A very powerful wizard, with such a high bolt power cap, would have very threatening bolts and may dissuade attempts to take a capital.

[Seravy]

Stronger bolts just mean two things
-Magic Immunity becomes even stronger
-If immunity is not available, then using many weak stacks is encouraged instead of one strong stack. So, my 9x9 spearmen will all get slaughtered by the lighting, while I doom bolt/flame strike/wave of despair etc every enemy? That...doesn't matter. People hasn't been using this sort of tactic too often because it's not comfortable and not very fun...but it does work. If all my units die no matter what I use, might as well use the weakest.

(and this would make Life (healing) and Nature (regeneration) capitals unbeatable while every other one will be still easy.)

[Nelphine]

Given how much testing I've done with strategic combat, and given my own love of overland spells, I've put a lot of time into games where my goal has always been to win battles despite the enemy combat spells. One side effect of this is that until recently, banished wizards still 'used spells' in strategic combat.

For me, overland spells, especially curses, are the most dangerous thing an AI has. Yes allowing conbat spells will be a nice quality of life change for the AI. But without overland spells, they stop being a threat. It may take longer to beat them once banished than it does now, but without overland spells, they effectively freeze in place. Their next capital won't have nearly as good defenses, they won't spell blast your crusade/Armageddon/create artifact, they won't warp node you. Unless other AI distract you, or unless the AI is far far larger than you after the banish, banishing will still mean the AI is done.

I'm not trying to stop the thoughts here in this thread, I'm simply trying to bring to light other problems, so that the suggested solutions won't be seen as things that will fix all the problems.

[Nelphine]

IF we want the game to be about slowly clearing the empire down and only taking the fortress at the end, I would suggest taking zitros idea, and expanding on it.

A ridiculously extreme version could be something like: for every point of power granted by nodes, the wizard gets that much extra casting skill in combat at its fortress, and the first man's spent up to that amount is free. Similarly, every population the wizard controls, will absorb one point of damage, or add one to a resist check until the resist check is a pass.

So a wizard with 3 nodes generating a total of 60 power, would get 60 extra casting skill and 60 free mana when fighting in its fortress.

And a wizard with 8 cities with a total population of 120, would ignore the first 120 damage its units took in that battle (and every time the wizards units made a resist check, they would auto pass, until those 120 points were gone, but each resist check would reduce the 120 by the same amount needed to pass. If a resist 5 unit rolled a 2 against something with -4 to resist, then it would pass, and use up 7 of those 120 points.)

Obviously a completely overpowered method, but you get the idea. You can put in mechanics (well actually I have no idea if you could fit that in at ALL, its a little obscene), that would defend the fortress, if you really want the fortress to be a different fight.

And if someone really was far more powerful, they could still take the fortress early.

[zitro1987]

IF we want the game to be about slowly clearing the empire down and only taking the fortress at the end, I would suggest taking zitros idea, and expanding on it.

A ridiculously extreme version could be something like: for every point of power granted by nodes, the wizard gets that much extra casting skill in combat at its fortress, and the first man's spent up to that amount is free. Similarly, every population the wizard controls, will absorb one point of damage, or add one to a resist check until the resist check is a pass.

So a wizard with 3 nodes generating a total of 60 power, would get 60 extra casting skill and 60 free mana when fighting in its fortress.

And a wizard with 8 cities with a total population of 120, would ignore the first 120 damage its units took in that battle (and every time the wizards units made a resist check, they would auto pass, until those 120 points were gone, but each resist check would reduce the 120 by the same amount needed to pass. If a resist 5 unit rolled a 2 against something with -4 to resist, then it would pass, and use up 7 of those 120 points.)

Obviously a completely overpowered method, but you get the idea. You can put in mechanics (well actually I have no idea if you could fit that in at ALL, its a little obscene), that would defend the fortress, if you really want the fortress to be a different fight.

And if someone really was far more powerful, they could still take the fortress early.

Not bad, but rather complex. In my opinion, we're getting closer to a partial or full solution.

How about the following well-rounded combination of 3 very different effects, strengthened by 'power' and having a very high cap of 1000 power? These effects are quite weak at the beginning of game.
Lightning Bolt - Strength of 5+ 0.03*power (+1 per 33 power) - Cap would then be 35
Extra Skill - +0.25 * power - Cap would then be +250 skill
Extra stats among defenders - Melee&Armor: +0.0066*power +1 per 150 power) Ranged&Resistance: +0.005*power (+1 per 200 power) - caps would then be +7 melee / +7 armor / +5 resistance / +5 ranged (I'd probably give an initial +1 to melee and armor stats on top for early game help)

Early game - 70 power : strength 7 bolts, +17 skill battle, +1 melee/armor stats
Mid game - 200 power: strength 11 bolts, +50 skill battle, +2 melee/armor , +1 ranged/resistance
Mid-late game - 400 power: strength 17 bolts, +100 skill battle, +3 melee/armor, +2 ranged/resistance
Mid-late game - 600 power: strength 23 bolts, +150 skill battle, +5 melee/armor, +3 ranged/resistance
late game - 800 power: strength 29 bolts, +200 skill battle, +6 melee/armor, +4 ranged/resistance

[Catwalk]

I agree with the problem stated, and I'm glad to see you've changed your mind on this issue. I also agree that the best fix is to make banishment less severe. Trying to fix it with stronger bolts is the wrong way to go and the other solutions above are also unnecessarily complex. Seravy's first idea is actually perfect, it is a moderate penalty which hampers the target without disabling him.

Also, Guardian should be removed and it's effects applied to all fortress battles. That's an easy fix which is easy to understand and which has a substantial impact on the difficulty of fortress combat. As a bonus, it also frees up a retort that noone ever picks. Guardian is valuable for the AI and largely useless for the human player, that's not good design. I feel that city curses are the same way. They're potent against the human player and ineffective against the AI.

[Juffos]

Spell of Returning being a 500 mana spell only gets you one or two turns without the enemy wizard in late game - this makes fast moving kill stacks almost obligatory. I would consider having the spell be cast at, for example, 50 + 1/4 current skill mana per turn a better option. This way early banishments wouldn't be such a game over for the AI and late banishments not a slap on the wrist.

However, with Spell of Returning taking longer to cast, I would allow combat casting during the banishment, but only with halved/thirded skill and a huge range modifier.

I wonder if the code can allow stuff like this.

[Catwalk]

Variable SoR makes a lot of sense, if it's possible to do it. I'd rather base it on turn than skill, but I can see an argument for making it skill based as that means it takes roughly the same number of turns throughout the game.

[Seravy]

Rather than variable cost SoR, an easier solution is to add this rule :

-While banished, a wizard can spend 100 mana per turn on casting Spell of Return instead of their casting skill.

(this way it always takes the same amount of turns.)