Breaking the Majority’s Silence: A Job Criticism Thread

[Jeeqbit]

That's just not how MMO development works =(

Yoshi-P suggested it takes a mere 2 hours to create a Dragoon action in the noclip documentary. I also recall that one person made lots of WoW classes as well. It seems pretty common that not many people are needed for it.

https://www.youtube.com/watch?v=aoOI5R-6u8k&t=257s

He said if "for example, we were creating a job. Let's say the Dragoon job. At first glance, you might think it's just one single task, creating the Dragoon job, but it's not. Because of course you have the job, but you have the different actions that are assigned to that particular job. And so you would need programming for a single job action. So we'd figure out how much time it takes to do that. In order to do that you would need the specification documents. So say, for example, we take a look at the task of creating a specification document. It's two hours to create. Then a request for review is sent to me, and then my review takes about 15 minutes."

So let's multiply 30 job actions by 2 hours. That's 60 hours. Let's multiply 15 minutes by 30 job actions. That's 7 and a half hours for a total of 67 hours or 8 work days lasting 8 hours each. Let's multiply 8 work days by 19 jobs and we get 159 work days or 31 work weeks or 8 months. Then we return to the fact they've had over 2 years and have multiple people working on them, drastically reducing the timeframe.

That whole example is extreme, because most job actions remain untouched anyway. They usually just make a few minor changes and add 3-5 new actions or traits.

I agree with you that it seems like designing things should take a long time, but it doesn't seem to in practice because the other teams handle the other tasks such as icon design, UI function and QA. This is also backed up by the apparently small job design team and the examples Yoshi-P seems to have provided, which may or may not have been accurate since they were examples, but the small job design team does seem like it would back it up doesn't it?

[SieyaM]

Yoshi-P suggested it takes a mere 2 hours to create a Dragoon action in the noclip documentary. I also recall that one person made lots of WoW classes as well. It seems pretty common that not many people are needed for it.

https://www.youtube.com/watch?v=aoOI5R-6u8k&t=257s

He said if "for example, we were creating a job. Let's say the Dragoon job. At first glance, you might think it's just one single task, creating the Dragoon job, but it's not. Because of course you have the job, but you have the different actions that are assigned to that particular job. And so you would need programming for a single job action. So we'd figure out how much time it takes to do that. In order to do that you would need the specification documents. So say, for example, we take a look at the task of creating a specification document. It's two hours to create. Then a request for review is sent to me, and then my review takes about 15 minutes."

So let's multiply 30 job actions by 2 hours. That's 60 hours. Let's multiply 15 minutes by 30 job actions. That's 7 and a half hours for a total of 67 hours or 8 work days lasting 8 hours each. Let's multiply 8 work days by 19 jobs and we get 159 work days or 31 work weeks or 8 months. Then we return to the fact they've had over 2 years and have multiple people working on them, drastically reducing the timeframe.

That whole example is extreme, because most job actions remain untouched anyway. They usually just make a few minor changes and add 3-5 new actions or traits.

I agree with you that it seems like designing things should take a long time, but it doesn't seem to in practice because the other teams handle the other tasks such as icon design, UI function and QA. This is also backed up by the apparently small job design team and the examples Yoshi-P seems to have provided, which may or may not have been accurate since they were examples, but the small job design team does seem like it would back it up doesn't it?

So is he saying to create the documents for review, which I assume would not include the actual programming but just the estimated time for it, would take two hours to create and then be sent in for review. Which then would either be given a go ahead and started or could be sent back with notes and then have to be resubmitted through this review process again. Even if programming is included in this two hour time, your estimate is assuming all actions and changes are approved on their first submission which probably isn't true.

[Jeeqbit]

So is he saying to create the documents for review, which I assume would not include the actual programming

Right, but when people talk about the job design team, they are talking about the job design team, not the programming team. So this piece of information is relevant.

your estimate is assuming all actions and changes are approved on their first submission which probably isn't true.

Even so, the only jobs that get 30 new actions are new ones. The old ones get 3-5 new actions/traits, with hardly any changes usually, unless they are being completely revamped. That on its own would reduce this estimated time to a fraction of what I put and then if you split in half between 2-4 people it would become even more fractional than that to design them.

So when people say a few people isn't enough for the job design team, even the most extreme estimate I come up with doesn't fit the 2 and half year span between expansions.

[Maltothoris]

Expect it does show trhat they are short on ideas. Here was this interview for endwalker https://www.ign.com/articles/final-f...hida-interview

All in all, Yoshida says development for the Reaper class took about three months from conception to fleshing out, while development on Sage took half a year.

It took them half a year to make a class that was 60 percent like SCH.

[Jeeqbit]

It took them half a year to make a class that was 60 percent like SCH.

60% is generous. It feels more like 90% SCH to me, as someone who played SCH for years before it.

The main difference is the pet heals are applied by your attacks instead, you have more clear control over who those heals affect due to Kardia, and some of the heals and Aetherpact have become heal potency increases instead to buff Kardia, a gap closer, an attack instead of Chain Stratagem, Aetherflow for heals and attacks being separate, and needing to unlock shields and dots with Eukrasia. All of that is so minor and everything else has something resembling it on Scholar.

They must be referring to the initial idea for the weapon, the aesthetic, animations, icons, models and lore of it, or have treated it as a background project, to take that long on it, because as you said it's functionally more of a Scholar clone than some people realize.

I did some comparing for anyone interested and well, they are the same.

