Author Topic: Lusches AH Stats Megathread  (Read 88172 times)

Offline McShark

  • Nickel Member
  • ***
  • Posts: 363
Re: Lusches AH Stats Megathread
« Reply #330 on: July 29, 2013, 03:22:25 AM »
Hi Lusche,


first of all thanks for this thread and the amount of work in it regarding all available data,  :salute

Regarding WW I and WW II ( as being the first wars with aerial combat ), several sites state that about 95 % of all aerial victories were scored by only 5 % of fighter pilots.
Of course, in AH we get new planes whenever we like ( and most important, we do not die...  ;) ) which was quite different in RL but I do wonder if there is any similarity regarding the number of victories scored by fighter pilots in AH?
Is the 'rule' as it has been seen in the great wars in any way applicable to AH?

Thx in advance
« Last Edit: July 29, 2013, 04:31:14 AM by McShark »
The problem with the world is that the intelligent people are full of doubts, while the stupid ones are full of confidence. -Charles Bukowski
Gleams the blade,Shines my Honor
Tour 19 - 163 McShark
Tour 163 -      Barkhorn

Offline Lusche

  • Radioactive Member
  • *******
  • Posts: 23828
      • Last.FM Profile
Re: Lusches AH Stats Megathread
« Reply #331 on: July 29, 2013, 06:56:13 PM »
first of all thanks for this thread and the amount of work in it regarding all available data,

Thank you :)

Regarding WW I and WW II ( as being the first wars with aerial combat ), several sites state that about 95 % of all aerial victories were scored by only 5 % of fighter pilots.

Yes I've read something like that too, with differing numbers. But no actual statistic on this (which doesn't mean there ain't one out there)

Of course, in AH we get new planes whenever we like ( and most important, we do not die...  ;) ) which was quite different in RL but I do wonder if there is any similarity regarding the number of victories scored by fighter pilots in AH?
Is the 'rule' as it has been seen in the great wars in any way applicable to AH?

The fact that we don't really die or get captured, but simply up a new plane makes this a bit difficult to answer. Yes, there is a minority of players getting a disproportional high share of kills. But this is only to some extend the result of talent/skill, but to a very high degree a matter of time they are putting into AH (which in turn does also have an effect om their skill level again.

So how to measure that? Let me present a few approaches to that:
(Note: All numbers presented are based on the active pilots of tour 160, which are defined by having at least one kill or death in that tour)

Pure volume of kills?
Adding together kills in Fighter as well as Attacker mode, the top 5% of the active pilots had ~35% of the kills. But as explained above, most of them had also simply been the most intensive players with the highest amount of hours

K/D in fighter mode?
The median pilot had a K/D of 0.5 (meaning 50% had a better and 50% a worse K/D). Only bout 25% of all active pilots usually has a K/D of 1 or better. To get into the top 5% a K/D of about 3.3 would have been necessary.

K-D?
Now this is an interesting approach which I only  started to consider relatively late. It's a kind of 'impact' of a pilot on the arena, reflecting both K/D as well as effort (time/sorties) put in. Bad pilots can get a high kill total by simply flying 200 hours/tour, and someone just being lucky on his single sortie can get a 'awesome' K/D. Simply subtracting deaths from kills eliminates sheer luck as well as pure suck, IMHO.

A quick examination of tour 160:



The top 5% have at least about 100 kills more than deaths in that two modes




You see, nothing is really directly answering your question, at least not in comparison to the real world. But maybe it gave another insight nevertheless.  :old:
Steam: DrKalv
Elite:Dangerous: Snailman

Offline McShark

  • Nickel Member
  • ***
  • Posts: 363
Re: Lusches AH Stats Megathread
« Reply #332 on: July 30, 2013, 05:27:28 PM »
Lusche,


thanks!  :aok

I will dig a bit in where I read about the high victory numbers of top aces. Will be back on that one.

I expected it to be tough one regarding the " not dying and coming back factor"

Your K - D approach is interesting. I would like to add:

Regarding the factor "time invested in AH ", some fly a lot, some less. Flying a lot gives you a higher chance collecting victories ( or get shot down, depending on skill ). An " average " pilot spends an " average" time in fighter / attack mode. Is it possible to extract the K-D in a given amount of time online?

