Tactical Air War

[Roo5ter]

Maybe LG team understand the things. Don't you think so?

All they have to do is make a response.  It's been about a week since this really picked up.  I think that is more than enough time to take a peek, look at it in detail and make some sort of response which has not happened.

 

Don't you think so?

[SDV_DecabrisT]

I think that you not only don't understand why it was a mistake for Blues to attack tanks (as reds does). You don't even understand the rules how statistics is calculated.

Edited February 27, 2017 by SDVr_DecabrisT

[Tuesday]

I think I'd like to give LW a try on this server. I'll probably have a hard time since I lose more IL-2s than I get back home, but I'd like to see things from the other side on this server.

 

To switch sides, do I just update my team to LW in my profile, or do I have to create a new profile?

 

So if someone doesn't care about stat logging, they can always fly the plane they want?

 

You just have to change your game name at il2sturmovik.com in order to play for the other side. 

[19//Moach]

I find it particularly strange that those same individuals who (and/or closely related thereto) have been seen in the password protected ALSIC channel (why so secret?) are also the only ones to become strangely aggressive at the disclosure of these unnatural data points, and also, this is the very same small group that holds such unusual scores, which the vast majority seems drastically unable to match, or even reduce the manyfold gap in score between them  -- I do not mean to point any fingers, but to point out observable facts - so I don't see the reason for this aggressive attitude being called for at all here 

 

 

that there's not been any response or sign of the LG members in any way for a while does not imply they do not care, let alone disagree that there is a problem - most probably means they have not been able to attend the server for the time being, and thus all there is to it is a matter of time

 

 

please to not attack players with unfounded claims towards what they do or not understand - this is particularly rude, and has a very unfriendly overtone when read the way it does - hopefully there is language barrier here, but it's the second time this has happened, and under the very same context as well

Edited February 27, 2017 by 19//Moach

[Roo5ter]

Well Moach, there is the first rebuttal.  I would alert Chappy but so far the rebuttal is 'you dont understand data' yet offers no details or substantiations.

[SDV_DecabrisT]

the manyfold gap in score between them

 

The answer is simple: attack and stay alive. That is important for statistics.

 

 

please to not attack players with unfounded claims towards what they do or not understand - this is particularly rude, and has a very unfriendly overtone when read the way it does - hopefully there is language barrier here, but it's the second time this has happened, and under the very same context as well

 

I see unjust and unjustified claims to the player. And attempts to fit data that he does not play fair, when he plays well.

Should I remain with friendly overtones?

Edited February 27, 2017 by SDVr_DecabrisT

[SDV_DecabrisT]

 

 

but so far the rebuttal is 'you dont understand data' yet offers no details or substantiations

 

Do I have to explain how statistics is calculated? Spare me please from this.

[DerNeueMensch]

[...]

 

anyone with a score lead from the second place which is larger than the sum of all the score differences across rest of the top five, is an extremely serious red flag - and would certainly fall under investigation on just about any serious sporting activity that you can think of

[...]

 

You said it yourself, data may be the cause of investigation, but is in no "serious sporting activity" sufficient for a ban. It's just data. That is why there is no need to "refute data".

Edited February 27, 2017 by DerNeueMensch

[19//Moach]

The answer is simple: attack and stay alive. That is important for statistics.

 

 

 

I see unjust and unjustified claims to the player. And attempts to fit data that he does not play fair, when he plays well.

Should I remain with friendly overtones?

 

 

whatever meaning this was supposed to have is more than not lost getting lost in translation... the little that comes through does not really add anything to the discussion, and seems quite out of place

 

no claims have been made towards any individuals - yet this seems to be an attempt to defend someone specific - albeit without producing any solid arguments to such defense, when attempting to deny largely self evident questionable results

 

 

and now this,

 

Do I have to explain how statistics is calculated? Spare me please from this.

 

 

 

again, completely out of place verbal attack towards Roo5ter, implying he doesn't "understand" with a condescending tone -- this does not help one bit, the overly defensive-aggressive stance is one which hints at "having something to lose" - if there is a genuine explanation for the absurd results the numbers are showing, it needs to be brought forth in a clear form, without resorting to arguments that seem based solely on attacking others with senseless assumptions of their understandings

 

we all understand very well how statistics are calculated

 

 

 

 

again, the data shown is presented only as grounds for admins to investigate - no specific accusations have been made

Edited February 27, 2017 by 19//Moach

[Leutnant_Artur]

We're reading your posts. I don't know what to tell you to be honest. Some of you saying that some guys are cheating but what evidence do you have, stats ? Is it all, stats ? If so then I'm a cheater too 'cos I've made streak of 42 kills in last campaign and finished in second place in overall.

