Hero Pattens

[Softmints]

Patterns? What are those?

A pattern is a term used in programming (or design in general) for a common solution to a common problem. A "hero pattern" is a guideline for designing heroes, which is always tricky. I figured I'd share a little about the patterns I use when designing heroes.

• Theme: Unless the theme is important for design reasons (Eph or Aquasoul), a hero should incorporate at least two thematic elements to avoid being a bland "fire hero" or "lightning hero".
  
• Ability Effectiveness: Abilities should never just do X. They should ideally scale between 0 and <some high value>, with typical effectiveness X. Examples are Judicator's new Trial By Fire replacing Judgement Ray: his new ability has greater potential, but less guaranteed effect.
  
• Synergy: If an ability does not interact with at least one other ability on the same hero, it does not belong on that hero. Examples of good synergy are Emperor, Chimera, and Geomancer; examples of bad synergy are Namek, Sivaz, and Stumpy.
• Item Choices: Every hero should have at least two viable directions to build in. If a hero has more than one "core item", something is probably wrong.
  
• Levelling Choices: Similarly, there should be at least two viable orderings for levelling up abilities. If one is dominant (as Psybeam was), something is probably wrong.

What do people think about these patterns in general? Are there any you would like to see added, or removed? Which heroes do you think are straying from the patterns?

[alyssandra]

Hero design is one of the most interesting things one can do. Every game has a pattern, as you've mentioned, and an "archdesign" that all heroes must follow so that they fit into their specified game. What I mean by this is, for example:

Age of Myths has three classes of heroes: ranged, caster, melee. It is meant to be balanced because each have (literal) pros and cons versus each other. Whether or not people think this type of system works, I don't know. Casters get their late game damage and effectiveness from activatable items with spells (Silence, Holds, Fan of Knives, Holy Light). Auto attack damage is generally high for all three classes and contributes to a majority of the damage.

In Tides of Blood, auto attack damage is incredibly weak, it may take 2-3 heroes to take down one with a combo of spells, heavy focus on aiming and fluid skill usage. Teamwork is a must, and you'll never see one hero completely dominate and steamroll a team.

In AotZ, it's mostly spells doing the work, with no emphasis on auto attack. Lots of flashiness, with little care for actual balance and fluidity. However, using some spells and combos can feel fun and rewarding, admittedly. But items make everyone equally effective, and damage is fair among most heroes.

The point I am trying to make is that each game listed has it's own "archdesign" and the systems and heroes in each game are unique to that general pattern. A DoE hero wouldn't work in DotA, a ToB hero wouldn't work in AotZ. It all feels relative to each other, and that any one specific hero belongs with his peers in that pool. DoE actually had a problem where a hero didn't feel like it belonged in DoE, so it was removed.

However, where I think RoW falls short is the split path of heroes. I never agreed with heroes like Kukulza and Rasputin being in the same game as an Iofur, Pythoness, Stumpy, etc. Kuku does not feel like he belongs near an Iofur. This creates a natural imbalance, because if Kuku can deal as much (usually more) damage than the others without needing to aim his skills, then the design vision falls apart (in my mind).

As for these listed patterns, I don't mind heroes not having synergy in their own kit. There are countless heroes with no synergy, not just in RoW, that are some of the funnest around. It may create some short term entertainment, but there are only so many times one can Rock Prison and Ancient Power before he or she begins to want something more satisfying.

"Levelling choices" is an interesting point. DotA and League seem pretty good to the extent that they allow for elastic skill builds. I imagine this is because of the larger playerbase, and the fact that metas become established and changed around.

But at the same time, Pyth is so focused on Arc Arrow, that I couldn't imagine maxing a skill like Blindside. For Abra, why not max Thresher over his passive or charge? I think RoW falls short here as well.

