[UPDATED] Weekly Stats, PvP Trends

[thorgold 91]

Pretty interesting. I assume that most players don't post on the forums, so we probably don't get a good sense of how rampant pvp is just based on the forum posts. I would imagine that most new players die to zombies a lot. I know I did.

I think this is the primary cause of all the bitching. People don't come onto the forum to say how they DIDN'T get shot, or killed by zeds, or starved, they come to bitch about campers in cherno.

It's a psychological thing - we remember the bad more vividly than the good. Your character may have 100 deaths, but the 10 that were murders are the only ones that stick out because they were so negative.

You're trying to see trend as a function of time. Yes, you have a lot of data points, but you only have two data points for your time (that's your domain). To see an accurate trend, you need way more than two "time data points" if you will.

The need for more than two points of time only applies if you want to analyze the trend over the entire period. What were the statistics for July 10th? 11th? You're asking for snapshots of those times in order to create a less linear chart, and that's understandable. Did banditry increase day one, but then have radically lower values that averaged out to the final data? We don't know. However, that doesn't invalidate the data we have. Instead of a curve, we get a line - a line that is correct in its endpoint (the percentages presented in his analysis), if technically inaccurate in the intervening time period.

We have an OVERALL trend, and for any linearization of data all you need are two points - the start and the end.

Edited July 14, 2012 by thorgold

[ZedsDeadBaby 2287]

dont forget people suiciding to get to a better spot.

Don't forget we already fucking covered that holy shit read the thread it's only one fucking page.

[Whattteva 32]

I hope you know it's not a good idea to get into a statistics debate here. Especially with Zed'sDeadBaby

I'm not trying to pick beef with anyone. I'm just trying to point out something that I think is a little off. Don't get me wrong, his data is really interesting and innovative. I'd definitely like to see where it's going. But, wouldn't you think it would be better for everyone if things constantly get improved? Questioning others is good. We'd still think the solar system is geocentric if people just never question anything. Science would get nowhere.

A collaboration of people's feedback always make things better, that's how open source and even this mod that we are testing works. Rocket may be the expert, but that doesn't mean our feedback won't make the mod better.

Edited July 14, 2012 by Whattteva

[outlawxxable 10]

hmmm,.... I don't believe this. I still think PvP is what causes most deaths

[Dingus (DayZ) 429]

I'm not trying to pick beef with anyone. I'm just trying to point out something that I think is a little off. Don't get me wrong, his data is really interesting and innovative. I'd definitely like to see where it's going. But, wouldn't you think it would be better for everyone if things constantly get improved?

I'm sure ZDB will keep and eye on it and provide more data later. At the moment it is a perfectly fine sample that allows him to make the presumptions he's making. He did not say "it will keep going down into the future." He is saying that for these two months, PvP has decreased, and his sample is more than sufficient to make that statement.

[kosh 5]

That's pretty interesting, please keep updating this :)

With the huge influx of noobs from the mod's awesomeness + steam sale, the data is going to be pretty skewed, especially compounded by the new zombie aggro. My friend and I just joined last weekend, and we have two more friends downloading Arma as we speak. Who will each probably bring in a couple more.. O_o EDIT: point being, each new player has to be good for at least half a dozen zombie deaths, especially with the new aggro.

Edited July 14, 2012 by Kosh401

[ZedsDeadBaby 2287]

I'm not trying to make long-term predictions; the trend isn't even that dramatic.

The reason it is significant is because it coincides with a constant din on the forums proclaiming the end of DayZ due to "constant deathmatching" and decrying the game for being "PvP only" and "not really about the zombies," etc., etc.

"Bandit skins removed? Now everyone shoots everyone all the time!"

"No more starting gun? Now bandits will camp the coastlines and murder people en masse!"

"Remove sniper rifles the game is just a snipe-fest!"

etc., etc.

So, yeah, I'll admit on its own a 7-day gradual downward trend isn't particularly scandalous, but when held in contrast to the constant calls of doom and gloom here on the forums I think it shows that reality doesn't quite match up with the protests.

I mean, I guess that shouldn't be surprising, but whatever. It is what it is.

point being, each new player has to be good for at least half a dozen zombie deaths, especially with the new aggro.

Well, that's kind of my point.

If the game were a "constant deathmatch" or "just like CoD" as people are wont to claim here, then a massive influx of clueless players would be fodder for the slaughter and murder rates would go through the roof.

But they're not, 'cause new players are generally stumbling around scared and alone and dying to zombies or environmental stuff, which is what these seem people seem to claim they want more of in the game...

Edited July 14, 2012 by ZedsDeadBaby

[Whattteva 32]

I'm sure ZDB will keep and eye on it and provide more data later. At the moment it is a perfectly fine sample that allows him to make the presumptions he's making. He did not say "it will keep going down into the future." He is saying that for these two months, PvP has decreased, and his sample is more than sufficient to make that statement.

