Whoops! Actually, muder rate is betwen 2 and 6%, not 16%

jasonrohrer · 2018-06-06 16:32:26

So, I forgot how this data is stored in the lineage server. killer_id is -1 if there is no killer, and if there is a killer, it's the id of the killer. But it can also be negative, storing the id of the non-human deadly object (snake or whatever). So I was counting all violent deaths as murder by accident, which gave the 16% figure. Here's the real data from as long as the lineage server has been running:

.
+------------------+---------+---------+-----------+-------------+
| DATE(death_time) | over_20 | murders | non_human | murder_rate |
+------------------+---------+---------+-----------+-------------+
| 2018-05-11       |    1342 |      36 |         0 | 2.68 %      |
| 2018-05-12       |    2186 |     104 |         0 | 4.76 %      |
| 2018-05-13       |    2394 |      63 |         0 | 2.63 %      |
| 2018-05-14       |    1863 |      78 |         0 | 4.19 %      |
| 2018-05-15       |    1816 |      65 |         0 | 3.58 %      |
| 2018-05-16       |    1654 |      63 |        58 | 3.81 %      |
| 2018-05-17       |    1531 |      83 |       229 | 5.42 %      |
| 2018-05-18       |    1549 |      66 |       252 | 4.26 %      |
| 2018-05-19       |    1791 |      81 |       314 | 4.52 %      |
| 2018-05-20       |    1891 |      90 |       317 | 4.76 %      |
| 2018-05-21       |    1693 |      81 |       208 | 4.78 %      |
| 2018-05-22       |    1693 |      63 |       270 | 3.72 %      |
| 2018-05-23       |    1570 |      80 |       222 | 5.10 %      |
| 2018-05-24       |    1500 |     100 |       223 | 6.67 %      |
| 2018-05-25       |    1577 |      66 |       259 | 4.19 %      |
| 2018-05-26       |    1995 |      89 |       338 | 4.46 %      |
| 2018-05-27       |    2062 |     101 |       372 | 4.90 %      |
| 2018-05-28       |    2058 |      67 |       390 | 3.26 %      |
| 2018-05-29       |    1955 |      46 |       337 | 2.35 %      |
| 2018-05-30       |    2039 |      39 |       360 | 1.91 %      |
| 2018-05-31       |    1953 |      44 |       359 | 2.25 %      |
| 2018-06-01       |    2096 |      29 |       359 | 1.38 %      |
| 2018-06-02       |    2252 |      43 |       386 | 1.91 %      |
| 2018-06-03       |    2320 |      75 |       441 | 3.23 %      |
| 2018-06-04       |    1950 |     114 |       333 | 5.85 %      |
| 2018-06-05       |    1833 |     105 |       292 | 5.73 %      |
| 2018-06-06       |    1105 |      43 |       184 | 3.89 %      |
+------------------+---------+---------+-----------+-------------+

So it's not nearly as bad as it sounded.

Here are all deaths/murders, not just those over the age of 20:

.
+------------------+------------+---------+-----------+-------------+
| DATE(death_time) | all_deaths | murders | non_human | murder_rate |
+------------------+------------+---------+-----------+-------------+
| 2018-05-11       |       6582 |      69 |         0 | 1.05 %      |
| 2018-05-12       |      11742 |     176 |         0 | 1.50 %      |
| 2018-05-13       |      13591 |     119 |         0 | 0.88 %      |
| 2018-05-14       |       9433 |     126 |         0 | 1.34 %      |
| 2018-05-15       |       8928 |     128 |         0 | 1.43 %      |
| 2018-05-16       |       8321 |     114 |       275 | 1.37 %      |
| 2018-05-17       |       7122 |     133 |      1300 | 1.87 %      |
| 2018-05-18       |       7519 |     109 |      1506 | 1.45 %      |
| 2018-05-19       |       9675 |     151 |      1764 | 1.56 %      |
| 2018-05-20       |       9539 |     183 |      1667 | 1.92 %      |
| 2018-05-21       |       7246 |     128 |      1325 | 1.77 %      |
| 2018-05-22       |       8277 |     128 |      1508 | 1.55 %      |
| 2018-05-23       |       7266 |     149 |      1242 | 2.05 %      |
| 2018-05-24       |       7728 |     194 |      1473 | 2.51 %      |
| 2018-05-25       |       8030 |     122 |      1477 | 1.52 %      |
| 2018-05-26       |      11727 |     148 |      2230 | 1.26 %      |
| 2018-05-27       |      11391 |     186 |      1995 | 1.63 %      |
| 2018-05-28       |      12303 |     147 |      2350 | 1.19 %      |
| 2018-05-29       |      11602 |     100 |      2094 | 0.86 %      |
| 2018-05-30       |      13434 |      87 |      2220 | 0.65 %      |
| 2018-05-31       |      11787 |      83 |      2081 | 0.70 %      |
| 2018-06-01       |      12584 |      88 |      1997 | 0.70 %      |
| 2018-06-02       |      14040 |      90 |      2445 | 0.64 %      |
| 2018-06-03       |      12900 |     127 |      2668 | 0.98 %      |
| 2018-06-04       |       9138 |     204 |      1858 | 2.23 %      |
| 2018-06-05       |       8018 |     199 |      1497 | 2.48 %      |
| 2018-06-06       |       4716 |      85 |       871 | 1.80 %      |
+------------------+------------+---------+-----------+-------------+

