ESO – Design Issues

While I am happy to have a respectable parse, I think of an interesting design issue this system causes for ZoS.

ESO is the first game I’ve gotten seriously into the end-game. I know WoW had mythic levels, scaling raids, and I participated somewhat, but was usually a healer and not trying to maximize my DPS char.

I don’t know what the spread is in other games, but here, there is a 10x or more damage difference between novices and experts. Heck, there is a 10x increase in my own dps since I started playing again, and a 5x difference from when I became interested in more than overland content roughly 6 months ago!

Some of that is gear and level and so on, but all the same, I think you could take a novice and expert and give them an identical char (say on a test server) and the expert will easily do 3 or 4 times as much damage, possibly more. In this situation, some players will barely parse 30K. Other will reach 60K, while a smaller group will do 90K or 120K; entirely based on their skill at performing the rotation (weaving and uptimes) using the identical chars as the other player!

The experience these groups have against the content will be VASTLY different. The 30K group will face every mechanic, perhaps struggle or fail. The longer the fight, the more time for mistakes and the more resource drain on supports. The 60K group will proceed carefully and progress, the 90K group won’t have much trouble, and the 120K group will burn stuff down so fast it is almost trivial to them Both the 90K and 120K group will seek vet trials, the 60K group will find it difficult and challenging, the 30K group has no hope.

Also there is boss design where something happens every 10% health lost – e.g. Bahsei in Rockgrove, who summons adds every 10% health lost. The 40K group can wear him down and deal with adds, where the 120K group might actually need to throttle themselves or they will have multiple adds from killing him too fast!

I think that’s why we see the normal, veteran, veteran hard mode, and then trifecta (speed run, no deaths, vet hard mode) splits. Offer a difficulty level to each group that they can opt into.

Anyway, it’s been fun “getting good” and I still have a ways to go. It took a lot of work to go from 20K to 90K and it’s probably even harder to get to 110K. But it’s a fun goal to work for. I feel like I’m in a good spot, doing decent damage and also decent at healing. I need more practice tanking, and I plan to address that as well. I’d just like to be competent at each role to give myself max flexibility trying to find a slot in a group.

ESO – Damage Per Second

MMO players are familiar with damage meters and measuring DPS.

In ESO, if you use a target dummy, the game will report your DPS after the fight is over. This method is essentially the same as the player using a stopwatch and dividing afterwards, except the dummy reports in-game. I never fought a dummy in WoW but it’s probably the same there.

Many MMOs allow add-ons and DPS meters are some of the most popular downloaded/installed add-ons available. WoW has many (and probably more), LoTRO has DPS Meter, ESO has CMX, etc. FF14 has them too!

In FF14, as I understand it, using DPS meters to shame or embarrass another player is a bannable offense, starting with a multi-day suspension progressing to loss of account. The official policy is 3rd party tools are not allowed; they are tolerated as long as the results are not used in-game. The FF14 dps-meter situation is along the lines of Fight Club: don’t talk about it in public.

It fits in what I think is the overall atmosphere of the game – track your DPS for your own information, perhaps if you are in an extreme trials progression group you opt into getting measured to identify needed improvements for the group overall, etc. But if you are a player in the roulette system, you aren’t there to be ridiculed or harassed (also, maybe you are trying a new job out and aren’t 100% familiar with the skills and rotation) so the FF14 devs have zero problems banning offenders. And frankly, that is awesome.

In ESO, the guilds I am in request a certain threshold of damage before taking you as a participant in vet trials. The way it works for normal trials is typically: sign up, be level 50 and cp 160+, and have some coherent gear (two five-sets and a monster set). A few spots are reserved for leaders and experienced players, which means 8 or 9 spots are up for grabs by anybody on a first come first serve basis.

When you have the requested gear, DPS level, and sometimes trial clears, you become eligible for vet trials. In this phase, there are often various tiers that trial leaders request when they post a run. Something like “tier 2 supports (meaning tanks and healers with uptimes thresholds and a larger pool of gear to mix-n-match), tier 2 DPS with clear, have 4 spots for tier 1 DPS with no clear”. The specifics obviously vary.

A common threshold is 60k DPS. I’ve also seen 40K, and then +/- 5K from either of those. When I took my first parse, it was barely above 10K. Yes, 10K. It appeared to be an impossible hill to climb.