As for ability effectiveness, I don't mind a spell being a point-click damage spell that doesn't need to be aimed; it just needs to fit within the games chosen "archdesign." In RoW, we have some aimed spells, then some spells like Dash and Heretic wards. I think it needs a firm stance on abilities before it can really grow in this way.

In conclusion, ToB heroes work for ToB, AoM heroes work for AoM, and I'm not sure some RoW heroes work for RoW. There isn't one ToB hero that can three shot another with auto attacks, because it is a spell based game that requires dependability on others to accomplish goals. AoM has actual counters to each approach a player takes, as bad of a reputation the game may have. An AotZ or DoE has this flashy style, where ecclectic heroes can really do whatever they wish, but no one hero feels completely out of place or necessarily at a disadvantage.


RoW is way too much of a mixture of many aspects, and should probably be put in a single direction. I'm not saying a  hero like Rasputin or Kuku should be "deleted" but redesigning the usual suspects would certainly help improve gameplay, atleast in my opinion.

[Gahn]

Part of the difficulty in leveling up abilities is that the cost increase per level discourages you from leveling up a single ability unless it's extremely powerful (Arc Arrow) or a passive.

I agree with Sep that some RoW heroes don't feel very "RoW-ish." A lot of the offenders, in my opinion, are Agility heroes (Kuku, Vamp, Assassin) because the abilities that they have, combined with the late game scaling provided by agility, tend to force the skill requirement on their opponents.

The one thing I'd like to see, would be to have the Effectiveness scale made a little extreme (assuming a hit with the ability), making the low effectiveness instances more powerful, and the high effectiveness ones a little less extreme (looking at you Arc Arrow).

[Softmints]

I think many of RoW's older heroes were good for their day, but haven't aged gracefully into 2013, where my patterns are more demanding and your expectations are higher.

@alyss: Heretic and his ilk will be addressed eventually. As has been the trend, I tend not to focus on stable heroes like him, even if they aren't good for the game.

Abra, Kukulza, Miney, Stumpy, Garry, Sivaz, Shadowdancer, and Pyth all fall under this category. Eph and Aqua are more immediate concerns right now.

Synergy is a creative limitation that I embrace because it helps me make better heroes, not because I think synergy > no synergy (though exceptions are rare).

Gahn: I would consider Assassin and Vamp to be fairly high skill-requirement heroes. Certainly late-game they can get out of hand, and some abilities have room for improvement (Guillotine, Toxin Bomb) but in general? My current plan for Arc Arrow is to improve the cc from a wound to something else, so it's more useful at setting up kills for her allies (at any range). I think that accomplishes what you wanted for Arc, but I don't know that it'll be my solution/approach in general.

My intention with growing mana costs was to encourage spreading points around, but allow people to buy +mana so they can use high-levelled spells earlier. How would you like to see abilities scale in future? Implement influence asap and have decreasing cooldowns? Decreasing inf-costs? Less aggressive mana costs?

[Gahn]

The current system works reasonably well, provided that the hero doesn't have a passive. Passives are largely the culprit because they provide all gains, and no cost increases (even if those increases are sometimes minimal, such as ignition).
I don't mind how abilities become less efficient in terms of total effect per mana spent, but the current increases can be too steep, depending on the ability. For example, Ty's Attract increases in cost from 70 to 205 from level 1-4, but only increases in damage from 120 + (1.1) to 180 + (2.3). Admittedly, the damage at level 1 is pretty damn high, but that's still a 192.5% increase in cost for a 50% increase in the base damage, and a 109% increase in the attribute damage.

Inf would be cool to have implemented, because it would make the ability costs easier to tailor to each hero, because you'd be dealing with percentage of pool instead of flat amount. It would also be a lot of work to implement, because it'd significant changes to the item, and possibly to the attributes (would Int become AP?).

Two other possible ideas would be to either change ability leveling so that you basic abilities needed levels 1, 4, 7, 10 instead 1, 3, 5, 7 in order to be leveled up, or introducing static Attribute modifiers for all abilities, so that even level 1 abilities would always scale well (but would probably need significantly reduced based damage at earlier levels). These are just thoughts that popped to mind, so I figured I'd toss them in.

