We first do ~ Final Attack - Final Block = 100 - 40 = 60% Damage
Then ~ 60% of 50 damage (Longsword) is 30
Then we have ~ 30 (Damage Taken) - 20 (Absorption) - 10*2 (AT) = 0 Final Damage Taken (would be -10, but attacks don't heal)
So the character is hit, but just got cut or a bad bruise.
Now if I have to clarify this even more... I'll cry
I've seen this stated two ways. One is the way stated above (I'll call it Method A), the other is as Dargoking mentioned (Method B), broken down as follows:
Final Attack - Final Block = 100 - 40 = 60
Difference - Absorption = 60 - 20 - 10*2 (AT) = 20 (this is the % dam)
20% of Longsword Base Damage 50 = 10 Damage
I cannot read Spanish and haven't bothered trying to use a translation program, but Method B is almost certainly correct. Method A makes it nearly impossible to to do any damage with weapons that have a low base damage, and is worse at stopping damage than the original system against weapons with high base damage. Method B works out exactly like the original table (excepting the wonky part of the low AT, low % section) except it gives everyone a free 2 AT. The way absorption is worded indicates that it's made to be a simpler way to calculate damage without a major change to the way AT protects you. Method A above makes a pretty large change in the way armor protects. It makes small, mundane attacks even less consequential, and it makes the big hitters hit even harder. I can't see any reason why that would be a desired change. The big hitters are lethal enough.
One last example to clarify my point:
Using Method A, let's look at a skilled Acrobat Warrior who completely specializes in throwing lots of stilettos. They have a base damage of 40. For fun, let's look at a level 4 Acrobat warrior who is maxed out in attack/def with 10 Dex (150 attack), Mastery in Sleight of Hand (can throw 7 stilettos per turn at -60) and ambidextrous (one more throw at -10). This guy should clearly be able to tear to shreds the poor level 1 guard wearing a breastplate (75 block, 4 THR AT), right?
Let's up the stakes for the AW to really decimate this guy, let's assume every roll he rolls 30 points higher than our poor guard. This guy should be toasted.
Final Att 120 120 120 120 120 120 120 170
Final Def 45 15 0 0 0 0 0 0
Dif 75 105 120 120 120 120 120 170
Pre Abs Dam 28 40 48 48 48 48 48 68
Abs 60 60 60 60 60 60 60 60
Final Dam 0 0 0 0 0 0 0 8
EDIT: This table is a mess, sorry. I don't really have the time/patience to figure out how to format it correctly. You can still figure it out without too much trouble if you really want to.
So, in all that, the guard was outclassed in every way, was beaten by a decent margin on every roll, and was successfully hit 8 times, yet he only ends up taking 8 LP of damage? Because he's wearing a simple breastplate? I personally think that just cannot be right. Honestly, I'd love someone fluent in Spanish to confirm, but as far as I'm concerned if Method A is RAW, then I'm houseruling to Method B and calling it a day. I've used both recently and B is very preferable, equally to my lightly and heavily armored players and monsters.