This is a story of how I went from 10K to where I am parsing today. 🙂

My 10K parse was with a build I put together myself, drawing from various sources such as Arzyel, Lucky Ghost, Alcast, Hack the Minotaur, Deltia, Xynode… and more are out there. I would look over common skills and gear, and then work towards that. I also subbed in skills I liked, added more defense or self healing to be comfortable.

Stage 1 – 10K

In this phase, I was looking over suggested skills and gear, deciding what I liked, working on collecting the sets, having fun playing overland content, delves, and dungeons, but also interested in seeing what a trial was like.

I knew 10K was essentially minimum damage you could get by randomly hitting random skills. To sum up, it was terrible as far as group content was concerned.

This level of parse was against my own regular target dummy, the ogre we got for free from some event. It is better to parse against a trial dummy because that one supplies all the buffs and debuffs you can expect in a trial – that is, it gives group effects while fighting solo. It even spawns the occasional “restore resources” synergy, just like you would get in a trial from other player casting orbs.

The rule of thumb is you parse 2x against a trial dummy (21 million health) versus a regular dummy (6 million health). The trial dummy is preferred as a better test of your sustain and ability to solo fight through 21 million health.

So as a rough estimate, my 10K orge parse was the equivalent of a 20K trial dummy parse.

My rotation: use all the skills on front bar, swap, use all the skills on back bar, swap, repeat.

Stage 2 – 30K

To get to this level, I self-optimized my builds. Started to look at various DoT skills, and improve my rotation. I moved to fighting the trial dummy in a guild hall, which is more or less a free 2x. So I really moved up from parsing 20K to 30K through self-optimization and using a trial dummy.

ESO has a concept of light attacks and heavy attacks that I don’t remember from any other game. Basically, a light attack is a quick attack you can do for a bit of damage; a heavy attack takes ~2 seconds to channel, however it returns resources, and may trigger other effects. Both take no resources (magicka or stamina) to do.

In most other games I’ve played, in between using your skills while waiting for a global cooldown, your character auto attacks. As a player, you are only concerned with hitting your skill rotation on global cooldown.

In ESO, you can do something in between that cooldown – light or heavy attack. The ESO cooldown is 1 second, faster than any other game I’ve played, so it comes up frequently.

What helps gets you higher DPS is a combo of “light attack weaving” (putting one light attack in between each skill) and “animation cancelling” (starting the light attack and immediately cancelling the associated animation with another skill). As I understand it these are separate but related concepts and the playerbase generally refers to doing both as “weaving”.

Light and heavy attacks also help gain Ultimate, a resource needed to use Ultimate skills. If you never light or heavy attack (and don’t heal others or rarely block) you might wonder why your Ultimate charges so slowly or never…

None of this is explained well in-game.

There are a ton of videos on youtube on how to weave and animation cancel: Alcast has a nice summary, Xynode has a great video that helped me a lot.

So to get to this stage, I fixed up my build, started to practice weaving my rotation. By weaving well, you more or less sneak 2 skills into each cooldown. Before I didn’t grasp the animation cancelling portion and was doing a light attack and waiting a second, rhythmically performing skill, LA, skill, LA, etc. each taking a second.

My rotation changed: start with DoTs, use one or two spammable skills, use ultimate when it was up (now that I was light attack weaving I actually could use my ultimate far more often), swap bars as needed, and weave.

Stage 3 – low 50Ks

For this jump, I had to throw out my build and get a better one. For that I went to SkinnyCheeks (check the DPS guides tab). His builds and the kind of builds linked to by Liko cross into the high-end DPS builds. I migrated to DragonKnight as my main by this point, so I pored over the info.

I had enough Champion Points to pick the suggested stars/perks. I changed my Mundus stone. My gear was now a mix of decent dungeon sets, so the suggested gear was a new goal to strive for. I participated in my guild’s newbie trials series and got a few trials in every week, collecting gear.

I looked over the skill bar setup with great interest, and copied it. For skills/morphs I didn’t have, I worked on getting them.

My initial parse in this stage was in the low 50’s. I remember when I broke 20K with my own frankenstein build and was ecstatic that I essentially doubled my DPS by working on weaving and rotation. Now I got a 150% increase on top of that and it blew my mind.

New rotation: DoTs and bar swapping, use spammable until time to refresh DoTs.

~55K parse

Stage 4 – low 70Ks

I had a problem with Skinny’s build/rotation – sustain. I’d blow my resources out trying to do it and need to heavy attack for resources, or drink a potion, but ended up losing time to heavy attacks since I couldn’t sustain.