[Grumbletok]

Should probably have read and commented on this earlier, but haven't had the time... Anyways.

I agree with you Alyss, that some heroes doesn't feel exactly like they fit ATM with the theme of RoW, which is supposed to be high-skill, high-reward. I had a pretty long discussion with Soft about this before, as it falls under the category of "Mage vs Warrior", from way back in games. I think one main solution, in this case, would be to remove that heroes primary attribute grants Attack, and instead just give heroes flat Attack gain per level, plus damage you buy from items. It would make it better for several heroes to build into their abilities when they build stats, rather than just building because it makes them so good at just hitting the opponent in the face. For an example, I don't think Kuku's abilities are wrong in RoW, they demand aiming and are fun to use. The problem is when those abilities only becomes a compliment to his autoattacks. I find Abra to be more of a problem, when it comes to design, since he demands SO little skill, and always has. He's always been my number one recommendation to new players who want to try a "easy" hero, because he's currently all about "Build massive amounts of Strength. Become neigh immortal AND gain amazing amounts of damage. Wreck havoc.", and this without the demand of skill which should be required. I simply don't think it's good if, let's say, Lothar's, is strictly better on one hero than another, not because it's unique, cleave component is better, but because it actually gives more in raw stats than it does to the others. Lothar's isn't made to be a +40 AD and health, reg and cleave on that, it's made for it's +20, health, reg and cleave. Removing the "primary attribute" would solve a lot of problems like this.

I don't know if I missunderstood something here, but I don't think a BUFF to arcarrow is what's in order, Soft. (My current plan for Arc Arrow is to improve the cc from a wound to something else, so it's more useful at setting up kills for her allies (at any range)), but rather her problem is that Arc Arrow is TO attractive ATM. I'd rather see it gain possibly a little bit better CC, with a pretty high damage reduction, and a buff to her other abilities, so that Arc Arrow stops being the hands-down best leveling up choice.

Personally, I dislike the extreme mana costs a LOT. If an ability is supposedly SO STRONG that it needs to empty half your mana pool, so you can't cast all your other spells, then it's power should be scaled down to fit a reasonable mana cost, not just going with "Well, if I overprice it it'll be balanced". It's one of my main problems with DotA and it sadly stands here in RoW as well. I want to be able to cast my spells and be active, not spend a majority of my time in the game just saving up mana to be able to use one more ability and then nothing more until I've reg'd half my mana pool again. I would prefer to just see all spells getting a normal amount of mana as a cost. With that I mean, NO non-ultimate over 120 mana, no ultimate over 250 mana. If a spell is to broken like that, then that spell should be nerfed to fit the cost, not the other way around.

As for passives, I would actually like to suggest making LoL as a role-model here. Each hero got one passive, which levels up by your hero level. This would make passives much less of a "I level this to keep my mana costs down" option and more of something bound to each hero. This would also mean that no hero should have more than one passive. I know this maybe change the amount of spells on some current heroes, etc etc, but, it's a suggestion, and I think it would be good for the game. Maybe some abilities could be reworked into passives, and so forth?

As for the point with some spells being obvious leveling choices over others, I think it's a problem that a lot of games have, in the early days of a hero, and then it should be fixed pretty fast. Some heroes could definitely use a look into on this point. I can understand if it was some wombo-combo hero, which was based on that you level on ability and then you have tohers circle around that to 100%. But, I don't think most heroes are like that in RoW, if any. However, some heroes got abilities that just take up to much attention. Some examples, from my point of view, would be:

Attract, on Ty. Sure, you might want one more owl or something, etc etc, but Attract is the strongest one to simply put your effort in.