 

Why is LW loosing ? Answer is : there's no teamplay and some of you said it already in this topic. If you don't believe me just go in the evening to TAW teamspeak and see how many guys are sitting on allies channels and compare numbers to axis.

 

Maybe Khaton will have something more to say.

[SDV_DecabrisT]

 

 

for the absurd results

 

What is absurd? Streak of 60? That is not absurd, is good enough and achieved under certain conditions.

[Roo5ter]

 

What is absurd? Streak of 60? That is not absurd, is good enough and achieved under certain conditions.

 

You are not up to date on everything clearly.  You need to go back to around page 83 or so and look what Chappy posted as well.  A number of anomalies are worrying whereas 1 is not particularly interesting.  Moach actually laid it out in one of his recent posts in that same manner - aka the note for the bombers.

 

Many anticheat programs especially in FPS look for statistical anomalies.  We don't have that in IL2 and that might be pretty hard to have in IL2 but they actually ban players when there are statistical anomalies to whatever degree.  Fairfight is a great example of this.  You can address 1 of them by saying a streak of 60 is possible and I wouldn't say you are wrong in the least bit.  That is just one part of a much larger puzzle.  I encourage you to pick apart everything Moach has said.  It would be much more reasonable than trying to latch onto one thing and then get snarky while ignoring the rest of the data.

We're reading your posts. I don't know what to tell you to be honest. Some of you saying that some guys are cheating but what evidence do you have, stats ? Is it all, stats ? If so then I'm a cheater too 'cos I've made streak of 42 kills in last campaign and finished in second place in overall.

 

Why is LW loosing ? Answer is : there's no teamplay and some of you said it already in this topic. If you don't believe me just go in the evening to TAW teamspeak and see how many guys are sitting on allies channels and compare numbers to axis.

 

Maybe Khaton will have something more to say.

The Luftwaffe losing is something for sure, but this is one part of it.  As I noted, the Luftwaffe would lose anyways but that doesn't address any of the other things that have been brought up.  If the Luftwaffe was going to lose anyways it doesn't justify someone on Red doing things outside of fairplay.  At least to me it doesn't.  I don't mind losing fair and square but if the LG stance is that blue would lose anyways then I think the problem is being ignored to the detriment of the campaign.

Edited February 27, 2017 by Roo5ter

[Monostripezebra]

Hi, I´m one of the blotted out names.. blotting out names with a reference to the place where they are written in clear and that everyone here knows is a really moot point. just sayin.

 

and I also made my science degree a long time ago. So let´s have a look at where you are all wrong:

 

1st: statistical correlation is not a proof for causality. THis standard textbook phrase of any statistics guide should be considered more... If someone needs a grip on that have a laugh at the nice book "spurious correlations".

2nd: the assumption of a bell curve like distrubtion without specific parameters of the graph function is not data based. At all. Especially when comparing aples and oranges: Player success could have a considerably different distribution picture if you add variabls like technical equipement (18 frames small monitor without headtracking and high end systems... etc..and time played/experience or game odds: if red is consistantly outnumbered, a red player would have far more statistical chances to meet an enemy, and make more decisions on taking or leaving the fight. So netto, he has absolutely numerical more chances for favourable conditions.. as long as there is still some element of choice (even though running is definately easier in blue planes, ingame). Technically, that makes any non-data based assumptions on what a statistical normal ratio between the players might be highly absurd. Some people spend an exponentially greater amount of time and money on the game then others, who is to say that this wouldn´t result in a 3x higher kill margin? All I can say as a social scientist is that the probability of these guys haveing a normal job and social life is smaller then average...  ;=)

 

Infact that would also be a real world example: Why where there some extremely high scoring aces during the war? because they had to serve the entire war and had no difficulties encountering the enemy. cue video link to the standard vid posted on flightsimforums to this topic here... Oh, and yes: Hartmann was cheating! ;=)

 

3rd) it is really cherry picking to look at absolute numbers in similar orders of magnitude and say "for bombers it´s ok, for fighters not". Without even looking at the time spent per "success".

 

 

I´ve been around flightsims for long and I have seen a lot of looong killstreaks. It is definately possible. I see you like gaming and theorie: so here is a nice field: if you want to see what difference setup (in regards to network tech) can make, have a chat with the "Open Arena" (former Quake, now open source) comunities: it is one of the oldest games and has spawned even new game modes out of bugs (additive movement speeds). It will teach you a lot on how games are not liniear in skill progression...

