Making the Best Use of Recount, Part 6: Healing Done

This is the sixth in a series of posts about Recount, an add-on for World of Warcraft. It gathers and reports on data taken during combat.

Segment List:

1. The Introduction
2. Display Window Basics
3. Damage Done Details
4. The DPS Report
5. Damage Taken and Friendly Fire
6. Healing Done

Healing Done

The Healing Done summary chart is very similar to the Damage Done chart. Players are listed by their total healing done, with the largest healing volume at the top and the lowest at the bottom. Each bar shows the player’s name and the total amount healed. In parentheses next to the total amount healed are a healing per second (HPS) value and the percentage of total healing done by all players which was contributed by that player. Total healing done and the HPS number include overhealing done.

Effective Healing Detail Window

Clicking on a player’s name brings up his Effective Healing detail report, which is the first of four healing done detail windows.

Much like the Damage Done detail window, it’s split in half horizontally. The top half is a summary of the effective healing done by each of the spells the player cast, where effective healing is all the healing that actually restored lost health — everything that wasn’t overhealing. Spells are listed by which spell healed the most hit points in total. After the name of the spell are three columns: the Count column, which shows how many times the spell healed during the data recording session; the Heal column, which is the total hit points healed by the spell, and the % column, which is the percentage of the player’s total effective heals contributed by that spell. Each spell is assigned a color which is used to create the pie chart on the left.

You can select an individual spell in the top half to see further details about it in the bottom half. The bottom half further divides the total healing done by the selected spell according to the different kinds of spell hits performed. The spell which was selected for the above picture was Renew, a heal-over-time (HoT) spell which cannot crit, so all it does is “Tick” once cast, periodically healing the recipient for a small amount. A different spell, such as Flash Heal, might show “Hit” and “Crit” under the Type column, reflecting data for each. Other columns show minimum, average, and maximum effective healing done per heal the spell performed, as well as the number of heals performed and what percentage of the total healing done with that spell came from that spell hit type.

Overhealing Detail Window

This detail window covers all the healing which was not effective healing, and is also split in half horizontally. The top half is a summary of the overhealing done by each of the spells the player cast. Spells are listed by which spell overhealed the most hit points in total. After the name of the spell are three columns: the Count column, which shows how many instances the spell overhealed someone during the data recording session; the Overheal column, which is the total hit points overhealed by the spell, and the % column, which is the percentage of the player’s total overheals contributed by that spell. Each spell is assigned a color which is used to create the pie chart on the left.

You can select an individual spell in the top half to see further details about it in the bottom half. The bottom half further divides the total overhealing done by the selected spell according to the different kinds of spell hits performed. Other columns show minimum, average, and maximum overhealing done per heal the spell performed, as well as the number of overheals performed and what percentage of the total overhealing done with that spell came from that spell hit type.

Healed Who Detail Window

This window is split in half horizontally. The top half lists the party/raid members on whom the player cast his healing spells, with the people receiving the most healing at the top. This list includes any pets healed. After each character or pet’s name are two columns. The Healed column lists how many hit points were healed in total for that character/pet, and the % column shows what percentage of the healer’s total healing done (both effective and overhealed) went to that character or pet. Each character/pet is assigned a color, which is used to represent it in the pie chart on the left.

Selecting a character or pet from the list brings up a detailed list of healing done to that character/pet by each kind of healing spell cast upon him. Spells are arranged here according to which dealt the most healing. After the list of spell names are two columns, with the Healing column listing the amount of healing done to the selected character/pet by that spell and the % column listing what percentage of the total healing done to the character/pet which was done by that spell. Again, each is assigned a color for use in a pie chart on the left-hand side.

There is no way to determine how much overhealing was dealt to a specific character.

Time Spent Healing Detail Window

This detail window is split horizontally into two sections. The top section lists the characters and pets healed by the player according to how much time each character/pet was under the effects of the player’s healing spells. After each character/pet name are two column entries. The first, Time (s), is the time in seconds that the character/pet was under the effects of the player’s healing spells. The second, %, lists the percentage of the player’s total time spent healing which was devoted to that character/pet. Colors are assigned to each character/pet for use in creation of the left-hand pie chart.

The bottom half of the detail window lists the healing spells used on the character/pet highlighted in the top half of the window according to which spells were in effect on the character/pet for the greatest amount of time. After the name of each healing spell are the Time (s) column, which lists the amount of time the highlighted character/pet was under the effect of that spell during the time the data was recorded, and the % column, which lists how much of the total time the selected character/pet was under the effect of the player’s heal spells was time spent under the effect of that spell.

