Jump to content

How Is Kdr A Valid Stat?


37 replies to this topic

#1 SiorAlpin Wolf

    Member

  • PipPipPipPipPipPip
  • The Predator
  • The Predator
  • 255 posts
  • LocationU.K

Posted 23 May 2013 - 05:59 AM

I ask this question in the hope that it can be re-addressed and perhaps a more realistic stat be put in its place.

Kill to Death ratio to me doesn't seem to be a stat that can be taken seriously, what i mean is i can run around in a jenner with 1 small laser and get 8 kills and only do a max of 100 damage, so all the players that put in the hard work bringing the enemy mechs to the point of destruction, i come along in my little jenner shoot them once and i get the kill. So my KDR increases for the better for me and does nothing for the other players that actually used skill.
There are many scenarios where you can shoot a mech once and it dies through no real skill or talent (unless its a head shot ofc) on the pilots behalf (the one that got the kill that is) just the fact that is was at the point of destruction anyway , does that mean that the pilot who dealt the killing blow should have his stat increased? because i don't, on the principle that it was someone elses hard work that brought the mech to that state.

(Im not saying that jenner pilots or light mech pilots are skill less by the way its a scenario k )


So a suggestion that could work, instead of the KDR being a whole number (integer) it could become a fraction (floating) so, every player that is involved in the destruction of a mech gets a fraction of the kill.....

A mechs health starts at 100% and there are 2 lance group drops at the moment equal to 8 players which would mean that if all 8 players were involved in bringing down an enemy mech then each player will get 0.125 of a kill. I think it will also help to prevent mech hugging and reduce the friendly fire damage, i mean mech hugging is done for one reason only......to ensure you get the kill.....

basically its 100 divided by the number of players involved in killing the mech, this would (as i see it) balance up the KDR stat for all levels of players, and maybe a Headshot stat should be introduced for those players that like a definitive stat.

And thats phase 1 just to start

The final phase would be to calculate the amount of damage each participating mech did and award the appropriate score to the percentage of damage done upto the point of the destruction of the mech, with various bonus's such as causing an ammo explosion, destroying a leg, destroying the main weapon, destroying the engine etc all of these carry a certain percentage bonus, so for example

If you were engaged in a battle with a mech with 2 of your team mates (the percentages in this example are not exact as i dont have the damage model) you leg the mech and did 10% additional damage:-

your score is 15% for legging plus your 10% damage gives you 25% of kill equating to a KDR of 0.25

This can get complicated and im willing to sit down and work it all out, but it's pointless if the idea isn't very popular, okay so you have to work harder for your kills but the rewards are there, i just think that the current system doesnt give a true account for a players activity on the battlefield. It could even help with the match making system as a players skill will be more creditable and accurate.

I can certainly see that it will change the way we aim our shots, now most of us just go for the core because it gives us the 1 KDR if we get the kill, but if the system was changed we would be looking for the head shot or something else as well as the engine, but as it stands the engine is the most valuable part of the mech as far as components are concerned so why are we coreing mechs if we want c-bills from salvage?

Anyway i will leave it there for now, but if you think a system like the one i suggested above for KDR is worth looking into then please speak up, at least it would give me foundation to write up a test sheet to present to the dev's, where it goes from there who knows but we can try right!!

#2 Raso

    Member

  • PipPipPipPipPipPipPipPip
  • The Sickle
  • The Sickle
  • 1,298 posts
  • LocationConnecticut

Posted 23 May 2013 - 06:05 AM

KDR lets all the bros out there know who's the alpha bro. This is vital in a social situation as it lets the group know who gets the largest portion of wings, Jägerbombs and Doritos by comparison of the size of their epeen. More epeen means more wings and first dibs on potential mates, bra.

