*** Reminder - This is not an argument, or a contesting of someone else's information. This is simply my results of GCD testing for various characters I have leveled to 75 so far.***

http://www.mrrabbit.net/misc/swtor/swtorglobalcooldown.php

Scroll down on that page for the chart of characters leveled and tested so far.

1. No Flagship Alacrity Bonus

2. No Alacrity Bonus in Gear

Note that several easily have 10-40 excess points in Alacrity Point Stat - i.e., they're wasted points - were not necessary to establish the targeted GCD.

Enjoy!

=8-)

I've updated the chart for:

Agent Operative - Medicine

http://www.mrrabbit.net/misc/swtor/swtorglobalcooldown.php

In previous patches, when trying to achieve a 1.3 GCD via the Test Dummy, a particular problem arose.

You don't get a "clean" 1.3 GCD when establishing the minimum number of points necessary to meet the Alacrity Rating % break point or GCD Tier.

In other words, when doing a test, the majority of the ability intervals will be in the 1.28 to 1.30 range...

...but interspersed among them will be a significant number of 1.31s, 1.32s, 1.33s all the way up to a rare 1.4.

In other words, a closely set 1.3 GCD character is going to experience quite a few slowdowns - i.e., 1.4 GCD intervals.

Typically what I have done in the past is add 1 augment more to minimize this.

This is what I've done with my Operative Healer for 6.0.

At 3355 Alacrity Points he triggered a 7.0 Energy Regen Rate. Channeled casts showed a -.1 cast reduction time as expected...

...but raw GCD Results kept coming in at 1.3066 to 1.31 which is a 1.4 GCD. Starparse Combat Log showed 6 intervals in the 1.31 to 1.4 range.

Adding another 228 Alacrity Augment bumped him to a solid 1.3044 and only 3 intervals that would be treated as 1.4 were present.

This character is in the opinion of some carrying 90-100 excess points in Alacrity Points for a 1.3 GCD.

Bottom Line: Bioware does not make achieving a 1.3 GCD easy.

Fortunately, at 294 Gear Rating, I was able to retain a 1600 Critical Rating which isn't too bad all things considered.

Next up will be my Marksmanship Sniper.

=8-)

Looking forward to the marksman sniper

Quick question. Is it possible to use the training dummy in the Rishi Strong Hold if I set it too operations and do I use the booster terminal.

Also, you need to hurry up and get that 306 gear

In other words, a closely set 1.3 GCD character is going to experience quite a few slowdowns - i.e., 1.4 GCD intervals.

The most likely explanation for this observation is because of human error, and natural delays in a human being pressing keys, or latency in the button press reaching the server ... not because of anything intrinsically different between levels of alacrity rating.

That's why a testing method which doesn't hand count ability activations of the basic attack, but rather, uses StarParse to count the number of attacks over a given amount of time that is sufficiently long to statistically exclude deviations in human performance is a superior method.

The most likely explanation for this observation is because of human error, and natural delays in a human being pressing keys, or latency in the button press reaching the server ... not because of anything intrinsically different between levels of alacrity rating.

 

That's why a testing method which doesn't hand count ability activations of the basic attack, but rather, uses StarParse to count the number of attacks over a given amount of time that is sufficiently long to statistically exclude deviations in human performance is a superior method.

Is there a better method to eliminate those possible anomalies?

The most likely explanation for this observation is because of human error, and natural delays in a human being pressing keys, or latency in the button press reaching the server ... not because of anything intrinsically different between levels of alacrity rating.

 

That's why a testing method which doesn't hand count ability activations of the basic attack, but rather, uses StarParse to count the number of attacks over a given amount of time that is sufficiently long to statistically exclude deviations in human performance is a superior method.

1. You know I'm using Starparse.

2. You know I'm using Starparse in local Parse mode.

3. You know Starparse audits the local client log file.

4. You know I'm micro spamming a button for a minute duration...not seconds...but a minute.

AND

...you know that I'm not presenting an argument or contesting someone else's data.

