Hi everyone,

Lately I have been having fun with this gem, but something is bothering me: I can't quite understand combat statistics.

Why is my Guardian Spirit that has 5 swords losing against a Swordsman Squad that only has 2 or 3 swords and 1 or 2 shields? (Can't remember it for sure)

From the manual I get that the number of swords represent a 30% hit chance on every sword, still I should be having more probabilities to damage than they do to damage me. Are the stats on the unit card calculated for EVERY model in the squad?

Yep.

Note: This is my (current) understanding of how combat works.
That 6 figure swordsman squad with 3 attack is throwing 6 shots of 3 attack each at your spirit each time the two units take swings at each other. Your spirit is defending against each attack with its full shield count, but that is a lot of attacks incoming. On the flip side, your spirit is fighting at full "strength" until it dies; the swordsmen will become less and less effective at dealing damage as more figures drop. Your spirit may take a bunch of damage at first, but it should start taking less and less as the two units exchange blows.

As for your spirit damaging the swordsmen: each figure in the unit will get to apply their full shield count against the incoming damage. When your spirit swings, the first sword swinger in the unit steps up to defend, if he dies then the next guy steps up with HIS full shield count, and so on. Lets say your spirit gets a full 5 hits of damage against the 2 def/1 hp swordsmen: the first guy succeeds at blocking two damage (all he can block), takes one damage, and dies. The second guy now only has to deal with 2 incoming damage; he might manage to block it all, but lets say he completely whiffs his block, takes one damage, and dies. The third guy now uses his full 2 shields to try and block that last point of damage; whether he succeeds or fails, the 4th guy won't have to worry about dying (yet).

Edit: Also, note that each sword represents a BASE 30% chance to hit (and each shield represents a base 30% chance to block). If the unit has +1 to hit, then each sword will have a 40% chance to hit.

These +1's can stack up, too. A High Elf (+1) Elite (+1) Swordsman with Magical Weapons (+1) and enchanted with the Holy Weapon spell (+1) has a 70% chance that each of its attack icons will result in a hit.

Each +1/-1 To Hit represents a step of 10% in the To Hit chance. I like to think of it as a scale of 1-10+, where 1 (10%) is the minimum (hard capped) while 10 or above (>= 100%) represents all attacks always hitting. Anything above 100% doesn't provide any benefit beyond being a buffer against some effect that would otherwise lower the unit's accuracy.

Thanks, that covered everything :)

It doesn't apply here, but keep in mind that certain abilities such as Thrown and First Strike allow the attacker to deal their damage first, rather than simultaneously. This can substantially reduce or eliminate any damage taken in return. (Note that this only applies when they're the attacking unit, not the defending one.)

This is one of the nasty things about Chaos spawn in close combat. Their abilities often work on both offense and defense and will work before many other abilities.

When the Chaos Spawn initiates a Melee Attack against an enemy unit, or an enemy unit initates a Melee Attack against the Chaos Spawn, the creature will first unleash a barrage of Special Attacks.
Note, the Chaos spawn has a melee strength of one, if that is reduced to zero then none of the other abilities will fire.
Please note that most of these effects are resisted by the target at -4.
The Chaos Spawn's special attacks are as follows:
◾ A Doom Gaze, which delivers exactly 4 points of Damage, regardless of the target's statistics.
◾ A Death Gaze, which can kill any enemy figure that fails to Resist it.
◾ A Stoning Gaze, which can kill any enemy figure that fails to Resist it.
◾ A Cause Fear attack, which will prevent any figure that fails to Resist it from dealing Melee Damage back at the Chaos Spawn.

Note I was "qutie sure", but wrong, in what I posted below. Should have tested! danhanegan and Bookwyrm627 are of course correct.

I am quite sure this is incorrect - the defending unit rolls its defense for all its shields once, regardless of how many attacking figures inflicted the damage. Were the guardian spirit rolling defense rolls for each attacking figure separately, the OP's guardian spirit would be mopping the floor with the opposing swordsmen (winning in 6 rounds with 0 damage on average). What actually happens in that example follows:

(simultaneously)
The swordsmen roll their attacks: 6 figures, 3 swords each, so 18 swords, for an average (unmodified) of 5.4 hits.
The guardian spirit rolls its attack: 1 figure, 10 swords, for an average of 3 hits.

The guardian defends: 1 figure, 4 shields, blocking on average 1.3 hits. In an average first round, the spirit takes 4.1 hits.
The swordsmen defends: each figure, with 2 shields, blocks an average of 0.6 hits. So in an average round, the first swordsman figure blocks 0.6 hits, then dies with the next 2. There is 0.4 damage "left over" which the next swordsman figure should usually block, unhurt.

In the second round, the guardian fights as above, but the swordsmen now have only 5 figures, so do an average of 4.5 hits... the spirit should take on average 3.2 damage, and kill 1 more swordsman. Next round, spirit takes 2.3 damage (grand total of 9.6 hits), and kills one more swordsman. Last round, the spirit dies, killing a fourth swordsman. Swordsmen win.

This is assuming no bonuses on either side, and average rolls of course.

The wiki indicates otherwise. My game experience also indicates otherwise: as I recall, generally a guardian spirit is up to the task of downing a unit of low experience swordsmen (the 3 swords indicate the swordsmen are likely recruits), though it has been awhile since I've tried to pit swordsmen against guardian spirits. With 6 attacks of 3 str each though, that is a lot of random variation, giving the swordsmen a good chance of hurting the spirit before dying.

Right, and the wiki is right (though I can't see the "physical damage" page right now for some reason, but the comments make its content clear). I'm wrong... but in my defense, so is the manual. :)
Thanks for the correction!