Lusches AH Stats Megathread

[Lusche]

Lusche, I'm curious, could you plot the statistical usage of perk units as their costs decrease, or is the data simply not available to you?

You mean decrease during playe due to changes in the perk modifier? That I can not do. I only can track overall usage, but not really the perk modifiers ingame, which are different for all 3 countries as well.

On availability of data in general: I have no more acess to AH data than any other AH player. I only see the same stats as everybody else.

[Tank-Ace]

Yes. I was curious to see how the usage changed as the perk price decreased

Specifically, I was wondering what the curve looked like; I suspect its fairly smooth up until a point, and then sees a decent spike at that point. My thinking is that there seems to be a decent group of players with small perk reserves, who either are unable or unwilling to up, say, a 262 before it dips below that 100 perk mark.

This also supported in some measure by real world marketing tactics; $19.95 sounds better than $20.00, and $99.95 sounds better than $100.00.

[MK-84]

Yes. I was curious to see how the usage changed as the perk price decreased

Specifically, I was wondering what the curve looked like; I suspect its fairly smooth up until a point, and then sees a decent spike at that point. My thinking is that there seems to be a decent group of players with small perk reserves, who either are unable or unwilling to up, say, a 262 before it dips below that 100 perk mark.

This also supported in some measure by real world marketing tactics; $19.95 sounds better than $20.00, and $99.95 sounds better than $100.00.

     I'm one of those people.  In my quest to wreck as many of my favorite aircraft as possible (the 262) I will often switch sides for the perk modifier so that I can earn perks more quickly, or so that I can afford to fly one.

[bozon]

I cannot roll a 262 during a high perk multiplier because ENY lmiter kicks in at the same time. More then once I wanted to roll a 262 to fight other 262s regardless of the perk cost and just couldn't.

[Vinkman]

A few minutes ago I was asked by a player "Is it true that most perks are made by bombing GV instead of shooting down other planes in air combat"

A question that might be interesting for others as well, so I give the answer in public:

IMHO,no. Not even close.

Taking all LW arena kills for the 8 finished tours in this year so far, we can calculate the theoretical raw perk gains (at perk bonus 1.0, no landing modifiers and current ENY settings) for each plane and each kill type. (As you know, perks gained = (ENY killer / ENY killed) * perk bonus * landing modifier)
This way, we get the theo. perk gain for P-51s killing other P-51s, Tigers, B-29s and if we sum it all up, the perk gain for fighter killing planes, fighter killing GV, bomber killing planes... and so on.

Result:

(Image removed from quote.)

The actual perk gain is different from that, as it depends on the modifiers on time of the kill, but the relation between categories will not be changed radically. And of course, bombers will have a much higher perk gain by the results of their bombings, but I can't even estimate a number for that.

It has to be noted that not only the number of A2A kills is much higher than the number of A2G kills, but also the perk gain per kill is less for A2G kills, as GV have generally a higher ENY than planes.

Excellent. Does this mean you can also calculate the base perks by player? In previous discussions, the community had expressed what the effect of including perks gained as part of the scoring formula would do the ranks. Any chance you could calculate the top 50 player ranks for the past tour or two with Perks included? 

[Lusche]

Excellent. Does this mean you can also calculate the base perks by player?

Theoretically I could apply the same approach to a players kills and resulting theoretical perk gain.

There are three problems with it however.
First, at this moment it would take quite some effort on my side.
Second, I can't apply this to rank, as perks and scoring do not match up: Someone having 250 kills in the pony and maybe 100 perks theoretical gain from it... but how did that distribute between fighter and attack score and thus rank?
But second, and more importantly, when examining and judging a single players perk exploits, you simply can not ignore the modifiers anymore this player was subject to at the time. In my opinion it works for getting the 'bigger picture' on big questions like the one I answered a few posts ago, but is less meaningful on an individual level.

[Zacherof]

Lusche is it possible to collect data such as the average amount of time players are online?

[Lusche]

Lusche is it possible to collect data such as the average amount of time players are online?

Time online? No, not really. Game tracks the time in different modes, but tower map staring or manned gun time and such does not appear in the stats.  

[Zacherof]

Time online? No, not really. Game tracks the time in different modes, but tower map staring or manned gun time and such does not appear in the stats.  

what about accuracy of ordanance and rounds?

[Lusche]

what about accuracy of ordanance and rounds?

Look at the score pages... that's exactly the stuff I can track 

[Zacherof]

Look at the score pages... that's exactly the stuff I can track 

[Vinkman]

Theoretically I could apply the same approach to a players kills and resulting theoretical perk gain.

There are three problems with it however.
First, at this moment it would take quite some effort on my side.
Second, I can't apply this to rank, as perks and scoring do not match up: Someone having 250 kills in the pony and maybe 100 perks theoretical gain from it... but how did that distribute between fighter and attack score and thus rank?
But second, and more importantly, when examining and judging a single players perk exploits, you simply can not ignore the modifiers anymore this player was subject to at the time. In my opinion it works for getting the 'bigger picture' on big questions like the one I answered a few posts ago, but is less meaningful on an individual level.

there's no way to separate Attack from fighter? I didn't realize that. That would confound any answer, but what if it were just assumed for now, to be fighter? Are you curious how the perks earned might affect the rank? It has been suggested that it would make a big difference. 

[Lusche]

That would confound any answer, but what if it were just assumed for now, to be fighter? Are you curious how the perks earned might affect the rank? It has been suggested that it would make a big difference.  

Thats too many variables to be any meaningful. When someone has 100 kills in fighter mode and 400 kills in attack mode, I just can't take all theoretical perks from that 500 kills and see how they would change fighter ranking. And I still don't know the perk bonus & landing modifiers, which make a huge difference on a individual scale.

[Vinkman]

Thats too many variables to be any meaningful. When someone has 100 kills in fighter mode and 400 kills in attack mode, I just can't take all theoretical perks from that 500 kills and see how they would change fighter ranking. And I still don't know the perk bonus & landing modifiers, which make a huge difference on a individual scale.

The landing modifiers alone are huge. Yes I agreed. Too bad. 

[Lusche]

The Big Three - top perk ride kill shares

Lusche,

  I have a question - What was the percentage of 262 keels in 2003?

Regards,

While at it, I can as well take a look at the kill shares of the all top three perked fighters of AH over the years. Me 163, Me 262, Tempest - all three planes had always K/D's way above the competition, being the most dreaded killers in AH land, all time.

So here they are. Keep in mind that the intital kill shares of each plane are always low because the planes aren't usually introduced in the beginning of a year. The earliest year where they were all in from the start was 2003:




Interesting to see how the 262 and the Tempest swapped places after 2010.
The cumulative chart shows that the cumulative kill share of these three 'monsters' has been relatively stable for the past three years at ~3% of all fighter kills.