Seismic Spike, Obelik. Seriously, this spell just get SO much better from you maxing it up. Sure, I can sort of understand wanting to max another spell first, but mostly, I think Seismic Spike will be the best. This is MY opinion.

Telekinesis, Psychic. I don't care what anyone says about Psybeam, Telekinesis is such a powerful tool, and leveling just makes it SO much stronger, I don't really see why it wouldn't be maxed out first.

Thunderbolt, Gyatso. Yeahhh.... I guess Nimbu is the the challanger here, but really? It's all about the Thunderbolt.

Carving Scythe/Switcheroo, Flisk. Okay, pretty tough, but I think Carving will ocme out on top, since you want the damage to spike early, but I can see Switcheroo to be maxed out first as well. Either way, any of thse will be the dominating one, none of the other spells are really attractive to get higher fast.

EXES! DOUBLE EXES!, Kukulza. Eherm, I mean, Cross Slash. Yeah. This is such a strong ability, and I don't really feel the benefit of maxing the others over it. POSSIBLY Abscond since it's mana cost doesn't increase, but meh.

Triumvirate, Mr.Winterchill. Yeah, I mean, really? Possibly the shield if you need to be on the defensive, but... Nah. This ability IS thou one of the few I feel could really become a center-of-the-kit spell, which was meant to be maxed first. But, atm, it's just so much better to do so, rather than a necessity.

Perforate, Ogake. I don't know if this is unusual, but it is without doubt my number one choice on Ogake. You just gain SO MUCH from leveling it, and it IS his best spell, IMO. Not coutning Mortal Mastery, since that spell is... Yeah, it's in a class of it's own. And an ultimate.

Thresher, Abra. Thresher, Thresher, Thresher, Thresher, Thresher, Thresher, Thresher, Thresher. Abra's only skill-based skill, and the only thing that is attractive to level up, really.

Arc Arrow, Agnes. This has been mentioned before.

Ricochet, Shaka'Zhan. This is for me, who play Autoattack Troll. But really, you don't NEED to level his other skills, they're pretty powerful as they are.

Shadow Ball, Nadir. I've tried to avoid heroes I don't play alot, but this one just feels so obvious to me. I don't know why you would max the others, if anyone can tell me, please do.

TAXES! TAXES! AH-HA! AH-HA! ... No, really, I dunno about Emperor, I just love this thing. :3

So, yeah. Those are the abilities I feel I can pretty safely say that they dominate their heroes a BIT to hard. Pretty many, eh?

Oh, and I apologias if I seem less serious, the further down you go along my post, I just get random ideas that pop in, this post IS meant to be taken seriously.

[Dracula]

[Slightly Off-Topic]
Honestly, the interesting thing I've noted as RoW's gone through the last year or two of development is that on the whole, heroes' abilities (at least, the heroes that aren't new-ish) have generally tended on the whole to stay relatively faithful to the RoW of two years ago, except with some numbers change. Even then, there have been a lot of heroes complained about over the last few months, despite the fact that they are not really significantly different in balance than they were in a prior age. Kukulza was definitely considered strong back on the old terrain, but he'd only changed fairly minimally between then and when several people began highlighting him as a balance problem recently instead of simply "good".

I feel Tempest suffered more or less the same fate, and there have been several other heroes and nuances of gameplay that, despite apparently being completely non-problematic on the old terrain, suddenly surfaced on the new terrain. So if the skills/stats on most of these offenders are functionally identical to how they were before, either people just randomly decided not to acknowledge there was a problem until now, or something else changed...
(My bets on the latter!)
[/Slightly Off-Topic]

While there are some heroes that have some very obvious (literally, no other competition) skill build orders, I don't think the problem in that respect is as universal as it seems. Take Psychic, for example. I don't think I've ever maxed Psybeam or Telekinesis first. For my playstyle, it doesn't give nearly as much benefit as maxing Barrier (50 second CD at level 1 vs 20 second CD at level 4), which has infinitely more utility and can be utilized for significant amounts of damage or rescuing in ways your opponents will never see coming, unlike Psybeam. Possibly some of the most satisfying kills I've ever had in the entirety of my playing RoW have been casting Barrier on enemy Voidmasochists or towers to send critically-weakened enemies flying into the trees as a finisher. Of course, I completely accept that I'm an odd one out in this case, and almost everyone else just prefers to harass with Psybeam, but I'd hope my games as Psy provide at least a corroboration of such a methodology's viability, especially since I do this without even bothering with the ever-loved Lothar build.

