How much of a power spike is acceptable?

[Quertus]

Personally, I'm in the "hungrily look forward to power spikes", "notice such spikes IRL", and "believe that reality is stranger than fantasy in that regard" camps. People IRL often have huge spikes - probably the most commonly discussed one is "seeing the elephant". Losing suspension of disbelief over power spikes arguably just shows a lack of understanding of the real world, let alone a lack of immersion in the game world in the first place.

I love the rewarding feeling of advancing to where I'm rolling two dice for my DR (to put it in D&D terms) instead of just one, or earned enough XP to reach the next level (D&D 2e) or take the next class (WHF) or afford to build the next magic item (D&D 3e, WoD Mage). Those sudden jumps feel good in a way most point buy just... doesn't. The endless treadmill of just keeping up is just not as interesting as sudden jumps.

And, when the party levels at different times, that sort of jumps-and-starts makes for a really interesting, "what can we do now?" minigame.

However, if we are talking about gaining a dramatic amount of power in some levels of D&D, that is bad, because you tend to stay the same level in each session, and you tend not to want some sessions to be too easy out of the blue and some sessions to be too hard out of the blue.

I don't think that's universally true. I mean, things suddenly being easy is a nice reinforcement of "you just leveled - leveling matters". And, personally, I'm all about the old-school D&D "wtf is game balance?" mindset where encounters aren't "CR appropriate", and it's up to the party to determine if it's something they want to tangle with or not.

So, to me, you've just defined a feature, not a bug.

developing complexity as you gain levels is good. Developing complexity is what brings people back session after session

Is it? I'd never thought of it that way before.

Sudden empowerment really exists.

Amen.

Some spikes like these are indeed "natural" in the sense that they mathematically follow from premises of the system and what the system is trying to model. It's not a big problem as long as the new & improved character can still be challenged to some extent by people on their previous level. F.ex. if level 5 Fighter can still be beat by 3 level 4 Fighters, things are still within reason. Bad things are usually going on if level N Fighter is suddenly untouchable to any number of level N-1 Fighters or similar.

... This might just be a style difference between us. Personally, I hate systems like WoD, where the max 5 dots in a skill not only might well have you falling against a noob with only a single dot in said skill, but have you doing so inordinately often. I don't expect someone who just picked up D&D yesterday to be schooling a veteran optimizer on character creation with any regulatory.

As I hate the point buy treadmill, I'd almost argue that I only care about differentiating things that matter - "Level 7" magic decks should stand very little chance against "Level 8" magic decks, for example, and, unless they specifically target the deck's weakness, "Level 9" decks should be right out. Otherwise, if they're able to compete against each other reasonably, why put them in different tiers?

[Frozen_Feet]

Quertus, read again what I wrote. I didn't say "people of N level should be challenged to some extent by complete beginners". I said they should be challenged to some extent by people of their previous level.

Magic decks make a poor example for what I was talking about. Like I said previously, if you're 5% faster in a race, you don't win 5% often, you win all the time. That's by-product of the nature of the contest less than a power spike.

Let's consider the example of fighters again. In a duel, when both participants are fresh, it might be a 5th level fighter has 100% win rate over 4th level fighters. But when examined closely, it is found the 5th level fighter is always left with just 1 hitpoint. This means the 5th level fighter has not spiked dramatically in power, the small difference in initial conditions just makes a big difference in the outcome. They are still challenged by people of their previous level and we can construe that against two 4th level fighters, they would not be winning 100% of time.

Let's take an alternate example. Suppose a 2nd level fighter can fight three 1st level fighters in a row. A 3rd level fighter can fight three 2nd level fighters in a row. A 4th level fighter can fight three 3rd level fighters. A 5th level fighter can fight three 4th level fighters. So on and so forth. This kind of progression satisfies my condition for a working power curve while still having a notable leap in power at each level. Even if a 6th level fighter can suddenly fight seven 5th level fighters in a row, or a 7th level fighter can only fight two 6th level fighters, things remain within reason. How many 1st level fighters a 5th level fighter can fight in a row isn't a question that's asked, nor is the answer important. The answer could be 81, it could be arbitrarily large. What's important is that there are no arbitrarily large gaps between adjacent points in the same continuum.

Tiers shouldn't be talked in this context because tiers are rarely based on or form a scale continuum. Tiers may be exclusive categories with any numbers associated with them not really signifying any clear mathematical relation between them.

