What's wrong with high-level/epic play?

[Upper_Krust]

Howdy bykov!

Out of curiosity, is there an ETA on v.6?

Thanks!

I may actually release that in a few weeks although I was planning on having it as an appendix in my last 3.5 book Godsend.

I may just give away the v6 bit for free. I should have probably released it ages ago. Its not a massive difference from v5 although I do think the changes are important.

 

[Kerrick]

I may actually release that in a few weeks although I was planning on having it as an appendix in my last 3.5 book Godsend.

This means we'll see it sometime around April.

I may just give away the v6 bit for free. I should have probably released it ages ago. Its not a massive difference from v5 although I do think the changes are important.

You should. All the others were free downloads; I didn't see why you were holding this one back for inclusion into Godsend.

 

[Runestar]

With regards to v5, I still find the explanations very confusing, especially the cr/EL table. Why is it that for monsters of up to cr14-15, the EL is higher than the cr? What does it mean that a cr6 monster is considered an EL11 fight?

Can someone clarify this for me once and for all? Thanks in advance for your help.

 

[Beginning of the End]

It was, I think, EN Publishing's Four-Color to Fantasy that introduced the concept of "hyper-rolls."

A hyper-roll is where, on a d20 roll, for every full +20 of bonuses you have, you remove +20 from the total bonus and add in another d20 to be rolled. For example, a fighter with a +42 (total) bonus to his attack roll would instead just have a +2 bonus and roll 3d20. This helps to keep results relatively lower, and makes the dice stay relevant at higher levels. It might be a good idea to introduce to epic-level play.

That's statistically busted. A character with +19 (1d20 + 19) will get higher results on average than a character with a +20 (2d20 + 0).

Average of 1d20 + 19 = 10.5 + 19 = 29.5
Average of 2d20 + 0 = 10.5 + 10.5 = 21

And the bell-curve on 2d20 will depress the high- and low-end results, as well.

There are basically three problems with high-level play:

(1) Complexity of prep due to the multitude of abilities. While some of these problems can probably be ameliorated in various ways, I don't think there's any clear-cut solution: Part of the appeal of playing at those high levels is to have that high power level.

4th Edition works around this problem by simply narrowing the focus. You don't really get more powerful types of abilities -- you just get the same abilities with bigger bonuses. This solution can work in 3rd Edition, too, but in a very real and meaningful sense it's a cheat. You're not really offering high-level play, you're just offering bigger numbers.

(2) Save-or-die effects. The link also has the solution I play under. These are independent problem, but it's exacerbated by...

(3) The limited range of the d20. If you allow a differentiation of ability, it doesn't matter what randomizer you use -- the differentiation of ability will eventually out-strip the range of the randomizer.

I have yet to see any good solution for this. The ELH tries to simply flatten out everyone's progression (but does so in a way that causes all kinds of wacky problems). Another way might be to use a "mega-roll" system for characters with sufficiently high bonuses (where the type of die switches out), but this creates noticeable and meaningful points of discontinuity.

I suppose you might be able to mute the discontinuities by instead switching to multiple dice. So when characters reach level 15 everyone starts rolling 2d20; and when characters reach level 30 everyone starts rolling 3d20. (Those numbers are pulled out of my ass.) The bell curve would mute the discontinuity, but you'd still be looking at some weird probability behavior.

4th Edition obviously fixes this by simply preventing characters from being meaningfully differentiated from each other. Everyone is basically Doc Savage and advances in all areas of skill and ability in a largely lock-step fashion. (Although if you only did it on attack rolls and saving throws -- which are the only real problem areas anyway -- the probability oddities might not be that bad.)

It should be noted that the extant epic-level rules of the ELH are broken in many other ways while doing little to actually fix these fundamental problems.

 

[Kerrick]

There are basically three problems with high-level play:

(1) Complexity of prep due to the multitude of abilities. While some of these problems can probably be ameliorated in various ways, I don't think there's any clear-cut solution: Part of the appeal of playing at those high levels is to have that high power level.

That is a difficult problem to find a solution for.

(2) Save-or-die effects. The link also has the solution I play under.

Pathfinder does something similar, except SoD spells like FoD deal hit point damage. Reaction to that "fix" is rather mixed - some people like it, and some think it's goofy. The rule I use is that a failed save = reduced to -1d4, -1d6, or -1d8 (depending on spell level), and then the alt death and dying rules kick in. It hasn't been playtested, though. I also boosted the level on all save-or-die spells (the real ones) by 1 - none of them should be below L6, IMO.

I also went with ongoing Dex damage for flesh to stone, because it was a cool idea (not my own).

These are independent problem, but it's exacerbated by...

(3) The limited range of the d20. If you allow a differentiation of ability, it doesn't matter what randomizer you use -- the differentiation of ability will eventually out-strip the range of the randomizer.

I came up with The Rule of Three, but it received a lot of negative feedback (mostly "Why bother when I can just scale the DCs to the PCs' scores?"). *shrug*

I have yet to see any good solution for this. The ELH tries to simply flatten out everyone's progression (but does so in a way that causes all kinds of wacky problems).

