when does CR/EL not work?

[Mustrum_Ridcully]

IMO, CR/EL never works.

Anytime situational factors affect encounter difficulty, or noncombat encounters play a significant part in the game, or PCs contribute unequally, I think the system fails. Anytime there are different power levels among PCs or enemies, the math gets hard.

I just ad hoc it.

I would phrase that differently. It doesn't work as the lone arbiter of encounter difficulty. It gives you a baseline to use, and than you have to adjust, based on your party composition and personal experience. That's the point were the CR/EL system fails - it doesn't give you anything concrete to make the adjustments.

 

[S'mon]

I find it a fairly poor system. In particular it doesn't address NPCs at all. NPCs 'real' threat level (CR) is usually far lower than their class level, the only exception is pre-buffed Wizards. NPC Fighters and Rogues are far weaker. Among monsters, the lack of any system for CR means listed CRs are very erratic; melee brute monsters tend to have CRs too low compared to spellcasters. Dragon CRs are just ridiculous.

 

[Jhaelen]

All in all it works surprisingly well in most cases. It breaks down at the very beginning (CR below 2) and when using the 2xCR = EL+2 formula too often (with encounters of 12+ creatures).
There are also some oddities with mixed CR groups, where altering the order of adding the creatures results in different ELs.

What doesn't work well at all are the different methods of advancement: by HD, template or class. It's just not possible to give a formula that works equally well for the incredible variety of creatures that exist.

CR is also problematic because there are no good recommendations on how to figure out an appropriate CR except by comparing a monster to existing ones. Something similar to the 'DM's Helper' Excel sheet which gives average values for HP, AC, attack & damage, etc. is incredibly useful in that regard.

Finally some things are frequently forgotten when figuring out CR / EL: party size, action points and point-buy method used (if any). All of these have an effect which has to be considered.

So, to sum it up:
CR / EL works well if sticking to standard parties / monsters. For anything beyond that extra effort is required for the method to stay somewhat useful.

I'm not convinced, btw. that the 4th.ed. approach to return to adding up xp values to create 'appropriate' encounters will be better in any way. You'll run into exactly the same problems as in the current edition.

 

[Sunderstone]

I say 'general' so big because CR/EL goes with the assumption of 4 PCs - Fighter, cleric, wizard, rogue, with appropriate treasure, and standard spells, skillsets, and abilities(most people forget these last three).

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.

I am totally in agreement with these two comments.
Its a guideline not the concrete system the naysayers want it to be. Die rolls can make or break some encounters.
Also splatbooks, imho, went a long way to destroy CR/EL rules. As Goblyn stated the CR/EL rules take into account a balanced 4 PC party with typical abilities (again its a guideline).


Delving into the above a little deepr....
I think most Player's/DM's who are so quick to accept these splatbooks dont realize what balance issues these splat classes can cause. Evidence of this is in just about every forum, and more so in "My party got TPK'ed" threads. Take a look as to what their party was made of before the TPK. Its almost always missing an important core role like "Cleric" (for main healing) or "Rogue" for trapfinding and are replaced by the more hip classes like Warmages, Scouts etc. The other party makeup of the "My party got TPK'ed" crowd does contain the critical classes but multiclassed too early with another class or in some cases multiple classes.

To tie up my point without deviating too much is that the CR/EL rules can seem to be useless when most people these days seem to run non-standard classes (and in some cases highly unbalanced) or an unbalanced group lacking a critical "Core" role like healer, trapfinder etc.

This is why Im against splatbooks in general. Its great to have nice options, but most dont really take into account game balance effects on CR/ELs of monsters.
WotC seems to throw out these half-thought out books for the masses without considering how they seem to screw up their own core rules (in this case... CR/ELs), then they tell us the game is broken and wait for 4th edition.

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.

Good Point. This is one way that "experienced" DMs can fix the problem, and definately helpful for the splatbook using crowd.