My bad. I realized this after re-reading the post. Why does people get so highly defensive around here? Well, I suppose, it's understandable with the general attitude of people around here.

[Whattteva 32]

I'm not trying to make long-term predictions; the trend isn't even that dramatic.

The reason it is significant is because it coincides with a constant din on the forums proclaiming the end of DayZ due to "constant deathmatching" and decrying the game for being "PvP only" and "not really about the zombies," etc., etc.

"Bandit skins removed? Now everyone shoots everyone all the time!"

"No more starting gun? Now bandits will camp the coastlines and murder people en masse!"

"Remove sniper rifles the game is just a snipe-fest!"

etc., etc.

So, yeah, I'll admit on its own a 7-day gradual downward trend isn't particularly scandalous, but when held in contrast to the constant calls of doom and gloom here on the forums I think it shows that reality doesn't quite match up with the protests.

I mean, I guess that shouldn't be surprising, but whatever. It is what it is.

It's my bad, too. I didn't read your post carefully enough to understand that. Personally, I'd love to see how this plays out. Keep up the good work!

[Yeung-Jin 28]

I'm not trying to pick beef with anyone. I'm just trying to point out something that I think is a little off. Don't get me wrong, his data is really interesting and innovative. I'd definitely like to see where it's going. But, wouldn't you think it would be better for everyone if things constantly get improved? Questioning others is good. We'd still think the solar system is geocentric if people just never question anything. Science would get nowhere.

A collaboration of people's feedback always make things better, that's how open source and even this mod that we are testing works. Rocket may be the expert, but that doesn't mean our feedback won't make the mod better.

My goodness, I'm not used to being taken so seriously. I'm sorry it looked like I was taking a jab at you, I've just seen the carnage a thread has become because of a stat debate with ZDB before. He gets pretty angry at anyone and anything. Especially if you're wrong.

Edited July 14, 2012 by Yeung-Jin

[Dingus (DayZ) 429]

Why does people get so highly defensive around here? Well, I suppose, it's understandable with the general attitude of people around here.

It's not defensive, you were just wrong on an internet forum. You were corrected and now it's over. Don't worry about it.

Edited July 14, 2012 by Dingus

[Whattteva 32]

My goodness, I'm not used to being taken so seriously. I'm sorry it looked like I was taking a jab at you, I've just seen the carnage a thread has become because of a stat debate with ZDB before. He gets pretty angry at anyone and anything.

It's cool. We got our differences resolved in a civil and rational matter just a few posts before yours. Look, it IS possible for things to be resolved in a calm manner if people keep the discussion with cool heads and without name-calling right? I think the reason a lot of threads degenerated to the point that you describe is cause people start name-calling and spouting profanities and the likes... maybe.

[Whattteva 32]

It's not defensive, you were just wrong on an internet forum. You were corrected and now it's over. Don't worry about it.

I misunderstood the intention, but I don't think what I pointed out was wrong as I explained in the post right after it... but whatever... I guess that's why things tend to degenerate around here. No one likes to admit their own faults. I'll take the blame on this one, so things don't escalate any further.

Edited July 14, 2012 by Whattteva

[Dingus (DayZ) 429]

I misunderstood the intention, but I don't think what I pointed out was wrong as I explained in the post right after it... but whatever... I guess that's why things tend to degenerate around here. No one likes to admit their own faults. I'll take the blame on this one, so things don't escalate any further.

I don't mean to make you feel dejected in any way. I'm sorry things came off negatively.

[Yeung-Jin 28]

Group hug, Bitches!

[grizzle 21]

Don't forget we already fucking covered that holy shit read the thread it's only one fucking page.

Before you get all bent out of shape, you should realize that the post you are quoting reeks of inaccuracy.

"The stats script on the main page is filtering out deaths within a timeframe that I consider respawns indeed."

What does that mean? Can't they track respawns because someone hit the respawn button? Doesn't seem like it. I'm sorry, that explanation is just way too vague and it clearly is just a guess. You lose.

They should publish the metrics and calculations that make up these statistics, otherwise they're as believable as any other numbers I could pull out of my ass.

[Whattteva 32]

I'm really interested to see how these stats trend over the next few weeks and see if the same trend continues or maybe we could even get fluctuations. I would think fluctuations would be more interesting cause there may be other factors that we have not yet discussed in here yet. What are your expectations?

[NFK 102]

I'm only a few weeks into this game, and these stats made me think about my deaths. Most of them were early on, and due to zombies. I've never starved to death, but hunger/thirst/broken legs have driven me to take stupid risks that resulted in getting eaten by zombies, so environmental effects have been a factor.

I've been been killed by another player exactly two times, and never on spawn. I have murdered exactly one other player, mostly out of fear and circumstance, not from a plan to go find players to kill. My current strategy focuses on avoidance, as I currently play alone so no external comms/communities to rely on. I imagine a lot of solo players also do this.