Yet you seem religiously determined to position yourself as the person who is "right" and the "last word".

Why?

Do you have an investment in this?

=8-|

Looking forward to the marksman sniper

 

Quick question. Is it possible to use the training dummy in the Rishi Strong Hold if I set it too operations and do I use the booster terminal.

 

Also, you need to hurry up and get that 306 gear

Believe it or not, even though I do HM OPs and FPs and occasionally do a few NiM fights...

...I'm always the last person to gear up. I'll even pass on all BiS bonus gear drops -I let everyone else have it.

I gear up taking my time...and for 6.0, my plan is to get all toons to 296, and stop at 298.

I don't really "need" that 306 gear.

My son on the other hand...Mr. Pyrotech...well...

Is there a better method to eliminate those possible anomalies?

Delays in human processing are inevitable. But, in the OP's original thread on this topic I commented on how I have, and would recommend, doing the testing to ultimately arrive at similar conclusions.

I think the more precise way is to run the test for a sufficiently long period of time, rather than counting ability activations. StarParse will count the number of ability activations for you and give you the stop time so that you can calculate the APM. The one it reports after analyzing the combat log will be artificially lower because of the several seconds of delay until the training dummy exits combat, so you need to use the time of the last hit of the basic attack to calcuate the true APM. APM will be statistically significantly different with a 1.5s GCD than a 1.4s GCD after doing about 84 seconds worth of testing on each. You need to run the test for about 4:33 if you are checking the 1.3s GCD because the least common multiple of 15, 14, and 13 is much higher. You also don’t need to have StarParse running in the background because you can just have it parse the combat log offline. Running fewer programs in the background means less chance for latency causing errors or delays in your button presses getting registered. There’s more chance for errors in counting ability activations so it’s better to set a timer or stopwatch and let StarParse count for you as I did in my test linked above.

Basically, dipstik's thread from PTS found the standard error of measurement and the standard deviation for similar testing setup. Using that information, along with calculating the least common multiple of 13, 14, and 15, you can estimate that you should see a statistically different number of APM at 84 seconds and 4:33 based on whether you are testing the 1.4s GCD or the 1.3s GCD.

The interesting part about the OP's post is actually this:

This character is in the opinion of some carrying 90-100 excess points in Alacrity Points for a 1.3 GCD.

In other words, the OP is suggesting that because of human performance error and latency, that carrying a few extra points in Alacrity Rating may help alleviate that error. That is an interesting hypothesis. However, dipstik's testing linked does not suggest this is the case. There was no statistically significant difference in APM between two different levels of alacrity that are above a given GCD threshold on PTS.

strawman

Never attributed the occurrence of under targeted GCD performance to anything....just made my own adjustment based upon observation.

Knock it off with the strawmans please..that's very dishonest.

Attributions elsewhere...

=8-)

did you ever took into account that there is a delay for the game writing the log and a delay of starparse reading it?

its not about human failure. 0.0XX Sec of delay is totaly in range for processing caused inaccuracy. that doesnt change your actual GCD.

the "religion" you are reffering to is called math. wich simply states: if you get X % of alacrity, your GCD will be reduced by that amount rounded up. the "magical" numbers everybody tells you are the exact numbers at where you hit the exact X % reduction you need without wasting points to GCD rounding.

some test samples with X amount of hits and Y amount of inacgurate GCDs is not math, its displaying the technical borders the game is bound to when processing and writing your attacks all the way into the combat log and starparse.

precise formular are no magic or religion and cant be proven wrong by an experiment bound to the same conditions, thats how math works. therefore you are totaly right, there is no argument.

If what the OP is writing is egregiously wrong will you guys please add useful information that actually proves his info is wrong and explain why?

How does this help with argumentative posts that don't have any technical information to compare with the OP's?

I realize Dipstik is regarded highly here, but that doesn't mean other people cannot provide useful or helpful information on topics Dipstik already wrote about.

