Game Design - Why percent based stats are bad for player development and future content + how to fix it

[Mcadams]

Hey guys, 

 

Like many of you, I plan on making a game with Intersect in the very near future. While waiting for the source to come out, I've been doing a lot of preliminary game design, which includes making classes, skills, crafting recipes, and something I've been spending a LOT of time on, stat balancing. Great art and content are very important for a good game, but if your game, at its core, is unbalanced and lacking fundamentals, it's not going to make for a good player experience.

 

One thing I've been working on recently are secondary stats, or the stats that your character gets only from items, and not from leveling. This can be stats like Critical Strike, damage reduction stats like Armor, or any custom stats you can make up. Let's take Critical Strike, or Crit, for example. Usually, Crit gives you a % chance to do double damage when you attack. Let's say you cap Crit in your game at 40%, so with all the best gear and buffs and whatever, you can at max have a 40% chance to Crit. Let's say in your game you have 5 Equipment slots, so each piece of gear can give up to 8% Crit. So a person with all 5 pieces of the best gear has 40% Crit. Seems all fair and balanced, right?

 

Here's where the problem starts.

 

For the rest of the development of your game, that gear at that time is going to make you reach the cap for Crit. So what happens if you want to release new, stronger gear? Well you could raise the cap of Crit to say, 60%, but then that gets out of hand. You can't have people running around critting everything and dealing double damage more often than not. Eventually as you make stronger items, the cap for Crit needs to go even higher, and higher, until everyone is too overpowered and content becomes too hard to balance and gets out of hand. You could just keep the cap at 40%, and all of the best gear only gives you 8% each at best, but then it doesn't really feel like your character is getting stronger with the new, "better" gear. It also makes the old gear viable, because you can reach the Crit and whatever other caps you have without ever upgrading.

 

So how do you fix this?

 

Instead of having a percentile number (40%) determine the hard cap for your stats, you use a exponential value. In this case, let's use 40. So now, each 1 point of Crit = 1% Crit chance. When you have 40 Crit, you'll reach the cap of 40%. Now, future content comes out, and you want to make stronger gear. Now you raise the Crit cap to 100. So now, players need to get 100 points of Crit to reach the 40%, and the 40 Crit that use to give 40% now only gives you 16%, making the old gear less powerful and allowing you to balance your items and your game without going back and changing them. In addition, the new pieces of gear that give +20 Crit each will make your character actually feel stronger, because your Crit chance is increasing back up to that 40%.

 

This, however, creates a new problem.

 

Say your level cap was 50, and you're increasing it to 60. You can't just have a global Crit cap for every level, or it creates this problem.

 

Level 50 Player with 40 Crit fighting a Level 50 Monster -> 40% Crit chance

*Level Cap is increased to 60 and Crit cap is increased to 100, and the player gains a level*

Level 51 Player with 40 Crit fighting a Level 50 Monster -> 16% Crit Chance

 

"What the hell? I just leveled up and got WEAKER!"

Well, yeah, you sorta did. Since the Crit cap was increased to 100, your existing gear was made worse so that the new gear could be better than it.

 

How to fix this

Instead of having a 'global' Crit cap, have the Crit cap increase only for people leveling up past the old level. So, for levels 1-50, you need 40 Crit for a 40% chance to crit on monsters your level. and for levels 50-60, you need 100 Crit for a 40% chance to crit on monsters your level. So now, a level 50 fighting a level 50 with 40 Crit still has a 40% chance, so he won't feel like he got weaker because you had to up the Crit cap.

This means:

max Crit points on level 50 gear - 40 = 40% chance to crit on a level 50 monster

max Crit points on level 60 gear - 100 = 40% chance to crit on a level 60 monster

Using this system, you can easily balance old, current, and future content, by simply upping the maximum caps for the new gear and levels you're creating, without any detriment to the player. They'll feel like they're getting stronger as they progress through new content, and they'll still feel strong when fighting older, lesser monsters.

 

I hope this quick writeup gives you guys some insight into my philosophy on game design and how to better manage your player's growth for the now and for the future. If you have any questions or would like some parts cleared up feel free to let me know in a reply or a PM. Thanks for reading.

 

 

[Zahvi]

Nice. This could be very useful information for people at AGD. Thanks for sharing!

[SkywardRiver]

The way I handled it (because stats in Skywardens don't have the typical 255 cap) at 300 Agility you'll have about 35% crit rate. any stats past that grant minute but noticeable gains in crit rate. At 30000 agility (far higher than any player is ever going to reach) you'd have 100% crit rate. So any points past 300 grant noticeably smaller scales of crit rate, but not so small as to turn the player off to investing in Agility for Crit based rewards.