[Dingus (DayZ) 429]

What I noticed last night in the two different servers I was in, different than normal, was the percentage of zombies killed out of those totally spawned. The new changes to zombie aggro mechanics clearly is forcing more players to engage them. I would imagine this would cause a downward trend in pvp. This is something that would be better represented in numbers for next week though.

At one point there was under 200 alive for 400 spawned on the server I was on. Usually barely any are dead.

[jonahcutter 51]

What these figures show most clearly is the mod is enjoying a huge influx of new players learning the mod while also having more aggressive zeds.

It also doesn't show the type of pvp that is happening, and its affect on the overall gameplay. Sniper pvp is far different from cqb pvp. I'd like to see as much pvp as possible, but of the close-quarters variety. Where everyone involved is also at risk of zed aggro. Long-range "pvp" (often simply getting sniped with little actual pvp involved) adds little (perhaps nothing) to the zombie apocalypse gameplay.

Remove long range rifles/scopes high-RoF weapons from the game and you still have an intense zombie apocalypse mod with pvp as a core construct. Perhaps even more intense because players must get closer to each other to engage in pvp. Remove the zeds and you've got a pretty bland sniper vs sniper mod.

When talking about death rates, breaking it down to simply zed vs pvp won't give you an accurate view of what is actually happening. It ignores other key factors. When talking about pvp in this mod, talking about the type of pvp is key. A death by getting sniped by someone who is hiding under a bush in the treeline is not the same type of gameplay as a death by a close quarters pistol/shotgun battle while all combatants are aggroing zeds.

Edited July 14, 2012 by jonahcutter

[orb917@hotmail.com 117]

*edit, realized i was wrong bout that quote being relevant

great thread zdb, good to see some number crunching. hope you keep doing this.

Edited July 14, 2012 by p1n34l

[Wulfear 6]

Dunno if it was mentioned, but people hitting respawn to 'start where they want' is probably skewing the shit out of non pvp deaths...

(seems it was, oh well)

Edited July 14, 2012 by Wulfear

[TheMachine 803]

Sorry Zed, no one cares about facts.

They personally, died - therefor there is a problem with PvP in the game.

No ammount of cold, hard facts are going to change their minds and stop the constant crying.

I think the more the game develops, the more PvP will become incidental to the game.

At the moment our group hunts players at the NWAF, there may come a day when a PvP engagement is incidental to our visit there, and not the focus of it.

I imagine that'll be when we can contruct our own 'bases' or something along those lines.

Until then, the river of tears will continue to flow.

[Ragdrazi 7]

Doesn't surprise me. I've died maybe 15 times in Day Z, only 3 were at the hands of other players. I tend to avoid the "OMG PVP" debates, because I honestly never saw it as much of a problem. It exists, but it doesn't seem as ubiquitous as some angry threads make it out as.

Good work ZDB.

Last I checked, that was just about the average.

[lev 39]

Edit: So I don't have to mention this 15 more times in the thread, deaths by respawn are not counted among these stats (source).

I have been taking snapshots of the stats table occasionally for the past week or so now, just to see if any simple trends emerged.

It has been nearly a full week, or will be tonight, so I thought it would be interesting to compare stats from 7 JUL to today and see what has changed besides the obvious massive influx of players +166,214 in one week or an average of about 24,000 new players per day.

I did some basic analysis on the trends surrounding PvP in the game, what with that being such a hot-button issue around here. The conclusions seem to indicate that despite an incredible influx of new players to the game, deaths by murder, bandit activity and PvP in general are on the decline, as a percentage of total. Without getting too editorial, this seems to contradict a prevailing notion that new players are being led to the slaughter at the hands of hordes of bloodthirsty sniper-bandits camping the coastlines, as well as the prevailing wisdom that new players are themselves becoming bandits in droves.

Here are the stats I derived. Sorry for the image but this board doesn't seem to like formatted text very much:

You did the statistical analysis incorrectly. The numbers you have gotten are a comparison between survival percentage ending in murder WITH a bunch of new players versus survival percentage ending in murder WITHOUT a bunch of new players. That is the most you can accurately say without starting to make a ton of assumptions.

What you need to do is collect more data or provide more data you have collected so you can do a rate analysis and see if murder rates are increasing or decreasing with respect to the increase of players. The reason you have to do this is because there are confounding factors that are hard to discriminate when presented with only 2 sets of data and changing population over time (some of which is not even counted for - e.g. players who have stopped playing). I could easily present the counter argument that your numbers demonstrate that new players die to other sources (bugs, falling, starvation, zombies, suicide, etc) than PvP. This is further backed up by the fact that new players know less about the game so are more likely to die to things that more experienced players can deal with. Another reason you need more data is simply that when presenting only two sets of numbers to represent the entire population you can easily misrepresent certain factors and numbers due to the fact that you only presented two snapshots of the system.

Another small tip: you should try to state your assumptions with regards to what the numbers in the data set represent. (e.g. do survival attempts include currently alive players? dead bandits? and other issues of that nature so that your statistical analysis can be verified)

Edited July 14, 2012 by Lev