D&D 3.5: You Must Be This Tall to Ride the Ride

[Ziegander]

Say what you will about him/her, Basket Burner is right about a few things regarding optimization, viable strategies, and monsters in D&D 3.5. The monsters are over there. They do have stock statistics, and without a baseline degree of effectiveness, PCs either will or will not be viable when we discuss D&D using in the "common" tongue. Sure, all games are played with some form of houserules. But, since we cannot know what houserules are being used, the only way we can discuss D&D together is by discussing it as written.

What one DM does with his D&D game and how he makes the game fun for his players is irrelevant to this discussion. Arguing that tanking is viable in one DMs game because he does "X, Y, and Z" to make it so, is akin to having an argument with someone in one language but trying to prove your point using only words and terms from a language that the other person doesn't speak. It's pointless and doesn't work.

So, yes, objectively, there is a determinable "height" on the "You Must Be This Tall to Ride the Ride" scale for D&D. Which is to say that, when measured against the capabilities of Team Monster, there are numbers that Team PC must be able to produce with their passive scores and their die rolls in order to survive for any extended period of time.

In this thread, I would like to examine exactly how tall one must be to ride the ride that is D&D 3.5. To do so, I will be using the following tools:

The Hypertext d20 SRD - The actual, factual statistics of stock, D&D monsters can be found there, as can all other relevant combat rules, and other important bits of information. We can literally look at the values we need to meet and/or beat in order to play the game.

D&D (3.5) Monster Finder - Using this, we can filter out sources of monsters, and sort them by CR. This should prove helpful in determining the numbers that PCs must produce in order to play the game at every level of the game.

As far as the criteria that should be used to judge the relative "height" of PCs, well, I could come up with my own, but I think it would be much more useful for everyone if we agreed to a set of criteria as a community.

What this thread should show, eventually, if we can avoid baiting and flaming each other for long enough, is that there is at the very least a baseline of PC competency require to play the game at all. Beyond that, it should show that below this baseline, playing is ostensibly impossible without houserules or other DM intervention. It's possible that this thread can show that for certain values over the baseline, playing becomes unrewarding for being too easy, predictable, and/or boring.

If we're lucky, this thread might even provide insight into something like the fabled "sweet spot" of optimization, which is sure to be subjective between every single D&D player and DM ever, wherein the PCs are able to play the game and be effective enough to not die and yet remain challenged and engaged in the game such that they, gasp, have fun.

[Ziegander]

Reserved for data.

[Ziegander]

Reserved for data.

[Ziegander]

Now that I've posted this, I'm sure there's been some exploration of this exact subject that's been done, probably a few times before by more qualified statisticians than I. Does anyone know of any similar threads started in the past, either here or on other forums, and if so, might you link to them, so that we can examine any material that's already been presented?

[caelic]

I've seen people tackle the issue before, but they tend to get bogged down in the sheer number of variables.  I think you're going to need to define your terms and categories of performance very precisely.  Start with the basics and work your way outward.

[Zionpopsickle]

The very difficult issue with these kinds of projects is that you are attempting to set out criterially generated guidelines, but not all these criteria necessarily arise axiomatically from objective goals but can be based upon complex premises that are themselves based upon other criterial goals.

In less technical terms (which will lack some of the precision), what you are attempting to set out as guidelines may themselves be dependent upon other guidelines, each which may require different amounts of both adherence and weighting.  The result is that there can be hidden synergistic effects which change the weighting and level of deviation that is acceptable from guideline to guideline based upon how well a character fulfills other, semi-related guidelines.

These kinds of synergistic effects can be easily seen in how they arise between miss chances, mirror image and AC, in which the interaction of a monsters AB and number of attacks can outweigh the overall maximal damage or weighted average damage such that given a large sample of even CR monsters the most threatening monster to said character is not the monster that is objectively the most threatening outside of the context presented.

