What's Your Line Between PO And TO?

[JNAProductions]

See title. Give examples, please.

[Mike Miller]

I don't see a line there. It is a very clear difference. It is a ten foot space.

Practical optimization does things effectively, efficiently, and utilizes combined abilities to their maximum.

Theoretical optimization breaks the game and is not playable. See pun-pun, bobafeat, Priya the prismatic princess, etc

[ryu]

See title. Give examples, please.

Is it theoretically possible for you to lose against an opponent for which it is theoretically possible for them to be beaten? If so you are guaranteed to be within PO. Could be real high PO but still PO. TO is when you can just say no to literally everything as opposed to just a hilariously large number of things.

[OldTrees1]

Practical Optimization is optimization limited by what the DM in question is and is not allowing in their campaign.

Theoretical Optimization is a subset of theoretical exercises that replaces the DM with an absent DM (RAW = Rules) or with a specified alteration (RAW + X = Rules).

----

As such there are 2 lines that divide PO from TO.
Line 1: If there is no group, then it is not PO.
Line 2: The multidimensional power level cap the DM is using as part of their attempt to optimize the enjoyment of everyone involved.

----

When I am DMing that 2nd line varies somewhat depending on the range of character concepts the players want to play and the style of campaign I want to run. Generally the cap is somewhere around high tier 3 (versatility cap) and beneath rocket tag (power cap).

----

But different DMs (and even different campaigns by the same DM) would have different caps.
Test of Spite was PO. Lord Drako was TO(Please do not derail. I needed unintuitive examples).

[tiercel]

Anthropic principle: the world is the way it is because of the way the rules of the world work.

To wit: is this campaign set in the Tippyverse?

If it is not, then rule exploits that would result in a Tippyverse must not work in this campaign (or else the world would already BE a Tippyverse).

Put yet another way: "anything you can do, I can do better." PCs who want to try an infinite-anything or instant-win condition should probably consider why some BBEG has not already used/is not already using such an exploit (given that a BBEG is typically of higher level and less morally constrained than most PCs).

Heck, this last perspective is usually the basic cornerstone of "gentlemen's agreements" in games which I have played in/DMed.

So, more specifically in my experience:

No infinite anything

No free wishes. (Either a genuine wish is a quest reward, or something you level up enough to straight-up cast, within the printed limits, or you are likely looking at a Monkey's Paw situation.)

In my experience: no 9th level spells generally. This doesn't fix everything by a long shot, but 9th level does seem to have a lot of "goodbye world as you thought you knew it" headache material.

Metamagic reduction tricks are generally limited (if nothing else, to the level of chicanery players want to see applied to incoming BBEG fire); I've not been in a game with Divine Metamagic at all, for instance.

Limited immediate actions. Classic actions like Feather Fall or immediate-action defensive boosts have generally been OK, but pure no-sell actions (Abrupt Jaunt, Greater Mirror Image, Celerity, Craft Contingent Spell, etc.) not so much (again, if nothing else, because PCs prefer to not face BBEGs who can just shrug "no" at will)

No 15-minute adventuring days

No economics exploits. D&D economics is mostly a hash anyway and "do not look behind the curtain" background to the primary focus on adventuring, not designed to stand up to players seeking an Economic Victory condition.

No physics exploits. Catgirls aside, these often degenerate into rationalizations for Arbitrarily Large Numbers that are only loosely related to heroic fantasy play.

[Darrin]

When your build has a negative impact on or negates the agency of the other players, you've crossed The Line. As a consequence of that, it depends a great deal on the group.

[SimonMoon6]

For me, the line is something like, "Would I be embarrassed to tell the DM what I'm about to do?"

If so, don't do it.

[Jormengand]

For me, TO is leagues apart from what I'd do in a real game. Sure, I can hack the universe not only to get infinite wishes, but to become the recipient of an epic spell which I don't even need to know (and no-one else does either), sure I can destroy every last creature and object on a plane of existence, and sure I can do this at level 1 (as a truenamer to add insult to injury, or at least for a +5 to a relevant check so I can't fail it), but that's not really fun. I mean, if I do all that then the universe just packs up and goes home. The story never happens. The great and terrible deity Dee Emm sheds a single tear, and then... probably kicks me out of his game.

Better I think to go around playing a character that cool stories are made of, by picking abilities which are effective but also which fit the theme of the character in question.

