Aces High Bulletin Board

General Forums => Aces High General Discussion => Topic started by: lazs1 on December 31, 2001, 08:23:00 AM

Title: I know I've mentioned this before...
Post by: lazs1 on December 31, 2001, 08:23:00 AM
but.. I would like to see an "average" stat or scorecard...  With all the new guys it would be especially useful and... I think it would be helpful for a lot of the guys who have been around awile too.

Average K/D K/S K/T hit percentage, kills per sortie etc.   New guys could have something to "shoot" for.  Give them an idea where they are doing good and where they are doing poorly.
lazs
Title: I know I've mentioned this before...
Post by: pimpjoe on December 31, 2001, 09:51:00 AM
wouldnt that be the .score command?
Title: I know I've mentioned this before...
Post by: eskimo2 on December 31, 2001, 09:54:00 AM
Good idea.
I bet the average K/D is about 0.6 to 0.75
If someone got credit every time someone else died, the average K/D would be 1.0
But crack-ups on take off, landing and middle of no-where augers take a pretty heavy toll, not to mention auto-acks.

eskimo
Title: I know I've mentioned this before...
Post by: Broes on December 31, 2001, 10:05:00 AM
www.hitechcreations.com (http://www.hitechcreations.com)  ->Community -> Score -> Either Stats and Score

TADA!

Broes
Title: I know I've mentioned this before...
Post by: Dead Man Flying on December 31, 2001, 10:31:00 AM
Broes, he's referring to an arena average for various stats.  What you just described only provides individual information.

I'd also like to know the scores up to three standard deviations from the mean.  This way, players could compare themselves not only to the average, but also to various percentiles above or below it.

-- Todd/Leviathn
Title: I know I've mentioned this before...
Post by: AKDejaVu on December 31, 2001, 11:55:00 AM
AKDejaVu
Title: I know I've mentioned this before...
Post by: AKDejaVu on December 31, 2001, 12:07:00 PM
I'd rather see individual category rankings/lists.  That way each individual could set a goal as to how high up in the rankings for each individual category he wants to get.  Averages really dont tell you anything <edit: except for gunnery %>.

AKDejaVu

[ 12-31-2001: Message edited by: AKDejaVu ]
Title: I know I've mentioned this before...
Post by: eskimo2 on December 31, 2001, 12:32:00 PM
AKdejaVu,

As I was saying, the average K/D must be less than 1.

For a kill to be awarded, someone must die.
But each death does not result in someone being awarded a kill.  Augers and ack deaths create more deaths than kills, overall.

Some folks are almost always killed by other players.

But others are killed by other players, and they also die in ack, while taking off overloaded from CVs, shmuck up landings, ditch/captured in enemy teritory, hit the ground 2nd after a HO, etc and no one is credited with a kill on them.

Because of this, some folks may die 1/2 of their deaths without someone else being awarded a kill.

I personally often die without someone else getting credit, for many of the above mentioned reasons.

Therefore, a K/D of 1.0, is well above average.

eskimo
Title: I know I've mentioned this before...
Post by: AKDejaVu on December 31, 2001, 01:00:00 PM
Ah eskimo... you are right to an extent.. but then it becomes an issue of what technically incorrect set of data you use to deterimine the numbers.

Both are in error.  Many kills are not tallied either due to proximity kill or vehicle kills while on fighter missions for one... kills not tallied on the other in many instances where they should be.

But... a pilot is going to know that he simply crashes too much.  Really... the kills per times shot down are more relevant.  And.. as said above... that will always be 1.

AKDejaVu
Title: I know I've mentioned this before...
Post by: Vector on December 31, 2001, 01:05:00 PM
Quote
Originally posted by AKDejaVu:
I'd rather see individual category rankings/lists.  That way each individual could set a goal as to how high up in the rankings for each individual category he wants to get.  Averages really dont tell you anything <edit: except for gunnery %>.