#3 SiorAlpin Wolf

    Member

  • PipPipPipPipPipPip
  • The Predator
  • The Predator
  • 255 posts
  • LocationU.K

Posted 23 May 2013 - 06:10 AM

lolz, so if you do all the damage to a mech and took alot of damage yourself doing so, your quite happy for me to come along 1 shoot the mech and kill it, not only taking the KDR point but the doritos, wings and jagerbombs so my epeen grows and yours stays the same? bra

#4 Josef Koba

    Member

  • PipPipPipPipPipPipPip
  • Wrath
  • Wrath
  • 527 posts

Posted 23 May 2013 - 06:22 AM

I definitely see your point. Sometimes it's frustrating waxing a mech hard only to miss out on the kill. But honestly I don't mind that much; I get my fair share of kills. I do wish, though, that they'd track assists, which is in my mind indicating that I got a share of a kill. All in all I don't much care who gets the kills; I pay attention to the damage dealt and the match score. In the end, I want my team to prevail and my personal kill stats come secondary. If a Jenner comes by and pops the Atlas I've worn down with my twin AC/20s, that's fine with me.

#5 Dan Nashe

    Member

  • PipPipPipPipPipPipPip
  • Philanthropist
  • Philanthropist
  • 606 posts

Posted 23 May 2013 - 06:37 AM

Statistical average now sortable by mech.
If you focus fire more, pick your targets well, don't run off by yourself, and don't facehug a jaeger, it goes up, on average. If you do mst of the damage to most of the mechs that die, you will get mostt of the kills. You'll still have 0 kill 700 damage games, but we're looking at averages here.

The fact of the matter is, as I've gotten better my kd has gone up. But I also try to not compare my hgn sniper to my atlas brawler to my 150 kph scout. I also know that my kdr is a lot lower in 8 mans.

So anyways, it's a valid stat but people are too simplistic and will routinely misstate its value and misuse it.
I.e. my sniper should have a higher kd than any atlas brawler of similar skill because my job is massive dps and killshots, whereas the atlas leads the charge and mixes it up. I also get to use cover at all times, he doesn't.

And of course, when a eapon is bugged to be too good, I use it ruthlessly. That will of. Course inflate my stats too.

At least for that chassis.
Also human memory is awful, so when I just feel like killing all the things, I like to have a stat that shows what has worked for me.
This is doubly true now that w/l is particularly unreliable and is designed to trend to 1/1 for even well above or below average players. Of course, tracking by mech means I win 75 percent of the time in some assaults, and lose 75 percent of the time I'm others.

Edit: more stats is food. I'd love to see assists too as additional data. Along with damage taken at the end of match screen :-).

Edited by DanNashe, 23 May 2013 - 06:39 AM.


#6 Hayashi

    Snowflake

  • PipPipPipPipPipPipPipPipPip
  • Bridesmaid
  • Bridesmaid
  • 3,395 posts
  • Location輝針城

Posted 23 May 2013 - 11:41 PM

KDR is not a good guide even should assist damage be figured in because dealing damage in the first place is not the only role of Mechs.

I'm far better at the HBK-4G than in the COM-2D for instance. Also bear in mind about 50% of my games in the 2D were in 4mans, whereas all of my 4G games are in PuGs.

Posted Image

How well you do as a whole is best reflected by WLR at the moment. If the XP allocation is fixed a bit better to take account of other roles like tanking that has the best potential to be a good statistic, but at the moment WLR really is the best gauge we have in spite of its flaws (not taking into account whether you drop alone or in coordinated 4mans for instance). That's kind of why ELO is based on WLR (though ELO being not very well implemented is another issue in itself).

#7 Shakespeare

    Member

  • PipPipPipPipPipPip
  • Legendary Founder
  • Legendary Founder
  • 429 posts
  • LocationGainesville, FL USA

Posted 26 May 2013 - 04:25 PM

Agreed. KDR does not describe one's ability to win in any real way. Most people will tell you that WLR is more indicative, followed, perhaps, by damage done/accuracy.

#8 Xeno Phalcon

    Member

  • PipPipPipPipPipPipPipPipPip
  • Bad Company
  • 2,461 posts
  • LocationEvening Ladies

Posted 26 May 2013 - 04:37 PM

KDR is a simplistic way for a player to gauge themselves, This allows players who are accustomed to other games such as call of duty to more comfortably relate to MWO and have something familiar to strive for. For these sorts of players the all important epeen is vital for their self esteem and with out it would likely wither and die.

*Checks own stats*

KDR MEANS NOTHING! ITS A USELESS STAT AND SHOULD BE REMOVED ASAP!

(o.o)

Edited by Xeno Phalcon, 26 May 2013 - 04:37 PM.