Christoffer · 2018-06-06 16:40:30

That’s a relief. Still, it would be good to increase the difficulty in doing killing sprees. Killing is sometimes anti-griefing, but mass killing is always griefing of the worst kind.

Alleria · 2018-06-06 17:05:24

Flintstone wrote:

That’s a relief. Still, it would be good to increase the difficulty in doing killing sprees. Killing is sometimes anti-griefing, but mass killing is always griefing of the worst kind.

I can't remember the last time I saw a mass killing spree... Carrot spongeing (even with OP berries) is still a more pernicious form of griefing than mass killing, since it's less detectable, nonpreventable, and more disastrous.

I don't see there being a "murder epidemic" like some have claimed - it's a non-issue.

Last edited by Alleria (2018-06-06 17:05:37)

Hstrike · 2018-06-06 17:38:26

The average murder rate over this period is: 3.93%

As you can see, despite the recent hike, we are still experiencing a declining pattern of murder.

EDIT: spreadsheet online here.

Last edited by Hstrike (2018-06-06 20:33:13)

jasonrohrer · 2018-06-06 18:04:11

Thank you for making that graph!

thirdplanet · 2018-06-06 18:50:18

woah that is going to be fascinating to keep watching... why the murder-spike over labor-day weekend ?? more blood-thirsty players at that time?

it would also be interesting to look at murder rate vs. absolute number of players... when there are just more players around, does the whole rate go up? like an over-crowding type of thing?

breezeknight · 2018-06-06 19:00:14

that clarifies it then

i was last killed on mai 27th, i was one of those insignificant 186 kills, it's mere 1.63 %

since emotional involvement in the gameplay is not intended, so looks like there's no need for me to be upset then
no need to change anything also, everything's just fine

how bureaucracy puts everything in the right perspective !
i feel much better now, thank you

thirdplanet · 2018-06-06 19:15:41

using "total deaths" as a substitute for "total number of players playing" (since everyone must die), murder rate vs. total number of players is thus. I'd have to do a regression to determine correlation...

(ignore the graph title I forgot to change it lol, should read "all deaths vs murder rate")

Last edited by thirdplanet (2018-06-06 19:16:38)

thirdplanet · 2018-06-06 19:26:03

So there was a spike in number of players and a depression of murder rate following the update that introduced new characters and roses. Players peaked and murder rate trough'd following corn/berry/well overhaul update. Players tanked/murder rate spiked shortly after though.

Don't have any good theories as to why in particular, I just love data analysis haha. Conclusion: needs more data.

Last edited by thirdplanet (2018-06-06 19:27:01)

Glassius · 2018-06-06 19:40:05

Regression graph is awesome. It showes, when we beat the environment and build sustainability, other players (called here griefers) are becoming a threat. Just like it was in history This way, I don't think murders must be tweaked. It is just another challenge!

Christoffer · 2018-06-06 19:43:44

Is it possible to remove all baby deaths from the graphs? Would be interesting to see if that changes anything.

Rebel · 2018-06-06 19:53:12

ayyyyyy a fellow data scientist.

YAHG · 2018-06-06 19:58:23

I find it satisfying to see that people have just been bitching.

thirdplanet · 2018-06-06 20:20:27

Rebel wrote:

ayyyyyy a fellow data scientist.

I would literally love to have access to a bunch of the raw data from OHOL, whip out Rcommander and study how things like certain updates and player load affect life span and type of death...............

@jason hmu, grad school doesn't start up till fall and I have a lot of pent up data-crunching energy to spend............

TrustyWay · 2018-06-06 20:27:05

So it goes down instead of growing like commies in asian countries, gr8

Last edited by TrustyWay (2018-06-06 20:34:01)

thirdplanet · 2018-06-06 20:28:47

Hstrike wrote:

The average murder rate over this period is: 3.93%
As you can see, despite the recent hike, we are still experiencing a declining pattern of murder.
https://i.imgur.com/kbCLGJT.png

Rsquare value is too small to be considered significant over such a short period of time tho! So its not fair to say definitively that murder rate is decreasing... behavior of murder rate is inconclusive given the current data.

Last edited by thirdplanet (2018-06-06 20:29:47)

Hstrike · 2018-06-06 20:29:45

jasonrohrer wrote:

Thank you for making that graph!

