Class only skills ended in 3.5. Only 3.0 had "X" on the skill table. Cross class skills are absolutely important, because they guarantee niche protection.
Surely there are better ways--particularly when some so-called "niche protection" actively hinders gameplay, because if you
don't have those things, your party is now screwed.
Skill points are important becauseallow the players to feel rewarded in investing them and making choices, and avoid the flattening "if you are strong X, you are skilled X in Y".
Skill points are terrible because they outright punish anyone for picking bad choices, and continuously punish classes that are robbed of the skill points they need. Consider the Fighter, which gets a measly 2+Int mod per level. Even if you aren't dumping Int (which you have no reason to not do so
other than skill points), 3 skill points literally only gets you enough to Jump, Climb, and Swim...and that's it.
"Making choices" is an ephemeral illusion with skill points. If you aren't pumping those skills to maximum or near-maximum, because of the incredibly harsh and bad way the 3.x skill system does DCs, you're hosed. If you choose to spread around your skills, for most skill uses you could attempt you'll be so far behind that you have a less than 1/3 chance of passing a typical check. Failing 2/3 of the time
sucks, and actively discourages any effort at branching out or doing novel/quirky/divergent play.
Capping cross-class skills at only half ranks
guarantees those skills will suck more and more with time. Being at an effective -5 penalty at level 10 means you literally fail 25 percentage points more often than someone who doesn't have that. Good game design usually puts average player success rates around 60%, because humans are weird and don't actually feel things are evenly matched unless they win somewhat more often than they lose. Eating a -25% penalty means you go from 60% to 35%. In other words, you fail almost twice as often as you succeed,
exactly as I described above.
The
idea of skill points is great: reward organic growth! The
practice of skill points never worked--not in 3.x, and not in PF1e. Prestige classes just made this all the worse, because they used "put points in crappy skills" (like "take crappy feats") as a gatekeeping mechanism for being
allowed to do a cool, fun thing in the first place.
Now, addressing what I feel, respectfully, are misconceptions about feats. While I think that the base combat in 3e should allow more actions:
1) You can totally attempt a disarm without the feat.
But it will fail more often than it succeeds. Which is punishing that behavior. That's kind of the point. Failing more often than you succeed encourages you to never, ever do that thing.
More importantly, you can attempt maneuvers on flat-footed enemies (not everyone is going to have combat reflexes) with no repercussions.
I can count the number of times a 3.x DM has classified enemies as "flat-footed" on one hand.
Or on enemies which already used their AOO. Teamwork helps in that regard. So do miss chances and AC optimization, which are totally possible in 3e.
But now you're
requiring that other people have already optimized their characters before you can presume to do something not-very-optimized on yours. This logic may work for a single character, but it breaks down as soon as
everyone starts thinking that way. You necessarily get a bizarro world where everyone assumes everyone else has optimized, while not actually optimizing themselves, leading to, as stated, higher failure rates than success rates, which results in players choosing not to do that thing.
2) This one is trickier and I can completely understand the sentiment behind it. I would probably be completely OK with a crafting system based on skill ranks as an example (I mean ranks substitute for prerequisites like caster level). I have to admit tho that homebrewing I found myself re-introducing the crafting feats because of the above (certain type of player enjoys the choose/reward cycle) and because it would allow an hyper-production from the players that had this with no sacrifice. What I did is that for Arms and Armor etc I allowed skill ranks anyway in place of caster level to allow warriors to craft their own armor.
So...you agree that there is a problem, and have intentionally altered the game to try to address it? This seems to be entirely conceding the point. Feats act as gatekeepers, shutting out cool actions people could attempt. Skill points act as gatekeepers, shutting out cool actions people could attempt. You have had to change the rules to try to break this pattern, but one of your players dislikes it when things can be attempted without having to pass through the gatekeeper first.
Your 3) & 4) points are for 4e and 5e? In case they aren't, as follows
3) In 3e incantations do not require feats to use. I am in fact quite shocked by that. A similar activity, Use Magic Device, requires skill points. with an hybrid system as proposed above, it looks like only Artificers, Rogues, 3e Warlocks (they were a different thing) and Bards can use them, remember that in older editions only Thieves had a % roll for scrolls, the others had to use spells from their list or give up, unless I am mistaken.
I'm not sure what you're talking about here. Certainly, UMD is present--and essentially guaranteed to fail for anyone who isn't pumping in skill points and, preferably, getting other bonuses too. The DC to cast from a scroll is 20+CL. If you don't have the spell on your list, you have to roll
another UMD check to emulate a class feature. And since many people casting from scrolls won't have the necessary ability scores (e.g. a 5th level Wizard spell requires Int 15 to cast), you have
yet another UMD check for THAT, too.
