State of the Game: Personal Player Feedback and Stats on PvE Content [No Parsers Plz]

[Delmontyb]

Also, will this system take into consideration gear disparity? What about utility abilities and simply bloating numbers to get a good score? If it doesn't, well, I can't say this is a good system to implement.

Very good points and questions, personally, the way I would handle this is not to break it down and worry too much about gear score, etc, sure gear will impact your performance, but since you are just competing against yourself, having better gear helps you have better performance.
The other factors are the system would be for pre-maids because I could see issues with random groups, players going for a "high score" and upset that someone isn't giving them more buffs, or something.
And as for the data center aspect, it would look at that content, and rate you as a % of the job, or like others have said, a score on how you did. You can do this by removing utility from the equation, and many other options. It's more of a math problem from how I look at it.

[Sebazy]

@Dzian: This is precisely why I suggested to keep the system as simple as possible. Every single angle you approach is going to be a compromise that has a flaw in some way. Trying to cover every possibility is just going to end up with a bloated mess that’s far too complex for it’s own good. Thus it’s easier to just keep things simple, straighforward and make no pretenses about what this is. It’s not a precise metric. It’s not a parser. It’s just a simple rough score to give Little Timmy a hint as to if they are doing well or should be trying harder.

Having your potential score lowered because the tank pulled as slowly and safely as possible is fine in my eyes. You did less damage because your group didn’t pull in any kind of an optimal manner. Thus it’s reflected in your score. The dispairty isn’t that huge TBH. Tanks that can’t hold agro or keep mob packs together are far more problematic in my experience.

@JunseiKei - Read my post please? Potency per second ignores gear entirely. Bloating numbers? What do you even mean by that? Damage is damage as far as dungeons go, there are very few opportunities to pad on targets that don’t matter and they typically pale compared to what can be done on trash.

Again, please remember that no system is going to be perfect. Rather than just discrediting and decrying these sort of suggestions. Let’s try to keep this a positive discussion? Don’t like something? Rather than picking out contrived examples where it might not work, let’s try suggesting something that will.

[Delmontyb]

There's just to many variables involved

I would say Hard to implement, but for your questions, one metric could be mechanics, etc...
The big part of the solution would be that this content would be for pre-maid parties, and while yes, you would get different numbers, and your example of healers is spot on... the numbers aren't there to try and have you beat them, they are there for you to be informed.

For healers, you can track healing, and overhealing... over-healing being anything that was healed that didn't need to be.

I think the trick here is for random groups because if each person was trying to beat some score, it would add stress, but in a pre-maid, it could offer a lot of valuable feedback.

[Sebazy]

I honestly don’t think that raw healing is viable to track, even if you purely go off effective healing done, there’s no baseline and normalising the damage taken isn’t going to give a meaningful result either. It’d be better to track and grade a healer on their cooldown and oGCD usage Vs the time taken IMO. Getting good use out of abilities like Assize and Tetra in action is a fundamental part of good play, even in a healer that might not DPS especially well.

[JunseiKei]

@JunseiKei - Read my post please? Potency per second ignores gear entirely. Bloating numbers? What do you even mean by that? Damage is damage as far as dungeons go, there are very few opportunities to pad on targets that don’t matter and they typically pale compared to what can be done on trash.

I had read it, but that doesn't mean I fully understand how it'd work out. Gear and utilities still increase these potencies. IE: Take two WHMs and give them the same gear - change ONLY the weapon damage. The WHM with the higher weapon has higher heals, higher damage. Now add a NIN providing trick. If it's more than 4-man things we're talking about here, add in an AST or a SMN using Garuda's Contagion. This all leads to 'number bloat' - numbers achieved through utilizing compositions of jobs to give them an edge. Potency per second, though, makes it sound like if you found one person's rotation that gives you 'max score,' then you'd simply need to copy it.