As you can see in the image above, heal-over-time spells such as Renew dominate the top of the spell list for each character/pet. Furthermore, the total time spent healing a character/pet is the sum of each individual spell’s time in effect on the character/pet. The data fails to take into account the fact that multiple healing spells from one player can be in effect on a single character/pet at one time. For example, a priest can cast Prayer of Mending and Power Word: Fortitude on a character who is already under the effect of Renew, and land Lesser Heal on the character at the same time Glyph of Power Word: Fortitude takes effect from the protective bubble being popped. The time a player spends healing is therefore of limited usefulness.

Flaws in and Potential Abuses of the Healing Charts

Each healing class in WoW has its own unique array of spells. Subsequently, they play very differently from each other and have different strengths and weaknesses. While any healing class, given a good enough player, can heal just about any situation, certain classes are better for certain healing jobs than others. The strongest examples of this are the paladin and shaman classes. Thanks to Beacon of Light, a good paladin healer can raid heal and still keep a tank up, but a paladin really excels when all he has to do is focus on one or two people. A shaman can keep the raid up all day with chain heal, but single-target healing is tough for him.

Another consequence of the differences in strengths and weaknesses from class to class is that certain classes are doomed to do a lot of overhealing in spite of their best efforts. Any healing spells which apply healing to a character at max health are said to have overhealed the character. This can occur when a healing spell is cast on someone who is already at full health or when a healing spell heals more health than the character is currently missing. Paladins who rely heavily on Beacon of Light often overheal their beacon; chain heal does its best to target people who need health the most, but can only jump so far and sometimes heals people who are hardly injured at all; and druids lay down heal over time spells which last for a given duration without regard for the character’s current health. Analyses of overhealing done should take these differences into account, a task which is made easier with the help of the detail windows.

Then there are the discipline priests, which specialize in Power word: Fortitude. They come up low on the Healing Done summary chart during raids because they’re too busy preventing damage to heal very much. I’ve often seen ignorant leaders of PUG’d raids show disdain for and/or kick discipline priests because their HPS was below some numerical standard. Discipline priests are invaluable allies, able to keep the party or raid alive for a long time… but they’re inaccurately represented by Recount’s data.

I must confess, the Time Spent Healing detail window seems silly to me. While the data contained therein is an important part of calculating the HPS number, the only reason I can see for having it publishable is to find out which of the DPS are taking up too much of the healers’ efforts. You can find that out just by listening to the healers, though — they’re usually pretty good about telling DPS when they’re taking inordinate amounts of preventable damage.

The good that comes of the Time Spent Healing detail window is an understanding of how HPS and DPS are calculated. Casters who rely heavily on heal-over-time and damage-over-time spells will have a much higher value for the time they spend healing or damaging than players whose healing/damage mostly come from one-hit abilities. Since healing per second and damage per second are both calculated by dividing the total healing or total damage done by the amount of time spent on doing it, they will come out with a lower HPS/DPS number. At a cursory glance, classes such as fire mages, affliction warlocks, and restoration druids may look weaker than other classes for that reason alone, regardless of whatever class nerfs Blizzard has in place at any given time.

An Early End

My interest in WoW has waned. In Blizzard’s attempts to make end game content more accessible to as many people as possible (and therefore increase the number of subscription months they sell), they’ve made end game content less challenging. The game, in short, bores me now. I had hoped to finish this series of posts in spite of that, but the Magic 8 Ball in the back of my head says, “not likely”. I had to force myself to finish this post. There are a couple of key features — like the real-time graph and the death detail window — that are to be left out, which is a real shame. I may go in and do those later, anyway, if I work up the internal motivation. (That would occur in summer 2010 at the earliest, since I just landed a part in TBA Theatre‘s production of The Sound of Music on top of my full-time job and three independent study courses.)

That said, if you look through the other posts in this series, you’ll get a feel for how most of the rest of the data windows and charts work. All of the basic data summaries are similar in format. Clicking on player names in any data summary chart brings up details windows with what should, by now, be familiar formatting (with one notable exception off the top of my head).

Keep an open mind, apply a little logic, and remember that data has to be interpreted to be of any use. You must keep both class and boss mechanics in mind when looking at the accumulated data in order to get a clear, realistic picture of what happened during the battle. A good player is more than the sum of his gear score and his DPS/HPS number; he understands what his character can do and when to use each skill in his arsenal to maximum effect. Recount is a tool which can help raid leaders and players shore up their weak spots, not a scoreboard.

Showing Up to a Raid is Good Etiquette

Some World of Warcraft players have a chronic problem with accepting raid invitations and failing to show up to them. It’s a problem I’ve seen before in Dungeons & Dragons, as well. The player is of the “it’s only a game” mentality — since it’s only a game, it doesn’t matter if he shows up or not.

