Threat conversation

[Freehugs]

That's great info SLi, thanks.

[SLi]

(05-15-2012 08:01 PM)Freehugs Wrote:  That's great info SLi, thanks.

Thanks

I'm still not sure if 2.75 meters is the set "melee" range as seen by the game mechanics or if it changes for each boss. It's just some information that I worked out and thought I'd share.

I highly doubt that it would be a unique range for each boss, however it could possibly be x% of the hit box.

It would be interesting to get some in game testing logs on boss fights to see if it is a static range.

However, it would likely require the tank (or off dps/tank with taunt) to record ranges and times at which testing taunts were done during the encounter as I don't think there is any information in the logs which specifies range.

This could be done reasonably easily by pressing taunt and Prt Scr (Print Screen) at or about the same time. It should record a rough estimate of the time as well as the range from the boss being taunted.

Ideally, this would need to be done while they were not doing damage to the boss to get a true measure of the % threat generated purely by taunt from their combat log.

I will also retest with AOE taunt, see if it works in the same way or not.

- Updated my post with preliminary testing torparse results.

[Caeden]

Nice work SLi

[Kaedis]

Excellent work! +Rep.

[SLi]

Wow, ty all for your comments and for the +Rep.

I've updated my original post with AoE taunt testing on my Powertech.

It would seem to confirm the 2.75 meter threshold on the Operations Training Dummy. Additionally, it would seem that the AoE taunt functions in the same manner as the single target taunt for % threat increase.

Hmm, now to test if there are any limitations on their ability to stack.


Also, a quick thought on ways to get in game testing as I mentioned previously.

If the Tank or DPS doing the testing pressed taunt and Prt Scr (Print Screen) together, it should record a rough estimate of the time, to the nearest minute, as well as the range from the boss being taunted.

Ideally, this would need to be done while they were not doing damage to the boss to get a true measure of the % threat generated purely by taunt from the combat log.


I think an interesting side note is the significant increase in threat from using taunt further than 2.75 meters.

If you look at this log and note the difference in total threat between the following runs.

2.74m (9:53:07 - 9:57:36) Total threat = 165571, Average TPS = 616.34

Ability threat generation.

Total No. of Hits: 787

585 Rapid Shots - Threat: 82694 - Average TPS: 307.83 - Average Threat: 141.36
5 Sonic Missile - Threat: 38304 - Average TPS: 142.59 - Average Threat: 7660.8
74 Ion Gas Cylinder - Threat: 18584 - Average TPS: 69.18 - Average Threat: 251.14
117 Shocked (Tech) - Threat: 17430 - Average TPS: 64.88 - Average Threat: 148.97
2 Rocket Punch - Threat: 4020 - Average TPS: 14.96 - Average Threat: 2010
2 Rail Shot - Threat: 3358 - Average TPS: 12.5 - Average Threat: 1679
2 Explosive Dart - Threat: 1181 - Average TPS: 4.4 - Average Threat: 590.5




2.76m (10:04:55 - 10:09:17) Total threat = 260654, Average TPS = 993.1

Ability threat generation.

Total No. of Hits: 763

5 Sonic Missile - Threat: 143685 - Average TPS: 547.44 - Average Threat: 28737
575 Rapid Shots - Threat: 77357 - Average TPS: 294.73 - Average Threat: 134.53
65 Ion Gas Cylinder - Threat: 15405 - Average TPS: 58.69 - Average Threat: 237
112 Shocked (Tech) - Threat: 15246 - Average TPS: 58.09 - Average Threat: 136.13
2 Rocket Punch - Threat: 3747 - Average TPS: 14.28 - Average Threat: 1873.5
2 Rail Shot - Threat: 4096 - Average TPS: 15.61 - Average Threat: 2048
2 Explosive Dart - Threat: 1118 - Average TPS: 4.26 - Average Threat: 559



The two runs were very similar, though not identical, and there would be variation due to crit and the number of Ion Gas Cylinder PROcs.

However, the difference in threat between the two, and in particular the increase in threat to Sonic Missile, is significant.

[Sith_Happens]

(05-18-2012 01:38 AM)SLi Wrote:  If the Tank or DPS doing the testing pressed taunt and Prt Scr (Print Screen) together, it should record a rough estimate of the time, to the nearest minute, as well as the range from the boss being taunted.

An easier way would be to set the screenshot key in Fraps to be the same as your taunt key.

[cmf]

(05-15-2012 06:12 PM)SLi Wrote:  From reasonably extensive testing on the Operations Training Dummy I found the threshold for Taunt changing from 110% to 130% threat increase was 2.75 meters.

