Jump to content

Win to Loss Ratio or Kill to Death?


99 replies to this topic

#81 Dozer

    Member

  • Legendary Founder
  • 289 posts

Posted 04 July 2012 - 09:24 PM

View PostDemona, on 04 July 2012 - 02:01 PM, said:


There's no I in Team, but there is a ME.


And that ME should turn itself upside down and become a WE :P

#82 Sean von Steinike

    Member

  • Elite Founder
  • 2264 posts
  • LocationUSA

Posted 04 July 2012 - 09:58 PM

Stats are generally used negatively I have found and lead to bad play, poor team work, kill stealing and kill farming. Best not to have them at all or at most, just privately available to the player.

#83 Leeroy Nimoy

    Newbie

  • Legendary Founder
  • 4 posts
  • LocationArkansas

Posted 04 July 2012 - 10:38 PM

View PostFLAKPANZER, on 03 July 2012 - 09:32 AM, said:

...when I see the two statistics mentioned in the title, I tend to prefer playing with
players that have a higher win to loss ratio than I do players with a higher kill to
death ratio.

The the former can indicate someone who just gets lucky by being placed on good
teams, it can also indicate a player who willingly sacrifices his own character/mech
for the benefit of the team...which of the above do you look at as being most important
when you judge other players?

I think the most important thing to try and remember is that there is no one stat that will give you all the information you need about a player. Taking into account time played, kill/death, win/loss, among other things will give the clearest picture of that player. There is a lot of talent out there that dies a lot, and there are a lot of horrible players with some amazing gear/stats just because they have friends that do great things. Holding on to that fact when you're looking at another player - or even yourself - is possibly the most important thing.

View PostSean von Steinike, on 04 July 2012 - 09:58 PM, said:

Stats are generally used negatively I have found and lead to bad play, poor team work, kill stealing and kill farming. Best not to have them at all or at most, just privately available to the player.

I couldn't agree more! E-Peen is bad, just saying.

#84 Rejarial Galatan

    Member

  • Elite Founder
  • 4070 posts
  • LocationOutter Periphery

Posted 04 July 2012 - 10:49 PM

View PostSean von Steinike, on 04 July 2012 - 09:58 PM, said:

Stats are generally used negatively I have found and lead to bad play, poor team work, kill stealing and kill farming. Best not to have them at all or at most, just privately available to the player.

could not agree more! public stats encourage all sorts of nasty behavior. Show me my stats so I know how to improve my self, but, I have 0 need to know your stats, or yours, or yours, or yours, or yours, ad nauseum. I see those stats as a big ol beacon to the undesirables club. Why? because if you do not have a specific KDL or WL then you are worthless. now before someone says, if i find a group that wants that, then, they are not the ones I want to play with. I am forced to PUG A LOT. And as a result, I tend to get stuck in random groups with people like that. Point is, there is more to a PvP game than how many kills to deaths I have or how often I win or lose.

#85 Demona

    Member

  • Elite Founder
  • 883 posts
  • LocationOutreach

Posted 05 July 2012 - 01:49 PM

View PostDozer, on 04 July 2012 - 09:24 PM, said:


And that ME should turn itself upside down and become a WE :P


If turned upside down wouldn't that make EW?

#86 Bodha

    Member

  • Legendary Founder
  • 519 posts
  • LocationAtlanta

Posted 05 July 2012 - 02:01 PM

View PostSean von Steinike, on 04 July 2012 - 09:58 PM, said:

Stats are generally used negatively I have found and lead to bad play, poor team work, kill stealing and kill farming. Best not to have them at all or at most, just privately available to the player.


stats for self improvement... GOOD
stats for e-peen... BAD
stats public no privacy option... BAD
stats with a public option...OK
stats with access like EVE API... BETTER

stats are like guns... they dont hurt people. people using them do. People who shout out the most in online communities about stats tend to use them from e-peen, while many ppl who are quieter tend to use them a bit more constructively.

As a final note. Stats are not the primary driver behind player killstealing farming and other bad player actions. Those players who do that kind of stuff heavily do those things b/c they are still focused upon the idea that higher k/d ratios mean they are doing something right. Whether k/d ratio is tracked or not will not change those kinds of players. So please stop saying things like stats lead to these bad actions.

#87 Dozer

    Member

  • Legendary Founder
  • 289 posts

Posted 05 July 2012 - 02:02 PM

View PostDemona, on 05 July 2012 - 01:49 PM, said:


If turned upside down wouldn't that make EW?


Nope that's upside down and transposed :)

#88 IronGoat

    Member

  • Elite Founder
  • 534 posts
  • LocationSolaris VII, trying to scam hip actuators from that seedy Liao Rep...