This is not to say that this is a fools errand, as compilation of data and discussion of the philosophy of how we use this data to come up with relevant criteria for character making is a valuable intellectual exercise that should give a lot of insight into the character making process.

I just think that it should be acknowledged that the likelyhood of coming up with a single clear objective picture is probably pretty slim just because there are numerous variables that have very sensitive interactions.  To comically extend your OP analogy, "Yes, you do have to be this tall to ride, but only on Tuesdays if you are wearing a hat.  On Wednsdays that it is raining there is instead a two drink minimum."

[Jackinthegreen]

Something to note in this is that we can't use examples from any campaign unless it has been proven that there were no house rules in play for the encounter in question, including the DM playing monsters to the proper degree.  This requires full details on the PCs and NPCs such as stat blocks and grid positioning throughout the fight.  One cannot make a properly informed judgment without that information.  The examples I've seen so far are lacking those, hence why I disregard them.

As far as creating such encounters to test, we'd have to agree on certain standards like stat generation.  Is 28PB the norm? Is 32?  What resources are available to each side?  Can each side actually use those resources effectively?

[Childe]

How do commoner cattle stampedes fit into this? I'm not sure I understand how you seek to set the height bar.

[DonQuixote]

I have to say that I'm intrigued.  As a DM, I fall pretty firmly into the "do X, Y, and Z to make everyone viable" camp, but that's because I have a fairly relaxed group of players.  The wizard/initiate of the Sevenfold Veil in my current party can be put in a group with a monk without making the monk irrelevant because he, as a player, respect's the monk-player's right to be a monk.  Thus, I can accommodate the monk without worrying about a sudden prismatic cyclone taking everything out.

However, as a player?  Dear God, I would never stat up a monk, especially not in a Core game.  There are a hell of a lot of non-viable choices in D&D 3.5--most of them, hilariously, in the Player's Handbook.

To address Zionpopsickle's point, I think that the best solution would be to actually set up a range of "ideals" for each level.  Pulling the metaphor in a different direction, while you must be above a certain height to enter the theme park, each ride itself has a different cutoff height.  Thus, you might have different categories--dragons alone should be set off by themselves, and you might subdivide further into types of encounter.  This might help players with particular kinds of DMs--"Oh, I'm in this DM's game and she likes to send undead at us.  I should be looking at section A." or "Hrmmm, this DM likes to throw oozes, so I should be looking at table B."

On the flip side, DMs could use this resource to figure out what kind of challenges they could use to allow different players to shine.  If one player's type of build does very well against Challenge One, but not quite as well against Challenge Two, Challenge Two can be used to give the spotlight to players of other builds.

Moreover, a scale of this sort might let us know when builds are too tall for a ride--if you're taking a single round to blow through the combat that was supposed to take half the session, you're achieving the roleplaying equivalent of thwacking your head on the rails above you.  You shouldn't bring Pun-Pun to a Cloaker fight, but--sometimes--it's less obvious than that.

I'm imagining something like what an actual sweet spot looks like.



Sniped:

As far as creating such encounters to test, we'd have to agree on certain standards like stat generation.  Is 28PB the norm? Is 32?  What resources are available to each side?  Can each side actually use those resources effectively?

I see no reason that separate data sets couldn't be generated for the "standard" point buy values listed in the Dungeon Master's Guide.


Sniped AGAIN:

How do commoner cattle stampedes fit into this? I'm not sure I understand how you seek to set the height bar.

While it does start to dip its toe in the issue of houserules vs. "as-written," I think that projects like this are allowed to invoke the rule of Common Sense every now and then.  If it's something that most people wouldn't allow in a serious game, I don't think that it needs to be a part of the data.

[Childe]

So, if we're disallowing the cattle stampede, where are we setting the "You Must Be This Short To Enter" bar that precedes the Ride bar? Because this sounds awfully arbitrary if we're asking what the minimum is but disallowing things that clearly aren't Pun Pun but are still beyond first expectations (a cursory glance at the game, without system mastery).