If the goal is simply self improvement, there's no need to compare it against other people. The scoring system only need remember your score per instance and possibly give tips at the end. IE: Been through a few dungeons where a DRG didn't use Heavy Thrust at all (terrible, I know). At the end, the scoring would say something skills not properly utilized: Heavy Thrust (we know the teams play the game they work on, so it's not like they are clueless to what should be used). Something similar could be done for AoE pulls (where far too many melee opt for single target, despite there being 6+ mobs); at the end, scoring could say actions not properly utilized: melee AoE rotation (number where it's more efficient to swap to AoE rotation).

[Sebazy]

I had read it, but that doesn't mean I fully understand how it'd work out. Gear and utilities still increase these potencies.

Gotcha, I'll clarify a bit.

You're thinking of DPS or HPS. Potency per Second is a little bit different and is a touch easier to understand as it's simply using the Potency values found on the in game tooltips. For example, if I cast a Stone IV, that's 250 potency, with our gcd at 2.5 seconds, that cast is 100 Potency per Second or PPS. Holy across 3 targets is a smidge over 150 potency per second etc. It's straight forward enough to multiply this for self only buffs to account for cleric usage and such. Gear does not factor into this whatsoever.

Trick Attack actually wouldn't have any effect on this because it's increasing the damage received by a target rather than adding a percentage to your potency.

Healing gets a little more complex as just calculating a healers raw healing potency is entirely meaningless and has little to no relevance to good play. Maintaining Medica II for an entire dungeon would have many groups wanting to kick me, there's absolutely no way I would want any system to encourage that kind of play. There are other options but all of them have their own downsides. Scoring a healer's actual healing done (aka ignoring any overhaul) is going to be massively inflated by an underperforming group taking additional damage so scoring that isn't really worthwhile. Whereas normalising the groups damage taken to a baseline and then comparing the healer's healing done vs that is kind of pointless. Either you healed them enough or people probably died, don't need a ranking to see that

Thus the most foolproof way I can think of to score healing in particular would be to simply tally up the healers oGCD (stuff like Tetra and Assize) and cooldown (Shroud, DS etc) usage and compare it to the time the dungeon took. The more oGCDs and cooldowns you use relative to the duty timer, the better you score.

I like this approach because it doesn't utterly discriminate against a healer who isn't confident enough to go full tilt on their holy key just yet. But the good form and practices this scoring system would encourage will buy them the time, space and confidence to work on their awareness and potentially their DPS as well.

If the goal is simply self improvement, there's no need to compare it against other people.

My concept relies on comparing the players score vs a historical set of averages of players that have previously run that dungeon. This isn't to throw up a leaderboard or some such, the player would never get to see this, at least not initially with how I envisioned things. Rather it's so that the system would be self maintaining. If SE change a job, the averages will very quickly reflect this. They wouldn't need to go in and adjust all the values themselves on a mountain of old content. So yep, whilst it might be a little confusing, my idea is precisely that you'll only ever see your own score. It just uses data from other people's scores to work yours out.

Been through a few dungeons where a DRG didn't use Heavy Thrust at all (terrible, I know). At the end, the scoring would say something skills not properly utilized: Heavy Thrust (we know the teams play the game they work on, so it's not like they are clueless to what should be used). Something similar could be done for AoE pulls (where far too many melee opt for single target, despite there being 6+ mobs); at the end, scoring could say actions not properly utilized: melee AoE rotation (number where it's more efficient to swap to AoE rotation).

Agreed, something like this would be a fair bit more complex to implement, but I do agree that it'd be worth doing. The amount of players that play in a fundamentally flawed manner is shocking.

A fun fact, I recently levelled RDM from 50 to 70 in nothing but dungeons, trials and roulettes. I never play DPS roles, I'm still pretty horrible with my cooldowns, but I looked at the tooltips, did the basic math and I think I'm reasonably on point with how I handle my AoE. However I went from 50-60 in ~110-130 gear and never once did anyone out dps me. At 60 I had a 270 set which I wore all the way to 70. I only got out dpsed once in Doma. If it was a close run thing then I wouldn't even think about it, but it was pretty typical for me to be doing 2 or 3 times more damage than my co DPSer. Pretty crazy that it's like that on near enough every single dungeon across a sample size that must be approaching 100 dungeons.

