Hi there,

 

with the newest update to Star Parse, there is a percentage display in the raid tab, that says, as a dps, 94%. What does that mean? i read in the notes, that it is a comparision with other raiders. But what does the percentage say?

 

Thanks for your answers.

Edit:

Well, Ok. It's not really clear what it means. The description leaves a lot of room for interpretion.

Edited May 28, 2015 by Sindariel

How about reading the change log next time instead of just clicking it away?

 

Hint: Click the "?" in the upper right corner of Star Parse and read what's written there!

 

Yeah, well: "While raiding, your DPS, DTPS or EHPS is automatically rated against others playing the same class."

 

I don't know, but to me this is not a helpfull information. We've been discussing this in our raidgroup and all said that they don't see the point in it. We're in a parser to gather information about our performance with our group and our tactic in our specific situation. Where does performance of others with probably different group setups, tactics and gear, help with that?

 

We agreed to ignore that number, but I have the nagging feeling, that those numbers could become a problem in groups. "You were performing at 41%. The boss is dead but you bad must not come with us again" - To be a bit drastic. All groups need to know is, if the raid is doing enough dps (defined by hitpoints of boss and enrage timer), if not who does not, analyze based on this information why and try again (same for DTPS and perhaps EHPS).

 

 

We agreed to ignore that number, but I have the nagging feeling, that those numbers could become a problem in groups. "You were performing at 41%. The boss is dead but you bad must not come with us again" - To be a bit drastic. All groups need to know is, if the raid is doing enough dps (defined by hitpoints of boss and enrage timer), if not who does not, analyze based on this information why and try again (same for DTPS and perhaps EHPS).

 

 

Right now this seems buggy, depending on who is checking the parse your rating can differ between 91 and 33%^^

But I'd rather see this feature as a motivation to get better:

Like "Oh I've got 3,5k dps and thats 70%, that means there is room for improvement because other dps with the same discipline are able to achieve more...

Anyways, from the few encounters we had tonight (torque, master blaster and underlurker) it seems like if every dps is below 50% you'll hit enrage and if you're above you're doing enough damage for this encounter.... Or as tank, if you are below 50% you're not using your deff cds good enough or you stay in voids too much...

Sounds very similar to some of the log comparisons in WoW, but built into the parser offbase From the sound of it, it looks like that is your rating out of 100, 100 being the best I'm assuming, compared to others who have done the fight. I would also assume this is class/spec based, but I'm sure Ixale can elaborate further if I'm accurate or completely off base Edited May 28, 2015 by SnickerJew

Its a rating about either your damage taken(while in tank spec), effective healing done(while in a heal spec...this could also just be general HPS but I would assume it's comparing EHPS instead), or your damage done(while in a damage spec). Ixale would have to comment on whether it's comparing just logs uploaded to parsely or if starparse has been collecting log data for this in the background or not. If it's just parsely logs, I would expect tanks and healers to get wild variations on their ratings when compared to DPS just because there are probably way more people uploading DPS logs than tank and healer logs to parsely, which means a few more or less DTPS or EHPS could shift your rating around a lot more percent than a few less or more DPS would. Plus if you thought comparing DPS logs with some lucky crit rates was bad, I can imagine tank DTPS log comparison with all the RNG that comes with passive mitigation and cooldown timing is 100 times worse.

 

Basically, if you go into a fight as a DPS, clear a boss and get a 5%, you are probably doing something very wrong. If you get something like 90+, you know your class and are trying your hardest to get as much DPS as you can(or aren't busy with mechanics while the rest of your group is free casting). It is also definitely based on which class you are instead of just comparing you to every class in the game as well. Tonight in some fights I would do better damage than someone else and yet be in a slightly lower percentile than them. For example, I was in a lower percentile than the merc in my raid on Torque HM even though I did way more DPS than him just because that fight is so much easier to do damage on in Marksman than an IO Merc. Not sure if it's also based on which discipline as well.

Edited May 29, 2015 by OMGITSJAD

Hello there,

 

sory for dropping the numbers there without any further explanation - I am actually working on a proper FAQ/tutorial web resource right now, so that should address all questions about this feature and possibly others as well.

 

Anyways, as mentioned in this thread already, the aim of this raiting is to provide an indicative (!) comparison to others playing the same class on the same encounter. It is only class (and role) dependent, so disciplines are not taken into the account as of now (although they are being recorded as well, so I can add such feature in the future even retrospectively).