Thanks OP for trying to be helpful.

If what the OP is writing is egregiously wrong will you guys please add useful information that actually proves his info is wrong and explain why?

 

How does this help with argumentative posts that don't have any technical information to compare with the OP's?

 

I realize Dipstik is regarded highly here, but that doesn't mean other people cannot provide useful or helpful information on topics Dipstik already wrote about.

 

Thanks OP for trying to be helpful.

what exactly do you want to know?

the alacrity treshold are as follows:

1.4s - 7.143%

1.3s - 15.385%

1.2s - 25%

1.1s - 36.364%

1.0s - 50%

these values are based on simple percentage calculation taking the rounding to the first decimal into account.

stat formulas for all stats including diminishing returns can be found here:

https://docs.google.com/spreadsheets/d/1R20FOfta3jXS-zQqs40jcKU8zEou5fujMW_-71-micI/edit#gid=0

put the percentage values from above into the alacrity formular to get the exact number needed to reach each treshold, done.

what OP is seeing and describing are simple inaccuracies caused by technical boundries of processing your inputs by the game, including human failures, delay from client to server, process timings on the server, read/write actions, transactions. even the fastest system is bound to the laws of physics, especially time. just because you are hiting your ability exactly every 1.4sec, which is btw. impossible, even if done by a script/bot, it doesnt get processed at the server every exact 1.4secs, because even machines need time to "do" things, that simple.

in short: even an infinit amount of samples of abilities not being processed at the exact expected interval, doesnt mean the math is wrong. it just means you can neither process, nor messure the inputs 100% accurate.

what exactly do you want to know?

the alacrity treshold are as follows:

1.4s - 7.143%

1.3s - 15.385%

1.2s - 25%

1.1s - 36.364%

1.0s - 50%

 

these values are based on simple percentage calculation taking the rounding to the first decimal into account.

stat formulas for all stats including diminishing returns can be found here:

https://docs.google.com/spreadsheets/d/1R20FOfta3jXS-zQqs40jcKU8zEou5fujMW_-71-micI/edit#gid=0

 

put the percentage values from above into the alacrity formular to get the exact number needed to reach each treshold, done.

 

what OP is seeing and describing are simple inaccuracies caused by technical boundries of processing your inputs by the game, including human failures, delay from client to server, process timings on the server, read/write actions, transactions. even the fastest system is bound to the laws of physics, especially time. just because you are hiting your ability exactly every 1.4sec, which is btw. impossible, even if done by a script/bot, it doesnt get processed at the server every exact 1.4secs, because even machines need time to "do" things, that simple.

 

in short: even an infinit amount of samples of abilities not being processed at the exact expected interval, doesnt mean the math is wrong. it just means you can neither process, nor messure the inputs 100% accurate.

Okay...gonna try one more time...but to be honest it's like talking with an software engineer who proclaims that computer science is all about the code, after I've informed him that computer science is first and foremost a behavioral science....and all the while he is describing all the behaviors he is trying to duplicate or mimic in his code - and all the behaviors he is troubleshooting - not realizing the whole time that there is no argument to begin with.

So again.

1. I'm am not contesting dipstik or anyone else.

2. I'm simply providing samples of my characters of where they are at.

3. I'm also noting where clearly they have a small amount of excess points in alacrity for their tier.

...and of course I'm reminding people to watch out for the Guild Flagship alacrity bonus and gear alacrity bonuses - as they do skew things a little.

So what does the above mean?

1. There's no need to defend me - as there is nothing to defend.

2. There 's no need to attack me - as there is nothing to attack.

About the only barb I have really tossed out is against those who keep throwing around a "magic" alacrity number around for all dps classes, all heals classes without any differentiations provided where needed.

We all know...

The tiers have an associated percentage break points...they have on a class and spec basis an associated amount of Alacrity points...

So for the last time, knock it off with the Strawmans please....

Creating a position that someone did not take - and then attacking it - in an effort to say "They're wrong!" is an argumentative fallacy...especially considering when an argument has not even been made.