AKDejaVu
[ 12-31-2001: Message edited by: AKDejaVu ]

S!
I'm looking forward to see your top 10 on each plane -stats coming up again. Was really good job. I've only one question: when?
 :)
Title: I know I've mentioned this before...
Post by: AKDejaVu on December 31, 2001, 01:17:00 PM
Quote
I've only one question: when?

That's not really up to me.  I'm patiently waiting.

AKDejaVu
Title: I know I've mentioned this before...
Post by: batdog on December 31, 2001, 01:24:00 PM
I like to use the STATS area to look and see how I'm doing in my favorite ride. It gives me a good idea of how I'm doing I think.

 xBAT
Title: I know I've mentioned this before...
Post by: eskimo2 on December 31, 2001, 01:25:00 PM
OK,
I was curious and just had to look.
In Tour 22, in all 4 categories (fighter, bomber, attack and GV) I had :
Bails - 10
Captured - 34
Deaths - 271
Disco - 3 (= 1.5 deaths)
Starter Deaths - 4 (we all begin with 1 death in each category so K/D can't be infinite)

Total: 320.5 deaths
- 33 ship gunner deaths =


On the stats page however;
My total deaths in all is 249
but, 20 of these death were awarded to ships gunners.
(These 20 deaths do not impact anyone's fighter, bomber, attack or GV score or ranking and therefor should be subtracted from my total deaths by other players.)
249 - 20 = 229

So,
229 eskimo deaths awarded to other players (in fighter, bomber, attack or GV).
Divided By:
287.5 eskimo's score recorded a death.
Equals;
79.6 % of eskimo's deaths are awarded to other players.

Conclusion;
If everyone behaved like eskimo, the overall average K/D would be about 0.8

eskimo
Title: I know I've mentioned this before...
Post by: eskimo2 on December 31, 2001, 01:31:00 PM
Quote
Originally posted by AKDejaVu:
Ah eskimo... you are right to an extent.. but then it becomes an issue of what technically incorrect set of data you use to deterimine the numbers.

Both are in error.  Many kills are not tallied either due to proximity kill or vehicle kills while on fighter missions for one... kills not tallied on the other in many instances where they should be.

But... a pilot is going to know that he simply crashes too much.  Really... the kills per times shot down are more relevant.  And.. as said above... that will always be 1.

AKDejaVu

We do get credit (as far as K/D goes) for proximity kills, don't we?

GV kills in fighter mode are definitly not recorded K/D wise, good point.

eskimo
Title: I know I've mentioned this before...
Post by: Drex on December 31, 2001, 01:37:00 PM
If everyone behaved like you eskimo, we would all be drunk and homeless.

Drex
Title: I know I've mentioned this before...
Post by: Dead Man Flying on December 31, 2001, 01:40:00 PM
Quote
Originally posted by Drex:
If everyone behaved like you eskimo, we would all be drunk and homeless.
Drex

So you're saying that Frenchy acts like eskimo?

 :D

-- Todd/Leviathn
Title: I know I've mentioned this before...
Post by: AKDejaVu on December 31, 2001, 01:56:00 PM
OK.. quick background:

Score Page:Stats Page:The score page will almost always show you with fewer kills and more deaths than the stats page.  The exceptions are not even worth mentioning since they rarely happen.

AKDejaVu
Title: I know I've mentioned this before...
Post by: bowser on December 31, 2001, 02:55:00 PM
I was never very good in math.    :)
Can somebody explain why the average k/d would be 1?

Say you had two players, player A and player B.  They had 27 fights.  Player A kills player B 19 times, and player B kills player A 8 times.

Player A: 19/8=2.375
Player B: 8/19=.421

Avg. k/d=1.398

Where am I going wrong?

Also, Warbird players use something call the Avin ratio which I always thought was a good measuring stick:
 http://warbirds.jackedin.com/ (http://warbirds.jackedin.com/)

bowser

[ 12-31-2001: Message edited by: bowser ]
Title: I know I've mentioned this before...
Post by: eskimo2 on December 31, 2001, 05:01:00 PM
Quote
Originally posted by bowser:
I was never very good in math.     :)
Can somebody explain why the average k/d would be 1?