[dman11235]

Honestly, I think that most characters will be able to handle stock CR monsters played with stock tactics.  When I did the Terrasque test a few years ago, we found that core only, everyone pulled their own weight relatively well.  The Wizard was necessary to actually defeat it, and ended up Plane Shifting it to the Plane of Chaos or something.  The bar ends up being incredibly "low", at least comparatively.

Ultimately, while this thread will almost certainly prove to be useful, I believe that use will not be finding the point where a character is unviable in a campaign without DM intervention.  You do have to define "viability", however.  I'm assuming the definition of "survive 4 encounters a day using resources" found in the DMG.  There's also the question of inter-party help, and whether or not the weaker classes are actually doing well or if it's the strong classes making that possible.  Only thing I have on that is the Terrasqu test, and annecdotes of casters helping other in campaigns.

[Ziegander]

I definitely agree that there are many complex situations and that attempting to arrive at a singular conclusion will be difficult at best. It is my hope that this thread generate many different conclusions that are all clearly determined and presented. It should be able to be used by DMs and players as a resource, and thus any collected information should be both factual and well-displayed.

When I agree with BB on the concept that PCs must be "this tall to ride", I don't mean that in the sense that only a certain build is capable of playing the game. The necessary "height" that I hope we can ascertain for PC characters is not a singular value, and perhaps not always a numerical one either. I recognize that the necessary "height" actually encompasses many, many variables, multiple different attributes and statistics, and complex interactions between gameplay and game rules.

But I posit that, if we agree to discard all subjective data, there is sufficient objective data within the game rules as written to evaluate the statistics of Team Monster and the statistics of prospective PCs. But we don't necessarily need to build player characters and play through encounters or game modules in order to do this. All we really need to do is generate numbers and consult the rules that are already given. This is not to say that creating characters has no place here. At some point it may prove useful to stat up some "generic" characters to show where they fall on the various "height" bars.

As far as setting the height bar, the most basic way to establish the baseline would be through the mechanism of "are you able to kill the enemies or not?" And the, the baseline would be the lowest set of statistics required. I know that's an overly simplified metric, but I think it's entirely mathematically possible to determine that baseline that would be required to survive X encounters of average monster statistics of a given CR. This would give us a starting point for the discussion, and enable us to do much deeper analysis.

For example, we can average the statistics for things throughout monsters of every CR, and we can even determine % chances per round that an "average" CR X monster produces some spell or special effect. From there we can generate the lowest possible numbers that PCs will need to survive however many encounters are required to gain a level. After that's established we can discuss the different ways the PCs can actually arrive at those numbers, which will presumably be myriad, and that's a good thing.

EDIT: And, no, I don't think that would be any sort of "ideal conclusion" to the thread. The whole, "average stats encounter" stuff is just a way to get the ball rolling, get us looking at things from the correct perspective, and begin the discussion. Like I said earlier, from the vantage point this would set us at, we can begin much deeper analysis of the rules and the statistics.

[Jackinthegreen]

An average for monster statistics has its place, but to set a minimum bar we should start with monsters that are weak within their CR.  We'll likely find several that should actually be lower than what they are, or others that should be raised of course.

[Ziegander]

An average for monster statistics has its place, but to set a minimum bar we should start with monsters that are weak within their CR.  We'll likely find several that should actually be lower than what they are, or others that should be raised of course.

Well, that's kind of venturing into deeper analysis. Just determining what monsters are "weak" for their CR seems like it would be a daunting task. The only way to determine what monsters are weak would be to first establish an average baseline to compare them to. Also, all that would do is give us a way to say, "you can, theoretically, play the game if you can beat these monsters and only these monsters, but that is only if your DM goes out of his way to challenge you with only monsters from this list." Which makes that data fairly subjective.

[Zionpopsickle]

I definitely agree that there are many complex situations and that attempting to arrive at a singular conclusion will be difficult at best. It is my hope that this thread generate many different conclusions that are all clearly determined and presented. It should be able to be used by DMs and players as a resource, and thus any collected information should be both factual and well-displayed.