Keep it up  :rock

McShark

The problem with the world is that the intelligent people are full of doubts, while the stupid ones are full of confidence. -Charles Bukowski
Gleams the blade,Shines my Honor
Tour 19 - 163 McShark
Tour 163 -      Barkhorn

Offline Lusche

  • Radioactive Member
  • *******
  • Posts: 23828
      • Last.FM Profile
Re: Lusches AH Stats Megathread
« Reply #333 on: July 30, 2013, 05:43:32 PM »
Is it possible to extract the K-D in a given amount of time online?

So that you end up with something like (K-D) per hour? Yes, that's technically possible, though my first thought is that it would not be much more significant than simply K per hour, which is already shown on our scores. You can't get very high k/h if you are getting shot down often

And of course K/H (as well as a theoretical (K-D)/H) is very much depending on playing style and even at what time of the day you are playing at. Off-peakers do often find themself chasing single cons while a prime time palyer can more often hurl himself into a furball. High alt bomber hunters (like me) will have a traditionally lower K/H than those never exceeding 10k...
Steam: DrKalv
Elite:Dangerous: Snailman

Offline Randy1

  • Platinum Member
  • ******
  • Posts: 4215
Re: Lusches AH Stats Megathread
« Reply #334 on: July 31, 2013, 07:13:53 AM »
This K/D curve reminds me of some ranking oddities that can occur when flight hours are minimal. A squad mate flew a bomber mission for the first time this tour in a B17.  He hates bombers.  Another squad mate talked him through the run down to even tell him how to open the doors.  He landed nearly 8000 damage points bombing a clean town and was ranked 185 in bombers.  ;)  I doubt if we could get back in a bomber again.

Offline McShark

  • Nickel Member
  • ***
  • Posts: 363
Re: Lusches AH Stats Megathread
« Reply #335 on: July 31, 2013, 09:36:29 PM »
Hi Lusche,


I do believe the circumstances do not allow us to prove the theory due to the difference in logged hours.  :confused:

I think your K-D as well your graph regarding player impact is as close as we can get.  :aok

Still did not find that article regarding the great wars... Will come back on that. Also looking about for a statistic backing this up. Must be somewhere  :headscratch:

Thanks again,  :aok :aok :aok

cya up!
The problem with the world is that the intelligent people are full of doubts, while the stupid ones are full of confidence. -Charles Bukowski
Gleams the blade,Shines my Honor
Tour 19 - 163 McShark
Tour 163 -      Barkhorn

Offline Lusche

  • Radioactive Member
  • *******
  • Posts: 23828
      • Last.FM Profile
Re: Lusches AH Stats Megathread
« Reply #336 on: August 03, 2013, 07:04:29 AM »
Fighter usage by year of introduction


This quote
Quote
LWA IS aces high and I do not want "late war" to become "last day of WWII" arena.
from the recent F4U-4 thread inspired me to take a look at the current fighter usage distribution again. This time not based on ENY or nationality, but strictly on real world dates of operational introduction.
I took the service dates from Mr.Fork's sticky in the Aircraft & Vehicles forum.




True late war fighters (1944 & 1945) are currently holding 64% of all LW arena kills+deaths.

« Last Edit: August 03, 2013, 07:06:23 AM by Lusche »
Steam: DrKalv
Elite:Dangerous: Snailman

Offline Tank-Ace

  • Persona Non Grata
  • Platinum Member
  • ******
  • Posts: 5298
Re: Lusches AH Stats Megathread
« Reply #337 on: August 03, 2013, 12:19:53 PM »
Lusche, I'm curious. Would it be an inordinate amount of work for you to track K/D over time, such that a graph of your data would show the learning curve?

Perhaps take a random sample of players with at least 2 years experience, and plot their K/D over a 24 month period?
You started this thread and it was obviously about your want and desire in spite of your use of 'we' and Google.

"Once more unto the breach"

Offline Lusche

  • Radioactive Member
  • *******
  • Posts: 23828
      • Last.FM Profile
Re: Lusches AH Stats Megathread
« Reply #338 on: August 03, 2013, 12:30:43 PM »
Perhaps take a random sample of players with at least 2 years experience, and plot their K/D over a 24 month period?


The problem is: I can't track players. Only ID's. When I see a new ID in a tour, I have no idea if it's a returning vet, player that just had his ID changed or actually a new one. Also many players change their ID at least once early in their career.