[Florian]

Hm. "Splittermond" as a system has an interesting way to handle "power growth" and "power spikes.
It´s basically a skill-based system with some "tiers of play" as limiting factor, meaning that the "Beginner levels" have a hard skill cap of 6 and you must grow "horizontally" enough (total XP) to be ready to advance to the next tier, which raises the skill cap by +3, and so on.
In this system, feats are tied to skills and you get to chose on whenever reaching the cap for a skill, modeling having reached a stage of mastery for that exact skill, putting you above your peers a bit.

"Classes" exist, but work differently than what we´re used to from D&D, as any character can learn any skill and even "schools of magic" are expressed as skills. Choosing a "class" and "culture" will simply give some free feats in advance, fitting the overall concept. (I.e. a "Sea League" Elf Bladedancer will start with a free seafaring, chain weapons and either wind or water magic feat.)

[CharonsHelper]

However, the fact that power spikes are bad must be weighed against the fact that developing complexity as you gain levels is good. Developing complexity is what brings people back session after session, as they play with one option one session, and get to look forward to trying out another at the next level. Developing complexity tends to require that characters gain new options as they level up, and these new options tend to make power spikes inevitable.

So power spikes are bad, but a quest to eliminate them altogether is also bad.

I'm going to disagree - but I know where you're coming from.

Complexity is always bad.

But depth of play is always good.

Depth of play is purchased with complexity, and one of the hardest parts of being a game designer is to get the best bang for your buck on that purchase and know where it's worth spending to stay with your game's theme.

I do agree that increasing depth helps bring people back, and adding complexity a bit at a time (such as with a levelling system) allows you to add more over time, and therefore purchase more system depth.

But - some of that is probably just semantics.

[Quertus]

Quertus, read again what I wrote. I didn't say "people of N level should be challenged to some extent by complete beginners". I said they should be challenged to some extent by people of their previous level.

Magic decks make a poor example for what I was talking about. Like I said previously, if you're 5% faster in a race, you don't win 5% often, you win all the time. That's by-product of the nature of the contest less than a power spike.

Let's consider the example of fighters again. In a duel, when both participants are fresh, it might be a 5th level fighter has 100% win rate over 4th level fighters. But when examined closely, it is found the 5th level fighter is always left with just 1 hitpoint. This means the 5th level fighter has not spiked dramatically in power, the small difference in initial conditions just makes a big difference in the outcome. They are still challenged by people of their previous level and we can construe that against two 4th level fighters, they would not be winning 100% of time.

Let's take an alternate example. Suppose a 2nd level fighter can fight three 1st level fighters in a row. A 3rd level fighter can fight three 2nd level fighters in a row. A 4th level fighter can fight three 3rd level fighters. A 5th level fighter can fight three 4th level fighters. So on and so forth. This kind of progression satisfies my condition for a working power curve while still having a notable leap in power at each level. Even if a 6th level fighter can suddenly fight seven 5th level fighters in a row, or a 7th level fighter can only fight two 6th level fighters, things remain within reason. How many 1st level fighters a 5th level fighter can fight in a row isn't a question that's asked, nor is the answer important. The answer could be 81, it could be arbitrarily large. What's important is that there are no arbitrarily large gaps between adjacent points in the same continuum.

Tiers shouldn't be talked in this context because tiers are rarely based on or form a scale continuum. Tiers may be exclusive categories with any numbers associated with them not really signifying any clear mathematical relation between them.

Thank you for the clarification. My seeming straw man was actually more just me trying to paint where I was coming from than me actually misreading your position. But I find your definition of what it means to be "challenged" quite interesting.

The tiers bit... was a stretch. People (maybe in this thread?) were discussing gaining new maneuvers as an alternative to more attacks, thus making tiers of maneuvers a seemingly related topic.

But, yes, a maneuver that deals 2d6 as an upgrade from one that deals 1d10 is actually arguably more of the gentle yet meaningful progression you're discussing, while still "quantifyably different" like I require differentiated objects to be.

I'm going to disagree - but I know where you're coming from.

Complexity is always bad.

But depth of play is always good.

Depth of play is purchased with complexity, and one of the hardest parts of being a game designer is to get the best bang for your buck on that purchase and know where it's worth spending to stay with your game's theme.

I do agree that increasing depth helps bring people back, and adding complexity a bit at a time (such as with a levelling system) allows you to add more over time, and therefore purchase more system depth.

But - some of that is probably just semantics.

This, on the other hand, is an explanation that sheds whole new light on a post that I had taken differently.