Posted 05 July 2012 - 04:49 PM

NEITHER :P

Damage done (account avg or by mech which is better)Vs Damage recieved
KDR and W/L are too conditional on factors beyond your control

Play awesome and your team makes an error and you lose= W/L tanks
Work hard to get some kills and some wizz bot in a light floats by and finishes off all your hard work with a killing shot =your KDR tanks

IMO damage done vs damge recieved ( and then sorted by mech for fairness, smaller mechs will do less damage etc)

a good second is "Avg Points" earned per match ( divided by mech type) and this only assumes there will be a "points"system in place that gives us a result after each match...

#89 Leeroy Nimoy

    Newbie

  • Legendary Founder
  • 4 posts
  • LocationArkansas

Posted 05 July 2012 - 10:27 PM

View PostIronGoat, on 05 July 2012 - 04:49 PM, said:

Damage done (account avg or by mech which is better)Vs Damage recieved
KDR and W/L are too conditional on factors beyond your control

Play awesome and your team makes an error and you lose= W/L tanks
Work hard to get some kills and some wizz bot in a light floats by and finishes off all your hard work with a killing shot =your KDR tanks

To a certain extent you don't have all the control over damage done/damage received. There are factors - though fewer of them - that can change one or the other without your intervention. I think some nice mix of all the a fore mentioned statistic variants with given percentage values towards an overall score would be acceptable. To use round numbers and just as a base example:
  • Kill/Death = 25%
  • Win/Loss = 25%
  • Damage Done = 15%
  • Damage Received = 15%
  • Scouting Points (or other comparable stat) = 15%
  • Rank in match = 5%
Something akin to this could give you an over rating that is dynamic - we'll say updating weekly or monthly - and it will always compound backwards once and average out. Again using round numbers for an example, let's say it can give you a composite score of 1000 max. So last week you might have a composite of 900, the week before you had a composite of 900, this week you would have an average composite of 900, but if you then got 500 at the end of the week you would drop to 700. This is because it would only account for current week and last weeks composite. This is just one idea, and I will grant you (before the trolls do ti for me) that this paragraph is a little convoluted and could use some fine-tuning. However, I still believe that the percentage of an overall score idea above is pretty sound.

#90 Reoh

    Member

  • Legendary Founder
  • 958 posts

Posted 06 July 2012 - 12:02 AM

View PostBodha, on 05 July 2012 - 02:01 PM, said:


stats for self improvement... GOOD
stats for e-peen... BAD
stats public no privacy option... BAD
stats with a public option...OK
stats with access like EVE API... BETTER

stats are like guns... they dont hurt people. people using them do. People who shout out the most in online communities about stats tend to use them from e-peen, while many ppl who are quieter tend to use them a bit more constructively.

As a final note. Stats are not the primary driver behind player killstealing farming and other bad player actions. Those players who do that kind of stuff heavily do those things b/c they are still focused upon the idea that higher k/d ratios mean they are doing something right. Whether k/d ratio is tracked or not will not change those kinds of players. So please stop saying things like stats lead to these bad actions.


Kinda reminds me of "Damage Meters" in other games. As a raid leader they could be really handy to help us figure out things (other than actual damage) like who wasn't getting on adds or was that shouldn't. When certain people would be doing more or less damage, who might need some guidance or spec assistance to help them get the most out of their game (we were a progression raid group, not casual).

The problem with them was everybody started running them, and using them incorrectly by oversimplifying the results.

#91 IronGoat

    Member

  • Elite Founder
  • 534 posts
  • LocationSolaris VII, trying to scam hip actuators from that seedy Liao Rep...

Posted 06 July 2012 - 08:15 AM

View PostReoh, on 06 July 2012 - 12:02 AM, said:


Kinda reminds me of "Damage Meters" in other games. As a raid leader they could be really handy to help us figure out things (other than actual damage) like who wasn't getting on adds or was that shouldn't. When certain people would be doing more or less damage, who might need some guidance or spec assistance to help them get the most out of their game (we were a progression raid group, not casual).

The problem with them was everybody started running them, and using them incorrectly by oversimplifying the results.



i wish it wasnt true that stats lead to bad gameplay but in some arenas they do. like it or not

The XVM mod for WoT was used mostly to harass and belittle gameplayers with lower W/L Ratios then anything else.
i cant count the number of PUGS i was in, where some twit running this mod announced at match start his team were all "losers" and then chose to log out mid game since his team had "no chance" according to his XVM.
Or where a player made a sugestion about tactics or a comment and was shouted down because his xvm stats werent high enough to grant him the right to have an idea or opinion.
or where a better player was unfairly focused on by the enemy team to eliminate him early on.
if you can recognize a good ( or bad) player on sight because you've faced/worked with before then fine. but the ability to single out good or bad players ( and good or bad based on stats mind you) is not a function of the game and is a bad idea.

im ALL for having personal stats. but given a choice id rather have no stat tracking then public stats.
as a further note im also against any third party mod that gives the player info IN GAME, the game designers didnt already include ( such as the XVM mod does)


if you want to get the most out of a game, play it as the devs intended it and be active in the community its played in IMO