While it’s true that it’s only a game, such people fail to see is that it’s not about the activity you engage in. It’s about the people on the other side of the table — or internet, in the case of WoW. We could be planning a huge cake-baking session, for example, and if the guy bringing all the flour decides at the last-minute that he’d rather go see a movie with his friends, well… sure, we can get flour somehow. It’s likely to take at least half an hour, though, and that’s several other people’s time wasted because one person was inconsiderate. In business terms, if you wasted half an hour of nine people’s time at the example wage of $10 an hour, then that’s 4.5 hours or $45 (plus the extra taxes employers have to pay).

And in the case of a WoW raid, the absence of a key player (such as a tank or healer) can cause a raid to fail to get off the ground altogether. That, then, is nine people’s plans for the evening ruined. Players have to block of a 3+ hour block of time for raiding, usually. If someone simply doesn’t show up, opportunities have been missed for the other nine people to do other things that evening. Maybe another member of the raid got invited to see a movie, too, but since the movie and the raid started at the same time had to decline; by the time the raid breaks up 45 minutes after the scheduled start time, the movie is half over. If he’d known the day before that the raid wouldn’t happen, he could have gone to see the movie and still had a good time.

What it really boils down to is that you are making a time commitment to your fellow raiders. Forget the game — that’s just what you guys have chosen to do with your time. If a real emergency comes up now and again, that’s okay. Your raid group will understand (unless, of course, they’re assholes). But if you can’t give your raid leader at least a day’s notice of cancellation so he can try to find a replacement for you, you should prioritize your commitment to your raiding friends over last-minute invitations to hang out with your face-to-face friends.

Unfortunately, your face-to-face friends may get offended because “it’s just a game”. If that happens, you should either point out the dictates of courtesy to them and make them deal with it, or choose not to commit yourself to the raid group. Don’t, however, use courtesy as an excuse to let your face-to-face friends and significant other completely fall by the wayside. Keep in mind, also, that it goes both ways — if your raid leader sets up a raid on a day on which you already have plans with someone, tell him you can’t go to raid that day because of a previous engagement.

Be responsible. Be courteous to all of your friends. Have fun.

Making the Accessed More Accessible

Since my World of Warcraft posts are so popular, I just added an entry to the menu for WoW. That page links to all of my World of Warcraft articles. Enjoy.

Making the Best Use of Recount, Part 5: Damage Taken and Friendly Fire

This is the fifth in a series of posts about Recount, an add-on for World of Warcraft. It gathers and reports on data taken during combat.

Segment List:

1. The Introduction
2. Display Window Basics
3. Damage Done Details
4. The DPS Report
5. Damage Taken and Friendly Fire
6. Healing Done

The next two reports I cover, again clicking to the right in Recount, are not complicated. Their formats are similar to those of reports covered previously.

The Damage Taken Report

The Damage Taken summary chart lists players in order according to who took how much damage. Each player’s bar shows how much damage they took, followed in parentheses by what percentage of the total damage taken by everyone was taken by that player.

Clicking on a player’s name in the list brings up a detail window similar to those for the Damage Done report.

The window is split in half, horizontally. In both halves, each item in the list is assigned a color for use in the pie chart to the left.

The top half lists the mobs which damaged the player according to how much damage they did, with those mobs doing the most damage at the top. The Damage column lists the flat number of damage dealt by all enemies with that name for the duration of the data collection, and the % column shows how much of the total damage the player took that damage is.

The lower half shows more details about the damage dealt to the player by the mobs of the selected name. It shows a list of the mobs attacks, with the attacks that dealt the most damage to the player at the top of the list. It does not show details about the mob’s critical hits, misses, and the like.

Why the Damage Taken Report Matters

The damage taken report can be used as a tool for diagnosing threat problems (though — to be completely honest with you — it’s not usually necessary, since the kind of problems it shows are usually pretty easy to diagnose without any help).

The tank(s) should be on top with a rather large chunk of the damage taken. If a damage dealer is pulling aggro, a quick look at this chart can tell you who that is so that the player can adjust his rotation or what have you and prevent further problems. Even the best tanks have trouble competing with damage dealers in superior gear going all-out on mobs — sometimes the damage dealer needs to back off a bit. A look at the Damage Taken detail window for a player can help determine if the player needs to be more careful on certain kinds of fights or against specific mobs.

The Friendly Fire Report

The Friendly Fire report doesn’t usually have anything to show you. The only times friendly fire comes into effect is when a player is mind controlled or has one (or more) of their spells reflected. When there is data, its summary looks more like the Damage Taken summary than the Damage Done summary. It only shows the amount of friendly fire damage done by a player and what percentage of the total friendly fire dealt that said amount was.

In contrast, the detail window is pretty much identical to the detail window for a player’s hostile attacks under the Damage Done report.

The window is split in half, horizontally. In both halves, each item in the list is assigned a color for use in the pie chart to the left.