#9 Zordicron

    Member

  • PipPipPipPipPipPipPipPipPip
  • Overlord
  • Overlord
  • 2,547 posts

Posted 26 May 2013 - 06:55 PM

K/D ratio is for chumps. Maybe in CoD or something, it has some sort of value. In Mech combat: war of attrition- nuh. If you need Epeen affirmation, it might be useful to you. W/L ratio is only marginally better, you can have the game of your life with 6 kills and 1200 dmg on your dragon 1c and lose to a base cap. You can get blown away from a headshot with 30 dmg done and win the match in the end. Averages are great, but not accurate.

Honestly, the only stat that has any kind of use is accuracy, and that has to take into account the type of system fired. Lasers sprayed in all directions will detect a hit. LRM are inherently prone to miss a % of missiles, or even a whole salvo, or get shot out of the air by multi AMS, etc. If you are a player amused by laser light shows, well your acc is probably not representative either.

mostly I find the stat page useful to see how many matches I have run on each map and on which mechs. The rest is all epeen fluff IMO.

#10 p00k

    Member

  • PipPipPipPipPipPipPipPip
  • 1,661 posts

Posted 26 May 2013 - 07:00 PM

eh, try it. play >1000 drops, and tell me how you do

bottom line is, your kdr will average out over time. you're never going to "killsteal" everything, and there will always be people trying to kill you (the death side of kdr)

it might be abnormally high or low after ~100 drops, but over time it evens out

#11 BR0WN_H0RN3T

    Member

  • PipPipPipPipPipPipPip
  • Philanthropist
  • Philanthropist
  • 701 posts
  • LocationElysium

Posted 26 May 2013 - 07:31 PM

View Postp00k, on 26 May 2013 - 07:00 PM, said:

eh, try it. play >1000 drops, and tell me how you do

bottom line is, your kdr will average out over time. you're never going to "killsteal" everything, and there will always be people trying to kill you (the death side of kdr)

it might be abnormally high or low after ~100 drops, but over time it evens out

Over time in PUG matches it should even out, yes, because they are supposed to be random drops against equivalent skilled opponents. It's a bit meaningless if u compare a team-droppers kdr to a PUG-dropper.

#12 p00k

    Member

  • PipPipPipPipPipPipPipPip
  • 1,661 posts

Posted 26 May 2013 - 09:26 PM

View PostBrown Hornet, on 26 May 2013 - 07:31 PM, said:

Over time in PUG matches it should even out, yes, because they are supposed to be random drops against equivalent skilled opponents. It's a bit meaningless if u compare a team-droppers kdr to a PUG-dropper.

you're not supposed to compare two people's kdr
you're supposed to compare it to your own over time, and how others' are changing over time

#13 Denolven

    Member

  • PipPipPipPipPipPipPip
  • 511 posts

Posted 27 May 2013 - 04:00 AM

View PostSiorAlpin Wolf, on 23 May 2013 - 05:59 AM, said:

How is KDR a valid stat?

Most of the values are just there, because they can put them there. Meaning that they are things that are easily trackable automatically. Killcounts, maps, mechs, all easy to track. Human behavior not so easy to track, and usefullness of human behavior no chance to track.

The statistics themselves are just stuff that was measured. How you interpret it is up to you. Statistics in general are VERY limited. You can only measure a tiny part of a thing, and even that bit of data is already interpreted in so many different ways that it's almost impossible to say what the data means. In fact, there is a whole part of science that is dedicated to finding out about how good the chances are that the data actually says what we think it says (inferential statistics).

The other part of the story is human psychology. Everyone has a different one, but there are patterns. Those patterns exist - whether you like them or not. One of them is the ego. In order to achieve a financial success, companies try to listen to those patterns. And if putting the k/d ratio on top of the list is going to satisfy more people's ego, then they will do that. It's really that simple: give the people what they want. And what people want is ALOT less noble than what people say they want, or even what they think they want.

You can call them stat whores (or chumps) all day long - but the fact that the vast majority of people are like that, in one way or the other, is pretty much undeniable. We all have an ego, and we all like it to be pleased, even if we are not searching for that pleasure explicitly.