Complexity is usually bad; needless complexity is always bad. "Everything should be as simple as it can be, and no simpler".

[Max_Killjoy]

Tangentially related to my thread about gaining HP, here's another question for the masses: When a character advances in power (probably by gaining a level), how steep of a power spike is acceptable?

To illustrate with a well-known example: in D&D 5e, most martial classes get Extra Attack at 5th level, allowing them to make two attacks per round instead of one. This effectively doubles their damage output, in addition to the usual benefits of levelling up (HP and proficiency bonus). The result is that a 5th-level character is more than twice as powerful as a 4th-level one. A power boost this large doesn't really occur at any other level.

So what do you think of sudden power boosts like this? Is it weird and jarring, or is it just a natural part of the game? Why or why not?

And if you do think it's jarring, here's another question. D&D 2e and 3e both had mechanics to dampen this sudden power boost.
- In 2e, a low-level fighter gets one attack per round, a mid-level fighter gets one and a half, and a high-level fighter gets two. Having one and a half attack per round basically means that you make two attacks in turn one, one attack in turn two, then two attacks in turn three, etc.
- In 3e, every additional attack you get comes at a -5 penalty compared to the last.
In my experience, both of these are generally considered pretty awkward. Assuming you agree (which you're also free not to), would you rather accept the power spike, or deal with a mechanic like this?

I prefer a gradual, shallow, linear increase, to start out with some real ability and then go up at a measured pace from there. If my character is supposed to be X, I want to start out at something like X... not as someone who could one day maybe become X.

I actively dislike the the "zero to demigod" upward curve of 3.5e and the sort. The sudden spikes of "leveling up" both make that all the more jarring, and have have their own head-scratching issues of disconnect from most settings and "fictional facts".

Meh.

Players choose whether to suspend disbelief or not and the system has very little role in helping that outside of cases where a hypothetical system might put in something that is clearly insane for an argument's sake. You can as easily say that at level 5, fighters will have an epiphany or breakthrough that allows them to massively improve the number of attacks they can execute in a space of time, or say that a fighter is not considered level 5 until he has had that epiphany or breakthrough.

But to answer the much more interesting question of whether dramatic power spikes are good for gameplay:

I believe power spikes as you describe are, in general, not good for tabletop RPG games. They are great in shorter form games like RTS, MOBA, head-to-head card games even, because they open and close timings for skilled players to exploit. However, if we are talking about gaining a dramatic amount of power in some levels of D&D, that is bad, because you tend to stay the same level in each session, and you tend not to want some sessions to be too easy out of the blue and some sessions to be too hard out of the blue.

However, the fact that power spikes are bad must be weighed against the fact that developing complexity as you gain levels is good. Developing complexity is what brings people back session after session, as they play with one option one session, and get to look forward to trying out another at the next level. Developing complexity tends to require that characters gain new options as they level up, and these new options tend to make power spikes inevitable.

So power spikes are bad, but a quest to eliminate them altogether is also bad.

I'm going to have to disagree on both.

System can easily "break disbelief" for me, and that's not a matter of choice. What I can choose to do is play anyway based on other factors.

On complexity... what you describe is not what keeps me coming back.

[Vitruviansquid]

I'm going to disagree - but I know where you're coming from.

Complexity is always bad.

But depth of play is always good.

Depth of play is purchased with complexity, and one of the hardest parts of being a game designer is to get the best bang for your buck on that purchase and know where it's worth spending to stay with your game's theme.

I do agree that increasing depth helps bring people back, and adding complexity a bit at a time (such as with a levelling system) allows you to add more over time, and therefore purchase more system depth.

But - some of that is probably just semantics.

Yeah, that's sound.

I was sloppily folding the concept of depth of play into the concept of complexity.

Semantic, probably, but sound.

I'll reply to the other stuff when I'm at a computer and can wrangle quote formatting.

[Florian]

Depth of play is purchased with complexity, and one of the hardest parts of being a game designer is to get the best bang for your buck on that purchase and know where it's worth spending to stay with your game's theme.

I do agree that increasing depth helps bring people back, and adding complexity a bit at a time (such as with a levelling system) allows you to add more over time, and therefore purchase more system depth.

That leads to the question what the actual goal of the game is/should be, what you´re going to play and how you´re going to play it.
Contrast, say, "Mountain Witch" and "Lady Blackbird" to the typical more open-ended design of systems like D&D.
Edit: Black Crusade has the inbuilt goal of reaching 100 Infamy before 100 Corruption, as a different, non-scenario-based example.