The top half lists the player’s damaging abilities according to how much of the player’s total damage they composed, from greatest to least. This amount is shown in the Damage column as an exact number, in the % column as a percentage of the total, and graphically in the pie chart. The Count column tells you how many times the ability hit something, not how many times it was used. So a multi-target ability (such as Blizzard) would add to the Count total every time it hits a mob.

The lower half shows further detail for the damaging ability selected. It shows how many of the counted hits with the ability were hits, crits, misses, dodged, parried, resisted, or what have you. They are listed by frequency, with most frequent at the top and least frequent at the bottom. For each type of hit, the minimum, maximum, and average damage dealt is shown.

Why the Friendly Fire Report Matters

In boss fights such as, say, Yogg-Saron where you have the potential to have multiple players mind controlled at once, you can look at who’s doing the most friendly fire damage to determine who needs to be taken care of first.

In situations where you’re dealing with spell reflect, however, it can really only serve as a reminder to casters to watch out for spell reflect. It could also be used by an individual to track his progress in avoiding spell reflect.

Coming Up Next

Next time I start getting into the healing charts. They’re the charts most often looked at after the Damage Done and DPS charts — and, likewise, the ones most abused after the Damage Done and DPS charts.

Making the Best Use of Recount, Part 4: The DPS Report

This is the fourth in a series of posts about Recount, an add-on for World of Warcraft. It gathers and reports on data taken during combat.

Segment List:

1. The Introduction
2. Display Window Basics
3. Damage Done Details
4. The DPS Report
5. Damage Taken and Friendly Fire
6. Healing Done

A Quick Note

In my last installment I overlooked the most obvious part of the Damage Done window, which is the summary chart. It’s fairly straightforward, but I still should have covered it.

The chart lists all the players who did damage during the fight or fights whose data is displayed, with the player who did the most damage at the top of the chart. Each player is represented by a bar in the color which commonly represents their class. Overlaid on that bar are the player’s name and the total damage they did during the fight or fights for which data is displayed, followed by parentheses containing a decimal number representing the damage the player did per second while in combat and what percentage of all damage done came from that player. All of the summary charts Recount displays use similar formatting.

That said, you may have noticed that the person who did the second-largest amount of damage in the above screenshot did more damage per second than the person who did the largest amount of damage. This leads me into Recount’s second data report.

The DPS Report

The second data report recount offers us — assuming we’re proceeding from left to right using the navigational arrows at the top of the window — is the DPS report. DPS is an abbreviation for Damage Per Second, and is a measure of ouchies a player causes in a given amount of time.

I don’t know what happened to the image I had for this. I may add it in later, but I don’t think it’s really necessary. The DPS report looks a lot like the Damage Done report, except that the only numbers displayed on a player’s bar are the DPS number and their percentage of the total damage done (the latter of which is in parentheses). Players are ranked according to who’s putting out the highest damage over time according to Recount’s calculating formula.

Clicking a player’s name in the list brings up the same detail windows you get from clicking player names on the Damage Done report.

Why DPS is Not Valid as a Sole Measure of a Player’s Eliteness

If the player who throws out the highest DPS dies in the first few seconds of the fight, he can end up being last on the Damage Done chart. (Just for the record, this is a perfect example of why good gear doesn’t make a good player. He who does not live to bring his massive DPS to bear is as useless as bringing a sack of flour to the raid.)

Now, even if all players stay alive for the duration of the fight, you’ll see some discrepancies between the Damage Done and DPS reports. There are two possible reasons for this:

1. The player(s) with higher DPS weren’t attacking as much as the player(s) with the highest Damage Done. There are any number of possible reasons for this. For some fights you have to move around a lot, so a spell caster with few or no instant cast spells ends up being robbed of time in which he could be attacking. In a fight in which the melee fighters have to keep switching targets, they can potentially spend a lot of time out of range of anything to attack. The raid leader might have started the fight while someone was AFK, or the player could have disconnected mid-fight. The player may have had his face handed to the floor in one swell foop after catching the boss’s attention with a poorly timed critical hit. The player may have spent most of their time on crowd-control and/or interrupts. Etcetera.
2. Recount’s DPS formula doesn’t represent all classes (or specs, even) equally. To use an example from my own class, if a fire mage and an arcane mage do the same amount of damage in the same amount of time, the arcane mage will always show up with a higher DPS number. Why? It seems to have to do with the fact that fire magery is more DoT based than arcane magery. We fire mages — if we’re playing right — are always sapping life from our enemies somehow, which inflates our active attack time and lowers our DPS number. I don’t know what similar inaccuracies there are for other classes, but I’m sure they exist.

Coming Up Next

The next report on our list is the Damage Taken report.