PDA

View Full Version : Mortality rates by Level and Class for two Hardcore 90min time samples.



Thoggy
10-03-2019, 11:09 AM
Description.

Character death times, levels, and classes on the hardcore server for two 90 minute periods (3-4:30pm & 7:45-9:45pm EST Wed Oct 2, 2019) are reported.
Race, class, location of death, and type of death was not recorded. Please see raw data at the end of post.

Each sample was 90 mins long. The Day sample has 4 deaths missing class information.
Total population size was not accurately estimated. Only estimate was that Level 8 was the rough midpoint in the Social Tab "who" for both samples (sorted by level) during sample times.

Results.

Percentage of Deaths by Level.
The Day sample had about 76% (13 dead) at levels 2-10 and about 24% (4 dead) at levels 11-20.
The Night sample had 80% (16 dead) at levels 2-10 and 20% (4 dead) at levels 11-20.

Number of Deaths per minute.
The Day sample had 1 death per 5.3 minutes (90/17) while the Night sample had 1 death per 4.5 minutes (90/20).
The higher death rate for the Night sample might be due to a higher playing population.

Time between Deaths in minutes.
The Day sample had 17 deaths with the mean average time between deaths of 4.82 mins (SD=5.49, Sum of X=82).
The Night sample had 20 deaths with the mean average time between deaths of 4.50 mins (SD=4.45, Sum of X=90).
Individually observed Time between Deaths ranged from 0 to 21 minutes.

Character death by Class.
Considering both samples, there were 10 multi-class character deaths (nine with 2 classes, 1 with 3), out of 37 total dead (about 27%).
Note that 4 deaths have unknown classes.
The distribution of known single class deaths while ignoring level is:

Art = 4
Barb = 1
Bard = 1
Cleric = 2
Fav Soul = 1
Fighter = 1
Monk = 2
Paladin = 1
Ranger = 2
Rogue = 3
Sorcerer = 1
Warlock = 4


End Notes.

This is a partial snapshot of a single day about mid-way through the 1st Hardcore server.

In both samples, "significantly" (note stat not actually calculated, this is a guess) more deaths occur between levels 2-10 (~75%-80%) than 11-20 (~20%-25%).

Both samples had roughly the same time between deaths (the Day's sample slightly higher rate explained by a combo of chance and probably a smaller server population.

Edit: If one assumes 1 death per 5 min then on average there are 288 deaths per 24 hours on the HC server. At 45 days that is 12, 960 deaths.

Party wipes can be possibly identified by multiple deaths of same level characters all within 0-2 mins of each other (e.g. 3:28pm Day or 8:43pm Night).

Character class information is very limited by sample size and missing data but possibly indicates higher mortality rates for trappers (Arts, Rogues) than other classes.

Better estimates of mortality rates by class would need a greater sample size than provided by these two samples plus some measure of the living population by class.

Comparisons to mortality rates from the first week of the server and/or the last week would be interesting.
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ ~~~~~~~~~~~~~~~~~~~~

Raw Data.

DAY Sample 1. Oct. 2, 3:00 to 4:30pm EST.

Time of death, Level, Class, (Mins between deaths)
3:21, 4, Monk, (21 mins from start observation)
3:28, 19, Wiz 17/Rogue 2, (7)
3:28, 17, Fav Soul, (0)
3:37, 4, Fighter 2/Rogue 2, (9)
3:42, 2, Missing, (5)
3:43, 4, Missing, (1)
3:43, 3, Warlock, (0)
3:45, 8, Missing, (2)
3:58, 9, Missing, (13)
3:58, 8, Cleric, (0)
4:03, 2, Paladin, (5)
4:09, 15, Rogue, (6)
4:11, 8, Warlock, (2)
4:15, 11, Art 9/Rogue 2, (4)
4:16, 3, Warlock, (1)
4:21, 3, Ranger, (5)
4:22, 5, Monk, (1)

NIGHT Sample 2. Oct 2, 7:45 to 9:15pm EST.

Time of death, Level, Class, (Mins between deaths)
7:50, 4, Bard, (5 mins from start observation)
7:52, 6, Art, (2)
7:52, 7, Art, (0)
7:55, 10, Art 8/Rogue 2, (3)
7:55, 4, Warlock 3/Rogue 1, (0)
7:59, 7, Rogue, (2)
8:20, 3, Rogue 2/Art 1, (21)
8:20, 3, Warlock 3, (0)
8:27, 10, Druid 5/Art 3/Barb 2, (7)
8:31, 17, Art, (4)
8:32, 9, Ranger, (5)
8:37, 2, Sorcerer, (5)
8:39, 6, Rogue 4/Art 2, (2)
8:43, 11, Rogue, (4)
8:44, 12, Cleric, (1)
8:44, 11, Fighter (0)
8:51, 5, Art (7)
9:03, 4, Monk 2/Wiz 2, (12)
9:06, 9, Barb 9, (3)
9:12, 5, Art 3/Bard 2 (6)

Ordinary
10-03-2019, 11:21 AM
Description.
3:28, 19, Wiz 17/Rogue 2, (7)
3:28, 17, Fav Soul, (0)

This was my group that was in the end fight of WGU. Just as all the Shadowvars died and he says "Do I have to come down there..." he promptly knocked the whole party down with some AOE attack while a load of electrical traps went of on top of us. These two died, while the rest of us all got knocked down to between 10-25% of our hp.
It was something none of us had ever seen before happen in the WGU end fight (although I have seen the electrical effect, just not the knockdown). At this point in the proceedings we had not had any close scrapes at all.

The rest of us went on to complete.

Also, some interesting stats.

Thoggy
10-03-2019, 11:37 AM
This was my group that was in the end fight of WGU. Just as all the Shadowvars died and he says "Do I have to come down there..." he promptly knocked the whole party down with some AOE attack while a load of electrical traps went of on top of us. These two died, while the rest of us all got knocked down to between 10-25% of our hp.
It was something none of us had ever seen before happen in the WGU end fight (although I have seen the electrical effect, just not the knockdown). At this point in the proceedings we had not had any close scrapes at all.

The rest of us went on to complete.



Thank you for the additional information.

Paisheng
10-03-2019, 01:47 PM
I appreciate the snap shot, it is interesting, but no conclusions can be drawn.

For example more deaths occurred from level 2-10 than 11-20. One cannot infer that makes the lower levels more dangerous without overall population counts. It is likely many more folks are in levels 2-10 than the higher levels which would skew the results. If there were 100 people in lower levels and 20 died that is 20% death. But if there are only 20 people in higher levels and 5 die that is 25% death.

I do find interesting the wide range of different classes that are represented however. More warlock and artis are no surprise as those appear to be some of the most popular builds and likely again have a greater population.

Silverleafeon
10-03-2019, 02:21 PM
Weekends tend to be far more populated than weekdays, although weeknights are actually almost as busy.

This last week has seen a marked decrease in overall impression of deaths occurring during the weekdays.
It is thought by my guildies that some players have migrated back to the main servers, while the remaining are being more careful/adjusted/smart.