[Dzian]

@Dzian: This is precisely why I suggested to keep the system as simple as possible.

How would you do that in an actually useful way though?

Giving a player a grade is meaningless if they can't see why they got that grade what it's comprised of. or how to improve it.

If a player does expert today and gets rank a but tomorrow only gets a rank c despite thinking he's played the same. He needs to be able to see what changed. And there is no simple or obvious way to do that.

Today that player might be doing most of the damage because the other dps is playing badly thus he gets an A rank. Tomorrow he might play exactly the same but do proportionately less damage because the other dps is making a bigger contribution so only gets a C rank..

You need details. There are so many variables some many of which aren't too obvious or even quantifiable easily but they can have a significant effect. If you don't have details players will see major fluctuations in There grades and be unable to quantify why. And If they can't quantify why. The system is useless. May as well just be rng. Roll a dice Today you get a b..

I generally think this is why yoshi never implements them because its simply impossible to make it accurate or even useful.

Even ACT isn't accurate. Back doing Susan ex for example I could have some major variance in my dps. And act wouldn't know why. It might pick up a lower uptime but it wouldn't be able to quantify why that was. Was I being lazy. Was I jailed by rocks every time or targeted for lightning. Was Susan tanked in a stupid place...

Don't really have more recent examples as I've fallen behind on content and tired of the treadmill

[Remedi]

I think that the easiest solution would be for SE to improve on the mentor system.
Create a real Ui for the system that allows you to open the Game guide and on it there'll be guides written by players.

Now I already see the problem of ppl writing BS guides, yes and that's why SE should meet good players and create a good partenership of sort if you catch my drift.
This way players guide to the game would be less obscure and more easily accessible.

It would also promote a more cooperative work between company and players.

I realize that this has many problems, but I think could be a good middle road

[Sebazy]

If a player does expert today and gets rank a but tomorrow only gets a rank c despite thinking he's played the same.

In honesty, if it even just gets that person thinking about what might have gone wrong to cause that drop in rank then that's a big win in my book as that's precisely what I'm gunning for. I don't want to bury people in metrics, I don't want to make people feel bad, I just want people to think a little more about their performance in duties and maybe take a little more pride in it where possible. The first step towards that is some super simple feedback that simply states 'hey, you did good' or 'hey, please try a bit harder', as currently this is something that the game is lacking in unless you parse or clock watch.

Regarding your other points, again, I can't stress it enough, my vision at least isn't meant to be an alternative to parsing for PS4 players, not even close. That's completely missing what I'm aiming for. I'm not talking about having ranks that are akin to gold/orange etc in fflogs where a few hundred DPS can make the difference between a 95%+ or barely breaking 80%. If someone can go between a A and C grade without anything going wrong on their part, then the ranks are set up much too narrow.

The system is useless. May as well just be rng. Roll a dice Today you get a b..

I don't buy it, sorry. My variance as a RDM between good and bad groups is somewhere between 5.3k and 4.7k in swallows compass, granted a portion of that is potentially my own mistakes as well. As WHM, it's actually a lot narrower with groups needing to be especially bad to the point of completely running me out of cooldowns and MP. The only thing that really seems to impact me is a tank losing control of a pull. Again, I'm not talking FFlogs here. If we said that my 5.3k was a perfect 100, then I'd consider an A to be within 70-80% of that, B to be within 50-60% of that and so forth. Steps big enough that an extra large pull or the opposite aren't going to destroy your rank.

Even ACT isn't accurate.

Mentioning this in the same breath as ACT is entirely missing the point. This is not meant to be that. Stop thinking fflogs percentiles and start thinking 3 star scores in angry birds

[JunseiKei]

~snip~

Ah, thanks for the break down. Yeah, that does make a lot more sense and sounds fairer for such a system.