Author Topic: Top 10 stats being posted as I type (Read 1451 times)

AKDejaVu · « **Reply #45 on:** May 04, 2001, 12:59:00 PM »

Blauk, if someone with a 17:0 record wanted to get higher in the K/D.. they should have ventured one more sortie and risked getting killed.

The only difference in WW2 between a pilot that had 17 kills and lived and a pilot that 18 kills then died is the fact that the latter had one more kill than the other (statistically). Nobody can say wether the 17 kill pilot would have made 18.

BTW.. by your formula.. someone with a 20:1 record would be below someone with 11:0.

AKDejaVu

Replicant · « **Reply #46 on:** May 04, 2001, 01:30:00 PM »

Hola!

Well, I honestly thought that I had a crap tour last tour and didn't expect to make any top Ten, yet I made the Typhoon, Yak-9T, B26 and Pnzr

Well, not that it means anything but I'm pleased that my K/D is pretty good compared to the other peeps!

Salute AKDejaVu!

Regards

Nexx

BlauK · « **Reply #47 on:** May 05, 2001, 06:44:00 AM »

Originally posted by AKDejaVu:
The only difference in WW2 between a pilot that had 17 kills and lived and a pilot that 18 kills then died is the fact that the latter had one more kill than the other (statistically). Nobody can say wether the 17 kill pilot would have made 18.

Yes, in WW2 like in real life they had and we have only one life. However in AH we all have as many lives as we need.

In AH 18 kills and 1 death does not mean that all 18 kills were achieved before the death. Much more likely 9 kills, death and another 9 kills. Thus the guy is a 9 kill guy. It is not about predicting if someone would have got more kills.

That 18 kill guy is ahead in the other chart where you count the kills only. Is it not the idea of the K/D (or K/L) chart to make a comparison to one's own losses. If someone can perform sowell that he is not killed at all, he deserves to be noted. It is like comparing the living WW2 veterans and the ones who died in action. If one of each both had 10 kills, one lived, abother died... which one did a better job?

BTW.. by your formula.. someone with a 20:1 record would be below someone with 11:0.

Yes.. like mentioned above, 11:0 has better performance and he is still alive. 20:1 has had 2 lives and scored average 10 kills in each.

Just my opinion.

AKDejaVu · « **Reply #48 on:** May 05, 2001, 11:08:00 AM »

You talk only about average and not about risk. The person with less kills in an aircraft has risked less statistically. That should not be used to an advantage.

The +1 in the divisor is a common feature for hands-off programming or large number calculations. Its not a common feature for figuring out calculations below 100 (where most of these stats occur).

So... no point in going on any further... its not going to change.

AKDejaVu

BlauK · « **Reply #49 on:** May 07, 2001, 03:55:00 AM »

Originally posted by AKDejaVu:
You talk only about average and not about risk. The person with less kills in an aircraft has risked less statistically. That should not be used to an advantage.

Uhh.. risk?? What has K/D to do with risk? Kills dont mean how often someone has been in a battle and risked his life. Number of sorties has some info on that, but still no info on how many fights have happened. D means how many times one has been shot down, not how much you have risked. I just dont agree that 0 times is same as 1 times. Just as well 1 times could be same as 2 times.

Someone may have risked his life in 10 sorties, he has been good to avoid getting killed though and just been a bad shot.. has scored only 2 kills. Another guy flies one sortie, gets 2 kills and dies... 2:0 and 2:1. No.. it does not tell anything about risk.

The +1 in the divisor is a common feature for hands-off programming or large number calculations. Its not a common feature for figuring out calculations below 100 (where most of these stats occur).

You lose me here... I am only saing that 0 deaths is not same as 1 death... and that 1 life is not same as 2 lives.

So... no point in going on any further... its not going to change.

If you say so... but if someone has not been killed at all in some plane, your stats are not correct for him.

Skybax · « **Reply #50 on:** May 07, 2001, 04:27:00 AM »

Not trying cause any rifts here, but Top 10 caught my eye.

Check out Deft`s WB Scorepages

The different areas might give you some ideas.

---------------
328th Fighter Squadron
352nd Fighter Group.com
Blue-Nosed Bastards of Bodney

[This message has been edited by Skybax (edited 05-07-2001).]

BlauK · « **Reply #51 on:** May 07, 2001, 04:54:00 AM »

uhhh... ohhh.. truly pro stuff those WB stats

Did anyone notice that they use the +1 in their ratios

... ok I shut up

AKDejaVu · « **Reply #52 on:** May 07, 2001, 07:38:00 AM »

Quote

You lose me here... I am only saing that 0 deaths is not same as 1 death... and that 1 life is not same as 2 lives.