[Mcadams]

6 minutes ago, SkywardRiver said:

The way I handled it (because stats in Skywardens don't have the typical 255 cap) at 300 Agility you'll have about 35% crit rate. any stats past that grant minute but noticeable gains in crit rate. At 30000 agility (far higher than any player is ever going to reach) you'd have 100% crit rate. So any points past 300 grant noticeably smaller scales of crit rate, but not so small as to turn the player off to investing in Agility for Crit based rewards.

That's a good system too, as long as it's not a raw % you should be fine. You have to be careful without having soft and hard caps though, because it requires you to keep track of all the sources where players can gain Agility and regulate them accordingly

[Craig1020]

Interesting read, kind of similar to income tax in many ways, applying a percentage to the first amount with increasing rates and fixed intervals, love it!

 

-Craig

[JimineyWhiskers]

I feel like if done right, percentage based stats can actually be pretty damn fun and challenging. Some friends and I used to play an MMO called Rohan: Blood Feud. My favorite part of the game was crafting different builds./ Each class had a starter tree and two elite trees which you could choose one at level 50. The skills were interestingly set up, for example, a templar was a white elf class that used magic attack % to apply melee damage to opponents. You could get a buff that would raise the stat by 25% for some duration (like 30 minutes or so). What made it fun was deciding how to allocate stat points every level up. As a templar, if I spent more stats on INT, my magic attack would go up and my buffs would be more effective for offense. On the other hand, I could put some more in WIS for better heals since I'm technically still a healer (White Mage -> Templar). What this resulted in was people running pretty different builds of the same classes and honestly, it was balanced pretty friggin well. The good thing about percentages is that it allows you to work with smaller numbers everywhere else. A max level character would have a base stat leveled at around 500 lets say. With buffs, that number would grow and if you had an attack spell that said "Apply 150% of your INT as damage" you would fiend over getting that stat boosted as much as posible if your role was damage. I'd actually much rather prefer this system, especially for skills since using percentages for them allows your skills to scale with your character.

[Mcadams]

5 hours ago, AcidTumbleweed said:

I feel like if done right, percentage based stats can actually be pretty damn fun and challenging. Some friends and I used to play an MMO called Rohan: Blood Feud. My favorite part of the game was crafting different builds./ Each class had a starter tree and two elite trees which you could choose one at level 50. The skills were interestingly set up, for example, a templar was a white elf class that used magic attack % to apply melee damage to opponents. You could get a buff that would raise the stat by 25% for some duration (like 30 minutes or so). What made it fun was deciding how to allocate stat points every level up. As a templar, if I spent more stats on INT, my magic attack would go up and my buffs would be more effective for offense. On the other hand, I could put some more in WIS for better heals since I'm technically still a healer (White Mage -> Templar). What this resulted in was people running pretty different builds of the same classes and honestly, it was balanced pretty friggin well. The good thing about percentages is that it allows you to work with smaller numbers everywhere else. A max level character would have a base stat leveled at around 500 lets say. With buffs, that number would grow and if you had an attack spell that said "Apply 150% of your INT as damage" you would fiend over getting that stat boosted as much as posible if your role was damage. I'd actually much rather prefer this system, especially for skills since using percentages for them allows your skills to scale with your character.

Agreed! You might've taken my post the wrong way. Percentages are still very good for game design, especially in formulas, buffs, and other means as you were saying. The point of my post was why on percent based stats were bad, such as having Critical Strike, Evasion, Blocking, and other stats have only a % instead of a numerical value. I'm sure the "magic attack %" stat in Rohan had a base numerical value, or at least a hard cap that stopped players from abusing it.

 

Like I said, I do agree though. In my game for example, player's spells scale off either a percentage of their Spell Power stat, or a percentage of their weapon's damage, depending on their class. Static percentages are fine, but percentage based stats that are only percentages often only lead to trouble if you plan on adding new content.

[JimineyWhiskers]

6 minutes ago, MCADAMS said:

Agreed! You might've taken my post the wrong way. Percentages are still very good for game design, especially in formulas, buffs, and other means as you were saying. The point of my post was why on percent based stats were bad, such as having Critical Strike, Evasion, Blocking, and other stats have only a % instead of a numerical value. I'm sure the "magic attack %" stat in Rohan had a base numerical value, or at least a hard cap that stopped players from abusing it.

 

Like I said, I do agree though. In my game for example, player's spells scale off either a percentage of their Spell Power stat, or a percentage of their weapon's damage, depending on their class. Static percentages are fine, but percentage based stats that are only percentages often only lead to trouble if you plan on adding new content.

 

Ah I see where I made the disconnect! Makes total sense, percentage based stats = bad, I think we're on the same page, apologies for misreading that

[ningimbly]

Hmm i agree with most of your argument. However your still causing power creep by making old content worthless. Extra credits on youtube power creep it will help and theres two or three videos on it. And yes when given stat or percentile. Yes stat is much easier to balance