Resistance needs to be converted to resistance rating

"

gh0un wrote:

Its not an objective opinion that a % based system is harder to balance than a rating based system, its a fact.

Um no it's not a fact. Just because you aren't smart enough to balance 0-100 instead of 0-10000, doesn't mean it's so for everyone.

You didn't provide any working example of what makes it harder, even if you claim to have done so.

Your "proof" was completely broken, because you can't even decide what you want out of the ice-damage-boss. You claim it's easier to balance him, but you can't figure out even how you want him to behave.

You want him to do ice damage, but you somehow want him to not be weak against people who stack ice resistance - which makes no sense at all, and even if it did make sense, changing the wording 5% to 50 rating doesn't prove that it's somehow hard or not hard to balance.

Rating implies it can be scaled to the encounter. You're thinking of it as a linear 100 rating -> 1% or something like that which is still a percent-based system. What he's asking for is a rating based so that it's not having 75% in all and being set for the rest of the game, but rather having to increase it to keep the reduction up.

Okay so the argument seems to be that flat %res is too easy to max and makes it difficult to balance damage.

While rating may be a solution to the never ending quest for higher numbers, it actually doesn't solve the balancing issue. If a boss is designed to be impossible with less then 50% resist X, then everyone will have to have more then 50% resist X or they wont be able to progress.

If the damage is balanced against 0% resist, then cap resists make it trivial, while if its balanced against cap resist then having anything less then cap is senseless.

So in this sense the problem is uni-damage themes making balance difficult as opposed to flat %. Encounters could then be designed to be "heavy" on a few damage types instead of ONLY-DamageA for balancing issues.

That said a full flat % system could work beautifully if designed right, for example a simple system like this:

Str Armor = %Phys res/+HP
Int Armor = %Fire/Cold or Lightning/+ES
Dex Armor = %Block/+less HP &/or ES (Remove evasion from the game, its basically block and they'll never balance it anyways)

%res scales up to a certain point, then the next "tier" follows the same res pattern, but has higher HP/ES. IE:

Leather Cap (5% phys, +5 hp) 10 STR
Plate Helm (15% phys, +10 hp) 50 STR
Leather Cap MKII (5% phys, +25 hp) 20 STR
Plate Helm MKII (15% phys, +50 hp) 100 STR

So lets say full plate set could net 75% physical base, and a caster could wear the lowbie plate with 50 STR and get that, the same resistance as a warrior with 100 STR wearing higher tier plate but who has vastly higher HP total.

Capping all resistances could be designed to only happen for a 1 trick build with probably no DPS. This system would be easy to understand AND encourage each player to mix and match their equipment in unique ways suited to their builds.

The mis-matching equipment might even encourage fashion concious players to buy apparel items to fix min/max their looks in addition to their toons.

"

Zarondd wrote:

Okay so the argument seems to be that flat %res is too easy to max and makes it difficult to balance damage.

While rating may be a solution to the never ending quest for higher numbers, it actually doesn't solve the balancing issue. If a boss is designed to be impossible with less then 50% resist X, then everyone will have to have more then 50% resist X or they wont be able to progress.

If the damage is balanced against 0% resist, then cap resists make it trivial, while if its balanced against cap resist then having anything less then cap is senseless.

So in this sense the problem is uni-damage themes making balance difficult as opposed to flat %. Encounters could then be designed to be "heavy" on a few damage types instead of ONLY-DamageA for balancing issues.

Thank you for making my point better than I have been able to.

Lets try to end the % VS Rating debate with another round of simple examples, but first of all, let's list the thing we have to consider, because this is not a simple, one-sided thing.

1) First and foremost, we must aim for a system that makes the game FUN. This may seem obvious, but we tend to forget about this as we design systems.

In an ARPG, fun generally implies killing stuff, upgrading equipment and leveling up, all of which should bring your character some amount of power, because ARPGs are all about your power level.

2) But it should also be noted that while your power raises, the CHALLENGE should raise accordingly, albeit at a much slower pace. This is a fundamental rule in videogames. The more experienced the player gets, the easier the game gets. We want to challenge him with every tool we have, but we want to make it happen slowly over a period of adaptation.

3) We must aim for a system that makes the game easy to BALANCE, because finding balance in a game is inherently hard, we don't need to add to it.

4) We want to make things SIMPLE. For equal gameplay, simple is always better. As players, we are very, very dumb. Few players can be bothered with complex calculus and algorithms. If we liked that kind of stuff, we wouldn't be playing games, real life has much more to offer in terms of complexity.

So, we want a system that's FUN, CHALLENGING, BALANCED and SIMPLE. Obviously it's going to be kinda hard to make all of those happen, but we can easily compare % and rating based systems.

1) The FUN thing about %-based system is the idea that you can max them out. Everyone likes to see a little golden 75 next to their resists in D2 (their use of colour for this is genius, btw).

On the other hand, it is very easy to fall in a trap, because you very often get this pleasure only once in the game (or once per difficulty level, in the case of D2)... There needs to be a mechanism that has your resists drop momentarily over the course of the game, so you can feel that "resist hunt" urge again.

I think resist rating would do a hell of a good job for that, because it steadily drops as you level-up, forcing you to search for new gear once every couple of lvls, and gear upgrading is probably the most important mechanic in every good ARPG I've played.