I'm quite a visual person, so crunching numbers into graphs is nice, generally.
Keep sharing the data!

If you chat with thirdplanet, do not hesitate to reach out as well.

ALSO if anyone wants to keep making graphs, you can access the data on Google Sheet format here, which has some calculations in there too. I find the violent death rate at 19.62%, though.

Last edited by Hstrike (2018-06-06 20:30:12)

thirdplanet · 2018-06-06 20:47:31

Data in table form on Google sheet, including info about day of week (weekends highlighted) and info about content updates!

thirdplanet · 2018-06-06 21:06:05

I used the data for all, not just over-20. It looks like Hstrike used the first data set. I find no change in murder rate over time in the second dataset. Unless someone can see some glaring error I missed??

Last edited by thirdplanet (2018-06-06 21:08:37)

ZneoC · 2018-06-07 04:03:49

So ok, less then expected.

The the problem isn't "how many people died from murder" now, but actually "how many villages died from each murder". Which is a lot harder to collect data for that.

If only 2 % of lineages make it from village to city thenn even though it is only 1 (to kill the only female) or 5(ish give or take a mix of males/females) murders ton completely end something that had serious potential.

Questions:

1. What is the rate of murder per day(per hr, s, m)
As well as birth rate.

2. At what rate per hour is a healthy gen 3 or village created(i.e. has fire, surrounded by a surplus of food, starting metals tools or even farther depending on who you ask).

3. How many babies never get fed AND die?
3.A. maybe possible to see graphs that represent the #times fed before death.(to help determine how many successful lives there were)

4.How many home markers were placed?

5. Interesting to ask: how long after a knife gets made, does someone die?

6.what is the average murder rate per son? (When someone does kill in their life time how many die?)

7. Out of curiousity : what age to killers die at? 60? 45?

8. Can people in forums who have played many hours answer this question:?"when you start out in a village(not even), out of all your lives during a play session does an entire village (or all females) die from a killing spree directly, or indirectly resulting from murder , OR starvation caused by murder(I.E. player with a murder weapon hordes villages only food source)

If anyone would like to answer those, feel free. I am on an older mobile and it is difficult to research.

The reason I am asking this is to try and find that butterfly effect that causes villages to fail, babies to suicide, and people to grief.

I think these answers will help the community by seeing accurate data, of which, will help us find the solutions to our problems and help design a better game all together.

Sorry for long read, tried to keep it to the point.

One Hour One Life Forums

#1 2018-06-06 16:32:26

Whoops! Actually, muder rate is betwen 2 and 6%, not 16%

#2 2018-06-06 16:40:30

Re: Whoops! Actually, muder rate is betwen 2 and 6%, not 16%

#3 2018-06-06 17:05:24

Re: Whoops! Actually, muder rate is betwen 2 and 6%, not 16%

#4 2018-06-06 17:38:26

Re: Whoops! Actually, muder rate is betwen 2 and 6%, not 16%

#5 2018-06-06 18:04:11

Re: Whoops! Actually, muder rate is betwen 2 and 6%, not 16%

#6 2018-06-06 18:50:18

Re: Whoops! Actually, muder rate is betwen 2 and 6%, not 16%

#7 2018-06-06 19:00:14

Re: Whoops! Actually, muder rate is betwen 2 and 6%, not 16%

#8 2018-06-06 19:15:41

Re: Whoops! Actually, muder rate is betwen 2 and 6%, not 16%

#9 2018-06-06 19:26:03

Re: Whoops! Actually, muder rate is betwen 2 and 6%, not 16%

#10 2018-06-06 19:40:05

Re: Whoops! Actually, muder rate is betwen 2 and 6%, not 16%

#11 2018-06-06 19:43:44

Re: Whoops! Actually, muder rate is betwen 2 and 6%, not 16%

#12 2018-06-06 19:53:12

Re: Whoops! Actually, muder rate is betwen 2 and 6%, not 16%

#13 2018-06-06 19:58:23

Re: Whoops! Actually, muder rate is betwen 2 and 6%, not 16%

#14 2018-06-06 20:20:27

Re: Whoops! Actually, muder rate is betwen 2 and 6%, not 16%

#15 2018-06-06 20:27:05

Re: Whoops! Actually, muder rate is betwen 2 and 6%, not 16%

#16 2018-06-06 20:28:47

Re: Whoops! Actually, muder rate is betwen 2 and 6%, not 16%

#17 2018-06-06 20:29:45

Re: Whoops! Actually, muder rate is betwen 2 and 6%, not 16%

#18 2018-06-06 20:47:31

Re: Whoops! Actually, muder rate is betwen 2 and 6%, not 16%

#19 2018-06-06 21:06:05

Re: Whoops! Actually, muder rate is betwen 2 and 6%, not 16%

#20 2018-06-07 04:03:49

Re: Whoops! Actually, muder rate is betwen 2 and 6%, not 16%

Board footer