[BoutsofInsanity]

T.O. are the exercises where we use the rules to do something in game. Theoretically it's possible to make Pun - Pun the Kobold God. To make infinite loops, or that drowning can heal you if you are in negative hit points.

Case and Point: The Tippyverse idea.

P.O. is the idea that you can look at the rules and apply some sort of rules as intended restrictions on them. The idea that the rules aren't perfect, and the arbitrator (Traditionally the DM) is going to apply some sort of interpretation on them in regards to the game you are playing.

For Example: I attempt to Pun-Pun. The DM says "Congrats, you win, you are the best, and can go home. Now the rest of you players wanting to play the game, can roll up their characters in a more traditional fantasy setting."

Pun Pun is not a P. O. because most likely it won't be allowed at any regular table.

Another Example: I'm dropped into water at negative hit points. I begin to drown. I raise my hit points to 0. The DM states, no, don't raise your hit points, that's stupid. Keep them at negative. Then I drown.


P.O.: Things that most likely are going to be allowed at the table

T.O.: Things that are possible within the rules as written, but are generally a "**** Move".

[Gruftzwerg]

P.O.: Things that most likely are going to be allowed at the table

T.O.: Things that are possible within the rules as written, but are generally a "**** Move".

I think this sums it very good up.
But you have to keep in mind that most P.O. builds will still be denied on most tables, cause most groups play on lower optimization lvls. It's just possible that a group could play on that optimization lvl.

[flappeercraft]

To me PO is if it is not stupidly game breaking but still powerful for a normal game. So one shotting the tarrasque is PO for me but killing every single statted creature in the ELH in a round is TO.

[137beth]

My line is "does it assume the use of house rules?" For TO I assume the rules of the game are being followed. For PO I assume there is a specific nonempty set of house rules: the house rules we are playing with. And that includes rules like "no pun-pun."

[ryu]

My line is "does it assume the use of house rules?" For TO I assume the rules of the game are being followed. For PO I assume there is a specific nonempty set of house rules: the house rules we are playing with. And that includes rules like "no pun-pun."

The easiest way to prevent pun-pun from effecting your game is to simply state that pun-pun already exists and refuses to share his position as over-overdeity. It's not that there's any hard rule preventing you from pun-puning. It's that pun-pun KNOWS WELL IN ADVANCE and considers it aggressive action.

[Psyren]

The easiest way to prevent pun-pun from effecting your game is to simply state that pun-pun already exists and refuses to share his position as over-overdeity. It's not that there's any hard rule preventing you from pun-puning. It's that pun-pun KNOWS WELL IN ADVANCE and considers it aggressive action.

I'd rather just ban him. Hard to create meaningful tension about the fate of your setting with Super Tom Bombadil Pun-Pun running around.

[Cosi]

Pun-Pun doesn't effect your game because Manipulate Form doesn't do the thing people say it does. It grants an ability, and it gives examples of the abilities it grants. It grants abilities in line with those examples, not "anything someone printed somewhere".

[Bucky]

Is dipping Warblade TO? It wouldn't fly at a lot of tables.

[Cosi]

Is dipping Warblade TO? It wouldn't fly at a lot of tables.

This is a good point. "Would a DM allow it" is an appealing standard, but applied to real DMs it catches a great deal of stuff that is not appropriate (like ToB, or for that matter non-core content). You could say "would a reasonable DM allow it", but then you have to define reasonable.

[Telonius]

I think infinite loops are pretty clearly on the "TO" side of things. So would arbitrarily high anythings, chain-gating Solars, and any Wish-based shenanigans. Attempts to trade souls, Dark Chaos Shuffles, anything that requires a yo-yo-ing alignment, diplomancy cheese, and anything to do with Tainted Scholar or any of the other famously overpowered prestige classes are still in the "TO" side, though a bit closer to the line. Anything that breaks WBL (Pillar of salt, 10 foot pole ladder factory) or breaks action economy is closer to the line. (Breaks as in actually breaks, not just gives an extra attack here or there). A wizard creating his own demiplane with weird time flow would probably be the closest thing to the line I can think of. Beyond that, it really depends on the group.


I'd rather just ban him. Hard to create meaningful tension about the fate of your setting with Super Tom Bombadil Pun-Pun running around.