Actually.. you are the one making it different.

I'm saying that 6 kills and zero deaths will always rank higher than 6 kills and 1 death. Unfortunately, it is impossible to show a k/d fraction with zero in the denominator, so 1 is added for the sake of display. 6:0 is still displayed.

Where we vary greatly is under the premis that 6 kills and 1 death equates to a 3:1 k/d ratio.

Once again... 6:0 will win out... but if the pilot wants a better record than someone at 7:1 he should have RISKED more and flown another sortie. That's what the risk is blauk... the more its done.. the more the chance for an outlier. That's why there is a minimum on the number of kills required to get into the k/d list.

And as for deft's scoring pages.. they are completely automated... thus the need for the hand's off error avoider +1.

AKDejaVu

[This message has been edited by AKDejaVu (edited 05-07-2001).]

darling · « **Reply #53 on:** May 07, 2001, 08:11:00 AM »

heya

nice work DJV

One thing

Darling kills in P51D: 29
Darling deaths in P51D: 3

Is there a lower threshold to attain before being listed? Never mind me, just being silly

This gives one something to aim for after tour 16. Now just for trying to survive in the Pony

Regards
Dar

danish · « **Reply #54 on:** May 07, 2001, 08:47:00 AM »

Darling.
As it says; (#10 pilot's total kills)/2 <rounded up>

Unfortunatly the P51 is reasonable popular so #10 in the "Top Killers" list has 97 kills.Thus you would have to have at least 49 kills to make the list :=)
Thats a lot of kills if you only fly once in a while.But you might consider a change to the somewhat less popular B model (and of course slightly more difficult to get those elusive kills...).Looks like 15-20 kills might be enough.Might.

Good luck. ;=)

danish

BlauK · « **Reply #55 on:** May 07, 2001, 10:24:00 AM »

AKDejaVu,

One more argument...

You defend your current calculation with the risk aspect. You also say that pilots who have risked more and have got more kills would suffer with K/D+1. I have 2 examples for you which tell the opposite. In your current system the guys who have more kills suffer.... it is no dependent on kills, but on deaths.

Check 109G-2 K/D

wilbuz 12:2 = 6.0 (12:3 would be 4.0)
DB603 210:42= 5.0 (210:43 would be 4.88)

DB has risked more and killed more, still you rank him below wilbuz. Same concerns 1Wmaker1 and StSanta.

So it is not about kills like I said... it is about deaths. Those pilots who have only 1 death benefit most of K/D when comparing to K/D+1. When one has more deaths the difference becomes smaller... on the other hand more deaths also usually mean more kills, so the top killers suffer. But who suffers most are the guys with 0 kills since they are penalized as having died once.... executed for the sake of not having to divide by 0

This is confusing to me. Why do you want to use a formula which gives the 1 death guys the best deal, a bit worse deal for those who died killed and risked much more and the worst deal to guys who did not die at all?

You say you do it by hand... maybe that is why you dont want to change it. A huge job.. S! for that. If it was all in a database, all that was needed was to add +1 to some places.

BlauK · « **Reply #56 on:** May 07, 2001, 10:28:00 AM »

btw, my 109G-6 was 27:8 =3.38
Should I not be on the list too

Zigrat · « **Reply #57 on:** May 07, 2001, 10:29:00 AM »

blauk its his list let him do it how he likes

AKDejaVu · « **Reply #58 on:** May 07, 2001, 04:01:00 PM »

No blauk, my current system does not do what you claim.

The only real rule that needs special consideration is the zero death pilot. That's it.

K/D for people with deaths will all be recorded with 100% accuracy. The people with zero deaths will be limited by the total number of kills since there is no way to guess just how much farther they would have gone without a death. If they wanted to go farther, they would have risked more.

The minimum number of kills required to make the k/d list is the screen for the minimum risk required to make the list. The only time that doesn't work well is with lesser used aircraft where someone with less than 20 kills can make the list.

So.. I know why I am doing the way I am. I'm sorry that you seem to have such a difficult time with it. I really hope you get over it soon.

AKDejaVu

AKDejaVu · « **Reply #59 on:** May 07, 2001, 04:06:00 PM »

Quote

btw, my 109G-6 was 27:8 =3.38
Should I not be on the list too

Must have missed that on the list. Most of the K/D was done by hand after sorting... oops.

BTW.. can't believe that after all you've posted you are just now mentioning this.

AKDejaVu