How much defense is too much?

This is somewhat of a rant, I guess.

I exasperatedly exclaimed something like, "Oh for ****'s sake!" during our last session, as someone rolled a 13 and made a Parry by 2 . . . not counting any DB, without Retreat, and without any extenuating bonuses of any kind. Just a pure parry roll of 15 before all of the fun stuff that's piled on top of that - the PC also had a 2 DB shield, and a Shield spell of +2 or +3, and could potentially Retreat. Why the exclamantion? Just some general muttering about how dangerous the situation was. You know, followed by the revelation of a 98.1% of success and only a rule-required margin for failure or critical failure.

Honestly, I get why my players do the things they do - why they layer defensive buffs over high defenses over everything.

But still, as a GM, it's face-palm territory when people say their defenses aren't good and then you find out they have a 15 Parry, without any DB or Retreat, or say they're about to get overwhelmed and die when they're forced to fall back on Dodge-14 and can't get that all-too-important +3 for Retreat. It's the defense equivalent of poor-mouthing . . . woe is me, I only have a near maximal defense, not the maximum possibly allowed.

Like a 95.4% of success is too low, and only a 98.1% chance of success - on a guy who also has Luck and a pile of DR and high stats and additional bonuses to deflect any non-damage followup.

Anything that can potentially cause any harm get classed as "lethal." So monsters are put into two categories - harmless, or so dangerous they can only be fought with superior numbers, in a defensive posture, while buffed, with no possibility of allowing flanking attacks, with room to Retreat.

Personally I think the "we can't afford to get hit, because we can't afford any casualties" approach means excessive caution.

It also can feed an arms race. Massively layered defenses that cannot be reasonably dealt with means they must be unreasonably dealt with. A GM really has no choice besides, effectively, "I guess you all win" or ramping up the offensive power of foes and provide for effects that cannot be avoided. Otherwise it's just rolling dice to see which PC kills the enemy, and how long they take to die, without real challenge of potential failure.

Yet, again, this means a push towards caution. It further forces the game towards tactical combat considerations in all situations, since Retreat is held up as a requirement to survive and worries about denying a flank dominate discussions about what to do.

This is why I've mused a few times about hard skill caps, hard defense caps, and other bounded limitations. That way the arms race at least ends at some point.

But I'm not sure it would stop what drives me a little nutty about the whole thing - the excessive concern that any chance of failure means a high chance of failure, and describing anything less than maximal defenses as poor. One of my players can get a little smug about his high defenses, but at least he's owning the fact that he has them and that they're high. I just feel like a more reasonable acceptance of how defensively secure the PCs are, right out of the gate and then especially after they get more points and better gear would help. So would an acceptance that risk can be reduced but not minimized without costing some - perhaps most - of the risk-reward challenge that DF is meant to be. And I've said it before - people over-patch and they tend to hold up "without any casualties" as delve goals . . . and it costs a lot without getting the end-goal in return.

Just some thoughts.