when does CR/EL not work?

[Delta]

1) When modifying monsters, whether its with templates, added hit dice, class levels, or all three, the numbers do not always turn up accurate.

I'll also agree with that. The guidelines for upping CR for modified monsters were very sketchy -- frankly, there's no reason to think that such simple additions could possibly be useful (and obviously even less playtested than the original monster CR's in the first place).

Those kinds of modifications really should have been much more emphasized as tentative guidelines, not actual rules that some DMs seemed to delight in coming up with the most broken results, for some reason.

 

[Shades of Green]

I give out the full XP award if the party circumvents an encounter through stealth, interaction, or some other reasonable means that include a fair countermeasure on the party of the enemy. So you don't get XP for everything you teleport past, since you had essentially no risk of encountering those foes or being caught by them, but if you bluff a pair of street thugs into believing that your pistol is loaded and you can use that gun to perform delicate brain surgery on gnats at 50 paces and they take the better part of valor, you deserve the full award.

Yes, but the problem is with EL/CR as a way to balance encounters vs. the PCs, as a monster which is tough to fight is not necessarily as tough to sneak past or bluff.

 

[DragonLancer]

The CR system doesn't work, for all the reasons mentioned above. But it was a good try by the designers to find a way to balance encounters. I've generally stopped using it, and look to the monsters stats themselves to see whether I think it will be a suitable, weak or more powerful challenge for the PC's.

 

[(Psi)SeveredHead]

I find CR works as a nice guideline. For the numbers to really work, you have to have a party similar to the playtest characters, eg low point buy (25-28), magic item distributions by the rules, "the core four", etc. (For some reason, the playtest PCs in "Enemies and Allies" have some holes in their optimization; the halfling rogue, for instance, didn't take Weapon Finesse!)

Because most games don't fall exactly in that zone, in practice the numbers will be a bit off for each group. Fortunately, the DM can figure out just how far off with some experience with the group. IME, class composition makes the biggest difference (my experiences are with reasonable point buy; unreasonably high stats might make a bigger difference). For instance, I've twice been in groups without primary spellcasters, and that's a pain if you're facing a big brute like a CR 8 hydra or a troll barbarian. A mage could shut down the former with spells like Fear and the latter with spells like Otiluke's Resilient Sphere.

Combat situation (eg it's dark and crowded) affect the EL, and this isn't the fault of the CR system. PCs can also set up situations to make a battle easier for them, just like the GM can set up a situation to make things harder for the PCs. That's fine.


I have to say, I don't like the "four weak encounters per day" paradigm; I hope that changes in 4e.

 

[freyar]

Sounds like basically what I expected, I guess! Works as a guideline, some monsters and templates are off, adding hit dice isn't always accurate, and care is needed working with unusual situations (environmental conditions, party composition, etc). CR and EL have been invaluable for me as a relatively inexperienced DM, so I thought I'd ask around.

Delta, do you have the results of those brute fights tabulated somewhere? Might be interested to look at...

 

[Hussar]

We explored factors that affected EL fairly well in this thread. Once you become cognizant of these factors, EL works pretty well IMO. Yes, you have to recognize that your six man party is more powerful than a 4 man party. How much more? Well, that's a bit of voodoo calculation, but, generally about 1, maybe 2 levels. (I peg it at +3/4 per PC)

If you pay attention to the elements listed in that thread, I think you will find that EL works more often than it doesn't.

The thing to remember though, more than anything else, is that CR and EL cannot account for luck. A string of great rolls by the DM WILL kill the PC more often than not, regardless of the CR of the baddie. EL is not a safety net. Fighting an EL less than par is not a guarantee of success. However, it does let you guestimate with some accuracy beforehand.

 

[Philotomy Jurament]

I didn't find CR/EL to be significantly better or worse than the old HD + eyeball-the-special-abilities method of gauging such things. CR/EL might have worked better, for me, if my game was closer to the default assumptions (4 PCs, et cetera), but it never was , so I always had to throw a healthy dose of the "voodoo" into the calculations. In the end, I planned/balanced 3E encounters the same way I do it for other versions of D&D: look at the general power-level (CR or HD), and go by feel and experience, taking the special abilities into account.

