Survey: When do you attack an enemy unit?

[Chalid]

How big attack odds do you want if you attack with:

- Hero
- Regular unit
- Catapult/Collateral Unit

- When attacking a unit that is much stronger than the best unit you have (eg. Mithril Golem)?

- When facing a stack of enemy units that has much more units than you have in the surrounding tiles?

Disclaimer: The results of this survey will influence the behaviour of the AI. You were warned..

 

[kevjm]

Personally, with important (highly-promoted) units, I don't like odds below 80%. With unpromoted units, I don't like anything below 65%, they will promote faster this way. Perhaps you could set the AI to treat units differently depending on their level?

Keep in mind I play on Prince level and am no expert

 

[abman]

I do about the same thing. Never attack unless its above 65% (except siege weapons). With a highly promoted unit, I might not attack unless its above 90%. Of course that can mean nothing. Last game I lost a hero at 99.8% odds! I also play on Prince.

 

[Corlindale]

I'm far too reckless. In some cases "above 50%" is good enough for me, I have even made that choice with hero units, which has lost me many a Rosier or Kithra in the past. But that's mostly when attacking well-fortified cities, where it's sometimes hard to get odds better than that unless you drag tons of catapults along. Archmages help alot, though.
In "everyday combat" I usually try not to risk my heroes too much.

I also try to estimate the defensive situation at my square when attacking cities, so I won't make my units too easy prey after the assault, either by not attacking with some units or by attacking only with very good odds, or by having a High Priest with Heal at hand.

 

[woodelf]

I usually use my crappy old obsoleted units to soften up a stronger foe and hope to finish him off with my better units in that same turn. That backfires if I fail to defeat a unit after he won that many battles though.

I'll attack with above 60% odds with anyone especially if I can finish off a straggler. I always try to save enough for the impending counterattack.

I use Heroes to kill lesser units or finish off stronger ones. Very seldom do they get risked on initial sieges with <80% odds. I also never leave them to defend cities (although I know this is being fixed soon).

I play on Monarch.

 

[Jono]

-Heroes 75%-85%+ depending on the importance of the unit I'm attacking (eg. Orthus - 70% will do).
-Normal Units 35%-60% (depending on promotions, sometimes I don't even care if the unit dies).
-I don't use catapults really. I suppose odds wouldn't matter to me aslong as I do enough damage...
-80%ish unless he gets stronger with time (Orthus).
-Not high really... If I see an enemy stack on it's way, I usually send a large group of summoned units.

I play on Prince~Monarch...

 

[Sureshot]

I normally make sure my unit has a higher base strength and more relevant promotions and only attack if they receive no defensive bonuses (which is usually atleast 99%), but thats because I'm paranoid and can't survive losing units on the hardest difficulties. The only exception is if its 50% or higher and will save more units/things of importance, than if I do nothing.

On easy difficulties or in a position where I'm dominating the world I'll throw away units (especially catapults, but never heroes) if its enough to achieve my goal (like destroying all units in an enemy city).

 

[mindlar]

It depends a lot on the unit.

With summons I will attack at 0% if part of an invasion to soften up the enemy. 5% if it is a lone mage/conjurer in relative safety.

With the lowest level werewolves I sometimes use them as cannon fodder (i.e. anything > 0), but usually at anything above 40% because they are easily replaceable.

Heroes are 95%+ almost exclusively. Occasionally I'll take a chance around 70%, but not often.

Built units are typically used anywhere from 60%+. Each level increases that %. With a level 8-10 I usually treat them similarly to heroes.

 

[TheJopa]

I just wrote an article but accidentaly clicked on wrong place and it is all lost...
So, short:
Heroes: 90%+ (I don't want to loose them)
strongest t4 units: 80%
when attacking avatars: 50%+ but attack only if I'm sure to defeat it even if I will lose some units
Catapults: Doesn't really matter because I use them for collateral and bombard, not to kill
But it is worth to mention that odds in most of my battles are over 95% because summons, meteors and fireballs do their job and heroes and t4 units with city raider 1,2,3 easily take weakened defenders

 

[Kael]

But it is worth to mention that odds in most of my battles are over 95% because summons, meteors and fireballs do their job and heroes and t4 units with city raider 1,2,3 easily take weakened defenders

Thats the way it should be. IMO a strategy game should play like chess, the results are almost garunteed given the proper strategy. Without predictable results the player is unable to plan and the game suffers.