He's more or less the DM's avatar in my games. Basically, because he has an arbitrary amount of wisdom, he recognizes that he's actually just a fictional character in my game, and wants it to run smoothly. The less interference, the better. So far I've never actually had to break him out, but my players know that if a squirrel ever looks at them sidelong, they're in dangerous territory.

[OldTrees1]

This is a good point. "Would a DM allow it" is an appealing standard, but applied to real DMs it catches a great deal of stuff that is not appropriate (like ToB, or for that matter non-core content). You could say "would a reasonable DM allow it", but then you have to define reasonable.

While attempting to define "a reasonable DM" is initially attractive, it avoids existing DMs and the abstraction tends to a fictitious one size fits all. In the end it becomes a Theoretical standard rather than an IRL standard.

If the Optimization is to be for practical purposes rather than theoretical, then it needs to be tied to the concrete circumstance rather than an abstraction of related circumstances.


Despite the "Would THE DM allow it" being affected by real world considerations like differences in opinion between DMs (some of which being unreasonable), it does fulfill its specified purpose of detailing the constraints for the optimization of the character for use at that DM's table.



Yes: This does mean that Fighter is not PO at some tables (those that ban Fighter).

[Gnaeus]

Tough.

For me it's not too different than RAI v. RAW.

For example, kobold loredrake sorcerers for me are TO. Because I don't think that was intended to be a thing. I consider most early entry to PRC tricks to be TO, because it looks to me like game wanted PRCs to come online at level 5-6 and I don't think that's what precocious apprentice was for.

I think using planar binding to summon an ally to be PO. Used to summon 150 allies in the week before the adventure to be TO.

Those are some specific examples. But I discuss it further in my guide to practical optimization in my sig. I agree that it varies wildly by table.

[Sam K]

TO: Optimization based purely on RAW, without any consideration for fun, balance, common sense, intention, physical or anatomical possibility or DM behaviour.
PO: Optimization for a specific set of metagame conditions, such as the level of power of a party/game, DM preferences, player preferences and sources allowed.

TO can be Pun-Pun. But TO can also be assuming that you can use a PRC from one setting in another setting, that you will have access to roughly recommended WBL, and any other amount of factors that may differ from DM to DM and table to table.

[Psyren]

TO: Optimization based purely on RAW, without any consideration for fun, balance, common sense, intention, physical or anatomical possibility or DM behaviour.
PO: Optimization for a specific set of metagame conditions, such as the level of power of a party/game, DM preferences, player preferences and sources allowed.

TO can be Pun-Pun. But TO can also be assuming that you can use a PRC from one setting in another setting, that you will have access to roughly recommended WBL, and any other amount of factors that may differ from DM to DM and table to table.

I don't think assuming WBL is TO, but I agree with the rest.

Note that many setting PrCs do have adaptation sections or other language for you to use them elsewhere, or they're generic enough (e.g. Radiant Servant) that there really isn't anything exclusively tying them to the setting they came from. "Super Sun Cleric" is a concept I could see just about anywhere.

[Cosi]

anything to do with Tainted Scholar or any of the other famously overpowered prestige classes are still in the "TO" side,

I think calling e.g. Incantatrix inherently TO is wrong. There are very clearly fair things to be done there, and the difficulty in defining PO and TO is precisely that there isn't a hard divide between the class's fair uses and its unfair ones.


Despite the "Would THE DM allow it" being affected by real world considerations like differences in opinion between DMs (some of which being unreasonable), it does fulfill its specified purpose of detailing the constraints for the optimization of the character for use at that DM's table.

Sure, but it's also totally meaningless. If "PO" just means "your DM allows it", then you can never say "X is PO". Clearly people want to be able to differentiate between "options that are fundamentally fair, but more powerful than some other options" (e.g. playing a Wizard instead of a Fighter) and "options that are fundamentally unfair" (e.g. using shapechange to get all EX abilities of valid shapechange targets). You need some standard that allows you to do that.

[OldTrees1]

Sure, but it's also totally meaningless. If "PO" just means "your DM allows it", then you can never say "X is PO". Clearly people want to be able to differentiate between "options that are fundamentally fair, but more powerful than some other options" (e.g. playing a Wizard instead of a Fighter) and "options that are fundamentally unfair" (e.g. using shapechange to get all EX abilities of valid shapechange targets). You need some standard that allows you to do that.