 

let us analise the current way of things at TAW right now - according to our Bell Curve Theorem, shall we?
 

 

 I've blotted out individual player names to avoid breaking forum rules 

 

 

 

 

now - the conclusions to be drawn here are somewhat self evident - particularly in the case of the two top fighter pilots: note that "the subject" mentioned before occupies the first place, and has a threefold lead over the second

 

the runners-up from third to fifth place then serve as a "control group"  - demonstrating what the natural distribution of results should be like:  take notice, particularly of how the difference in scores actually decreases towards the top - as it is natural to do in any score based competition

 

 

anyone with a score lead from the second place which is larger than the sum of all the score differences across rest of the top five, is an extremely serious red flag - and would certainly fall under investigation on just about any serious sporting activity that you can think of

 

it is simply not natural to see this, and more often than not, this is alone reason to hold the competitor up as to provide his own proof of innocence (e.g. with a urine sample for a doping test, in physical sports, or with a thorough inspection of a vehicle, in motor racing) - remember, a competitive event is not a court of law, thus the "burden of proof" paradigm does not necessarily apply

 

 

TAW admins have been contacted about this already, with individual names forwarded in private, to prevent needless embarrassment and/or entertaining the offender's very possible desire for negative public attention (known in psychology as "narcissistic supply")

 

in order to help, it is imperative to observe forum rules, and refrain from public accusation at all times - you may freely do so in private messages to the author of this thread, who is in charge of the TAW server and is the ONLY person to whom the specific identity of the "subject" should be of any use.

 

so do your part - and send your message in private - keep the boards clean, and then we can avoid problems such as the thread being locked, and all such inconveniences that will surely arise should we take the "torches and pitchforks" alternative instead

 

 

further reading: http://www.pcgamer.com/hacks-an-investigation-into-aimbot-dealers-wallhack-users-and-the-million-dollar-business-of-video-game-cheating/ -- it is irresponsible to deny cheating as a possibility in any gaming activity

Edited February 27, 2017 by Monostripezebra

[StG77_Kondor]

Why is LW loosing ? Answer is : there's no teamplay and some of you said it already in this topic. If you don't believe me just go in the evening to TAW teamspeak and see how many guys are sitting on allies channels and compare numbers to axis.

 

Is that what the TS looks like in Euro high time? Because American evening time the English channel constantly had five or more pilots (with even more when Map #1 kicked off - lately it's died because, well - see for yourself). In addition, many of the Blue pilots use private squad TS's. Which sometimes I've jumped into and joined them. Sometimes we maintain comms just through game chat. So you're telling me, the Red side that has more Russian speaking players is VASTLY better at communication with non-Russian speaking Euros and Americans? 

 

Lack of coordination doesn't account for the abundance of first shot PKs. Lack of coordination doesn't account for Map #2 flipping in ~48 hours. Lack of coordination is such a small part of the much larger problem. 

 

But I guess we'll see if/when these guys switch to Blue to 'prove a point'. When in the end, they'll just end up proving mine. 

[Roo5ter]

Jesus Mono.  No one is saying that everything is impossible simply improbable.  No one said this was absolute proof, you are taking it out of context to write a long, pointless rebuttal.  

Coming into contact with more enemies is pretty important for making a higher score, also not being outnumbered is incredibly important for keeping a high score.  You address the pro's and not the cons to justify your empty point.  


Ok so let's go into logic here.  There are several statistical anomalies and the individuals in question refuse to address them or simply have a reasonable chat about it.  

Let's talk this out a bit since obviously we are headed nowhere fast.  Many AAA games do ban people on statistics alone, regardless of what Monostripezebra thinks is proof or not.  Moach has stated multiple times that you cannot prove someone is cheating in this game, this is pretty much a fact, yet Monostripezebra and one SDV guy completely miss this over and over.  

Moving on. 

Let's describe the situation here.
Someone has a entirely possible but improbable kill streak.
The group of players with that person also have pretty awesome statistics.  Not entirely impossible at all and like skilled players often band together, especially those with the same primary language in an international game.
Rate of kills for mainly air and eh.. somewhat for ground targets is concerning. 

Multiple anomalies without an explanation.  So what comes next?  Everyone freezes apparently.  No, that is probably not the best course of action.  The best course of action would be to have a productive conversation with those individuals, ideally arranged in advance so they can properly prepare to have a short discussion if there is a mild language barrier or go over some things that look odd and think about what they are doing that maybe other people aren't that leads to such success.  

