Thread: Statistics Series: Revivable (KDR/AKDR)

Now...I'm expecting to be corrected, and please do if I've missed something, but this utterly confused me -along with others posting it seems- and I don't think it's the best way to get the result you want.

Originally Posted by Zoraster

Please take your regular KDR and divide it from your result from last week (That was the alternative kill to death ratio or (WK+VK)/(WD+VD).

For a start, I think you meant KDR divided by AKDR, not from, since that's what you did afterwards. Although, luckily, no-one else caught that and they all followed your numbers, otherwise this would've been even more confusing...


So, this statistic is:


KDR
______

AKDR


What does that actually mean? - Calculations


Well, Zoraster explained KDR vs AKDR nicely before so here it is:

Originally Posted by Zoraster

Regular KDR is really an apples to oranges comparison. It's not measuring the same thing. It's measuring the number of kills you make (regardless of whether they get revived) to the number of times you die and don't get revived.

If this were a perfect world, we'd simply compare number of times you kill someone and they don't get revived (thus losing their team a ticket) to the number of times you die and lose a ticket. This is what matters. However, I don't know how to do that with the basic stats provided.

This AKDR, on the other hand, does provide an apples to apples comparison. It's simply the number of times you kill someone (regardless of revives) versus the number of times they kill you (regardless of whether you get revived)

Therefore, KDR translates to your Total Kills, whether revived or not, divided by your Unrevived Deaths.

AKDR translates to, as best as the stats allow for, your Total Kills, revived or not, divided by your Total Deaths, revived or not.

The alternative kills and deaths, ie, kills and deaths with guns + vehicles isn't quite equal to your total kills and deaths:

Originally Posted by Zoraster

Q: What's left out?
From what I can figure out, you're going to miss kills and deaths from orbital strikes, supply boxes dropping on your head, and deaths from teammates.

Even though they arn't quite the same: they are near enough to it. So ignoring that, it seems strange that total kills should be in the calculation twice.

Using algebra for a moment, make:

Total Kills = A
Unrevived Deaths = B
Total Deaths = C

KDR = A divided by B
AKDR = A divided by C

Therefore, KDR divided by AKDR is:

A / B
_____

A / C

A Is both above and beneath the divide by line, but can't be simplified as it is. To change it, let's invert both fractions involved (and correct it later) so it becomes:

B / A
_____

C / A

The A's now cancel out, so we are left with B / C
But then of course, it was inverted, so inverting the fraction back it becomes C / B


What does that actually mean? - Result


Substituting back into this gives:

Total Deaths
___________

Unrevived Deaths

This is why, turning the statistic into english, it becomes:

Originally Posted by flik2142

ratio of 1 => you are never revived
ratio > 1 => you are revived
higher ratios => you get loads of revives

Also:

Originally Posted by Zoraster

So it at least purports to show how often you're revived in comparison to how many kills you get. There may be a better way of getting at that number, but this was the first thing that occurred to me.

Unfortunately, it doesn't involve kills in any significant way.

My stats would be:

KDR = 1.221 (current)
AKDR = 1.10 (last week)

KDR / AKDR = 1.11

Alternate Deaths is as close to Total Deaths as the stats allow, so, near enough AD = TD

Total Deaths = 40,351 (current)
Unrevived Deaths = 37,444 (current)

TD / UD = 1.07


Summary


In any case...if you want total times you are revived, I think it'd be simpler to do Total Deaths minus Unrevived Deaths. Of course, everyone will have played for different lengths of time so you'd want to do Revived Deaths divided by Total Deaths.

As it is, this calculation seems both upside down and to be using unrevived rather than revived deaths. But as far as comparisons go...everyone's doing the same thing and >1 is better. I still like these new stats Zoraster, it just boiled down to bit of confusing wishy washy numbers to me.