Double Defense (AoD) and Feint

[dripton]

Originally Posted by T.K.

On the math discussion currently presented so far, I might be wrong but from what I could understand people are not considering that the extra Crit from keeping skill 16 completely negates the defense of the target...is that the case?

No, we are considering it.

It's just not that huge an effect. At skill 16+, you have a 9.25% chance of a critical hit. At skill up to 14, you have a 1.85% chance. That 7.4% difference is significant, but not huge. Now consider lowering an opponent's defense roll from 12 (74.07% chance to defend) to 10 (50% chance). That's a 24.07% difference, much bigger.

[Ulzgoroth]

Originally Posted by DouglasCole

P(Crit) + (1-P(Crit))* Probability of a Hit * Probability of Failed Defense

This appears to be wrong unless your P(Crit) is actually P(Crit|Hit)

[Tomsdad]

Originally Posted by dripton

No, we are considering it.

It's just not that huge an effect. At skill 16+, you have a 9.25% chance of a critical hit. At skill up to 14, you have a 1.85% chance. That 7.4% difference is significant, but not huge. Now consider lowering an opponent's defense roll from 12 (74.07% chance to defend) to 10 (50% chance). That's a 24.07% difference, much bigger.

Not sure about that, because it's not a like for like comparison (i.e simply 7.4% being less than 24.07%)

That 24.07% is subject to the defender's choices, he could say opt to retreat and get a +3 bonus that would more than remove that 24.07%.

A critical hit functionally removes defence as an option, and makes the defender a passive observer unable to effect the result with his own choices.

Imagine if he's been making combat choices that boost his defence. Defensive attacks, Defensive Feints etc, etc or if he'd been doing AoA's

It's also a balance that's dependent on the targets defence score, 12-10 is better reduction then a 9-7 for instance (but again that will also be dependent on the defenders choices)

Now don't get me wrong that's a bit of an esoteric distinction but it's one worth considering.

[Varyon]

Originally Posted by Anders

The thought strikes me that some pointers on when to use which attack option would be useful. If written by someone qualified.

Started a new thread for this, here.

[Peter V. Dell'Orto]

Originally Posted by T.K.

Do I DA beyond effective skill 16, thus loosing crit, or cap at 16 to keep maximum crit range and reduce critical failure range as well?

Right, that's the tradeoff. It's actually fairly straightforward:

- one approach maximizes the chances of penetrating defenses overall, even at a cost of critical hits (which, at the very least, bypass defenses, and in Advanced Combat have a table roll for further effects) and slightly more critical misses (which can be disastrous at times). This is usually done with lots of Deceptive Attack, close to the minimum skill - aka the "go to 12" approach.

- the other approach maximizes the chances of getting critical hits and minimizes the chances of critical failures, at the cost of less hits overall across the length of a battle. This is the "stay at 16 or 17" approach.

Both are independent of Feint, though - Feint can greatly boost both, if you have superior skill. Like I said, they stack, and attacks aren't always free and unlimited (as Parry U and double-dagger weapons users know.)

Of course, you can freely mix both - you can stay at 16-17 most of the time, then bust out a "go to 12" approach for a target that needs it or when the consequences of a critical failure are low/non-existent or when the upsides of a critical success just don't matter (or, when you can't hit skill 16 in the first place!). That's probably the most fair assessment of our most consistently successful (in combat) players in my games, that they err heavily on the conservative side until a situation comes up where the upsides of risk exceed the downsides of failure.

The ones who consistently err on the cautious side, though, tend to survive longer. I have a pair of brothers, one of whom is best exemplified by a recent comment that boiled down to - "12 or less? That's like 100% chance of success!" and the other is extremely conservative and careful. One is on his third character this campaign and his Weapon Master in the previous game died at least twice that I remember. Maybe three times. The other played one character in each campaign and neither needed Resurrection.

[dripton]

Originally Posted by Peter V. Dell'Orto

The ones who consistently err on the cautious side, though, tend to survive longer. I have a pair of brothers, one of whom is best exemplified by a recent comment that boiled down to - "12 or less? That's like 100% chance of success!" and the other is extremely conservative and careful. One is on his third character this campaign and his Weapon Master in the previous game died at least twice that I remember. Maybe three times. The other played one character in each campaign and neither needed Resurrection.

I agree that caution is usually good. It's definitely possible to be too reckless. All Out Attack is usually a bad idea, for example. Or turning your back on an enemy. Or Move and Attack because you're in a hurry to get in the first attack (at a 9), even though it means you'll be taking a full-strength attack back, with no parry or retreat possible.

Wear good armor. Buy Luck. Search for traps. Post a watch at night. Be invisible whenever possible. Carry extra healing potions. Learn Swimming. Floss twice a day. Do all the cautious things that make sense.

But it's also sometimes possible to be too cautious. If you're fighting an extremely dangerous opponent (say, a boss Giant with weapon skill 16 and a huge axe with damage high enough to pretty much automatically blow through your DR and put you at negative HP), then every turn you let it stay standing, active, and within melee range is basically at least a 10% chance of taking a hit that probably costs you the fight. (Unless you have unused Luck, in which case it's the second time that gets you.) It's Russian Roulette. If you can't run away and you don't have a tactic that lets you win from a distance, then you need to take it down as quickly as possible, before it rolls a lucky hit.

Against such an opponent, I'll take an extra ~1.4% chance of a critical miss to get an extra ~5% chance to do some damage, every time. If I critically miss, I'm Fast Drawing my extra weapon or defending at -2 next turn or something equally annoying but non-immediately-fatal. If I miss because of overly cautious tactics and the Giant critically hits, I'm likely making a new character.

[Peter V. Dell'Orto]

Originally Posted by dripton

Against such an opponent, I'll take an extra ~1.4% chance of a critical miss to get an extra ~5% chance to do some damage, every time. If I critically miss, I'm Fast Drawing my extra weapon or defending at -2 next turn or something equally annoying but non-immediately-fatal. If I miss because of overly cautious tactics and the Giant critically hits, I'm likely making a new character.

This is true - but equally some people will look at that situation and say it's better to maximize the chance of a critical hit. It's really a question of where you see the risk falling. And in the case of a skill 16 boss giant in my DF game, well, to get back on topic - I know our knight's player would use Feint as part of a Rapid Strike, and spike his defenses into the dirt first, regardless of how much Deceptive Attack he chose to use on the following attack. :)

[DouglasCole]

Originally Posted by Ulzgoroth

This appears to be wrong unless your P(Crit) is actually P(Crit|Hit)

Indeed. It winds up slightly changing the numbers for P(Hit) being close to 1, and can be pretty drastic if P(Hit) gets closer to 50%.

So I fixed it, and went back and did the analysis again. Answer unchanged: lacking any information about the nature of your foe, 16 is the overall best strategy. In most cases, choosing 16 will be within 5% of the mathematically best chance to hit per blow, and depending on your mental weighting of the import of critical hits (as was discussed upthread) this will throw your numbers one way or another. Choosing 12 over 16 is never more than just shy of 6% better by the math. It can be (in extreme situations) 20-25% worse, but that's if you're being silly and choosing 12 over 16 in a situation where your foe is defending at very low effective scores AND you're great.

Anyway, the spreadsheet is available for download if anyone wants it.