 

Have this private, confidential discussion over the matter where those people either can justify the numbers or can't.  They obviously do not want to publicly discuss success.  The public response has been, "Oh you think I am cheating, oh yeah funny stuff!  Haha, I so much better!  *popcorn*  Time to troll you instead of talk to each other like human beings!'  

Request live streams, not recordings from the individuals.  As low as quality as possible since performance can have a factor and watch these players for a bit.  If there is an ability for server admin to spectate live that would also be an less desirable but entirely reasonable alternative that doesn't require additional resource constraints from the accused parties PC or connection.

I dunno, I'm just a person shotgunning rough ideas but it seems like there is a lot that can be done, it is simply a matter of caring.  Surely there are other actions and discussions that could take place.
There is also the route of saying, "We can't prove it and are not interested in sinking the time into doing a investigation of the matter regardless of the information put forth.  Statistics are just that and they might be good enough for triple AAA games but they are not something we are interested in since they are not empirically proof in our eyes."  That is an entirely viable path and I think the community would appreciate just having that direction from the TAW team so we can go ahead and 'take it or leave it.'

Edit:  Just wanted to point out the idea that cheating is happening is not a fringe one https://forum.il2sturmovik.com/topic/27046-cheating-battle-stalingradmoscow/
Pand is reputable and respected in the community (though we must not let him know) and he thought it was concerning enough to take to the forums and create a poll which shows overwhelmingly 88% think it is entirely possible in multiplayer IL2.

Edited February 27, 2017 by Roo5ter

[SC_Manu653]

I know this is a 5 on 7 to win the campaign. But can we make a tug of war... finish all the 7 map. It's sad to not see the later planes.

I know, I know, I know you just have to win the map and bla bla bla . But in a war you fight until the end of the war not of the fight. 

[Roo5ter]

I know this is a 5 on 7 to win the campaign. But can we make a tug of war... finish all the 7 map. It's sad to not see the later planes.

I know, I know, I know you just have to win the map and bla bla bla . But in a war you fight until the end of the war not of the fight. 

How do you envision that working?  Having 13 maps where you take the current maps and they are     7 - 6 - 5 - 4 - 3 - 2 - (Starting map) 1 - 2 - 3 - 4 - 5 - 6 - 7 and you have to push to 7 to win?

 

I think its a solid idea though I do enjoy the off season of the campaign as well (so i can take care of other things and focus more on the campaign when it is live) and that one would take a long time!

[19//Moach]

Jesus Mono.  No one is saying that everything is impossible simply improbable.  No one said this was absolute proof, you are taking it out of context to write a long, pointless rebuttal.  

 

Coming into contact with more enemies is pretty important for making a higher score, also not being outnumbered is incredibly important for keeping a high score.  You address the pro's and not the cons to justify your empty point.  

 

 

Ok so let's go into logic here.  There are several statistical anomalies and the individuals in question refuse to address them or simply have a reasonable chat about it.  

 

Let's talk this out a bit since obviously we are headed nowhere fast.  Many AAA games do ban people on statistics alone, regardless of what Monostripezebra thinks is proof or not.  Moach has stated multiple times that you cannot prove someone is cheating in this game, this is pretty much a fact, yet Monostripezebra and one SDV guy completely miss this over and over.  

 

Moving on. 

 

Let's describe the situation here.

Someone has a entirely possible but improbable kill streak.

The group of players with that person also have pretty awesome statistics.  Not entirely impossible at all and like skilled players often band together, especially those with the same primary language in an international game.

Rate of kills for mainly air and eh.. somewhat for ground targets is concerning. 

 

Multiple anomalies without an explanation.  So what comes next?  Everyone freezes apparently.  No, that is probably not the best course of action.  The best course of action would be to have a productive conversation with those individuals, ideally arranged in advance so they can properly prepare to have a short discussion if there is a mild language barrier or go over some things that look odd and think about what they are doing that maybe other people aren't that leads to such success.  

 

Have this private, confidential discussion over the matter where those people either can justify the numbers or can't.  They obviously do not want to publicly discuss success.  The public response has been, "Oh you think I am cheating, oh yeah funny stuff!  Haha, I so much better!  *popcorn*  Time to troll you instead of talk to each other like human beings!'  

 

Request live streams, not recordings from the individuals.  As low as quality as possible since performance can have a factor and watch these players for a bit.  If there is an ability for server admin to spectate live that would also be an less desirable but entirely reasonable alternative that doesn't require additional resource constraints from the accused parties PC or connection.

 

I dunno, I'm just a person shotgunning rough ideas but it seems like there is a lot that can be done, it is simply a matter of caring.  Surely there are other actions and discussions that could take place.