Agreed. I think it is the same thing that just depends on the context.

If I want to talk about a particular PO thread then the reference to the DM in question is enough to say "X is PO for this thread" or "Y is not PO for this thread".

If I want to talk about PO in general, rather than a particular PO exercise in particular, then we could use the "typical DM" standard (or the "reasonable DM" standard).

I tend to focus on the former (since those are the threads I tend to be answering) but I should not be dismissive of the latter (especially in a thread asking about the former while being the latter).

[ryu]

I'd rather just ban him. Hard to create meaningful tension about the fate of your setting with Super Tom Bombadil Pun-Pun running around.

I mean deities, overdeitys, and various beings orders of magnitude more powerful than most adventurers are far from uncommon ANYWAYS. What's so off-putting about calling the most powerful one pun-pun?

[atemu1234]

I find the difference lies in what the other players do and how the plotline is affected by it. Playing an Incantatrix to the best of your ability? PO. Using that to get infinite spells and try to take on a deity in spite of the actual misson? TO.

[RoboEmperor]

TO: Whatever that skyrockets you way beyond your expected power level
PO: Whatever that makes you very strong, but still within reason.

Surge of Fortune + Vorpal Throwing Knife + True Strike lets you one shot everything. That's TO.
Planar Binding Ghaeles. Early Access to 7th level cleric spells. That's TO. Stops being TO when you are level 13 yourself. Other TO outsiders are Nightmares, Efreetis, etc.
Ubercharger is just as bad as the vorpal combo mentioned above.

Planar Binding is unique. It becomes PO or TO depending on the outsiders you bind, which means outsider choice is part of the optimization. You can optimize planar binding's charisma check and save dc all you want, it doesn't make your character stronger, just more reliable and safe, but if you optimize your outsider selection to the max, that's when it becomes TO from PO, so you intentionally don't optimize your outsider selection and instead pick one that fits the power level of your character. As mentioned before, ghaeles at level 11 is truly OP, but at level 13 its more acceptable but still really powerful, or is still OP if your game is not that optimized.

[Hackulator]

Would you want to slap someone/question your decision to invite them into your game if they asked to do the thing? If so, the thing is TO.

Likewise, if you look at a thing and know you would never even ask for it because its too stupid and broken, it's TO.

[Florian]

PO seeks to reach the boundaries while staying within the rules and level of comfort set for a given table.

TO searches for cracks and dysfunctions of the rules to overcome those boundaries. Part of it might or might not be deliberately misreading the rules or glossing certain things over.

[Sam K]

I don't think assuming WBL is TO, but I agree with the rest.

Note that many setting PrCs do have adaptation sections or other language for you to use them elsewhere, or they're generic enough (e.g. Radiant Servant) that there really isn't anything exclusively tying them to the setting they came from. "Super Sun Cleric" is a concept I could see just about anywhere.

I agree that WBL is a bit of an edge case, but quite a few builds assume that you will have access to roughly WBL and can access certain key items to play to your strengths and cover your worst weaknesses. For example, DMM builds tend to assume that you can get some items that boost your turning attempts (assuming you can stack nightsticks is a completely different assumption). Casters built on save-or-suck/lose/die spells assume that you can bump the DC of your spells. But yeah, it's an edge case as most TO builds work without access to WBL/specific items, but lacking them may leave them far less powerful than theory would dictate.

Most PRCs are easy to refluff or adapt, but you cannot assume it will be allowed. For example, my DM (who is fine with a draconic water orc RKV, and dipping warblade if I want to) was completely against me refluffing RKV to another religion, thus forcing me to play a lawful character for the first time ever.

But yes, these are fairly minor cases, I just listed them to highlight that TO doesn't just involve Pun Pun, there's a lot of far smaller assumptions that can still factor in.

[prufock]

In comparison to Practical Optimization, Theoretical Optimization does at least one of the following:

- makes in-game assumptions about the setting, reactions of the NPCs, and other things that require GM arbitration.

- requires calculations involving loops, infinite numbers, near-infinite numbers, or arbitrarily high numbers.

- requires questionable reading of unclear rules. If a rule is written in such a way that, whether considered alone or in reference to other rules, is not clear, theoretical optimization will assume the most liberal reading.