When done purposely and repeatedly - it's a tell about yourself as a person.

Also as an added note.

A couple hundred visitors have checked my chart...which is nice.

Thank you.

Also, I have had a few whispers from folks on Fleet @ Satele Shan giving me feedback.

Unfortunately I keep getting asked for a sample for Sin DPS - sadly it's gonna be a long time on that one - she's only Level 48 and getting her up leveled is just not a priority right now.

But in all honestly, you really don't need my sample if you know how to manipulate Alacrity points while monitoring Alacrity Rating %.

=8-)

Okay...gonna try one more time...but to be honest it's like talking with an software engineer who proclaims that computer science is all about the code, after I've informed him that computer science is first and foremost a behavioral science....and all the while he is describing all the behaviors he is trying to duplicate or mimic in his code - and all the behaviors he is troubleshooting - not realizing the whole time that there is no argument to begin with.

 

So again.

 

1. I'm am not contesting dipstik or anyone else.

2. I'm simply providing samples of my characters of where they are at.

3. I'm also noting where clearly they have a small amount of excess points in alacrity for their tier.

 

...and of course I'm reminding people to watch out for the Guild Flagship alacrity bonus and gear alacrity bonuses - as they do skew things a little.

 

So what does the above mean?

 

1. There's no need to defend me - as there is nothing to defend.

2. There 's no need to attack me - as there is nothing to attack.

 

About the only barb I have really tossed out is against those who keep throwing around a "magic" alacrity number around for all dps classes, all heals classes without any differentiations provided where needed.

 

We all know...

 

The tiers have an associated percentage break points...they have on a class and spec basis an associated amount of Alacrity points...

 

So for the last time, knock it off with the Strawmans please....

 

Creating a position that someone did not take - and then attacking it - in an effort to say "They're wrong!" is an argumentative fallacy...especially considering when an argument has not even been made.

 

When done purposely and repeatedly - it's a tell about yourself as a person.

 

i'll try again as well:

the behavior you are observing is simple messurement inaccuracy and therefor an adjustment to stats is NOT needed and wont change anything. period.

i explained why in previous posts.

btw. proving someone wrong != attacking them.

@Op

currently one class/spec can hit the 1.2 and 1.1 second GCD markers, for brief periods, within a reasonable build:

Lightning Sorcerer, after ramp up (=5%), and under the effect of Polarity Shift (+20%).... 1.0 *might* be theoretically possible with some external buffs, but not by itself (falls ~2030 points shy unless I'm forgetting some alacrity relics/adrenals)

i'll try again as well:

the behavior you are observing is simple messurement inaccuracy and therefor an adjustment to stats is NOT needed and wont change anything. period.

i explained why in previous posts.

 

btw. proving someone wrong != attacking them.

And again, I have provided no argument for which to be proven wrong to begin with. Simply provided observations, given samples, and noted my wasted points and adjustments that I make.

That's it...

This is the oddest thread ever.

It’s like some one testing the acceleration due to gravity at different altitudes with a grandfather clock while proclaiming that he is not arguing with Newton.

2. I'm simply providing samples of my characters of where they are at.

The one thing that is missing is to what end you are doing this. What is the purpose of this sharing? What are you trying to say or prove or indicate with these numbers?

What other people are saying is that you're using a method that is prone to errors, which is a simple truth as far as I'm concerned. So that diminishes the value and even the point of this whole exercise.

So really, my sincere question to you is what are you trying to achieve by posting your findings?

Cause at the moment I have no idea what the point is of your thread.

This is the oddest thread ever.

It’s like some one testing the acceleration due to gravity at different altitudes with a grandfather clock while proclaiming that he is not arguing with Newton.

People ask me, "Rabbit, what are you doing for your Merc heals for GCD."

I give 'em the exact answer...

Exactly what my character is.

No argument presented - no argument called for.