There is also the route of saying, "We can't prove it and are not interested in sinking the time into doing a investigation of the matter regardless of the information put forth.  Statistics are just that and they might be good enough for triple AAA games but they are not something we are interested in since they are not empirically proof in our eyes."  That is an entirely viable path and I think the community would appreciate just having that direction from the TAW team so we can go ahead and 'take it or leave it.'

 

Edit:  Just wanted to point out the idea that cheating is happening is not a fringe one https://forum.il2sturmovik.com/topic/27046-cheating-battle-stalingradmoscow/

Pand is reputable and respected in the community (though we must not let him know) and he thought it was concerning enough to take to the forums and create a poll which shows overwhelmingly 88% think it is entirely possible in multiplayer IL2.

 

all of these are very sane points - and requesting to watch a live stream from the persons which are being suspected is actually a very sound idea, which I heartily endorse - if they refuse, then it can be said they have something to hide, as it would be at least a courtesy to comply with any effort to keep the game fair and enjoyable to all

 

again, the main rationale for how we can draw the need for official attention - not "accusations", as some have incorrectly inferred - from statistics is outlined here: https://forum.il2sturmovik.com/topic/21029-tactical-air-war/page-80?do=findComment&comment=444416

 

anyhow, while streaks are nevertheless fickle, and not very substantial as statistics data, it is worth remembering that the difference between points across first and second places is what is being called out - not the amount of points themselves - there have been longer streaks, for sure, but it is odd that a fairly small and secretive group of players seem to be consistently scoring ahead of their runners-up by a very disparate amount, which is not seen across the rest of the ranks (which escalate fairly gradually, each with a small difference in score to the players listed above and below)

 

the scores themselves are not directly relevant to this analysis - the patterns seen in how they distribute across the playerbase is - and where this pattern is abruptly broken, and a large gap in scores appears - then that warrants a call for attention from those in charge.  

 

also - this has nothing to do with why the blue team keeps getting defeated -- scarce teamwork and improper use of their asset's strong points are far more meaningful issues to this end -- I never once said that the statistical abnormalities were to blame for that

 

 

 

I have also outlined a number of suggestions for gameplay adjustments which I have thoroughly considered, and believe that if introduced would make for a largely more enjoyable, less frustrating and much easier kept in balance next campaign

[Roo5ter]

Not sure how many in the community have seen this movie but it definitely shows some hard to explain things can easily happen in this game.  Obviously the development team has addressed vulnerabilities in the past and this video is 9 months old.  I will let it speak for itself.  The best of the clip is past 4 minutes.  

Weird things happen, draw your own conclusion.  Stuff like this is why a small sampling of the community (poll got blocked) turns up 89% who believe that cheating is a -possibility- https://forum.il2sturmovik.com/topic/27046-cheating-battle-stalingradmoscow/ Edited February 27, 2017 by Roo5ter

[Semir]

Not sure how many in the community have seen this movie but it definitely shows some hard to explain things can easily happen in this game.  Obviously the development team has addressed vulnerabilities in the past and this video is 9 months old.  I will let it speak for itself.  The best of the clip is past 4 minutes.  

 

---

 

Weird things happen, draw your own conclusion.  Stuff like this is why a small sampling of the community (poll got blocked) turns up 89% who believe that cheating is a -possibility- https://forum.il2sturmovik.com/topic/27046-cheating-battle-stalingradmoscow/

 

 

I recall someone claimed that certain people stopped playing for several weeks after the patch meant to "fix" such exploits has landed.

Edited February 27, 2017 by Semir

[coconut]

I've done the math, I can only invite amateur statisticians to do the same, in case I messed something up.

If you look at the kills per hour, the person with the 60 killstreak is among the 0.1% best pilots. I don't find that so unlikely that we would have such a person playing on TAW. There is actually another pilot that's in the top 0.01%

[Roo5ter]

I've done the math, I can only invite amateur statisticians to do the same, in case I messed something up.

If you look at the kills per hour, the person with the 60 killstreak is among the 0.1% best pilots. I don't find that so unlikely that we would have such a person playing on TAW. There is actually another pilot that's in the top 0.01%

Cool but if you are going to invite others to collab and check your work, you will have to show your work or at least detail the process as others have done.

 

Edit:  Actually you already confused me.  I see 1515 total records for pilots in TAW per statistics on the site.  1 / 1515 = .0006 or .06%.  That means the very first pilot on that leaderboard is THE top .06%.

