Wednesday, November 20, 2013

How hard is that fight anyway?

I have been slugging away at my system for figuring out how monster difficulty should be rated in Heroes By Trade.  Doing so has given me more insight into why 4th edition DnD did monsters the way it did.  In 4th ed. monsters were all given a level that corresponded with player levels.  12th level players should fight 12th level monsters, for example.  Also monsters were categorized by type - minion, regular, elite, or solo.  In theory this makes for a very tight system for creating encounters where monsters always have stats that are appropriate for the players.  In practice it makes the game encounters feel very artificial because random brigands you meet at level 20 are so drastically more powerful than random brigands you meet at level 1.  The world isn't populated with interesting things at that point; instead everything the players fight just happens to have the perfect stat spread.

The thing is that the system described above is just so convenient.  I was trying to assign static numbers to monsters that would work across a variety of player levels and it was a nightmare.  Monsters that worked fine as part of a level 25 encounter were utterly lethal when faced as a solo encounter at level 10 because the players just couldn't hit them or couldn't get past their armour.  There simply wasn't a single number I could assign to them that adequately reflected their difficulty.  This is the genius of the 4th ed. model - you have two dimensions to the difficulty (level and minion-solo) and that makes it so easy to figure out because you always know what level the players are.

The major challenge involves the solo monster problem.  That is, when one side in a fight has a single HP pool it is at a massive advantage.  The other side cannot focus fire to remove threats and focus fire is the primary strategy of any successful group.  The flip side of the coin is that a single monster often cannot make use of its debuffs and is far more vulnerable to debuffs than a group.  This means that monsters that have debuffs as a primary focus can be more deadly in groups and monsters that have big stats as a focus are powerful solo.  So far in testing it has seemed to me that the advantage of having a single HP pool massively outweighs any debuff advantage in most cases.

I finally had to give up on making a system that would work across all party sizes and encounter types.  My final attempt works as follows:  Each player has a Encounter Strength (ES) of Level + 6 and those are added together to determine the ES of the party.  I then model a monster fighting two players and figure out at which level the fight is even just based on damage dealt.  Then I calculate the monster ES based on the ES of two players at the given level.  This is all very reasonable but then comes in the kludge - any monster encountered alone should be considered to have a ES 25% higher than listed because of the previously mentioned single HP pool factor.  I mashed a bunch of encounters through my simulator and it seems to work out all right but it continues to irk me that I have not found a simple, clean solution.  I suspect that unfortunately such a solution does not exist as otherwise some edition of DnD would probably be using it.

I don't like having monsters be so artificial as they are in 4th ed. even though it is very convenient for the GM.  I also don't want them to be as ridiculous as 3rd ed. or Pathfinder where 'appropriate' encounters ended on round 1 and it was easy to be unhittable or unmissable.  The old days of monsters being incredibly random and oscillating between lethal and trivial was entertaining at times but eventually unsatisfying and the new model defies immersion.  If the perfect system does not exist though I will lean towards old school because I am convinced I can do a better job of the numbers than Gygax did back in the day.


  1. One time we played a 3rd edition module and the DM made rules about only one non-spellcasting class, armor restrictions (people were wearing chainmail) and other things just to make the module not a cakewalk. I played a neutral cleric who was dominating a shadow. As they say in the business, gg.

  2. I'm having a bit of trouble following.

    For random encounters, I understand what you mean. It's fake to have all random encounters scale with the players. I assume it's the GM's responsibility to balance that - the dragon flies overhead in the distance vs. the small squad of orcs stumbling into your camp or setting up an ambush.

    Your discussion of solo encounters confuses me. A single monster vs. a group of players? And it's to the monster's advantage because they are either alive or dead vs. a group of monsters where you can slowly weaken them over time and take out the bigger threats early? My experience (15+ years ago) was that solo monsters were at a huge disadvantage because of the concentrated fire - a lich could only cast one spell while the adventures threatened with range, magical and melee every turn. But I suspect HBT handles this some other way?

    Overall, I'm surprised so much effort is put into this by D&D and yourself, which suggests I'm missing some aspect, or perhaps played differently. Using Dungeon magazine modules, part of the fun would be seeing if my 4th level adventurers could handle a dungeon that someone else thought was appropriate. I never changed the encounter toughness, but I guess I was able to eyeball the adventure and sense what was appropriate. It feels weird to try and turn it into a formula.

    Now I'm curious. What does everyone else do? Thinking back, did I always use other people's adventures (with my own plot hooks and occasional encounter in between) so this work was being done without my knowing it? Did I just use HD? Would my adventures have been better if properly scaled so they didn't have to resort to the messy "leave the dungeon for a week, come back healed and oh look, nothing has really changed"?

    You've got me doubting my entire D&D experience!

  3. @MattV I am going to make a full post in response to your comment. Replying in depth the way I want to in this stupid text box is being annoying.

  4. "In practice it makes the game encounters feel very artificial because random brigands you meet at level 20 are so drastically more powerful than random brigands you meet at level 1." -- I do not think this ever happened in any game I played.

    You didn't meet up with "random brigands" at level 20, unless they where gizyanki raiders on an astral skiff raiding a caravan destined for the city of brass containing a fingernail from a dead primodeal.

    "That is, when one side in a fight has a single HP pool it is at a massive advantage. The other side cannot focus fire to remove threats and focus fire is the primary strategy of any successful group."

    An approach to fix this is to make focus fire seriously suboptimal.

    Basically anyone who has not been attacked is far more dangerous -- or, every single attack causes targets to become "suppressed" or "defensive".

    You could imagine a system whereby if you aren't attacked between turns, you get a second attack action on your turn (normally, that attack action is lost to defending yourself). Now, you want to *not* focus fire, but rather spread fire out (area attacks are highly important, and numbers have a quality of their very own).

    Then, keep careful control of attack/defence bonuses. If damage and HP goes up linearly/affinely (linear with a constant factor) with level, then they should go up logarithmically (as the hyperbolic component they add to power is approximetally exponential, and logarithmic growth in a term that contributes exponetially to power adds linearly to power).

    Make the advantage of numbers (whatever it is -- doesn't have to be a second attack if not attacked) match up roughly with that attack/defence bonus scaling.

    Then, the add-up-each-sides-HD strategy might work for balance. A big badass 120 HD dragon might be a reasonable challenge for 5-6 20 HD PCs, or a single 120 HD PC. The dragon has serious focus fire capabilities, but if the dragon uses it the PCs who aren't focus fired upon double in effectiveness.

    We could even add "total defence" that you go into *between turns*, costing you your next turns attacks. Then a focus fire attempt can run into the "total defence" problem, where the efficiency of your attacks drops off sharply if you attempt overkill on one target.

    And, at the same time, the army of 120 1 HD orcs can stand up to the 20 HD PCs to some extent (so long as the PCs don't have area attacks capable of killing 120 targets at once).

    In short, bounded accuracy to prevent wiff, suppression effects and reactive total defence to make focus fire suboptimal, and linear HP/Damage could make encounter balance a matter of "adding up hit dice". Which would be awesome.

    We could then have "1st level characters" have 1 racial HD, 1 background HD, and 1 adventuring HD, with each contributing about the same amount of offence as well.

    - AfN