Thats not to say that you don't have those occasional 50/60/70% matchups that leave you wondering if you should risk it or not. That part of the game too, but it shouldn't occur in most of the battles.

 

[bluehorn]

Heroes: Generally over 95% (my heroes are all craven)

Highly promoted reg units: Maybe 70% and up depending on the circumstances

Regular/unpromoted units: If I need to use these guys to mop up after the tigers/skelatons etc. have had their fill then I'll generally attack with any odds, say 40% and up

Catapults: Can't say I really ever use them to attack anything, they pretty much just bombard for me

 

[Samuelson]

If I'm attacking a city and I've brought along an army then I send in my weak units first even though I know they are going to die to soften up the enemy so my strong units can kill them. So basically the first units I send in no matter what the odds, even if it is 10% because I'm expecting them to win.

When not in a city it depends if I have any backup units to kill them if my first one dies. Generally 55%+.

 

[Nikis-Knight]

I'll risk more if the outcome is greater, i.e., an inquisitor against an enemy priest or a pivitol city. Sometimes I'll attack with an arcane unit if the odds are >80% and he needs a promotion. Heroes need to have 60 or so, but anyone with 10+ levels are probably saved til 90%. And ironically or not, the better the odds the more likely I'll quicksave before attacking--if they win at 99.5, I'm not about to let them keep it!

 

[Kael]

I'll risk more if the outcome is greater, i.e., an inquisitor against an enemy priest or a pivitol city. Sometimes I'll attack with an arcane unit if the odds are >80% and he needs a promotion. Heroes need to have 60 or so, but anyone with 10+ levels are probably saved til 90%. And ironically or not, the better the odds the more likely I'll quicksave before attacking--if they win at 99.5, I'm not about to let them keep it!

Cheater!!!

 

[Black Whole]

- Heroes: At least 90% (and create a savegame before I attack )
- Normal Units: At least 70%
- Catapults: I always would attack, even if there was a 0.0% chance
- Mithril Golem: I haven't played the game so far yet but I would throw my whole army against this unit except for my best units. Someday, he must die.
- Stack: I would run home, fortify and build mass units.

Disclaimer: The results of this survey will influence the behaviour of the AI. You were warned..

The fine-print sentences are always the most devious.

 

[Panda]

80+, especially heroes. Not that the situation always allows it.

@Woodelf: Nice avatar!

 

[Xuenay]

How big attack odds do you want if you attack with:

- Hero

95-99%, usually. 90% if I feel like pushing my luck.

- Regular unit

Depends on how many I have of them. 50-70%, usually, but if I have lots of units to sacrifice I might go lower.

- Catapult/Collateral Unit

Depends on the target and whether or not I think I have enough units in the stack to beat the opposing stack. If I can't kill the stack or bring it down to very heavy casualites, then there's no point in throwing units at it and giving it more experience. I haven't used catapults much, but if it seems like my summons will be able to weaken the opposition enough, then I'll use the summons at any 0%> odds. If I'm not confident of being able to take down most of the stack, then I want at least 30-50% odds for the summons.

- When attacking a unit that is much stronger than the best unit you have (eg. Mithril Golem)?

Odds high enough that I think I can take it down with the entire stack. Depends on whether it has friends and if it's a unique unit or not. If it's unique, it would seem smart to throw everything non-unique I have nearby at it to take it down. I might be ready to trade heroes in the correct circumstances, but only if there's no other choice - I can't use up my heroes taking down the heroes of one opponent and then be overrun by the heroes of everybody else.

If I'm facing much stronger non-unique units, stall and sue for peace ASAP.

- When facing a stack of enemy units that has much more units than you have in the surrounding tiles?