2) The CHALLENGE implied by resist systems is about overcoming elemental-based enemies. This is where the % system get hurt a lot. And it's really quite simple. %-based systems let you with very few design choices. You either design items/passives with a pseudo-fixed amount of resists on them over the levels(like the actual system), and that implies people will max out on resists VERY early in the game and make every subsequent encounter with elemental mobs a walk in the park, or you design it so that the resists % on items is very low at first, and goes up as you level, but that basically makes resist irrelevant for the first half of the game since, as has already been pointed out, 5% resist is too little a gain for it to be worth the sacrifice of a prefix/suffix slot.

On the other hand, a rating-based system is quite simple to handle. The players can keep high resists at all levels, but they cannot simply hold on to their lvl 11 rings forever, because low-level items would have low ratings and provide good resists only to low-level characters. The players simply have to upgrade their stuff as they go, or face important resists drops.

3) The BALANCING of this kind of system is something tricky. We want elemental fights to be challenging enough to have players equip resists. But we don't want players with maxed resists just rushing trough those all the time, and we don't want players with un-maxed resists to die instantly either.

We have shown before that %-based systems tend to end up as max-resist-cakewalks for this very reason: designers do not want to make the game too difficult for non-maxed players. So. How do we do that? We level the playfield between low and high resists. By using a rating system, we ensure that almost no player will be maxed at all times, we also never want players to have 0 resists, since this widens the gap needlessly (more on this later).

Basically what rating does for us here is stop players from maxing out their resists, unless they have capped their level as well. It makes the difference between the rich and the poor smaller, but still significant, and allows the designers to balance stuff around a middle ground of resists.

4) Last thing we have to deal with is SIMPLICITY. Here, as was pointed out numerous times, %-based systems are the clear winners. And this is not simply an issue you can dismiss by saying the 3 first points had rating as the winner.

Because in game design, simplicity is king. Even more so in F2P on-line games, because you want to appeal to the masses, and the masses hate complicated stuff. So, we quite clearly have to fix this, don't we?.

Here's what I would do with a rating resist system: make it scale with level like it should, but give it a cap and display the %resist for your level directly on the item. The idea is simple: make it clear to the player that rating and %resist are related, but make it equally clear that this is based on level.

Ex:
-----------------------
Plate Vest of the Inuit
At level 8
7% Damage reduction
8% Cold resistance
-----------------------
14 Armour
16 Cold resist rating
-----------------------

This way, you have a quick way to compare two items, and you still have the exact values somewhere nearby. The other thing I talked about is capping it. I think it would be best to have resist rating be a linear buff to your resists, both to avoid going into ridiculous numbers later in the game and to promote the idea (and inherent fun) of maxing out your resists several times over the course of the game. High level optimisation would then revolve around not wasting more prefixes/affixes than necessary on resists instead of a race to arms for high resist ratings on every item.

Heh... That was pretty long, wasn't it... Hope you don't mind my somewhat bad English as I'm primarily a French speaker. Oh, I totally forgot to talk about an important point.

When I talked about reducing the gap between highly-optimized gear and normal gear, I said something about making it hard to have 0 resist. That is quite important as the game will be balanced around an average resistance (except for end game content) and we don't want to kill players that forget about resists too much in the early-mid game.

So here's my solution: Make the stats give some resists. Have each 2 points in STR give 1 Fire resist rating, each 2 points in INT give 1 Cold resist rating and each 2 DEX give 1 Lightning resist rating. This gives both a sense of measure for what is actually 1 unit of resist rating and provides naked players with an inherent bonus to some resist(s) that adds diversity and closes the gap of optimisation that makes it so hard to balance resists vs Ele Dmg.

TL;DR: Read it, you lazy monkey, it's good food for thought about game design, ARPGs and player's needs!

Now I'm done. For real. Good night.

Nice analysis. This is what I feel the devs should read, and then try it out as an experiment.

So, to lay out the choices layed out here:

% is easy to understand, but not truly scalable.

Ratings are easily scalable, but people misunderstand them very forcefully.


I think the easiest solution would probably be a rating system much like armor in wc3; simple diminishing returns.

You should read my post. Around the end I give some justification as to why we should not use a system with diminishing return scaling for rating.

Basically, you want to let players max out their resists (because it feels good to them), you just don't want them to do it all the time without improving their gear, because that trivialises the elemental encounters for the rest of the game.

A rating system just increases the need for gear turnover (Ie once every 2 levels instead of 5), essentially just making the game a grindfest (while this maybe the namesake of dev team, I for one dont find it fun).

There is no fundamental change to CAP resists other then the need to regrind for gear to maintain CAP. This is because switching to rating from flat% does not change the amount of gear slots needed to cap. IF 1 ring can cap 1 resist, there is no difference between flat % and rating except for the need to regrind for gear more often.

This is why I say hard capped flat % is better. Ie you need at least 3 gear slots worth of gear + gear mods to 50% 1 resist. If it takes ALL gear slots to CAP 1 resist then suddenly passives, active skills and the bandit reward look alot better.

Hard capped flat% gear forces players to choose gear setups to min/max, while gear rating simply force players to grind gear often to maintain cap.

I agree that resistance should be a rating rather than %; it should be maxed over the course of the game, not from Normal difficulty onwards. Doing so would give more weight to resistance passives and the Bandit Quest option. It's not too confusing, armor already works the same way for physical damage, and it makes sense that resistances should work like magic armor. It also means you don't keep the same high-resistance trinkets that you've had since Normal Difficulty all game.