So I looked over the skills and selectively replaced a few, or cut them out of the rotation. Basically I shrank the rotation from all the skills suggested to fewer ones that had a good resource usage to damage ratio. I couldn’t make Molten Whip work so I used Stone Giant as my spammable.

I changed my food, using Lava Foot, because it was a single stat buff/sustain food. And the dummy doesn’t fight back so you can sacrifice health, defense, etc to parse higher. I changed the traits on my jewelry, from here on up you need the Bloodthirsty trait for the added damage.

I noticed that I regenerated a huge amount of resources when I used my Ultimate, so I made a point of working that into my rotation as a sustain method. It turns out the DragonKnight has a passive Battle Roar that is the reason.

With various tweaks, I got the modified rotation to where my sustain was sufficient. During the fight I would drain out of course, but by making a point of eating stamina sustain food (Lava Foot Soup-and-Saltrice), always grabbing the resource synergy the trial dummy provides, drinking the occasional tri-stat potion (free from daily login rewards so I have thousands of them), and using my Flawless Dawnbreaker Ultimate when it was ready, I was resource neutral. By the time my resources were nearly depleted I could recover a chunk and keep going. Essentially I stopped needing to heavy attack to replenish.

A trial dummy went on sale in the store and I bought it, so I could parse in my own house.

~72K parse

Stage 5- low 80Ks

By now I was wearing Arms of Relequen, Harpooner’s Wading Kilt, and had been using Kinras jewelry/weapons. For this jump I began subbing back some of the skills from the SkinnyCheeks build. Basically, my rotation expanded from 6 skills to 7 because I introduced an extra DoT into it. I was able to move from Stone Giant back to Molten Whip because an update changed resource usage, which helped – it used to cost 2000 magicka, and now is 1000 stamina and 1000 magicka (rounded off numbers of course).

I also finally farmed the Pillar of Nirn dagger so I changed from Kinras’s Wrath to Pillar of Nirn jewelry/weapons, the new meta. I made another skill sub, replacing a skill with Flames of Oblivion. I used Standard of Might instead of Flawless Dawnbreaker, which boosts DPS a bit because Standard of Might boosts damage including damage from the DoTs I am laying down.

Rotation is spammable and focusing on good DoT uptimes.

~83K parse

Stage 5 – low 90Ks

I am currently at this level, quite a climb from the 10K/20K days!

To get here from the previous stage, I essentially need 95% concentration on DoT uptimes. Start with DoTs and use spammable going back to refreshing DoTs. Well that sounds trivial but it is very tough for me to do. This level of DoT refreshing involves watching my skill bar constantly and refresh a DoT as soon as possible. It is mandatory to enable the setting that lets you watch timers on your backbar, or get an add-on that does it.

Back at the 70K level, I would generally refresh the back bar in one fell swoop – wait still all were expired then get them in one pass. That does not fly for reaching 90K. This means doing frequent bar swap, several in a row sometimes. It is a little exhausting in its own way. Bar swapping cleanly to go right into the weave takes practice as it is extra timing to get used to.

This level of DPS requires a good mastery of rotation, reflexively hitting the right button. Things like accidentally hitting the wrong skill or hesitating, and I can see the results in the real-time numbers CMX displays – a few seconds go by and DPS dips a bit. It’s recoverable but I can’t make too many mistakes without dropping under 90K.

Champion points matter but I went from ~50K and ~1700 CP to ~90K and ~1800 CP. By this point, I had enough to slot 4 for my build plus get the passive buffs from various unlocks, so those 100 CP didn’t play a part. I think it is somewhere around 1200 CP you can get all the key perks you need so I could have done this 500+ CP ago.

Some other changes suggested from resources in my guild – I swapped some Champion Point stars because I was at the crit cap and one buff wasn’t useful, while another was of even more benefit. Also swapped to a different monster helm for boosting one stat over another.

I am thrilled to parse in the 90Ks, even if the reality is more like this: out of 5 parses I will do 87K, 89K, 91K, 92K, 93K to average 90K but not always be in the 90Ks.

I cropped the pics from earlier shots to focus on the DPS number in the upper left. Here’s the full info the CMX provides for my all-time highest parse of 93583 damage per second:

~93K parse fight stats
~93K parse build/skills info

I’m no expert but I can extract the info I need from this. The good news is, I still have some upside available by tightening things up.