#92 Kurogo

    Newbie

  • Members
  • Pip
  • 8 posts

Posted 06 July 2012 - 08:16 AM

I agree that there is more to this game than just kill/death. And that scouts and support don't get fairly represented in the K/D ratio. In FFA matches then there are no roles and the K/D ratio is directly representative of W/L. But in Team Play, it seems that it is necessary to announce what your role is to the game before the match so that it knows what stats to apply to you. Brawlers scout about as much as scouts brawl. It is possible that a socut runs into an enemy scout and gets to kill it, just like it seems possible for a brawler to run into a lance of mechs that someone hasn't noticed yet. If the ratios are there, they should be the same for both roles. Apply this idea to any other roles and you get the gist of my point.

#93 Homadais

    Member

  • Elite Founder
  • 84 posts
  • LocationPort Orchard, Washington

Posted 06 July 2012 - 08:31 AM

When are people going to play for the fun of it and not the stats of it. Yes I like to win. We all do. Well welcome to the realdeal. If you win someone loses. If they win you lose. That is the fact of it. You will never always win. STATS do not tell the whole story of a persons ability. I do like the idead of personal stats that only you know but not everone else.

#94 Kraven Kor

    Member

  • Legendary Founder
  • 4369 posts

Posted 06 July 2012 - 08:33 AM

I think it should be something like Battlefield, to an extent.

I think you should be rewarded for, in pretty much this order:

Wins - Winning the match / war should be the most rewarding thing you can do.
Kills
Objectives Completed
Damage Dealt
"Events" like head shots, ammo crits, etc.
Credits Earned vs. Spent

#95 Khanahar

    Member

  • Legendary Founder
  • 396 posts

Posted 06 July 2012 - 08:40 AM

Of course, W/L is going to be more important than K/D in the long run, and gives a fairer shake to scouts and others who don't focus on racking up kills, but still help their teams.

On the other hand, K/D is of elevated importance in a no-respawn game; if you "lone-wolf" your way to killing half the enemy team, you may have just won the battle. Assuming your team didn't die in the meantime.

#96 Ricama

    Member

  • Elite Founder
  • 879 posts
  • LocationCanada

Posted 06 July 2012 - 10:58 AM

Any one stat can be cooked, you need to consider them all in some formula.

#97 Filth Pig

    Member

  • Members
  • PipPip
  • 39 posts

Posted 13 February 2013 - 04:32 AM

While you all can sit here and say that K/D doesn't matter and blah blah blah.. the fact remains. If I kill 4 of you every match, then my team has to do less to secure the win by killing the rest of you.

It is still a game about skill, and yes some people are good core seekers and get kills by just luck, however those that consistently rack up kills are the ones that I want on my team..

#98 KKRonkka

    Member

  • Members
  • PipPip
  • 115 posts
  • LocationFinland

Posted 13 February 2013 - 05:37 AM

HIgh winrate indicates that player is doing something right but has a own share of luck. Platoons tend to increase this. Kill to death ratio is in general unfair to medium/lights (although those 2d+3L ECM monsters can reap many enemies apart) and stock mechs, limiting force they can use to kill. Damage done > kill assists > kills indicates more to me.

A value of efficiency combined for example from
- average exp per match
- victory/defeat rate
- average damage
- average kill
- spotting (could be nice)
- average mech tonnage (heavier = better? Not always)
- capture points
- defence (could be nice as well, like BatWing mentioned in his post)
- played matches (how experienced)
Familiar thing from WoT and who uses xvm mod. Visible statistics in general is a double-edged sword - it may give you information how to expect enemy/friendlies to behave but can also "blind" you.

As this is a team based game it would be nice to see somekind of "fame" system implemented instead of staring of individuals like always, kinda roleplaying an event where two enemy groups encounter each other and may know about pilots and mechs in the area. A team based value like "noted" if team has above-average efficiency, "fearsome" when high k/d ratio" or "group led by veterans" if there are few more effective fighter among group.

Edited by KKRonkka, 13 February 2013 - 05:37 AM.


#99 Mongoose Trueborn

    Member

  • Veteran Founder
  • 612 posts

Posted 13 February 2013 - 05:41 AM

My win percentage is 93%. You are telling me that I'm a better player than someone with a 90% win percentage?

All I see there is that I would play with better people than the other guy and it's 0 reflection on actual skill.

#100 MadcatX

    Member

  • Elite Founder
  • 291 posts

Posted 13 February 2013 - 05:42 AM

The stat that I'd like to see is Assists. K/D ratio's work fine if your playing FFA games (and yes, some matches do feel like that), but in what is meant to be a team-based game, K/D alone isn't the best means of tracking a person's skill, and W/L ratio isn't really good if you PuG.





1 user(s) are reading this topic

0 members, 1 guests, 0 anonymous users