So now, in order to cast this spell, you have to already have identified the spell (UMD, DC 25+spell level), you have to emulate the class feature (UMD, DC 20), you probably have to emulate the ability score (UMD, effective ability score = check - 15, so if you're trying to cast even a 1st level Wizard spell and you only have Int 10, you need a check of
at least 26), and you have to then actually do the casting from the scroll (UMD, DC 20+caster level of scroll or other spell completion item). That's 3 or 4 checks, all at DC of 20+. Let's say it's a 10th level character, Cha 12, Int 10, UMD is a cross-class skill. They've invested all 5 points they can, per your rules so it isn't at a harsh rate. Their total bonus is 1+5=6. To cast even a 1st-level spell, they must pass the following checks (ignoring the DC 26 check to identify the spell in the first place, since they have almost no hope of succeding at that):
1. UMD DC 20, emulate the spellcasting class feature of the Wizard class. They succeed only on 14+, aka 35% of the time.
2. UMD DC 26, emulate the Intelligence score required to cast a 1st level Wizard spell. They succeed only on a natural 20, aka 5% of the time.
3. UMD DC 20+1 = 21, actually cast the spell from the scroll. They succeed only on a 15+, aka 30% of the time.
Even if we ignore the stat-emulation requirement, this is still needing two checks that fail around twice as often as they succeed. In other words, the actual chance of success is 10.5%. Almost 90% of the time, this character with cross-class UMD fails to cast the spell from the scroll--and that was leaving out the "must crit" part in the middle, which would reduce this to only ~0.5% success rate, ~99.5% failure rate. Someone who actually builds for good UMD--skill in class, high Cha (18), max ranks--can have a fairly reasonable chance of passing these checks two checks. Total skill bonus 17 (10 ranks + 4 stat + 3 in-class bonus) makes DC 20 checks pass 85% of the time and DC 21 pass 80% of the time, though the stat-emulation one is still dicey at only 55% just to emulate an Intelligence score of 11! Again ignoring the stat emulation check, that's 68% success chance, which is quite reasonable.
Now consider the middle-of-the-road. Something like UMD is a cross-class skill but you have Cha 18, and you have some other random bonus that brings your total to +10. You succeed on a DC 20 check 55% of the time, 50% of the time for DC 21 (the Int-emulation check succeeds only 25% of the time). That means, again, you only succeed 27.5% of the time; you fail almost three times as often as you succeed,
not counting the stat emulation.
This is what I mean by the game ruthlessly punishing anyone who doesn't pump max ranks into things. If you
do get max or near-max ranks in most things and
do have the in-class bonus (PF-style) and
do have good stat bonus for it, then you won't really have a problem with the ridiculously high checks and insistence on rolling several times before an action can succeed. If you
don't have all of those things though, then even if you have a lot of them, your overall success rate is abysmal and gets
worse as you gain levels. Especially if you're trying to do any cross-class skill.
4) Is this again 5e? In 3e, anyone can charge unless restrained or exhausted or in similar conditions.
Yes, but you shouldn't unless you literally don't have any other targets, because you're giving up 1-3 bonus attacks to do so. The hegemony of the Full Attack weighs upon us all, and is a huge part of why 3.x and PF1e are such static, unmoving games. (Spheres of Might is quite popular specifically because it attempts to break this hegemony by rewarding actions that
aren't Full Attacks.)
This, incidentally, is why Pounce is such a powerful ability, and why optimizers will make almost any sacrifice necessary to get Pounce. It completely changes the game, making charging actually worthwhile in most circumstances.
I overall gave more combat feats to warriors, and options to improvise some like the PF1e Brawler, to compensate for such backtracking. I wish the original designers did the same.
Does this not mean you are admitting that feats, as they are, act as gatekeepers preventing or at least punishing actions unless the player has heavily invested into them? Like it really seems like you're conceding the core point here. You had to give out more feats in order to compensate for the fact that not having enough feats makes these actions not worth taking.
Mounted charge in 3e is similar. With the feats you deal triple damage with certain weapons like the lance. But without the feat, you still deal double damage! No feats needed. Same thing for "brace" weapons (say an halberd or pike) dealing double damage on the receiving end. This creates very thrilling fights. It also allows me to recall what I wrote above: weapon selection is very important, and while a lot of weapons are kind of meh, many are just different tools, good or excellent, or sometimes just decent, for a given job.
Oh,
mounted combat sucks for completely unrelated reasons--namely, that mounts are
heavily penalized in most indoor environments, strongly discouraging their use.
4e was, frankly, dramatically better than both 3e and 5e for weapons though. 4e weapons actually have meaningful properties. Feats interact with those properties and with weapon types (e.g. axe, light blade, polearm, etc.) in interesting and rewarding ways without obviating the underlying value of the weapons themselves. 5.5e is trying to bring some of this back with its "Mastery Properties," but I'm frankly skeptical that they'll make any meaningful impact. I'll need to see them in play before I'll believe it. I actually have a like 85% complete "build your own weapons" homebrew that would work just as well for 5e as it does for 4e, that brings in lots of fun properties for existing weapons, many directly taken from 4e, some inspired by other sources or purely of my own invention.