80+%, if I can get those odds for the vast majority of my units and still leave a couple unused to defend against the upcoming counterattack. If not, retreat, regroup and get reinforcements.

 

[Sarisin]

As a warrior who has fought thousands and thousands of battles (mostly with raging barbs) here are five of my observations on attacking:

1. I think the Combat Odds are out of whack. I have lost way too many battles I was favored to win in the 90+% range. Plenty of them in the 99.X% amount and one, only one, memorable time I lost a battle when the odds were 100%!!!

2. I think the Combat Odds for 80% or lower should be saying those are the odds you will LOSE not win. That has really been my experience, and I will not chance an important unit in a battle 80% or lower.

3. Defending odds are also sometimes out the window I have found. For example, yesterday I was defending a key resource (copper mine) with a maxed out Warrior who had fortified the position to 25%. He even had 40% against mounted units. Along comes a lone Worg Rider and takes him out, has a movement point left and destroys the mine! Fortunately, I did a quick save (c'mon we all do this for key battles ) and decided instead to send my archer there - again maxed out on XP. He was killed too. Finally, I put a maxed out Hunter there, and, you guessed it, he was also dispatched by the Rider. Nothing to do but give up the resource and quickly kill the offending unit. But, the point is I have to think the AI's combat odds against any of those units was really, really low - yet he still won.

4. I agree with Kael that it should be more like a chess match where you set up the proper offense/defense and the right units with the appropriate attack/defend value. To then lose when the odds are overwhelmingly in your favor is frustrating. But, hey, I guess it history it has happened over and over again. I am just saying in this game, if you have a 99.9% combat odds of winning, you should win 999 out 1,000 battles. I don't think that is necessarily the case.

5. Lastly, I think the terrain features can be murder on the combat odds. Trying to kill a unit on forest/hills with another until is very tough. Maybe that is as it should be, but it sure can skew the combat odds.

And...I am still trying to find out a way to get those nasty barb Lizard Men early in a game who park themselves in the water waiting to pounce. They have perhaps the best terrain advantage of all!

 

[Kerrang]

Heroes - 85% chance or better to send them into combat. Depending on which hero it is, and how I am faring in the game, I may load up a saved game if I lose at that point.

Regular units - 50/50 chance is good enough for most regular units. If they are highly promoted I will wait until they have a slightly better chance (60%-70%) before they attack.

Collateral units - I will suicide collateral units against a stack all day long if it will increase the other units chances of winning. That is why I make them in the first place.

Against Mithril Golem - If another civ has the Mithril Golem, and I do not have the Meshabber of Dis (or visa versa), it is time to start a new game.

Outnumbered vs. a huge stack - It really depends on what units are in the stack. If the stack is all tier 2 units, and i have half as many tier 3 units and heroes in the surrounding tiles, I don't sweat it at all, I know I can take them. If heroes and tier 3 units are mixed in the stack, it is time to head for the nearest walled city and fortify.

 

[Brancaleone]

Last night i had a shield wall on grassland forested hills with over 30 xp, and my unit got reduced from strenght 12 to strenght 0.3 by a tiger.
strenght 12
+80% hills defense and +80% grassland defense, unit abilities
+75% terrain
+120% against melee

Man if you do some math youll se my unit had a strenght of almost 50 against a tiger with 3.6 strenght. Incredible. And two times i lost walled cities garrisoned by Battlemasters (dovielo axemen) by goddamn tigers. Whats wrong with those creatures, do they have a bonus against Brancaleone or something?

Im sure this kind of thing happens with everyone, but since the last update i never left a battle with more than 50% of my unit strenght, even if i had a chance of 90% or so. Did the mod became more "bloodly" or something? My heroes have to fortify and heal after dispatching every damn wolf!

With heroes, i always quicksave (you wouldnt belive the creatues that killed my heroes already), and always against little guys. With normal units, if its better than 50% its good for me, unless i have nothing else to protect him while he heals), i lose lots of catapults against cities and stacks, but thats why theyre for, and summons, i barely use them if not playing arcane civs.