Say you had two players, player A and player B.  They had 27 fights.  Player A kills player B 19 times, and player B kills player A 8 times.

Player A: 19/8=2.375
Player B: 8/19=.421

Avg. k/d=1.398

Where am I going wrong?

Also, Warbird players use something call the Avin ratio which I always thought was a good measuring stick:
 http://warbirds.jackedin.com/ (http://warbirds.jackedin.com/)

bowser

[ 12-31-2001: Message edited by: bowser ]

bowser;
all that realy matters is that there are 27 kills, and 27 deaths, 1:1

AKDejaVu;
I am pretty sure that proximity kills count as a kill, and help your K/D.  You earn no score points however.

"If everyone behaved like you eskimo, we would all be drunk and homeless.
Drex "

And unemployed, lazy and willing to waste the best years of their lives playing astupid computer game.

 :)

eskimo
Title: I know I've mentioned this before...
Post by: bowser on December 31, 2001, 05:32:00 PM
"...bowser;all that realy matters is that there are 27 kills, and 27 deaths, 1:1...".

Why are you using total arena kills/total arena deaths to determine an average k/d ratio for players?  In my example, I have an arena with two players.  One has a k/d of 2.375, the other has a k/d of .421.  How do you get an avg. k/d of 1?

bowser

[ 12-31-2001: Message edited by: bowser ]
Title: I know I've mentioned this before...
Post by: AKDejaVu on December 31, 2001, 05:57:00 PM
basically... when you average someone that getts 1 kill to 2 deaths with someone that gets 2 kills to 1 death, you should come up with a 1:1 average.  The ratios are averaged.. not the calculated %.

You add each side of the ratio and divide by the total.

If you don't, you run into the problem of those that can go infinately above one skewing the totals too far to one side.  Those below one can only go a total of 1 below it.

AKDejaVu
Title: I know I've mentioned this before...
Post by: bowser on January 01, 2002, 11:10:00 AM
I'll take one more crack at this, then drop it.   :)

I think the misunderstanding is in the terminology.  You seem to think the original poster was looking for a ratio of kills to deaths in the arena, which will of course be approximately 1:1.  In your calculations there is no averaging done, only a ratio of kills to losses.

I think what he was looking for was an average of "individual" player's k/d, not the arena as a whole.  So he could compare his k/d to an average player's k/d.

If you accept that each player has a k/d figure, you can then average those figures.

average individual k/d = pilot1 k/d + pilot2 k/d + pilot3 k/d.../number of pilots

bowser
Title: I know I've mentioned this before...
Post by: AKDejaVu on January 01, 2002, 12:08:00 PM
Bowser, you are not getting an average K/D when you do it that way.  You cannot average fractions with whole numbers and get valid results unless those numbers are measured.

When it comes to RATIOS (such as K/D), you must average as a ratio or the numbers are WRONG <unless all numbers are below 1>.  Its not a matter of how you do it, its a matter of doing it the correct way.

AKDejaVu
Title: I know I've mentioned this before...
Post by: Zigrat on January 01, 2002, 12:08:00 PM
i dont know why you gyus are so into scoring. pilot abilities are ranked in my mind not by the numbers the other guy earns but whether he can make me go "oh toejam" when we merge.

numbers are too easily manipulated.
Title: I know I've mentioned this before...
Post by: AKDejaVu on January 01, 2002, 12:12:00 PM
Zigrat, nobody is saying this should be a way to measure someone's ability.  Please don't turn it into that.

AKDejaVu
Title: I know I've mentioned this before...
Post by: eskimo2 on January 01, 2002, 02:34:00 PM
Zigrat;
It's a sickness.
I have always loved statistics and charts.
The fact that the AH score and stats pages come without explanation makes them all the more interesting to decipher.

eskimo