 

The numbers are based and all raiding data gathered over the past few weeks (that is everyone who is actively raiding using StarParse and has not selected "disable server storing"), so I expect them to get more relevant as the times goes by (as thousands of pulls are being evaluated in the process).

 

To be more specific, the following metrics are being measured:

1) for DPS specs, raw DPS output

2) for tank specs, raw DTPS minus self-gained shielding (e.g. absorb relic procs, Blade Barrier procs etc)

3) for healer specs, effective HPS plus all shielding done (sages/sorcs parsing with other raid members getting their bubbles counted properly will achieve higher numbers)

 

The most tricky part is to decide which parses to include - in WOW, the combat logging is potent enough to reliably tell wipes from boss kills, however, thats not 100% possible in SWTOR. Therefore for the time being, ALL pulls meeting certain time tresholds are being included in the calculations (e.g. 6 minutes+ on Monoliths, 3.5 minutes on many other bosses etc). I will keep monitoring and adjusting these limits continously so the data are as relevant as possible (together with implementing more reliable detections into the client itself to help identifying successful pulls).

 

Last but not least, everyone using tools like StarParse should be already aware the pure numbers can and are often misleading and their interpretation needs to be done with several things in mind - this new addition is no exception. I am definitely not trying to hinder groups by pressuring their members to achieve higher virtual numbers, but merely provide another insight and perspective into their output (that is, low healing numbers are not about incompetent healers, but more about group taking less damage than average etc). As long as the group gets the job done, the numbers are just numbers to amuse and possibly inspire.

 

As always, I am open to suggestions how to tune the feature, even possibility to forcibly disable it by raid leaders might be a thing if such desire arises.

 

In the meantime, happy parsing and thanks for your continuing support fo StarParse!

Hi there and thanks for the quick reply. I really appreciate it. The comparrision is a real cool feature. Yesterday, we recognised that the fight with Malapher was to short to get the percentages displayed, as he was killed under the mentioned 3:30 minutes.

Thanks for the quick answer Also, thank you for your awesome tool

As always, I am open to suggestions how to tune the feature, even possibility to forcibly disable it by raid leaders might be a thing if such desire arises.

 

I would really appreciate the option to disable it. These numbers might encourage raidmembers to play "for the numbers", ignoring mechanics in doing so. I know this should not be a thing, but as a matter of fact (at least in my raidgroup) it sometimes is. So the option to disable this would be great

Edited May 29, 2015 by Words_of_Wisdom

I would really appreciate the option to disable it. These numbers might encourage raidmembers to play "for the numbers", ignoring mechanics in doing so. I know this should not be a thing, but as a matter of fact (at least in my raidgroup) it sometimes is. So the option to disable this would be great

I bet the people who are inclined to play for numbers would keep that option enabled. I know I sometimes do that even without such a comparative number, thinking "can I get away with finishing this channeled ability while standing in stupid". Once at Grob'Thok (DF 3rd boss) I got the mining droid and proceeded to stand in the fire for several seconds while executing a channeled attack. When prompted about it, I said something along the lines of "do you know what moving does to my dps", which became a running joke with my raid group for a short while.

I would really appreciate the option to disable it. These numbers might encourage raidmembers to play "for the numbers", ignoring mechanics in doing so. I know this should not be a thing, but as a matter of fact (at least in my raidgroup) it sometimes is. So the option to disable this would be great

 

Rumor has it, he's working on an: Avoidable Damage tab. Hmm, must have gotten the idea from WoW parsers.

Dumb Question: Is it intended that the timers for push/pull on Revan only work partially?

Dumb Question: Is it intended that the timers for push/pull on Revan only work partially?

 

Yep. Revan got nerfed again this patch and the timers were delayed again. I forget if it were unstable abberations or the push/pulls again...

 

Either way, the push/pull mechanic has an obvious animation to it anyway, so you shouldn't need the timers for that.

This 'number' is directionally useful at best - and quite honestly, not something which I find useful by itself, nor while raiding (though I like having logged data available to compare after raids - both personally and as a group).

 

Personal damage done / DPS is effected by all kinds of things from player discipline, player gear / gear rating, overall length of fight, other player performance / damage done by others, classes present / raid buffs, not to mention things like ignoring mechanics / cheesing to pad numbers (which tend to happen more and more as a tier progresses) - all of which makes averaged comparisons between random players mostly useless.

 

All of the above are why I prefer to use logs / log sites to drill down into the details where useful data can be gathered (some of which just doesn't exist in TOR such as capturing gear rating), and actionable information on how and where to improve can be found by comparing similar parses and filtering out noise.