When I agree with BB on the concept that PCs must be "this tall to ride", I don't mean that in the sense that only a certain build is capable of playing the game. The necessary "height" that I hope we can ascertain for PC characters is not a singular value, and perhaps not always a numerical one either. I recognize that the necessary "height" actually encompasses many, many variables, multiple different attributes and statistics, and complex interactions between gameplay and game rules.

But I posit that, if we agree to discard all subjective data, there is sufficient objective data within the game rules as written to evaluate the statistics of Team Monster and the statistics of prospective PCs. But we don't necessarily need to build player characters and play through encounters or game modules in order to do this. All we really need to do is generate numbers and consult the rules that are already given. This is not to say that creating characters has no place here. At some point it may prove useful to stat up some "generic" characters to show where they fall on the various "height" bars.

As far as setting the height bar, the most basic way to establish the baseline would be through the mechanism of "are you able to kill the enemies or not?" And the, the baseline would be the lowest set of statistics required. I know that's an overly simplified metric, but I think it's entirely mathematically possible to determine that baseline that would be required to survive X encounters of average monster statistics of a given CR. This would give us a starting point for the discussion, and enable us to do much deeper analysis.

For example, we can average the statistics for things throughout monsters of every CR, and we can even determine % chances per round that an "average" CR X monster produces some spell or special effect. From there we can generate the lowest possible numbers that PCs will need to survive however many encounters are required to gain a level. After that's established we can discuss the different ways the PCs can actually arrive at those numbers, which will presumably be myriad, and that's a good thing.

EDIT: And, no, I don't think that would be any sort of "ideal conclusion" to the thread. The whole, "average stats encounter" stuff is just a way to get the ball rolling, get us looking at things from the correct perspective, and begin the discussion. Like I said earlier, from the vantage point this would set us at, we can begin much deeper analysis of the rules and the statistics.

I can whole heartedly agree with this post Ziegander.  I think that it is a very healthy methodology to assume that there will be a number of viable solutions.

To contribute a devils advocate position for a second:  I think that raw damage output is actually not all that important as a metric.  Reasoning:  Most parties possess a large number of ways to debilitate the enemy and can readily do so within the opening round of combat.  Thus, it doesn't seem that actually doing very large amounts of damage is that important because most of the combat will be decided by the ability to disable the enemy after which killing them is trivial.
 

[Jackinthegreen]

I can whole heartedly agree with this post Ziegander.  I think that it is a very healthy methodology to assume that there will be a number of viable solutions.

To contribute a devils advocate position for a second:  I think that raw damage output is actually not all that important as a metric.  Reasoning:  Most parties possess a large number of ways to debilitate the enemy and can readily do so within the opening round of combat.  Thus, it doesn't seem that actually doing very large amounts of damage is that important because most of the combat will be decided by the ability to disable the enemy after which killing them is trivial.

In my experience many casual groups tend to think only in terms of damage.  Using "most parties" would seem statistically incorrect given the average player base.  Specifying that the more powerful groups will have debilitating effects is probably more accurate, which further reinforces the notion that damage itself isn't as important.  It is still a viable tactic in many cases for both team player and team monster.

Well, that's kind of venturing into deeper analysis. Just determining what monsters are "weak" for their CR seems like it would be a daunting task. The only way to determine what monsters are weak would be to first establish an average baseline to compare them to. Also, all that would do is give us a way to say, "you can, theoretically, play the game if you can beat these monsters and only these monsters, but that is only if your DM goes out of his way to challenge you with only monsters from this list." Which makes that data fairly subjective.

You're right of course, finding the average would probably be the best start.  That itself is also a fairly daunting task since we'd need to record and compile all the data for the various monsters then get the averages for as many of the statistics as possible.  Average health, AC, saves, and spell resistance is where I'd start. 