 

[Phlebas]

I play a heavily houseruled game, and i still find the CR guidelines to be pretty accurate given the inherent variation you'll get in terrain, party composition etc etc etc

most encounters i've seen go horribly wrong were either due to

A - bad luck on the dice rolls - several 1's in a row as opposed to several 20's can turn encounters easy / scary in an instant (eg critical hit by a Brute monster, failed save by the cleric early in an encounter against undead)

B - Really, really bad decisions by the players ("What do you mean the potion of neutralise poison i had in my bag would have kept the thief alive?" "I try to jump over the minotaur zombie with the great axe!" "Everyone else is running? OK, I fire one last blast and then try to run")

Nothing a CR system can do with any of those. The only caveat i'd put in is that remember with all the possible variations the CR's are only guidelines

 

[Delta]

Delta, do you have the results of those brute fights tabulated somewhere? Might be interested to look at...

Well, I use the program anytime I'm evaluating a new brute-type monster (so the results are scattered all over the place, wherever I have monster writeups, and usually they're new monsters no one is familiar with). That said, attached is one analysis I did the other night of melee vs. ranged differences, for example -- melee fights seem to go up by about x2 --> +2 EL (because of limited combatants), while ranged fights almost go up by x2 --> x2 (because presumably all the enemies are firing all the time).

I find that 3 1st-level human warriors are evenly matched against a 1st-level NPC fighter (so CR 1/3?), so are 3 goblins (so CR 1/3, same as book), while a standard hill giant is evenly matched against a fighter of level 8 or 11 (ranged vs. melee; book says CR 7). The latter shows that you do have to degrade the output numbers at higher levels, because the simple combat brutes don't work so well against high-powered magic or flying characters, etc. (I've got a table written in my MM giants entry that says you can take the melee result and subtract 5, or the ranged result and subtract 2, and basically get the book CR for those giants.)

If I had more time I'd playtest some of FrankTheDM's really powerful creatures, incorporate an unreachable flying wizard & cleric, and see at what point those creatures kill 25% of the party (indicating, I guess, the proper CR). Again, I use the program as a starting guess, and then ideally run a couple of playtests with a stock NPC party of that level and see if it needs nudging up or down.

 

[Mustrum_Ridcully]

These are my experience with the CR / EL system:

1) Custom Monsters, Monster Advancement
Custom made monsters and advanced monsters are hard to gauge into the system. This also applies to custom monsters from modules our group played. The reason might be the lack of playtesting, but sometimes I also think it is also because people that designed the monsters didn't understand the rules well enough to see the flaws in their design.
To get this right, you _really_ need more eloberate rules or at least guidelines, especially if you know that most module designers and later MM writers don't have the resources to make such extensive playtests as in the base rulework. The experience made when play testing the rules the first time must be turned into a rule system and/or guidelines.

You might say that there are a lot of good guidelines described in the 3.5 MM and later MMs, but there are still limits.
Example: What spell (spell level) and caster level is appropriate for a monster of CR X, if you want it to have spell-like abilities? Some spells are easy (no Fireball CL 5 at CR 1. But what about Blasphemy at CL 20 with CR 15? Why is this different from, say, Cone of Cold?)

2) NPC Challenge and 1 on 4 fights.
The CR/EL rules for NPCs with levels don't work out well, even with non-associated class levels for monster.
Especially single-classed non-monstrous NPCs often aren't worth their CR.
You can't put a Barbarian 14 against a party with average level 10 and expect it to become a tough fight (even if no spellcasters are around). 4 actions per round vs one mean that this encounter isn't really worth more than CR 10. It changes a lot if you add some helping NPCs to get the same EL, but this is not what the system tells you in the face - and what is if you really want them to fight just one creature? Basically, the only answer here is use the same as was done for the Dragons - say your enemy is CR X, but it's actually CR X+3 (and hope that you didn't forget that some high level spells are based on level difference and might turn the whole thing in TPK)

3)
The basic assumption of the DMG that a party will mostly fight enemies with a CR around its party level didn' work out well, either. Maybe it is just because of the omnipresent Wands of CLWs, but unless the characters have to get through a lot of these encounters in a short time frame, they aren't particularly exciting, and thus you typically end up with fewer encounters with CR ranges of L +2 to PL +4, sometimes even beyond. A side effect of this seems to be that a lot of the newer modules or the dragon adventure paths (which so far I have all liked) are pretty deadly (not neccessarily TPK deadly, but deadly nonetheless) which is not always bad, but there are limits...)