When I go outside and observe, "Nice bright blue sky today..." I'm not arguing a particular shade. If someone argues as though I am - well....it indicates something about that person.

=8-(

And again, I have provided no argument for which to be proven wrong to begin with. Simply provided observations, given samples, and noted my wasted points and adjustments that I make.

 

That's it...:

incorrect, and inaccurate....

you are not just providing your observations, you have also provided your opinions as to the cause of those observations, which is a judgment and conclusion based on the the argument of your observations.

at least have the strength of your convictions dude.

PS

A "Nice Bright Day" is an opinion, not an observation of fact. If you want to claim refuge in science, then you can't use subjective descriptors

*** Another Update ***

Took my Bodyguard Mercenary Healer up to the 1.3 GCD Tier.

This move was extremely painful...took 2 hours to get it done.

Triggered a 5.8 /s Heat Dissipation Rate @ 3160 Alacrity Points . . .

But she kept coming in at 1.31 or higher in her Raw GCD Results.

So I added an augment...very small creep up after several test runs.

So I added another augment...for another very small creep up.

Finally, 5 Augments and 3 Head Piece Upgrades later . . .

. . . 5.9 /s Heat Dissipation Rate was triggered...

...and she come in at a solid 1.3017 to 1.3035.

She feels awesome rotation speed wise...but God Almighty, she took a huge hit in Critical Points all the way back down to 1672.

http://www.mrrabbit.net/misc/swtor/swtorglobalcooldown.php

I was planning on stopping this character at 298 Gear, but I'll probably put in 300 Level Critical Enhancements when they drop just to get some of the Critical Points back.

=8-|

Just looked at the table on your website. Man, these approximations are so painful to read... If you are doing so much work on this, why not learn how to do actual math and figure out the exact Alacrity values needed?

Now, let's put real math behind us and take a look at what you have written.

You are saying that your Pyrotech PT with 2038 alacrity is carrying 30-40 wasted points but at the same time Corruption Sorcerer only has 10-15 wasted points with 2577 Alacrity.

How is this possible, considering that both of these classes have no passive alacrity boosts and it simply makes sense that they should have identical Alacrity Rating values? Why does your Sorc need ~500 more alacrity points than your Pyro PT? Why does Bodyguard require ~200 less Alacrity than Corruption sorc? Again, it affects them the same way, so why do you need ~200 extra on sorc healer?

P.S. By the way, according to real math, your Pyro PT has 825 wasted Alacrity points. And Corruption sorc has 1364 wasted points. The optimal number for 1.4s GCD is 1213 Alacrity rating. Basically, the main thing you are missing is that the GCD values are rounded in this game. So, you do not need to go all the way to 1.39-1.40, if you a little below 1.45, it is enough. That's why your values are so much higher than what is actually enough for the GCDs.

P.P.S. And no, 1.6s GCD does not exist. Just think about what you are saying. At 0 alacrity you would have pure, unchanged GCD value. And in this game it is 1.5s. How do you get 1.6? By running negative alacrity? That's what you get when you are basing your calculations off of practical tests that have a lot of inaccuracies.

Uh-oh, okay. After reading this, I see there is no point trying to talk any sense into this guy.

Okay...gonna try one more time...but to be honest it's like talking with an software engineer who proclaims that computer science is all about the code, after I've informed him that computer science is first and foremost a behavioral science....and all the while he is describing all the behaviors he is trying to duplicate or mimic in his code - and all the behaviors he is troubleshooting - not realizing the whole time that there is no argument to begin with.

And again, I have provided no argument for which to be proven wrong to begin with. Simply provided observations, given samples, and noted my wasted points and adjustments that I make.

 

That's it...

 

you are right, you just provided wrong conclusions based of inaccurate observations. i´m trying to prevent people from taking them as facts or do the same mistakes.

call it an argument, call it sharing opinions, i´m fine either way.

still.......wondering why someone not trying to get involved into an argument uses exaggerations like "religion" or "magic" for pure facts based on simple math, tho......