Contrast with Heretic and skills like Grim Anchor and Jinx Totem, which have scaling aspects that are very rarely ever relevant. (When am I ever going to have time to charge up for the full 10 second chain in a case that isn't win-more? Will a few hundred more health points really matter against endgame damage?)

Nadir's a weird duck. Shadowball's his only viable skill going into endgame (the others are too situational/win-more, save for Cakewalk, which is usually used to defy death), but it has the least useful scaling out of his skills early on, as only the mana burned and cooldown improve. Conga Line is currently almost useless, but it becomes (insignificantly in most cases, admittedly, but still a better improvement than leveling Shadowball)more relevant to level than Shadowball early on when stat contributions to mana burn are near-meaningless. There's not much point in leveling Cakewalk besides cooldown - it doesn't do enough damage to outpace the ball as DPS, and if you're maxing it first for defensive purposes, you're sort of just walking onto a lane to get attacked before running away several seconds later. I always max the innate first, as it improves Shadowball's damage output more than leveling Shadowball does (more movement speed allows you to juggle more safely/more accurately, getting more hits in), and the movement speed increase is relevant enough to dramatically improve Nadir's response time to situations that require his presence on another lane. In terms of actives, though, yeah, Shadowball's his most frequently used active because it's his most reliable.

I don't mind extreme mana costs as a general trend, simply because many of RoW's abilities have such high potential reward that it would essentially turn RoW into Imba Wars if they became spammable. Admittedly, as applied to levels of spells specifically, some spells do seem inordinately expensive based on how much they actually do, as their potential is often much lower than, say, Telekinesis, for which an absurdly high mana cost makes sense, even at level 1. I think there needs to be a bit more consistency in appropriating mana costs of abilities according to the risk-reward spectrum, as well as taking heroes' mana pools into account.

As for synergy, I'm a bit torn. I like the idea of synergy, as it generally feels satisfying that my abilities work together if applied correctly to create a result superior to using both abilities independently at separate times - it encourages more investment by players, as well as skewing the game a bit higher on the risk-reward spectrum. Unfortunately, there are some synergies in the game right now that lend themselves extremely poorly to properly corresponding with risk-reward. Geomancer is one such example. Stone Prison into Ancient Power or Seismic Spike is a very, very stable and consistent synergy that has been proven to work very well, even with the presence of several Legion heroes in the game that have various means of countering it. The problem here is that it's both stupidly easy and very rarely risky on the part of Geo to perform this synergy. Unless the enemy has an escape, Stone Prison provides a guaranteed followup. There's almost no risk there (sans player error), and the skill requirement that would have to be present to justify the inherent lack of risk for such a high reward isn't present either.

A good and very fun example of synergy (although public opinion will probably disagree with me)is Psychic. Instead of Geo's guaranteed Ancient Power followup if Stone Prison lands on a viable target, we have things like Telekinesis, Barrier, and Miracle Eye. While Tele and Barrier have absolute limits (numbers) to their potency, their tactical effectiveness is potentially godlike if the player is simply creative enough. Besides the very obvious uses (Barrier or Telekinesis to save a friend/yourself or send a friend/yourself after an enemy), there are several very nuanced uses that many people just don't consider:

A: Barrier or Telekinesis to shove your opponent out of your line of sight, lining up a Miracle Eye.
B: Barrier on a troop next to an enemy for an unexpected displacement that can't be responded to in time.
C: The classic all-in "Barrier/Telekinesis Surprise!"
D: Barrier has interactions with many allies' abilities. It can be as simple as turning Ancient Power into "OMGWTFTEHPOWAH", mischievous as taking advantage of invisible allies, or as brutally complex as throwing one onto Kukulza just as he uses Cross Slash for a quite fantastic light-show across whatever lane he's on.