First, let me say very nice work so far - the data made available in the game's UI and log files make testing slightly tricky. Based on some tests I ran some time ago the center to center melee range did not vary from unit to unit (see below) – however, open reproducible tests demonstrating the same are sorely lacking. Unfortunately, you'll find further testing is going to become much more difficult

You may or may not have noticed during your testing that targets have a non-zero zero tool-tip radius, which is to say if you are within some fixed radius of a unit's center the tool-tip will display a distance of zero meters. Unfortunately, this radius varies substantially from unit to unit - take a look at the Tatooine world boss for a clear example. If you perform similar testing on that boss it will be clear that the tool-tip melee threshold is not universal - it's possible to stand at a tool-tip distance of 0m and still be outside his melee threshold.

The distance that you'll want to use is the center to center distance from the target to the tank - based on my testing the center to center melee threshold is universal. It is possible to measure center to center distances using the tool-tip distances, however the only method I'm aware of that could accurately measure the zero tool-tip radius is four point trilateration - you'd need to recruit three others and perform detailed measurements on every target you plan on testing. To make matters worse to confirm the center to center melee threshold holds on large targets you'd need to measure player and target positions using trilateration for each test, as the threshold will lie within the zero tool-tip radius.

I hope you find the above useful – I look forward to reading the results of your future tests

[SLi]

(05-19-2012 10:49 PM)cmf Wrote:  ....based on my testing the center to center melee threshold is universal

I guess this is a terminology thing but when you say "melee threshold" would you be referring to the melee hit box? That is, the range at which you can use abilities such as Rocket Punch and Assault?

From your testing you've found that the melee hit box is a set distance? That is good info, ty .

Being as close as possible to the the training dummy does indeed set my range to 0m on the tooltip. Would it be safe to say the melee hit box is always 4.0m from absolute centre distance regardless of tooltip?

(05-19-2012 10:49 PM)cmf Wrote:  The distance that you'll want to use is the center to center distance from the target to the tank

If the "centre to centre" distance varies, it "should" be based on the outer edge of a circle that expands based on the size of the boss.

I guess the next step, as you say, is to do some testing on a couple of larger targets and determine if the "melee" range is; a unique range per boss, a set % size of the "centre to centre" circle, a set tooltip distance of < or = to 2.75 meters, or a set range from absolute centre (like the melee hit box).

I would "assume" that if the 2.75 meter tooltip threshold is the same on 3 or more boss level targets, it "could" also be universal. However, testing on every boss in EC would be ideal as you say.

Unless for some reason it is indeed a unique range for each boss, I assume this could mean one of two things. Always taunt/melee DPS from max hit box range (where possible), or from 2.76+ meter tootip range.

[cmf]

(05-20-2012 05:43 AM)SLi Wrote:  

(05-19-2012 10:49 PM)cmf Wrote:  ....based on my testing the center to center melee threshold is universal

I guess this is a terminology thing but when you say "melee threshold" would you be referring to the melee hit box? That is, the range at which you can use abilities such as Rocket Punch and Assault?

From your testing you've found that the melee hit box is a set distance? That is good info, ty .

Being as close as possible to the the training dummy does indeed set my range to 0m on the tooltip. Would it be safe to say the melee hit box is always 4.0m from absolute centre distance regardless of tooltip?

What I called the "melee threshold" was the range at which taunt transitions from 130% to 110% - that is the distance you were measuring in your previous post. Based on my testing the center to center melee threshold was fixed at just under 4m I believe.

The melee hit box is most definitely not a fixed distance from the center of the target

(05-20-2012 05:43 AM)SLi Wrote:  

(05-19-2012 10:49 PM)cmf Wrote:  The distance that you'll want to use is the center to center distance from the target to the tank

If the "centre to centre" distance varies, it "should" be based on the outer edge of a circle that expands based on the size of the boss.

I don't believe the zero tool-tip radius was necessarily dependent on the size of the targeting reticle, however I can't remember why I thought that anymore. I could certainly be mistaken there.

(05-20-2012 05:43 AM)SLi Wrote:  I guess the next step, as you say, is to do some testing on a couple of larger targets and determine if the "melee" range is; a unique range per boss, a set % size of the "centre to centre" circle, a set tooltip distance of < or = to 2.75 meters, or a set range from absolute centre (like the melee hit box).

I would "assume" that if the 2.75 meter tooltip threshold is the same on 3 or more boss level targets, it "could" also be universal. However, testing on every boss in EC would be ideal as you say.

Unless for some reason it is indeed a unique range for each boss, I assume this could mean one of two things. Always taunt/melee DPS from max hit box range (where possible), or from 2.76+ meter tootip range.

That would be ideal I think I just wanted to make you aware of the issues you'll face in the above measurements - hopefully it will save you a bit of time.

[SLi]

Updated my earlier post with graphs from Torparse now that it's back online!!