Kardia/Summon Eos, Eukrasia and Pepsis/Emergency Tactics, Dosis/Broil, Eukrasian Dosis/Bio, Diagnosis/Physic, Eukrasian Diagnosis/Adloquium, Prognosis/Emergency Tactics+Succor, Eukrasian Prognosis/Succor, Ergeiro/Resurrection, Physis/Whispering Dawn+Fey Illumination, Phlegma/Energy Drain, Druochole/Lustrate, Dyskrasia/Art of War, Kerachole/Sacred Soil, Ixochole/Indomitable, Taurochole/Excogitation, Haima/Summon Seraph, Panhaima/Consolation, Rhizomata/Aetherflow, Krasis/Protraction, Pneuma/Dissipation, Holos/Fey Blessing, Zoe/Recitation, Soteria/Aetherpact

SCH:
Deployment Tactics
Chain Stratagem
Expedient

SGE:
Icarus
Toxikon (and Addersting)

[RitsukoSonoda]

60% is generous. It feels more like 90% SCH to me, as someone who played SCH for years before it.

The main difference is the pet heals are applied by your attacks instead, you have more clear control over who those heals affect due to Kardia, and some of the heals and Aetherpact have become heal potency increases instead to buff Kardia, a gap closer, an attack instead of Chain Stratagem, Aetherflow for heals and attacks being separate, and needing to unlock shields and dots with Eukrasia. All of that is so minor and everything else has something resembling it on Scholar.

They must be referring to the initial idea for the weapon, the aesthetic, animations, icons, models and lore of it, or have treated it as a background project, to take that long on it, because as you said it's functionally more of a Scholar clone than some people realize.

I did some comparing for anyone interested and well, they are the same.

Kardia/Summon Eos, Eukrasia and Pepsis/Emergency Tactics, Dosis/Broil, Eukrasian Dosis/Bio, Diagnosis/Physic, Eukrasian Diagnosis/Adloquium, Prognosis/Emergency Tactics+Succor, Eukrasian Prognosis/Succor, Ergeiro/Resurrection, Physis/Whispering Dawn+Fey Illumination, Phlegma/Energy Drain, Druochole/Lustrate, Dyskrasia/Art of War, Kerachole/Sacred Soil, Ixochole/Indomitable, Taurochole/Excogitation, Haima/Summon Seraph, Panhaima/Consolation, Rhizomata/Aetherflow, Krasis/Protraction, Pneuma/Dissipation, Holos/Fey Blessing, Zoe/Recitation, Soteria/Aetherpact

SCH:
Deployment Tactics
Chain Stratagem
Expedient

SGE:
Icarus
Toxikon (and Addersting)

Probably skill names as well added to that time to make SGE. For the most part RPR's are pretty straightforward.

Alot of SGE's design seemed to focus on flash and style. Probably an attempt to draw ppl into playing a healer role job by having it let players play around with Gundam style Funnels without fear of Char's Counterattack.

[Ayche]

As a praxis, do not try to guess what game development is like if you have not been a game dev yourself.

I have had laymen question work in my own field and they have always been clueless about it.

[Jeeqbit]

As a praxis, do not try to guess what game development is like if you have not been a game dev yourself.

I have had laymen question work in my own field and they have always been clueless about it.

Overall, in case people forget, I'm defending SE for only having a few people listed as working on class/jobs. The whole point I've been making here is that even some layman math suggests it's enough, as well as maths based on Yoshi-P's own words.

[Calysto]

What makes you say that "As for the Q/A, not breaking the game and "functioning as a whole", they most likely skimp on theses steps"? I don't see the evidence of that, if we look at Sage for example- it wasn't missing any critical skills on delivery, not was it far above nor way below the other healers. Did it get some adjustments late? yes- but they were relatively minor.

I would also agree that job designers likely don't start off from absolutely zero, however there is nonetheless a design/discovery process, reviews, presentations, discussions, approvals- before anyone gets into the actual detailed design, coding, QA, etc so no , I honestly doubt that 40 day estimate follows Square's methodology.

SGE at release was 90% SCH, that would explain why it was ok.
Wasn't 6.0 the only expansion with no job broken on release ? (because of homogenization mainly, so not really like it's a good thing <.< )

Simplifying down to flat potency only really works when you take a job in isolation. You need to factor in how jobs scale with each other as well, a task that in itself has spiralled as we've gotten more jobs over the years. I assume that this is the real reason why the whole piercing debuff thing was removed amongst other things.

As for them skimping, I think it's more a case of them diluting and homogenising everything to an extreme to minimise any risks of game breaking wombo combos springing up out of nowhere

It's the other way around. Jobs are already "tuned", so global increase is still balanced across the board.
I was not talking about "flat potency", they (probably) define that things will increase by like 20% (be that dps or healing or whatever, it's an example).
They know the nominal potencies. They know they can allow an amount to whatever; if it doesn't fit exactly, just drop new tier that compensate for the missing potency on existing actions.
Of course if it was done like that, you'd see early patches x.01 fine tuning the potencies.

[IDontPetLalas]

SGE at release was 90% SCH, that would explain why it was ok.
Wasn't 6.0 the only expansion with no job broken on release ? (because of homogenization mainly, so not really like it's a good thing <.< )



It's the other way around. Jobs are already "tuned", so global increase is still balanced across the board.
I was not talking about "flat potency", they (probably) define that things will increase by like 20% (be that dps or healing or whatever, it's an example).
They know the nominal potencies. They know they can allow an amount to whatever; if it doesn't fit exactly, just drop new tier that compensate for the missing potency on existing actions.
Of course if it was done like that, you'd see early patches x.01 fine tuning the potencies.

I'm aware of the similarity between SGE and SCH (and I agree) , how do you draw your conclusion from what I said? The discovery and conceptual design process still remains. QA still had to be done, balancing still had to be done- so where's the skimping? In addition, I could have used DRK as an example - I don't recall DRK as being "broken" on initial release. again, that doesn't mean that a job never required adjustments after its initial release. Now- if we are talking AST- I would say that it's Square's worst case of a job rework , however they're not exactly stellar when it comes to healer design.