[Anonymouswizard]

A good point has been brought up, we should consider start points and end points as well.

I'm not a fan of zero to demigod, or even zero to hero. I like my games to go from competent to hero or professional to hero. Or I'd rather begin at the point where we can reasonably complete jobs without GM help (in D&D terms roughly third level), and end at the point where we may become legendary or our story may fade in a decade (in D&D terms roughly tenth level). I'm also not concerned with getting too much in the way of new stuff, but I do want a slow increase in competence. If D&D has a gradient of 1 I want a gradient of about 0.125 or less.

Although I also like demigod to demigod. Thankfully most superhero systems can handle that power level.

[Quertus]

A good point has been brought up, we should consider start points and end points as well.

I'm not a fan of zero to demigod, or even zero to hero. I like my games to go from competent to hero or professional to hero. Or I'd rather begin at the point where we can reasonably complete jobs without GM help (in D&D terms roughly third level), and end at the point where we may become legendary or our story may fade in a decade (in D&D terms roughly tenth level). I'm also not concerned with getting too much in the way of new stuff, but I do want a slow increase in competence. If D&D has a gradient of 1 I want a gradient of about 0.125 or less.

Although I also like demigod to demigod. Thankfully most superhero systems can handle that power level.

If push comes to shove, I probably prefer the "demigod to god" range of play. But, regardless of the range, I generally prefer the snail's pace of, say, 2e D&D advancement over 3e's turbo advancement rate.

[Anonymouswizard]

If push comes to shove, I probably prefer the "demigod to god" range of play. But, regardless of the range, I generally prefer the snail's pace of, say, 2e D&D advancement over 3e's turbo advancement rate.

True, I'm thinking more about this problem now as well.

We have three variables to alter, power range (p), campaign length (l), and advancement rate (a), to discover our fourth quantity (power spike size/s).

For power range, let's assign a scale between 0 (no increase in power) to 10 (character becomes many times as powerful). I'm going to assume that this is a linear scale to make our maths easier.

For campaign length, let's say that we're going to measure it in months. We could do years, but I've personally not seen a multi-year game.

For advancement rate let's say we want to advance x times a month (or year if measuring it in years).

Now, we can put these variables into a formula where s=x(p/(l*r)), where x is our unit of advancement.

[Quertus]

True, I'm thinking more about this problem now as well.

We have three variables to alter, power range (p), campaign length (l), and advancement rate (a), to discover our fourth quantity (power spike size/s).

For power range, let's assign a scale between 0 (no increase in power) to 10 (character becomes many times as powerful). I'm going to assume that this is a linear scale to make our maths easier.

For campaign length, let's say that we're going to measure it in months. We could do years, but I've personally not seen a multi-year game.

For advancement rate let's say we want to advance x times a month (or year if measuring it in years).

Now, we can put these variables into a formula where s=x(p/(l*r)), where x is our unit of advancement.

Perhaps we should be measuring is sessions, or even hours of play, given how variable session length is?

Really, it's not about play length, it's about total power change, and number of increments to get there.

So, for 2e D&D, it's the power difference between level 1 and level 20, divided into 19 discrete steps. The size of each step, on average, is (power(20)-power(1)) / 19.

But, since not all of those steps are equal, I believe what this thread is actually concerned with is measuring just how big the largest such step is, and asking what people's max value for that step is. Myself, I'm much more concerned about what the minimal value of that step is. I like big changes!

For me, ascending from peasant to godhood is a perfectly fine step, so long as there is a good in-game reason for it.

Or it happens IRL. That'd be fine, too.

[Florian]

@Anonymouswizard:

Brrr! You´re jumping too far ahead for your own good, at least if this should be a serious shot at creating a system.

Cover the basics first by taking a clear stance how you want to treat the three basic cornerstones of Game, Simulation, Narrative and how you weight the individual aspects and how much influence they should have for designing the rules.

[Lalliman]

@Anonymouswizard:

Brrr! You´re jumping too far ahead for your own good, at least if this should be a serious shot at creating a system.

Cover the basics first by taking a clear stance how you want to treat the three basic cornerstones of Game, Simulation, Narrative and how you weight the individual aspects and how much influence they should have for designing the rules.

It should be mentioned that, despite having the same avatar, AnonymousWizard isn't me (the poster). I'm the one creating a system. I think AnonymousWizard is just abstractifying the problem as a thought experiment, and doesn't actually intend to somehow use this as the basis for a mechanic.

*Goes back to lurking impartially*