 

How is a pilot in the top .01%?  We would need 6x as many pilots on the list than we have now or 1515 x 6 = 9090 just to have a .01%

Edited February 27, 2017 by Roo5ter

[FMF*Mannock]

Hello gents!

 

Today I have a strange problem... I'm being disconnected from the server without any warning message or error message into the multiplayer menu :-/ it's annoying... is there anyone who have same problem?

[Semir]

Cool but if you are going to invite others to collab and check your work, you will have to show your work or at least detail the process as others have done.

 

Edit:  Actually you already confused me.  I see 1515 total records for pilots in TAW per statistics on the site.  1 / 1515 = .0006 or .06%.  That means the very first pilot on that leaderboard is THE top .06%.

 

How is a pilot in the top .01%?  We would need 6x as many pilots on the list than we have now or 1515 x 6 = 9090 just to have a .01%

 

 

Bell curve?

[Roo5ter]

Bell curve?

How would a bell curve influence that number?  There are only enough people to create a .06 % so you could pretend that the same person is the .01% because well they are but they are also the next 5 people as well then.  I don't understand how a bell curve would create something out of nothing.

[LLv24_Veccu_VR]

Hi all...

 

Axis is losing but with a good teamwork we can win this shit Just hitting those tanks before they hit ours...

 

We will win this, someday....

 

 

And he landed...

http://taw.stg2.de/pilot_sortie.php?id=21081&name=Ktif

[Semir]

How would a bell curve influence that number?  There are only enough people to create a .06 % so you could pretend that the same person is the .01% because well they are but they are also the next 5 people as well then.  I don't understand how a bell curve would create something out of nothing.

 

https://www.easycalculation.com/statistics/bell-curve-calculator.php

 

Calculate the mean score of players. Then the standard deviation. Then input the value in question and choose "right tailed". That will give you the probability of such or better score existing given the rest of the distribution. Assuming normal distribution of scores is of course a leap of faith.

Edited February 27, 2017 by Semir

[Roo5ter]

https://www.easycalculation.com/statistics/bell-curve-calculator.php

 

Calculate the mean score of players. Then the standard deviation. Then input the value in question and choose "right tailed". That will give you the probability of such or better score existing given the rest of the distribution. Assuming normal distribution of scores is of course a leap of faith.

I made a promise to quit poking at the issue publicly so I am going to have to steer away from this for now.  I think the originator of that can make his response how he got his data and respond to me if he so chooses but I will not be giving another reply on that topic.

 

Hi all...

 

Axis is losing but with a good teamwork we can win this shit Just hitting those tanks before they hit ours...

 

We will win this, someday....

 

 

And he landed...

http://taw.stg2.de/pilot_sortie.php?id=21081&name=Ktif

We don't need to focus on tanks, we must focus on airfields.  We MUST bomb the piss out of their airfields then just JU52 paradrop to take over during the next rotation.  We have an 80% capture rate for every airfield that we paradrop on if it is 100% damaged.

 

I believe the campaign is a race between the two sides.  The Russians are racing to get 400 tank kills and the Luftwaffe is racing to take airfields before that 400 tank mark.  The single most significant problem we are facing as Luftwaffe is a lack of teamwork.  We have some great moments such as 5 He111's in formation last night to bomb an airfield with some amount of fighter cover.  We do not have that regularly enough and we do not have the escorts to get the paratroopers to the target.

[Willy__]

Hi all...

 

Axis is losing but with a good teamwork we can win this shit Just hitting those tanks before they hit ours...

 

We will win this, someday....

 

 

And he landed...

http://taw.stg2.de/pilot_sortie.php?id=21081&name=Ktif

 

LMAO What a joke. 

 

Yeah, teamwork is the only culprit as to why the LW always loses... yeah right. Keep living on your barbie world reds. 

Edited February 27, 2017 by JAGER_Staiger

[Roo5ter]

Here are 4 shorts that tell the story of  Kondor, Haza, two others I can't remember off the top of my head, and myself doing a Ju52 run with a single F2 escort on Wednesday night.  This resulted in an airfield capture.

Our ingress to the target area with 5 Ju52 and 1 F2

Our 60 man drop

On the return run home we were surprised by an unmarked enemy AAA position.  Immediate action was a dive and apparently that position was on our nose.  My 52 was damaged and another was destroyed.  The F2 pilot struck the Maxim as we altered course

Bringing my wounded bird in with 2nd engine barely spinning.  The crosswinds and significant damage made the landing a bit rough as I could not get it to drop but all was well and we captured an airfield for a total loss of 1 Ju52.  Considering the Ju52 is a CM+1 bird, I call that a win.