- breaks established assumptions. For example, wealth-by-level is an established assumption.

- obscenity rule: I know it when I see it.

[Florian]

Note that many setting PrCs do have adaptation sections or other language for you to use them elsewhere, or they're generic enough (e.g. Radiant Servant) that there really isn't anything exclusively tying them to the setting they came from. "Super Sun Cleric" is a concept I could see just about anywhere.

That´s more a thing of socialization into the hobby. Around here, most people came into the hobby playing DSA, which has a very strict stance on fluff and crunch. Crunch is only there to model the fluff and if that leads to results that contradict the fluff, the fluff will always prevail. This makes "setting" the "hardest" rule that always stands above all other rules. It´s actually pretty common that your fellow players will say "Super Sun Cleric? Not part of the setting, we veto it out".

@prufock:

Can´t fully agree. Martial classes in PF have access to some very simple escalating 1-2-combos that will possibly chain up to n1. That´s by design.

[prufock]

@prufock:

Can´t fully agree. Martial classes in PF have access to some very simple escalating 1-2-combos that will possibly chain up to n1. That´s by design.

Examples? How do you know this is by design?

[gooddragon1]

I think PO is within authorial intent, whereas TO is outside authorial intent.

[ryu]

I think PO is within authorial intent, whereas TO is outside authorial intent.

I mean... Authorial intent is blaster healbot corpse BSF... The druid play test didn't even cast spells, shapeshift, or use the pet for combat... I don't think that's a standard most people will agree to.

[martixy]

Mine is very liberal:


Goes against established assumptions or clearly evident RAI.
Twists or perverts for its own convenience ambiguous rules.
Divergent loops.
Abuse of unintended side effects.


Let's take for example the classic Lightning maces crit fisher:
- Disciple of Dispater is TO, because while technically legal, it goes against how crits were redesigned in 3.5.
- Aptituding Lightning maces to high-threat range weapons is borderline TO, because it goes against RAI AND makes a +1 weapon property more powerful than the class feature it is derived from. Though I'm inclined to let that one stand, and also rule that whatever the item property can do, the class feature can also do.
- Let's say we allow Aptitude to work in its TO sense, even then the "crit fishing with lightning maces" loop I would consider PO, because it you cannot make it divergent, unless you use more TO.
That covers the first 3 points. DCFS is covered by the 4th.

[gooddragon1]

I mean... Authorial intent is blaster healbot corpse BSF... The druid play test didn't even cast spells, shapeshift, or use the pet for combat... I don't think that's a standard most people will agree to.

That may have been the original authorial intent, but I think with time and splatbooks their intent evolved. They just never anticipated people doing things like using manipulate form or exploiting no XP sla wishes.

[Florian]

Examples? How do you know this is by design?

Everything to do with Overrun, Sword n Board being geared towards TWF and Bashing Finish.
Core only or Core + APG have many things that will create automatic chain combos.
Now that could be unintentional, but later supplements brought us style feat chains that work and enhance exactly those combos.


I mean... Authorial intent is blaster healbot corpse BSF... The druid play test didn't even cast spells, shapeshift, or use the pet for combat... I don't think that's a standard most people will agree to.

As 3.0 was intended to be a direct continuation and upgrade of AD&D 2nd and the play testers handled their classes more or less exactly as they would AD&D characters, well....

[Hurnn]

For me the line is crossed when you automatically defeat any Cr +/- 5 challenge regardless of dice roll results, and or obsolete the rest of the party.

[Gnaeus]

For me the line is crossed when you automatically defeat any Cr +/- 5 challenge regardless of dice roll results, and or obsolete the rest of the party.

I see where you are coming from, but I think both definitions are flawed.

Definition 1 is reasonably easy to evaluate, which is good. But I could imagine a 17th level party of Wizard, Cleric, Druid, Archivist which beat it without undue shenanigans and which functioned passibly in play.

Definition 2 has 2 problems. 1. Obsolete the rest of the party is very vague. It's kind of like saying TO is what the DM says it is. It's not entirely false, but it's hard to discuss on a forum. 2. Imagine a 17th level party of ranger, monk, fighter, Druid. The Druid is TO by your definition just by preparing spells in the morning. Doesn't even matter which spells, he could obsolete the rest of the party with only AC + wildshape + high level SNAs. I think most of us would agree that there was an issue there, but I don't think that the issue would be that the Druid player was using theoretical optimization.

