This is Part III of a 4-part series on combat skill caps, defense caps, and fixed feint/deceptive attack rules. It's going to connect to DF in Part IV, but these rules should work perfectly well in straight-up GURPS and are intended for potentially broad application.
Let's look at a combined offense and defense, and related issues.
Putting it Together
Defenses are based on skills, so setting the offensive limit
Let's run some numbers and see if I like what I see.
Aaron A. Aardvark has Broadsword-16, Combat Reflexes, Shield-14, and a Medium Shield (DB 2), and Speed 6 (and Unluckiness, but that's neither here nor there.) He has Parry 12 + 2 DB = 14, Block 11 +2 DB = 13, and Dodge 9 + 2 DB = 11 with No Encumbrance.
He faces his opposite number, Betty Basher, with Axe/Mace-18, Combat Reflexes, Shield-14, and a Medium Shield (DB 2), but only Speed 5.5. She uses a knobbed club to avoid Parry U. She has Parry 13 + 2 DB = 15, Block 11 + 2 DB = 13, and Dodge 8 + 2 DB = 10 with No Encumbrance.
Aaron can crit-fish against Betty for a 3-6 critical on 16 or less. Betty defends with Parry 15, 16 with Retreat.
He can use Deceptive Attack for a 12 or less, but Betty defends with a Parry 15, 16 with Retreat, or 13 and 14 after the -2.
He can Rapid Strike with two 10s, and if both hit Betty defends at a 15 and a 13 (16 and 14 with Retreat.)
Aaron can also Feint Betty, but 16 vs. 18 gives him less than 50% odds of gaining an edge. It's not a smart play but it could work. If it does, he can both crit-fish at a 16 and force Betty to defend at an 11, 12 with Retreat. If that somehow works, that 12 or less Deceptive Attack is mighty tempting but risky.
Betty can crit-fish against Aaron with a 3-6 critical on an 18 or less - 17 and 18 still miss, of course. He can't hairsplit Deceptive Attack. Aaron defends with Parry 14, 15 with Retreat.
She can use Deceptive Attack for a 14 or less, forcing Aaron to defend at a 12, 13 with Retreat.
She can Rapid Strike for a pair of 12s, and if both hit, Aaron defends at 14, 15 with Retreat, and then 13, 14 with Retreat.
She can Feint Aaron, and with an 18 vs. a 16 she's got reasonable odds of success, althought not great. A success, though, can really turn the fight - Aaron defends at 10, 11 with Retreat . . . and it's easy enough to attack with Deceptive Attack to put this to 8, 9 with Retreat.
What if Aaron has Broadsword-24, instead? That's Parry 16 + 2 DB.
Aaron can Deceptive Attack at the Very Deceptive Attack level (needs a new name) on a 16 or less; this criticals on a 3-6, and Betty has to defend at an 11, 12 with Retreat.
Aaron can Rapid Strike for two 18s, or make the Deceptive as well for two 14s. If they both land, Betty has a 15 and a 13, or a 13 and an 11, if both are also Deceptive.
Aaron can Feint Better, and 24 vs. 18 means it's very likely to succeed. Betty has an 11 and a 9, or that followed by a Very Deceptive Attack forces to Betty to defend at 7 and 5.
All of that seems okay to me - the fixed Feint and fixed Deceptive Attack levels seem reasonable and workable. High skill still matters, a lot, but you don't really get to finagle it as well, so the defender still has a shot to defend.
I think I like the fixed numbers. I also think the skill cap of around 25-30 is probably a workable max, too, depending on the game.
Race to the Cap!
It's quite possible that people will race to the maxima. I feel the best way to deal with this is to make it a soft maximum on skill. Flat cost until the cap, and then a rising cost - double or even triple cost per level afterward. It's easily done with an Unusual Background if you don't like directly messing with skill costs, or want it to keep scaling up for each additional level.
Do you still need a DB cap?
I'd say yes. Although it's hard to get past DB 3 in Basic Set without supernatural aid or superscience, it's easy to get high single digits with GURPS Magic in play. The theoretical maximum of DB 19 calculated here is still possible, after all, and that's really in any game using the basic magic system and extended spellcasting for Magery above 3.
If active defense is limited, then it's logical to try to maximize your defense bonus any way possible.
Having a cap - I suggesed 8 DB - is reasonable. I also like the idea of Option 3 - best source - and DB means contact effects - Option 2. And I like DB from magical effects not working close combat.
In any case, preventing DB from rising above 6-8 has the nice effect of saying that defenses should probably reach around 16, and very high ones - thanks to DB, advantages, and skill - can't get much higher than 30 . . . 31-33 with Retreat . . . if everything is maxed out and skill caps at 30. Capping at 25 skill would make it more like a 6 DB cap and a 25 maximum Parry or Block before Retreat or weapon-based defensive bonuses. Look at the examples above - give either of them DB 5-6 instead of DB 2 and most of the value of the options are reduced. Give them DB 8-10+ and it's basically over. Equally, though, toss in Blur and it suddenly matters, even to Aaron with skill 26, as it changes options since crit-fishing is just one viable strategy. Equally, it's not critical to give out Shield to everyone because it's harder to really surpress defenses past a certain point.
Overall: I think the defenses that spill out of the likely skill levels are okay, here, and that a Defense Bonus cap is worth it. I don't like a fixed cap, really, athough it's simple compared to the other options. The option of "best source" makes a fair bit of sense, though it suffers from some logical holes, like Shield spells suddenly being mostly useless on shield-using fighters and two swords being the all-around best option for defenses if you have a Shield-casting buddy or can cast it yourself.
I think the numbers here show that you can still keep Feint and Deceptive Attack relevant with the more limited, fixed versions, and with a skill cap, so long as you also mind a maximum of DB in some fashion.
It's the Points, Stupid!
Not that anyone has been that blunt, but it's come up in the comments. And it's a broader game design question - if you're having issues of high skill and high defenses, it's because you set the point values of the campaign too high. It's the "sweet spot" idea - that certain games do well between levels X and Y, or points YYY and ZZZ, or some other limit. Outside of that, the idea goes, things break down . . . so set the game up on the bottom end of that range (or even below it) and work up . . . and stop playing or limit growth past the far end of it.
That'll work, for sure. But if GURPS is fine at 150 points and even 250 points but breaks down at 300, or 350, or 400, or some other number, then starting at 150 instead of 250 just puts the problem off. It's still coming - and in a point buy system, if it's a specific skill level, or defense level, or combination of factors that make it an issue, it can come as soon as someone puts their points to getting to that level.
If the problem is where you end up, I'd rather address where you end up than pull back the point where you start.
I think addressing the issues that cause a race up in skill, and having an idea of what skill levels are appropriate and building systems around that is the way to go. I think GURPS does perfectly fine at higher power levels . . . it's just that unlimited skill, therefore unlimited defenses, and the ability to freely leverage skill to smash defenses means you can get caught up in raising the numbers by not knowing when enough is enough. Defining "enough" and having rules that make it so may make it easier for the players and the GM to be on the same page in terms of what "enough" is.
Overall: I think I like what I see so far. Tomorrow I'll get into maximums and caps for DF in general.