Steam: DrKalv
Elite:Dangerous: Snailman

Offline Tank-Ace

  • Persona Non Grata
  • Platinum Member
  • ******
  • Posts: 5298
Re: Lusches AH Stats Megathread
« Reply #339 on: August 03, 2013, 12:47:26 PM »
Would asking for volunteers on the forums work? It would require that you trust them to be honest, but the group here is pretty good in general.
You started this thread and it was obviously about your want and desire in spite of your use of 'we' and Google.

"Once more unto the breach"

Offline Lusche

  • Radioactive Member
  • *******
  • Posts: 23828
      • Last.FM Profile
Re: Lusches AH Stats Megathread
« Reply #340 on: August 03, 2013, 12:52:11 PM »
I'm not sure what we would expect to come out of this. We would surely see the K/Ds climbing over time, but the sample size would be fairly low and most probably not really representative of the ordnary & average AH player, who is for the most part not active here on the BBS.
So that would be more anecdotal than really sound data.
On top of that, k/d is just that and for the most part not really that telling all by itself alone.
Steam: DrKalv
Elite:Dangerous: Snailman

Offline caldera

  • Platinum Member
  • ******
  • Posts: 6437
Re: Lusches AH Stats Megathread
« Reply #341 on: August 03, 2013, 12:58:54 PM »
Lusche, I'm curious. Would it be an inordinate amount of work for you to track K/D over time, such that a graph of your data would show the learning curve?

Perhaps take a random sample of players with at least 2 years experience, and plot their K/D over a 24 month period?

That won't end up being an accurate measurement.  If a player improves over time but changes his style from timid to reckless, he might end up looking like his skills are regressing.  Plane choices also would play into this.  A player with 2 years experience flying a Spit might well have a worse K/D two years later, when flying a P-47.
"Then out spake brave Horatius, the Captain of the gate:
 To every man upon this earth, death cometh soon or late.
 And how can man die better, than facing fearful odds.
 For the ashes of his fathers and the temples of his Gods."

Offline Lusche

  • Radioactive Member
  • *******
  • Posts: 23828
      • Last.FM Profile
Re: Lusches AH Stats Megathread
« Reply #342 on: September 03, 2013, 01:10:46 PM »
The dust has settled - the latest toys revisited

As the new planes & tanks are around for some time now, let's see how they are doing now that thew initial craze has subsided.
Has the Yak-3 really become the terror of the late war arena, as some have predicted?




My old usage/KD chart comes with a slight change this time: Usage is the sum of all kills& deaths, while K/D (x-axis) is purely based on air to air combat results. IMHO this allows for a slightly better comparison between pure fighters and jabos
As we can see, no matter how we judge the capabilities of the Yak-3, it's level of 'usage' and the results (in terms of a2a K/D) aren't very spectacular at all. (In terms of fighter 'usage', it's ranked #15)

The Yak-7b ended up as an "also-ran" plane. Not really that bad in K/D success, but quite rarely used at all (#57 out of 71). But who would have expected something different?


Now for the current tank affairs:


Same deal here, total usage (K+D) vs pure tank combat K/D

The Jagdpanzer 38 (t) has quickly become relatively popular and drained quite a bit of  'usage' from the Panzer F, complementing it as one of the main tank perk farmers. As in the past, the performance of the Tiger I is still unimpressive compared to the Panther VG (In direct combat, the Panther had almost a 2 K/D against the Tiger I)


Steam: DrKalv
Elite:Dangerous: Snailman

Offline Vinkman

  • Gold Member
  • *****
  • Posts: 2884
Re: Lusches AH Stats Megathread
« Reply #343 on: September 03, 2013, 02:09:01 PM »
It surprises me to see the F6F so high on the usage axis. I just don't come across that many of them in the LWMA. Also ou have one plane labelled as 12548 which is the Y-axis value. What is the plane?  :salute
Who is John Galt?

Offline Lusche

  • Radioactive Member
  • *******
  • Posts: 23828
      • Last.FM Profile
Re: Lusches AH Stats Megathread
« Reply #344 on: September 03, 2013, 02:16:08 PM »
It surprises me to see the F6F so high on the usage axis. I just don't come across that many of them in the LWMA. Also ou have one plane labelled as 12548 which is the Y-axis value. What is the plane?  :salute

With the Cv cruising just off their target, carrier planes will get many sorties (and much combat) in short time, while not being used much in pure land based combat. That's why the usage of bomb trucks like the F4U-1D and the F6F is quite high.

And 12548 was an oversight, it's the F4U-1D
Steam: DrKalv
Elite:Dangerous: Snailman