[Cosi]

I think PO is within authorial intent, whereas TO is outside authorial intent.

The authors explicitly changed wish and shapechange to be more broken in the transition from 3.0 to 3.5. Is authorial intent that you make infinitely powerful items with SLA wish? Because it seems to me pretty clearly intentional.

[prufock]

Everything to do with Overrun, Sword n Board being geared towards TWF and Bashing Finish.
Core only or Core + APG have many things that will create automatic chain combos.
Now that could be unintentional, but later supplements brought us style feat chains that work and enhance exactly those combos.
I'm not following you. How does this create near infinite combos or damage?

[Florian]

I'm not following you. How does this create near infinite combos or damage?

Simplified version: "Hit with your main hand and you get a free attack with your shield" + "Hit with your shield and you get a free attack with your main hand". Besides that you have to hit, this is a loop.
Now what did they add? "If you hit with both, you get two additional free attacks, once each with main hand and shield".
Now consider the wording. It says "shield", not "off-hand". The PO conclusion now would be to use two shields. Silly, right, can´t be what´s intended.... But: Now we´ve got the Dwarves Razor Shield that is actually supposed to be used in pairs.....

[prufock]

Simplified version: "Hit with your main hand and you get a free attack with your shield" + "Hit with your shield and you get a free attack with your main hand". Besides that you have to hit, this is a loop.
Now what did they add? "If you hit with both, you get two additional free attacks, once each with main hand and shield".
Now consider the wording. It says "shield", not "off-hand". The PO conclusion now would be to use two shields. Silly, right, can´t be what´s intended.... But: Now we´ve got the Dwarves Razor Shield that is actually supposed to be used in pairs.....

Thanks for trying to clear this up, but I'm still confused. I might need this laid out explicitly.

What is the feat that grants you "hit with your shield and get a free attack with your main hand"?
What feat gives you 2 free attacks if you hit with both?
Bashing Finish only triggers on critical hits, so unless you have a way to reliably make successful crits, it's statistically unlikely to hit near infinite levels. Is there some method you're using to circumvent that?

[Goaty14]

If you cite 3+ sources, you're probably going in the wrong direction
If you take less than 5 levels in a class; you're probably not using that class right (Cancer Mage)

[Drakevarg]

If you cite 3+ sources, you're probably going in the wrong direction
If you take less than 5 levels in a class; you're probably not using that class right (Cancer Mage)

That seems way too stringent. Some classes simply aren't worth it for the first few levels, even non-optimizing players can see that. And when, like me, you own well over a dozen splatbooks that cover a broad range of topics, why not utilize what you have available? Anything less is a wasted investment.

[JNAProductions]

Well, if you're playing, say, a Wizard, then you don't NEED to go splat-diving for power. (Versatility, maybe. Theming, total. But power? Nah.)

If you're playing a Fighter (or Fighter-esque character) then splat-diving can be virtually mandatory at some tables.

[Sayt]

Honestly I think the difference has devolved to PO being "What I think I ought to be able to get away with in a randomly selected playgroup" and TO being "What I don't think I can get away with in a randomly selected playgroup"

[Rijan_Sai]

My take:
Practical Optimization: Optimising your character (in a manor consistent with the rest of the table), with the sources available at the table, to be powerful enough to contribute but that everyone has fun.

Theoretical Optimization: Optimising your build to get the most power available using any sources, without regards (generally) to team play. Useful for finding combos, loopholes, etc. not usually noticed, but not really intended for actual play.

[prufock]

If you cite 3+ sources, you're probably going in the wrong direction
If you take less than 5 levels in a class; you're probably not using that class right (Cancer Mage)

Player's Handbook, Dungeon Master's Guide, and Monster Manual are 3 sources. I guess a wood elf rogue/wizard/uncanny trickster counts as TO?

[Gnaeus]

Honestly I think the difference has devolved to PO being "What I think I ought to be able to get away with in a randomly selected playgroup" and TO being "What I don't think I can get away with in a randomly selected playgroup"

For purposes of internet debate, I don't think that's "devolving". The only better functional definition is that PO is what my specific DM allows. I think your definition is meaningfully the same as "would most DMs see this as a problem". And that's about as accurate as we can get without knowing all DMs personally.