 

Oh and I'd also like to know if the comparisons are for just a specific boss, or is it broken down further by boss / raid size / difficulty mode.

 

I do love the tool - thanks for creating BTW - just not sure this new 'feature' is something of much value, certainly not without at least controlling for disciplines and boss / size / difficulty (if not already).

Edited May 30, 2015 by DawnAskham

Yep. Revan got nerfed again this patch and the timers were delayed again. I forget if it were unstable abberations or the push/pulls again...

 

Either way, the push/pull mechanic has an obvious animation to it anyway, so you shouldn't need the timers for that.

 

Not saying you're lying, kwerty, but i just checked all patchnotes back to 3.1 and the only nerf to revan that is mentioned is the unstable abberations time increase.

What has Bern nerfed exactly?

I would really appreciate the option to disable it. These numbers might encourage raidmembers to play "for the numbers", ignoring mechanics in doing so. I know this should not be a thing, but as a matter of fact (at least in my raidgroup) it sometimes is. So the option to disable this would be great

I agree that this option is not necessary, it only leads to creating entry barriers into raid groups. Yes, good raidleaders will know not to fully rely on this number alone but there are always bad apples.

 

If you want to improve StarParse, my suggestion would be to open up the DPS/HPS/DTPS tabs so you can see everyone's stats. For example, if someone takes 1,200 DTPS, that doesn't help me as a raidleader because I cannot see from which attacks he took damage; only that player can see it. Yes, he can tell me what he took damage from but from my experience, some players are overwhelmed by StarParse and it will easily take a minute before they figure out why they died. If you make those detailed data available for everyone, I can just quickly go through the group, see who stood in how many circles (based on how often he took damage from a certain attack) etc. and move on with the raid.

Edited May 31, 2015 by Jerba

Not saying you're lying, kwerty, but i just checked all patchnotes back to 3.1 and the only nerf to revan that is mentioned is the unstable abberations time increase.

What has Bern nerfed exactly?

 

According to datamined information (oh noes kwerty thats not completely ok!), the 3.2.1 patch also had a revan nerf again. Another thing that was left out of the notes was 3.2.1's bug fix to charged burst - it was hitting ~25% harder than it should have been before 3.2.1 was launched.

 

The exact change?

 

Further increased timers on the Unstable Aberations in the Revan fight to 36 seconds from 30 seconds.

Regard the bad apple raidleaders. Just be happy that you don't need to bother with them. Avoiding them is even easier with this feature.

 

Anyways. Ixale, could you elaborate what the rating is based on? You said its based on raw dps output but to what are you comparing the dps? I guess from the options I listed below its the 2nd.

I could list more options but I guess this shows what I mean and why I am slightly confused how you calculate it.

 

Option 1 - baseline=average dps / rating=based on dps

On average people are parsing 4k dps on a boss.

You parse 2k dps so your rating will be 50% (50% from 4k)

You parse 1k dps so your rating will be 25% (25% from 4k)

 

Option 2 - baseline=maximum dps / rating=based on dps

The maximum dps someone got is 4k dps on a boss

You parse 2k dps so your rating will be 50% (50% from 4k)

You parse 1k dps so your rating will be 25% (25% from 4k)

 

Option 3 - baseline=max dps / rating=based on the amount of people

The maximum dps someone got is 6k dps on a boss. Out of 100 people you got

5 parsing 1k dps,

10 parsing 2k dps,

15 parsing 3k dps

40 parsing 4k dps

20 parsing 5k dps

10 parsing 6k dps

You are parsing 2k dps so your rating is 15% (85% people are able to parse higher)

 

Nice feature btw

Quite fun to see tanking parses from swordsquadron when not doing a tankswap

All these ratings and even parse leader-boards are useless without taking gear into account. People will cry that their 192 character is not doing as well as somebody with 198 and 204 MH. Entering gear information for each character in StarParse and keeping it up to date would be PITA. I think its possible to have some approximate rating by heuristically inferring gear level by ability averages on crit, no crit and crit rate. Edited May 31, 2015 by Arunas

Hello everyone,

 

to address your questions:

 

Dumb Question: Is it intended that the timers for push/pull on Revan only work partially?

 

I havent seen Revan lately, but I am not aware the pull/push mechanics has been altered at all, so the original 15s timers should still work (Abberation timers per se have never been included in StarParse) - could you please elaborate what exactly doesnt seem to work and attach a screenshot/recording ideally? That would be greatly appreciated so I can fix it!

