XP Scaling for Level >10

Re: Co8 New Content: Moathouse Respawn

Perhaps the mod should feature a recommendation for an appropriate level cap? (e.g. 13-14 for average party size)

Part of the rationale is that the XP table caps at level 10, thus giving inappropriately large XP rewards for CL > 10.

I wouldn't force it or make it the default, though. E.g. solo players would probably need the higher levels.

 

Re: Co8 New Content: Moathouse Respawn

personally I'd love to see this change. I think a lot of great content is getting trivialised by the overpowered PCs you quickly end up with. Of course, you'd have to make those avatars a bit easier too - I think to a certain extent their massive difficulty encourages bloated characters in order to take them on, and that reduces the challenge on the way there. I'd be really excited to see this change.

 

Re: Co8 New Content: Moathouse Respawn

I really hope that you do reduce the XP reward for all of the post-Zuggtmoy Co8 new content, because, as Sitra Achara stated, the player is currently cheating by receiving more XP than the player is supposed to receive when the player's characters are above character level 10, so it would make the game more balanced and challenging (which are both good things). I will go to my grave pleading that Scather and Fragarach be either significantly nerfed or removed from the game because those 2 weapons are the main cause of imbalance in this computer game as it currently stands.

 

Re: Co8 New Content: Moathouse Respawn

Here's another idea - to simulate the effect of the XP reduction that should have been due to CL>10, one can reduce the creature's CR via script. A rough rule of thumb is that for every CL above 10, you should reduce the creature's CR by 0.7.

To see accurate XP rewards, use this:
http://www.d20srd.org/extras/d20encountercalculator/

The proposed method would be as such:

1. Insert a script into monsters' san_start_combat
2. The script checks for average party level
3. Script modifies the creature's CR (obj_f_npc_challenge_rating) accordingly

Example:
A Hextorite Assassin is a 15th level Rogue, currently with a -2 CR adjustment, i.e. total CR = 13.
Currently, any party whose level is >= 10 would receive a total of 9000 XP (before XP modifier, and dividing by group size) for killing the Assassin, no matter how much higher level they are.

What SHOULD be earned is:
Level 11: 6600
Level 12: 5400
Level 13: 3900
Level 14: 2800
Level 15: 2250
Level 16: 1600

So, instead, we modify the CR based on the party's level until we get close to the expected value:
Level 11: CR -> 12 ; 6000 XP (should be 6600)
Level 12: CR -> 12 ; 6000 XP (should be 5400)
Level 13: CR -> 11 ; 4500 XP (should be 3900)
Level 14: CR -> 10 ; 3000 XP (should be 2800)
Level 15: CR -> 9 ; 2000 XP (should be 2250)
Level 16: CR -> 9 ; 2000 XP (should be 1600)

You can see that there's a small inaccuracy in this method (usually about 10%), but it will largely get averaged out (sometimes you get more, sometimes less).

To counter tricking out the system by carrying along a couple of low level minions, I'd calculate party level according to the highest 50% or so. If someone is hellbent on tricking that out too, let them. The system should cover at least 90% of normal playthroughs.

In short, it's slightly ham-fisted, but unless someone manages a proper DLL hack, I think it's the best we got.

I can write the script and test it if there's interest.

 

I'm all for more difficulty, but as a cleric enthusiast, I enjoy access to my highest level spells and feats, which I might not be able to get if the xp is shoved down so much that we can't reach level 20 by the end of the game.

 

Progress update:

1. Verified that you can alter NPCs' CR modifier and that it properly affects XP given at the end of combat.
   *san_dying seems to be a good place for it. (it works even when NPC gets knocked unconscious and auto-dies when the fight ends)
2. Coded and tested average party level calculation that takes into account the top 50%
3. Coded and tested the math-y part that calculates which new CR is appropriate for a given party level and unaffected CR.

So basically what's left now is the book keeping part. i.e. how do you determine the unaffected CR.
Two options here:

1. Manually: Assign it for every NPC, be it explicitly through the function call argument, or some hand-coded lookup table.
2. Automatically: By calculating the CR according to NPC's CL + original CR modifier.
   This would necessitate using an internal flag to mark that the CR modifier has been altered. Alternately, you can rely on placing the script inside san_dying, given that usually you only die once (usually )

I'm leaning towards 2, jut wanted to hear if you think otherwise.

 

If you make them right, I think you're going to need a ton of new content to make sure there's enough XP to get a party of 8 to 20th level (without the obvious cheats at any rate).

 

Sounds good to me

 

Attached is the script (in combat_standard_routines.py) and example usage (in hextor_invader.py, plugged inside san_dying).

Note - script uses npc_set internal flag # 32.

 

So, being naturally wary of ToEE's vagaries, I decided to test it in a "real" scenario by applying it to the Drow. Of course it didn't work at first, but I found out why (after about an hour of hair pulling debugging)

It turns out that the py00381Drow.py script uses 4 spaces for indentation instead of Tab. When I inserted my Tab indented code, it refused to work at all (e.g. the drow death count didn't go up either). When I changed my script addition to 4 space indentation, it worked fine (new CR modification included)

I recommend that we conform to the Tab indentation style, because:
1. That's what Troika was using in the original scripts
2. Mixing the two styles breaks scripts, apparently.
3. It's less keystrokes...