[19//Moach]

the specifics of the curve and exact probabilities aren't really important - just use common sense, there's a smooth, gradual increase in player scores as you read the list from top to bottom - and this tends to get closer and closer near to the top, where competition is more intense -- this would be the expected disposition of any ranking of scores 

 

a red flag pops up when one, or a small group which is at the very top, appears to be ahead by a disproportionate margin, which represents a gap in the pattern - this would be an indicator that there is more than pure skill involved, as human prowess (in any sense) tends to be asymptotic rather than infinite - a gap in an otherwise homogenous pattern is thus a good place to ask questions, since there is likely some form of advantage (aka a force multiplier) which those above the divide are enjoying, which may or not be accessible and/or known to those below

 

whether or not the factor causing the gap is within or without the competition rules is not knowable by the numbers alone - but it is less than wise to leave it all but ignored, given the possibility does exist that it could be the case

 

 

still - that's not why the blues lose - that's another thing altogether

 

 

BTW: Kathon has PM'd me in reply to say he's had no time to tend to the server or the boards lately - so he will review this thread at a later time once he gets the chance

Edited February 27, 2017 by 19//Moach

[SDV_DecabrisT]

Many anticheat programs especially in FPS look for statistical anomalies.

 

You are looking for a black cat in a dark room. There are no statistical anomalies. It is not correct to compare streaks at least until all players do not work to reach the best streak. There are not many. (and even after the fulfillment of this condition it is not correct, because streak is quite random variable, even a novice can achieve good streak).

It would by fairer (not best but better) to compare K/D ratio, and more fairer to compare K/H ratio (where H is hour). The player, that you means, is not even first in the list. But this is not fit to your "picture of the world".

 

 

 

Multiple anomalies without an explanation.

There are no (in my opinion).

 

 

 

If you look at the kills per hour, the person with the 60 killstreak is among the 0.1% best pilots.

 

Great. I'm sorry - I'm not so good in english. This is even not the first my foreign language. I read too slow. I've read this only just now.

Edited February 28, 2017 by SDVr_DecabrisT

[=IL2AU=chappyj]

I'm sorry my friend, but i can't stay calm.

 

A couple of questions to your picture:

1. Do you understand what streak is?

2. Do you understans the relationship between streak and score?

 

3. What analise can you introduce, if you do not understand this all? And i see - you do not. 

 

 

 

 

Maybe LG team understand the things. Don't you think so?

 

 

 

You are looking for a black cat in a dark room. There are no statistical anomalies. It is not correct to compare streaks at least until all players do not work to reach the best streak. There are not many. (and even after the fulfillment of this condition it is not correct, because streak is quite random variable, even a novice can achieve good streak). It would by fairer (not best but better) to compare K/D ratio, and more fairer to compare K/H ratio (where H is hour). The player, that you means, is not even first in the list. But this is not fit to your "picture of the world".

 

 

hi decabrisT

You refer to using KD or KH ratios. and then you suggest that the subject is not very high or out of the ordinary based on these stats. 

The problem is these stats are conflated and inaccurate when they are mixed with a2g and a2a missions. But i'm sure you already know this. 

 

So, a few pages back I looked at kills per hour using a basic sample size of the first page of mission sorties for the 1st and 3rd top A2A killers. (at the time of writing the 3rd top killer has since slid to 5th and now the top several killers are all linked by association. ie: teamspeak)

What it showed was that for missions that were more than 5 minutes long, were using fighter type aircraft and had no A2G kills. the KPH stat of the top A2A killer was THREE TIMES the KPH of the 3rd top A2A killer (2nd top was ignored due to association with the 1st top A2A killer (teamspeak))

Edited February 28, 2017 by =WFPK=chappyj

[SDV_DecabrisT]

Oh peoples

 

Here is sample of 100 players (not all in the picture... good 39 players. That's enough). Where do you see the THREE TIMES???

I can built the bell curve on the data and see nothing extraordinary.

Do it yourself, and do it correct please.

 

https://yadi.sk/i/u1jqOW4D3EksY2

[=IL2AU=chappyj]

lol oh well decabris i thought because of your condescending tone towards moach that perhaps you were actually familiar with statistics. 

 

So i guess i was wrong. 

 

it is important that when you are looking at trends that you compare like for like. KPH and KD do NOT compare like for like because they lump a2a in with a2g missions. in other words. time spent in bombers dilutes the a2a statistics. You therefore have to FILTER your stats to look at A2A sorties. 

