The State of Machinist in DT - Let's talk about it (again).

[xbahax92]

I would really like to get some information and would be very grateful if someone could briefly explain the statistics to me. Do the data shown here truly reflect the job's performance 1:1?
I know that fflogs has many variables and isn't always the most accurate source of information, but I would still like to know if the data presented here is correct.

Because you still see opinions here and there saying that a lot of the uproar is just being exaggerated.
I’m not really familiar with fflogs.

[Supersnow845]

I would really like to get some information and would be very grateful if someone could briefly explain the statistics to me. Do the data shown here truly reflect the job's performance 1:1?
I know that fflogs has many variables and isn't always the most accurate source of information, but I would still like to know if the data presented here is correct.

Because you still see opinions here and there saying that a lot of the uproar is just being exaggerated.
I’m not really familiar with fflogs.

This image represents the distribution of performance of a job in all CLEARS specifically

The numbers beneath are a standardised scale to compare across the 4 fights, if you select stats for a particular fight you will be given it in DPS specifically

So yes this data is accurate and gives a general overview across the 4 fights, like you can “on average a median percentile MCH performs roughly equivalent to a RDM performing 1 standard deviation below average”. If you want specific fight data you can get the same box plot for a specific fight giving you its actual DPS

This is also measured in rDPS by default. Depending on the comparison you are trying to make c, a or even n DPS can be slightly better but that’s a bit beyond the scope of your question

[Aco505]

Data on the first weeks has to be analyzed carefully, as not many players have cleared the entire tier and M6S will skew some jobs due to the add phase. We will know how balanced they are with better accuracy as weeks pass and players start getting BiS.

[xbahax92]

Data on the first weeks has to be analyzed carefully, as not many players have cleared the entire tier and M6S will skew some jobs due to the add phase. We will know how balanced they are with better accuracy as weeks pass and players start getting BiS.

But doesnt that also change the view on the chart? We're having better gear and more mistakes are allowed to clear the content.
It gets easier for certain underperforming jobs to clear the content. And as of now there is not a single log in a m8s clear for MCH if I understood correct.

[Valence]

I would really like to get some information and would be very grateful if someone could briefly explain the statistics to me. Do the data shown here truly reflect the job's performance 1:1?
I know that fflogs has many variables and isn't always the most accurate source of information, but I would still like to know if the data presented here is correct.

Because you still see opinions here and there saying that a lot of the uproar is just being exaggerated.
I’m not really familiar with fflogs.

You have currently selected all bosses, this is problematic because M6S gets all of the metrics completely skewed by the heavy AoE add phase. Selecting M6S individually for instance, can also make sense to compare job performance within that specific encounter design.
Later fights like M7S and M8S are still too low on samples to give out any semblance of reliable data, so for now I'd suggest to just keep at individual fights.

You always need to take metrics with a spoon of salt of course, but if you want a fast metric to get a decent overview of job balance, go for the cDPS tab which includes both raid buffs you give and raid buffs you benefit from (it's close to the rDPS one, but more fine tuned to account for how every job plays within buffs).

[Supersnow845]

You have currently selected all bosses, this is problematic because M6S gets all of the metrics completely skewed by the heavy AoE add phase. Selecting M6S individually for instance, can also make sense to compare job performance within that specific encounter design.
Later fights like M7S and M8S are still too low on samples to give out any semblance of reliable data, so for now I'd suggest to just keep at individual fights.

You always need to take metrics with a spoon of salt of course, but if you want a fast metric to get a decent overview of job balance, go for the cDPS tab which includes both raid buffs you give and raid buffs you benefit from (it's close to the rDPS one, but more fine tuned to account for how every job plays within buffs).

Do note that cDPS is also known to double count some situations so can overestimate classes that are designed around buff feeding (I think it’s worst with AST and DNC)

[Valence]

The day we'll have perfect metrics pigs will fl.. owait-

[Aco505]

I don't think we'll ever get a perfect metric, which is why we have to look at them all to have an overall view of the situation, but then the type of content can also matter (e. g. the difference between PCT in FRU vs CODCAR or VPR in FRU vs M6S).

But doesnt that also change the view on the chart? We're having better gear and more mistakes are allowed to clear the content.
It gets easier for certain underperforming jobs to clear the content. And as of now there is not a single log in a m8s clear for MCH if I understood correct.

Of course, but there's also other factors such as statics funneling all their gear to specific DPS (which can benefit DNC and BRD due to their buffs but not MCH) or PF locking out jobs due to community feeling on their ability to increase their odds of success. Since M6S has walled a lot of people in both statics and PF, the amount of M7S and M8S clears is still going to be low.

Obviously, we can still analyze what's going on and realize that DNC performs better in M6S due to add cleaving and padding. However, as a gauge job, MCH is able to save as much of it as possible to then quickly deal with troublesome targets such as cats, the first yan or their own manta. This doesn't change the fact that the job has been in a terrible situation balance-wise since the beginning of the expansion but it does have its strengths.

[CKNovel]

Around 11 Machinists cleared on week 1.
In comparision, there's 286 Dancers and 135 Bards.

MCH is also waaay lower than any melee DPS, so much for that reduction of melee uptime.

[Supersnow845]

Around 11 Machinists cleared on week 1.
In comparision, there's 286 Dancers and 135 Bards.

MCH is also waaay lower than any melee DPS, so much for that reduction of melee uptime.

As anyone who has spent 0.0000005 seconds outside of JP would have told the dev team “buffing melees for uptime concerns” was always a stupid suggestion because NA refuses to even consider a strat if it doesn’t maximise melee uptime anyway. Like that god awful M7 strat that’s forcing all the casters and phys range out of boss range to make the mechanic easier for melee to hit positionals (because a positional is 100% worth more than 3 seperate jobs entire GCD)