[Unbeliever]

Say what you will about him/her, Basket Burner is right about a few things regarding optimization, viable strategies, and monsters in D&D 3.5. The monsters are over there. They do have stock statistics, and without a baseline degree of effectiveness, PCs either will or will not be viable when we discuss D&D using in the "common" tongue. Sure, all games are played with some form of houserules. But, since we cannot know what houserules are being used, the only way we can discuss D&D together is by discussing it as written.

I never thought this was particularly contentious.  My issues with BB's approach was that he was just making the numbers up.  That, or his copies of the Monster Manuals are radically different then mine are.  Viz., the tanking thread, where I was told that stock monsters have attack bonuses ranging from 10-20 higher than those printed in the books.  I'd suggest those printed in the books as a baseline, which I think of as uncontroversial. 

Have at it, though.  I'll echo other posters, though, in saying that if we're talking about 4ish players taking on stock monsters of CR X, especially throwing out ones that are under-CR'ed, the minimal amount of "firepower" (broadly-construed) is going to be pretty modest. 

Two suggestions that may actually be helpful.  First, the benchmarks should probably be on the back end, meaning target numbers, rather than build-based.  Although exemplar builds are always helpful measuring sticks, I find.  Much more helpful than just classes.  What I'm thinking is that Character A may have numbers that would not make them particularly viable alone, but put them in a party with that War Weaver and well ...

Second, it's important to get a decent definition of what "viable" means.  I'd probably qualify it as something along the lines of a decent change of winning (high probability at CR -1 or 2, maybe winning against a CR +1 or 2 with decent luck or tactics?) and not being able to do nothing against the enemy.  In other words, I'd suggest that there's a lot of space between saying "I win motherfucker!" to the monster and "viable."  A reasonable chance of losing encounters, especially challenging ones (whatever that means), is part of the gentleman's agreement concerning optimization restraint in the games I participate in.

P.S.:  Trailblazer has averages, if you want to start there.  Although would "iconic" monsters of various types, e.g., Beholder for a caster type, Giant of some flavor for a melee type, be more useful? 

[Bearchucks]

http://www.minmaxboards.com/index.php?topic=3472.0 This might be a good place to start.  There was a post a while back that flight was expected by ECL5, the ability to deal with grapplers by ECL7, reach by such and such a point...just the trends.

[Basket Burner]

I just would like to point out that stock values are often misleading.

For example.

Ask someone what the to hits of CR 20 monsters are.

Assuming they actually look at the books they'll give you 2 numbers around 30, 6 numbers in the 36-42 range, and one number of 57. Most people don't even do that much. Instead they go look at some so called baseline that doesn't even pass a giggle check such as the PF bestiary table (does not describe even PF stock enemies accurately), or that Trailblazer nonsense (any guide that tells you 9 numbers that are all greater than 29 average 29 can be ignored on principle).

So what's the problem with this?

The first two are pure casters. Their to hits don't matter. The next six are dragons, and 36-42 are the numbers they get from BAB and Str alone. Not their final numbers. The last one is going to be at least mostly accurate.

But say you make determinations without reading into things. That gives you an average of 39... which is almost an entire D20 lower than the high number. It also gets you hit on a 2 by almost every single melee enemy even if you forget they have feats, items, and spells entirely... and the few exceptions hit on a 3.

Not only is that not an accurate baseline, it's worse than no baseline at all.

Now let's say you look at this properly.

The two casters won't attack, so you ignore their attack stats.

The six dragons all have the wealth of an NPC of their level, spellcasting ability, and 10-12 feats. That 36-42 ends up closer to 50-60 in practice. The Tarrasque doesn't change much or at all.

That gives you a bunch of numbers in the 50s. Very consistent.

[Mooncrow]

The interesting thing about assuming every encounter is an elite encounter is that it indirectly helps out the more magic item dependent classes - having a WBL that's 3-4 times the listed can prop up just about any character concept you want to think about.