Non-holy pally healing gear

[swirly]

Alright so my problem before was that I wasn't sure what mp5 value to assume for the Insightful meta. 25? 35? 50? Which would be fair. I've come to the conclusion though that that was just me being dense. Thanks to conc I have the proc info and all the other numbers. For figuring out the stat values I've assume FoL spam and so I might as well do that for this as well. I'm going to count it as a 2s cast instead of 1.5s just to account for human error and other such things that would slow a person down. I've also been going on about how I don't know how to account for different gems being required, but like conc talked about in a previous post all that has to be done is calc an approximate loss from gem switches.

So then from all of that we end up with

Gem value =(12/int rating) + (((300/(20*2))*5)/mp5 rating) - ((red - blue)*2) - ((red - yellow)*2) + (2/mp5 rating)

Which assumes in the pointing of the gem a loss from 2 blue and 2 yellow switches, but a gain from a socket bonus of 2 mp5 (which I feel is the most common bonus amungst the gear).

This seems like a reasonable way of valuing the gem. So the spreedsheat has been updated with that info.

[swirly]

So while figuring out how to value the insightful meta I decided that counting FoL as 2s instead of 1.5s might be a good idea. Now maybe there are people who have good connections and good reaction times such that they can actually get 1.5s flash heals. My gut says that I'm probably always late on mine. Maybe not a full 0.5 seconds late, but I'm going to assume that for some calcs anyway to see what happens.

100 healing

100 * 0.4286 = 42.86 healing added to each FoL.

1542.86h / 2s = 771.43h/s

100 haste

New Casting Time = 2355 / 1670 = 1.41

The 0.5 that I am adding isn't shortened by the haste itself, so I'm just adding it onto the time. So New Cast Time = 1.91s

1500h / 1.91s = 785.34h/s

Haste - Heal = Gain from haste

1500: 785.34h/s - 771.43h/s = 13.91

Which means there is still a gain from haste, but it is about half of what it was at 1.5s.

Baseline: 1500h / 2s = 750h/s

Gain from base

100 Heal: 771.43h/s - 750h/s = 21.43h/s
100 Haste: 785.34h/s - 750h/s = 35.34h/s

So you are looking at a 1.65 gain from haste. That puts us at a 10 heal = 6 haste scale. Add 1 to be sure you meet the 1500h value and add 2 to account for the mana requirements of haste and you get 10 heal = 9 haste. So it seems that 10=7 is about where it is if you are perfect in your casts and 10=9 is about where it is if you are slow. I think I would rather have the spreadsheet error on the side of not being perfect and so I'm going to change it to 10=9.

So that puts the scale at

10 heal = 9 haste = 4 mp5 = 25 crit = 25 stam = 30 int

As usual, the spreadhseet has been updated to reflect this.

[WimpySmurf]

So while figuring out how to value the insightful meta I decided that counting FoL as 2s instead of 1.5s might be a good idea. Now maybe there are people who have good connections and good reaction times such that they can actually get 1.5s flash heals. My gut says that I'm probably always late on mine. Maybe not a full 0.5 seconds late, but I'm going to assume that for some calcs anyway to see what happens.

With the change to stopcasting you can chain cast FoL. All spell requests made are sent to the server and the first one made with 0.5s left on the cast is "queued" up to start casting immediately after the current cast is finished. So you could chain cast FoL at 1.5s.

-WimpySmurf

[swirly]

With the change to stopcasting you can chain cast FoL. All spell requests made are sent to the server and the first one made with 0.5s left on the cast is "queued" up to start casting immediately after the current cast is finished. So you could chain cast FoL at 1.5s.

-WimpySmurf

Do you have a source for this? Specifically "the first one made with 0.5s left on the cast is "queued" up to start casting immediately after the current cast is finished". Patch notes say "Client spell cast requests are now sent to the server even if your player is already casting another spell. This eliminates the need for /stopcasting in macros to compensate for latency." As far as I know there is no queueing.

Now the change does allow you to send your casts before they show as finished on your screen so that you can chain them much better. There is still human error involved. Using a mod like quartz you can get a general idea of what your latency is and so time when to send the next heal. Latency is variable though and so you aren't going to get it exactly right. Its still better than heal..wait for it to finish...heal, but it isn't going to be a perfect 1.5s cast rotation. Some would say "just spam the key so you can get it as close as possible", but my understanding is that if you send the heal early then your client ends up having to wait for the error message saying it was early and so you actually get set farther back in your timing. So what you end up doing is trying to time when you hit the cast button so that the command reaches the server just as your previous spell finishes. It isn't an exact thing though.

Now maybe I'm wrong about this. There could have been more changes since the patch I'm refering to. I've certainly not paid overly close attention to such to know for sure. The above explains my current understanding though. So please point me to a source that says otherwise if things have changed.

In addtion to all of that, there is also stuttering, having to move, delay from changing targets, etc etc. So even if one could reach a perfect 1.5s cast cycle, I think the 2s calculations are more accurate due to these other delays in the fight. I'm certainly open to hearing arguements otherwise though!

[swirly]

I've gone in and added some very basic drop/vender locations to the items in the spreadsheet. This way a person can judge what items they should be aiming for not only based off of score, but also off of price or availability. In addition to this I have added a second sheet to the document that has 4 sections for a person to choose gear. In each of these sections one can select a gear set. The document totals the points for the gear selected so that you can then compare the total points for each set. This makes it easier to see how different overall sets compare.

the spreadsheet

[swirly]

Since there has been some discussion about what cast time to use in the calculations, I decided to make it an option in the spreadsheet. So now there is a place to enter what cast time to use as well as a place to enter a delay. So for FoL one can enter 1.5 for the cast speed and either 0 or 0.5 for the delay depending on what you feel is accurate. So in order to make this actually have an effect I had to go in and change the ratings from hard coded numbers to equations. In the process of this I realized some of my previous math was horrible.

First off, when I was calculationg the ratio for haste/heal I was doing it using the gain from base values. It is more accurate to use the actual HPS values for that calculation. Doing that raises haste quite a bit so that the rating is 10 heal = 9.82 haste. This is with FoL hitting for 1500, which I have also added as an adjustable value in the spreadsheet. Due to this, I don't think there is a need for penalizing the haste rating to be sure you hit the assumed heal rating. There still may be need for a mana penalty though and so I've added a field for that. Currently I just have it set at a mana penalty of 1 which brings it to 10 heal = 10.82 haste.

I also noticed that my crit conclusion was really weird. I said that heal wins at 1500 and below, but looking back at the numbers they clearly show that crit is better from 1000 on up. So I don't know what I was thinking at the time. Instead of worrying about that though, I've just switched the crit rating value over to an equation as well which is based off of the value you entered for your FoL heals. That ends up giving 10 heal = 22.54 crit.

So the ratings now stand at

10 heal = 10.82 haste = 4 mp5 = 22.54 crit = 25 stam = 30 int

With some of then being equation based and so variable based on what you enter for cast speed and etc. It should also be noted that the Insightful calculations are also based off of the value entered for cast speed/delay. I also took out the the penalty on it for having to use different gems to meet the requirement. I figure that with the comparison sheet now, one can see what using the different gemmed gear does. So having Insightful penaltized in it's own rating would cause it to be penalized twice.

So basically more things have been converted to equations for excel to calculate so as to remove human error and allow for easier adjustments.

The spreadsheet