First, I can upgrade: gold glyphs, gold back-bar weapon, gold armor, gold jewelry (that is very expensive to do so it will definitely be last).

Second, my weaving average is 0.218s, lower is better. This is the average amount of time between a light attack and skill. Details for each skill are available, so lowering the average means working on firing off Barbed Trap faster for example, since that skill is a 0.3s weave. Meanwhile, Scalding Rune is 0.41s and even worse, Standard of Might is 0.57 seconds! OUCH those are really hurting my weave average and parse.

Third, examining the time column, and noting which DoTs aren’t getting refreshed as fast as possible. For example, Flames of Oblivion is 17.06 seconds, meaning I cast it just over every 17 seconds on average. But it is a DoT that lasts 15 seconds so I am losing 2 seconds of DPS due to letting it idle.

Might as well look all these up for my own benefit!

Noxious Breath26.1820~62.8
Venomous Claw26.1620~62.8
Barbed Trap26.1820~62.8
Flames of Oblivion17.0615~21.8
Scalding Rune31.6322~103.3
Standard of Might93.89??

There are subtleties here – both Barbed Trap and Scalding Rune have an arming delay, Carve can be refreshed early and stacked up to 30s. But I have long gaps and the largest offenders of the normal skills are Scalding Rune and Degeneration. Over a 4 min parse, those lost seconds are entire cycles of those DoTs that I’m losing.

Consider this, over a 4 min parse, 240 seconds, I would ideally do 12 casts of Venomous Claw (240/20), but I am only doing 9.2 (240/26.2). That’s what the Loss column is, how many extra skill cycles I could have under perfect refresh. Sorted this way, I need to work on Carve first, then Degeneration and Scalding Rune. Then reassess.

Standard of Might is a special case, as an Ultimate skill I need a different pool of resources to use it. So while it is up for 16s, I can’t actually cast it near that often. What I’ll need to do here is concentrate on using it ASAP every time, possibly compare with another parse where I use Flawless Dawnbreaker every time, and see if the damage buff for 16s is worth using FD at least twice as much (Ultimate cost is 125 vs 250). It could be I let SoM sit idle too long as well – at ~90 seconds I get 4 uses during my parse, maybe I could get 5?

My ratio of weapon attacks to skills is pretty good: 188 to 185. You want these about equal, meaning one light attack for every skill. If these are out of whack the reason is typically timing around the light attack – cancelling too early, or stated another way, firing off the skill too fast, before the light attack registers. Another timing issue to work on if this is the case.

Tighten all these up and I will see a DPS increase…. not sure how big though. I will find out. 🙂

Stage 6 – 100K?

93K is my peak now, going from ~10K to 93K is almost an order of magnitude improvement by a combo of gear, build, rotation, and practice. And what is mind boggling is there is still roughly a 33% increase available – the very best players can pull ~125K!!

I feel like I’ll be practicing at the 90Ks for quite a while, working on the various things I can see in my CMX data. And that’s completely fine with me. I’d rather always parse in the 90Ks than only peak there when doing several parses.

A side effect of parsing well is they take less time. Obviously. But the impact on practice is huge. It takes a bit under 4 minutes to parse a trial dummy at ~90K, so I can squeeze that in here or there or do 2 or 3 in a row. Parsing at 20K would take ~4.5 times as long, 18 minutes. In that era of DPS, by the time I finished a parse I didn’t want to do another and be locked in for another 18 minutes! Group up with a friend and they often need 5 mins to finish something, I can squeeze in a 4 min parse. But if the parse requires 18-20 mins, you gotta specifically put that time aside.

At this point you might ask whether or not parsing is a useful exercise. This comes up constantly. Yes, it is artificial, you have a build with low health, no defense or healing, fighting a dummy that buffs you, debuffs itself, supplies resources, and doesn’t attack or move.

However, it is generally accepted that parsing is the only usable baseline, because it is solo and the conditions are equal for everyone. If you can’t parse 50K against the dummy, there is zero chance you can parse 50K against a trial boss.

Practicing the DPS rotation has also paid off huge for my healing, because it is a similar process: rotate through HoTs instead of DoTs, constant casting, don’t refresh early, use spammable and/or resource recovery (heavy attack) when there is nothing better to do. For supports there is the concept of uptimes which I covered a bit in a previous post.

Parsing teaches you to get all your skills out there ticking away, NOT refresh early, and do something else in any downtime you have. In actual content, you will need to reposition and/or avoid mechanics so that’s your downtime filled.