The problem is, they have to flatten out at some point - the research I've done indicates that BAB advances faster than AC, so without a cap of some sort, you'll be rolling auto-hits. I did, however, figure out a solution: the cap doesn't kick in until your BAB hits +20, no matter what your level is. This eliminates the wonky problems and still ensures that BAB doesn't outstrip AC.

 

[Noumenon]

Anot

another way might be to use a "mega-roll" system for characters with sufficiently high bonuses (where the type of die switches out), but this creates noticeable and meaningful points of discontinuity.

The 1d20+10 = 1d20 + (1d10 + 5) hyper-roll I proposed upthread gives a character with a +9 a 75% chance to hit a DC 15 and a 5% chance to hit a DC 29, while a character with a +10 gets an 82% chance to hit DC 15 and a 14% chance to hit a DC 29 -- it's not awful.

All we need is to make "take 20" mean "take 20+5+5" so he won't try to hit DC 35, or perhaps just cap him at the +30 a normal fifth-level character with a +10 could hit.

 

[Upper_Krust]

Hey Kerrick mate!

This means we'll see it sometime around April.

Given my track record thats probably accurate.

You should. All the others were free downloads; I didn't see why you were holding this one back for inclusion into Godsend.

Back in the day I had more time to do stuff 'for free'. Now I barely have the time to do stuff 'for pay'.

Added to which it forces me to delve back into a system (3E) I no longer like.

 

[Upper_Krust]

Howdy Runestar!

With regards to v5, I still find the explanations very confusing, especially the cr/EL table. Why is it that for monsters of up to cr14-15, the EL is higher than the cr? What does it mean that a cr6 monster is considered an EL11 fight?

Can someone clarify this for me once and for all? Thanks in advance for your help.

Okay, its been a while since I went over it, but basically forget about EL until after you pick.

The number of PCs and their average level.

Lets say you have one PC of 10th-level. A 50/50 challenge for him will be a CR 10-11 Monster.

Lets say you have four PCs of 10th-level fighting the same CR 10-11 monster. With four PCs the EL of the CR 10-11 monster is reduced by 4. So it becomes a moderate challenge using 25% of their resources.

So the basic thing to remember is that EL (in this system) has nothing to do with the level of your PCs.

EL is a method for determining what the relative difference will be for your PCs if you add/subtract either monsters or PCs to the equation.

So.

1. Take the average PC level.

Lets say they average 50th-level.

2. That determines the average CR range.

The range (Level 50 falls under) is CR 48-55

3. On the Table what is the EL next to that CR.

CR 48-55 = EL 23

4. Then modify that EL by the number of PCs

Lets say you have 7 PCs, that means you need monsters of EL +5 (EL 28 in this case) to give the PCs a 50/50 challenge, EL 26 would be a tough challenge, EL 24 a moderate challenge and so on.

For a single monster we look at EL 28 on the Table and see that it coresponds to CR 112-127.

For multiple monsters of the same level we modify EL just like we did for the PCs. For instance a 50/50 challenge for those PCs would be between 32-47 Balors. Normal Balor = CR 20 = EL 18. 32-47 Balors are +10 EL, which gives us EL 28.

However, while (I hope) that explains v5, I should point out that all those calculations are incorrect where (the more accurate) v6 is concerned - basically because I made a flaw in the relationship between CR and EL. The initial equation was CR x2 = EL +4 (and thus CR x1.5 = EL +2).

But the correct equation is more like CR x2 = EL +8 (and thus CR x1.5 = EL +4).

 

[phloog]

I'm into math, and this idea of replacing each +20 with a new d20 instead is intriguing to me, because it keeps you rolling dice (which is one of the basic elements of traditional D&D).

What I'm wondering, though, is related to the complaint that a +19 is better than a +21 under those rules.

Has anyone ever tried replacing each +11 with a d20 roll? or some other method using dice in place of +X?

It makes it far more 'swingy' in terms of potential benefits, since your +11 could get you a 20 on the die. But on average you're going to be rolling 10.5.

A big issue for me with this approach is not the average, but the fact that each value is equally likely (assuming you aren't using my Player-Hated Golden Die of Mostly 20s).

Another idea that seems to appeal to me as an old Shadowrun player, but would take things a bit away from traditional D&D and might be too odd, is to introduce something more tightly distributed into the mix...like each +10 suddenly becomes +3d6...now you still average 10.5...you use +10 just because it's rounder and easier to remember, and it's less likely you get greatly penalized or helped...it's just still possible...there's a 1 in 216 chance of only adding 3, and a 1 in 216 chance of adding 18, but about 60% of the time you'll add 8 to 12.

With a +9, you'd get +9, with a +11 you'd get +3d6 +1, and you'd TEND to be better off...you MIGHT be lower than the +9, but you get the added benefit of the POSSIBILITY of much higher (or lower, true).

With the +20 => +d20, 95% of the time you're adding less than 20. That seems to be the INTENT, but it leads to this goofiness...any fix that replaces +X with +1dX will have this "Bonus just over the threshold is worse than just under" issue.

With this alternative idea, you still have a lot of the issue of the character 'cannot miss', but it's not as certain.

Just some quick Tuesday thoughts - - figured someone has to have tried / refuted this idea already, and was looking for WHY it is bad.