Of course, the list is hardly exhaustive. Even item interactions alone could constitute a list. Either way, all four options I listed above either require an element of risk, skill, or creativity simply not found in the toolkit of, say, Geomancer. A synergistic interaction that does not meet at least one of those conditions will usually lead to either just un-fun gameplay or all sorts of degenerate nonsense. Blindside could be argued to be a symptom of such a synergy relationship, often rewarding Pyth for what she benefits from doing anyway (instead leaving the majority of the application of the skill to Pyth's teammates, which is admittedly interesting but still makes it just a bland damage-booster for Pyth).

Basically, I see good synergies as being synergies that aren't bland, aren't necessarily even obvious, and actually take a bit of thought to set-up or execute. Good synergies result in gameplay that encourages high-risk high-reward outcomes, encourage initiation to achieve said outcomes, and provide some requirement of tactical understanding. While not all synergies (admittedly, very few) capture all of these qualities, I generally tend to rate a synergy's effectiveness based on how many they do capture, and how well they are captured. Not every kill needs to be the most impressive thing I've ever seen to be good or fun - Blizzard Barrage + Rollout is probably one of the most inherently straightforward synergies in the game, but the amount of skill and thought to actually capitalize on such a synergy consistently is enormous.

Something as simple as, say, Plunder + Cinder is not a good synergy. It requires no skill to set up (melee autoattack once, woo?) and almost completely ignores the risk-reward continuum, as well as being very anti-fun. Hexwerk Trap + Pressure Mine is not a good synergy. It requires minimal thought and is essentially "fire-and-forget", and is rarely ever fun to be on the receiving end of.

On the other hand, something like Particle Accelerator is an amazing synergistic tool that works both within Trek's own skillset and outside it with allies to create some nasty and incredibly fun results. It takes good planning, good aiming, and a decent resource investment (mana/cooldowns/time spent planting mines) for what could potentially be an extremely high reward - if your aiming and planning were correct. If not, you just wasted a decent amount of time and probably at least 33% of your mana pool for naught.

I guess I technically incorporated discussion of ability effectiveness into synergy discussion. Oops.

[Softmints]

Gahn: Putting points into passive abilities isn't a problem, but (like any build) if it's too common, there's probably something wrong. I don't think forcing players to spread points, or increasing the potential of one-point wonders is much help. The solution will be sensible tuning.


@grum: Switching Arc Arrow to be more cc-focused is not intended to make players try her other abilities, because everyone knows she has nothing else to fall back on. Once her kit is more complete, I can start to address how power should be distributed within the kit (but I can't do that until I have the kit in front of me).

Due to RoW's aiming philosophy, abilities must be moderately expensive to discourage spamming. I would prefer to be using Influence which is more dynamic, but for now all I can do is tune mana costs to individual heroes. If you are having difficulty with mana, try starting with -i2 courage -i adelaster.

I don't agree that all of the builds you mention are optimal, but there are definitely faults in some heroes because I haven't offered viable alternatives. I am working on making those alternatives available.


@drac: I am going to start calling issues like wonky scaling and dysfunctional ability sets: "2010 problems". Let's try and have all that ironed out by 2014.

I don't think Geomancer's synergy is a bad thing. If Stone Prison is too "risk-free", that's Stone Prison's problem and should be addressed separately.

Namek's Plunder, Miney's Hexwerk, and many more abilities are filler. I am actively working on replacing them; the changelogs reflect this. Would that it were faster!

It is interesting that the majority of the synergies you mention involve displacement.