So in short I would say: Yes, the actual value of that ratio is rather low - but people like talking about it, so let them have their fun. The validity comes from the fact that it can be measured, and that people care about it (for whatever reason, doesn't really matter).

#14 Aslena

    Member

  • PipPipPipPipPip
  • 138 posts

Posted 27 May 2013 - 07:43 AM

There are people that like to brag It won't make them stop if this stat is removed, they'll just pick another stat. I for one like to look over my stats from time to time, it's an enjoyable natural extension of the game that allows me to look at stuff when I can't actually play.

No single stat by itself really means much, however I can tell how I'm doing with my personal play style to a small degree. I personally would rather see them add to the stats rather than take away by adding things such as a today's stat list, this week's stat list ect...

#15 JudgeDeathCZ

    Member

  • PipPipPipPipPipPipPipPip
  • The Defiant
  • The Defiant
  • 1,929 posts

Posted 27 May 2013 - 10:05 PM

Higher KDR than 2.0 means you are killstealer...

#16 Ralgas

    Member

  • PipPipPipPipPipPipPipPip
  • Overlord
  • Overlord
  • 1,628 posts
  • LocationThe Wonderful world of OZ

Posted 27 May 2013 - 10:13 PM

It's all e-peen. i can pull 600+ dmg on 3 mediums atm with a 3x lbx build, but it doesn't mean i did anything worthwhile during the match. i can go 4 2 with 400 odd dmg in an ac/40 but it just means the other team wasn't smart enough to watch it's flanks or have someone with seismic. A light spending a match fulltime capping has it even harder as they look useless on paper.

W/L is your best stat, especially if you full-time pug as you've then done it handicapped.

Edited by Ralgas, 27 May 2013 - 10:15 PM.


#17 Dexter Herbivore

    Member

  • PipPipPipPipPipPip
  • Bad Company
  • 241 posts
  • LocationPerth WA

Posted 27 May 2013 - 10:25 PM

KDR, Gearscore... it's all the same epeen to me.

#18 amirite

    Rookie

  • 1 posts

Posted 28 May 2013 - 05:48 AM

No stat is a full summary of a player's skill. As in all games, you should look at a player's entire range of stats to determine how good they are. Why waste so much breath on KDR? Do you take issue with the way MWO tracks it's stats in general?

#19 Demosthones

    Member

  • PipPipPip
  • Overlord
  • Overlord
  • 71 posts

Posted 28 May 2013 - 06:03 AM

View PostEldagore, on 26 May 2013 - 06:55 PM, said:

K/D ratio is for chumps. Maybe in CoD or something, it has some sort of value. In Mech combat: war of attrition- nuh.

Honestly, the only stat that has any kind of use is accuracy, and that has to take into account the type of system fired. Lasers sprayed in all directions will detect a hit. LRM are inherently prone to miss a % of missiles, or even a whole salvo, or get shot out of the air by multi AMS, etc. If you are a player amused by laser light shows, well your acc is probably not representative either.



I would disagree that the only stat of use is accuracy. I regularly copy and paste the mech stats into excel. Then I add a column for damage per match. This lets you see if your mechs weight is proportional to the amount of damage you do. My best is the AS7-DDC, which is as it should be. I goup my mechs by wieght then look and see which chasis I do the most damage in. After that I think about the waepons loadout and why that particulair loadout worked. I can use this to make my less effective builds better.

#20 Warge

    Member

  • PipPipPipPipPipPipPipPip
  • 1,027 posts
  • LocationKiyiv

Posted 28 May 2013 - 06:12 AM

View PostJudgeDeathCZ, on 27 May 2013 - 10:05 PM, said:

Higher KDR than 2.0 means you are killstealer...

Don't be so sure. All my Hunchbacks have KRD ~2. And it's fair numbers...

View PostDemosthones, on 28 May 2013 - 06:03 AM, said:

I regularly copy and paste the mech stats into excel. Then I add a column for damage per match. This lets you see if your mechs weight is proportional to the amount of damage you do. ... I goup my mechs by wieght then look and see which chasis I do the most damage in. After that I think about the waepons loadout and why that particulair loadout worked. I can use this to make my less effective builds better.

http://mwomercs.com/...04#entry2346704





10 user(s) are reading this topic

0 members, 10 guests, 0 anonymous users