 

the filter is the following;  fighter type aircraft. no a2g kills. mission length > 5 minutes. 

 

 

so. i have done this again based on the current sample size of the front page sorties of the top A2A killers for 1st, 2nd and 5th for today. (3rd sample size too small, only 9 a2a sorties but for what its worth it shows the same thing. ) . 5th A2A killer was chosen because he was used in my previous sample. 

additional context. the top 4 are all loosely associated by teamspeak. (share a channel often)

 

so about the subjects. 

1st A2A killer completed 22 air missions on the front page of his stats for a total of 639 minutes. with 31 kills. 

2nd top completed 16 sorties for 593 minutes with 26kills. 

4th  16 sorties 461minutes 17 kills.

5th 24 sorties  completed  639 minutes for 10 kills. 

 

the first and  4th and 5th top killers were the only subjects to spend the majority of their time in A2A role during the sample period. but looking at the broader KPH as decabris would wish us to do, we can see that the subject 1 has diluted their A2A global stats throughout the campaign with quite a bit of time in bombers as well. whereas subject five has almost the same KPH globally as he does in my sample (good because it confirms my sample method)

 

the first and 5th killers were the subjects of my previous sample from a few days ago. 

 

stats:

 

1st top killer : KPH 2.92

2nd top killer KPH: 2.65

4th top killer 2.23KPH

5th top killer  KPH 0.94

 

in case that needs spelling out to anyone. thats now OVER three times the KPH between 1st and 5th. and still well  over DOUBLE between 4th and 5th. if anything the disparity has gotten worse since i did my previous sample. 

Edited February 28, 2017 by =WFPK=chappyj

[SDV_DecabrisT]

I can built the bell curve on the data and see nothing extraordinary.

Almost ))))

 

https://yadi.sk/i/rT39KEyT3EkxEb

[=IL2AU=chappyj]

and your AKH data is based on global stats still ?

A2g data in bombers is tallied to the flight time and dilutes the A2A stats. you can't even begin to compare them globally

there are so many fundamental flaws in your method its not even funny.

 

you have people in your sample with 12 A2A kills total for the campaign.

 

even if you go find the 35th and 76th ranked players. and insert them into the manual data (which arguably you can because at least those 2 players have only ever flown fighters so their global stats are ok although their sample size is quite low. I discounted the third top killer because I could only count 9 a2a sorties )

they STILL slot in a full 1KPH lower than the subjects mentioned 1st -4th in the top A2A killers list ie; you STILL have a significant cluster separated from the curve.

Edited February 28, 2017 by =WFPK=chappyj

[SDV_DecabrisT]

1. I've not said that my way the best is. I've said it is just better.

2. I don't have much time to do detailed analysis.

3. "A2g data in bombers is tallied to the flight time and dilutes the A2A stats"  - Yes, but in a first approximation it can be ignored and it does not give too much error. I can justify why.

 

...4. During the analysis I have some questions to another player. And he's not red one. )))

but I still do not accuse anyone. Statistics is statistics. Everything is possible.

Edited February 28, 2017 by SDVr_DecabrisT

[=IL2AU=chappyj]

1. I've not said that my way the best is. I've said it is just better.

2. I don't have much time to do detailed analysis.

3. "A2g data in bombers is tallied to the flight time and dilutes the A2A stats"  - Yes, but in a first approximation it can be ignored and it does not give too much error. I can justify why.

 

...4. During the analysis I have some questions to another player. And he's not red one. )))

but I still do not accuse anyone. Statistics is statistics. Everything is possible.

 

 

1: im sorry but it is not better.

2: neither do i. and my method is not ideal either but i dont see a way to get full filtered stats short of manual grinding the visible sorties list which is only ever the most recent ones

3: NO it certainly cannot be ignored. the more A2G missions completed the greater the dilution. some of the subjects in question have SIGNIFICANT time in A2g and i have shown twice now how it skews overall KPH completely. 2.9 in pure A2A but 1.6 in global stats. massive difference that cannot be ignored. 

4: nfi. not relevant to my investigation. if you have concerns then raise them for investigation just like these concerns are being raised.  (nothing in my analysis has anything to do with red or blue or nation or location)

[SDV_DecabrisT]

 

stats:

 

1st top killer : KPH 2.92

2nd top killer KPH: 2.65

4th top killer 2.23KPH

5th top killer  KPH 0.94

I'm sure here is a mistake. And I guess where... Do you really think that top 5 of players on the main site have killed more then others?

I have to disappoint you. The score, and number of kills is not directly related.

Edited February 28, 2017 by SDVr_DecabrisT