 

Oh and I'd also like to know if the comparisons are for just a specific boss, or is it broken down further by boss / raid size / difficulty mode..

 

Yes, the rank is specific for each boss, mode, size and class (not disciplines at the moment, however, only dps stances are included in the stats, so no tanks nor healers are affecting it, and vice versa).

 

I agree that this option is not necessary, it only leads to creating entry barriers into raid groups. Yes, good raidleaders will know not to fully rely on this number alone but there are always bad apples.

 

If you want to improve StarParse, my suggestion would be to open up the DPS/HPS/DTPS tabs so you can see everyone's stats.

 

As stated already, I added the feature knowing people who already use StarParse actually want to see the numbers (for basic sanity checks or even their amusement only), so this seemed as just another number into the mix. Also there have been plenty of "what shoud my damage/healing etc be on boss xy" questions on Reddit et al, so again, I thought I am helping there as well.

 

Regarding the detailed insight into other logs, sharing everything with everyone would be an overkill probably, but I am considering adding a "Death Recap" broadcast so when you die, your last X seconds would be sent to the raid leader (all?) for them to see what killed you. I will probably add something like that together with the "Avoidable damage" analysis in the upcoming weeks.

 

Anyways. Ixale, could you elaborate what the rating is based on? You said its based on raw dps output but to what are you comparing the dps?

 

It is actually percentile-based, so I guess the third option - what I do is basicaly:

1) sort all eligible parses by the respective output (dps, dtps-selfabs, ehps+abs)

2) calculate 1 to 99th percentiles (by taking every nth / count parse as the value)

 

So if your rating is 75%, it simply means 75% of the parses are below you and 25% is better. This approach is less susceptible to edge cases, on the other hand its not proportionate (being 50% does not usually mean the best can do twice as much).

 

All these ratings and even parse leader-boards are useless without taking gear into account. People will cry that their 192 character is not doing as well as somebody with 198 and 204 MH. Entering gear information for each character in StarParse and keeping it up to date would be PITA. I think its possible to have some approximate rating by heuristically inferring gear level by ability averages on crit, no crit and crit rate.

 

Well, yes, taking the gear into account "somehow" would be great, on the other hand, now I am simply stating "X % of the people on this boss currently in the game are doing more numbers than you are" - the bosses do not scale with the raid gorup either, so I believe it is still relevant to see you are probably hindering the group by not performing well, being it because your gear or just a lack of experience.

 

-----

 

Anyways, thank you all for taking the time to share your thoughts - as stated, this feature is experimental, so pleas bear with me, anything is possible (if it proves irrelevant or even harmful).

Random but related question:

 

With this information do you have available the number of uploaded by class. Or what I am asking is with this ranking system could we see of often an AC is brought to a fight compared to others. Would be interesting to see what the average raid composition is for certain bosses.

Random but related question:

 

With this information do you have available the number of uploaded by class. Or what I am asking is with this ranking system could we see of often an AC is brought to a fight compared to others. Would be interesting to see what the average raid composition is for certain bosses.

 

Hey, yes indeed - classes with disciplines are being stored as well, so any derived statistics are possible. Basically the following summary stats are being gathered (one entry for each player and combat):

 

- raiding group and character

- class and discipline

- boss, mode, size

- combat start and finish

- apm

- damage

- heal and effective heal

- damage taken

- damage absorbed

- heal taken and effective heal taken

- threat

- damage or healing in challenges, if any

- shielding done on others, if any

I havent seen Revan lately, but I am not aware the pull/push mechanics has been altered at all, so the original 15s timers should still work (Abberation timers per se have never been included in StarParse) - could you please elaborate what exactly doesnt seem to work and attach a screenshot/recording ideally? That would be greatly appreciated so I can fix it!

Well, don't know when it happened, but somehow the timers stopped showing at all, but heave timer still worked. After the second last patch, the push timer appeared again for me, but not the pull timer. Didn't change anything in the settings. I could upload a parse from Saturday , where we pulled Revan for ~4-5 hours if that helps?

Edit: Okay, I just spoke with one of our dps and for her both push and pull timers show, but only for third floor.

Edited May 31, 2015 by Torvai

Does it take into consideration where the damage was dealt? For instance, the high Bulo parses have, in some cases, upwards of 40% of the damage dealt to pirates, which gives a really high overall, but relatively low parse on the boss itself. Edited May 31, 2015 by ChroniKill