76 replies to this topic

[Dawnstealer]

Dunning Kruger Effect, on 26 March 2014 - 01:12 PM, said:

High damage and low kills is a sign a player is (a) not targeting injured components ( a poor shot © using "spread" weapons like LRMs - neither of which is an indicator of "good" play.

That, right there - I pride myself on wiping out components. Another thing that leads to my early death in matches is that I'm willing to sacrifice myself to take out a DDC's left shoulder. Why? Because that's where the ECM is that protecting the entire enemy team.

It goes down, I'm killed, but then their entire team is subjected to LRM fire and the match is usually over quickly after.

[T0ADKILLERD0G]

I don't know about you guys but my K/D and W/L are almost the same (PUG) xD - And I am fine with that I still have fun win or lose! We all need to remember its just a game...

[Water Bear]

I think what many people are really doing is equating K/D with skill. That is, if they have a high K/D they think they are good.

For me, I try to play to have fun. My K/Ds are usually bad but I have placed fourth in a tournament so I feel vindicated in my belief that an obsession with K/D is foolish, and isn't even a measure of skill.

I don't understand the point of concerning yourself with performance. Why bother playing with an arranged team? Is it that important that you win all the time? Who really cares other than you, and why do you care?

Edited by Water Bear, 28 March 2014 - 08:35 AM.

[Uncleclint]

Unfortunately there are still too few statistics to tell the whole story of someone being a good or a bad mechwarrior, you might get a hint in which direction you might be going, but in general there´s just not enough to analyze and draw real conclusions.

Let´s say you have whatever Mech with the configuration A and play the first 50-60 games without changing anything. By that time your Mech doesn´t have all those tiny little skills that make it turn faster, cool down more heat and you know the rest, plus you may need a few games to get the right feeling for this mech. The next hundred or so games, you change that configuration into B - something that suits your style of play with that specific mech a lot better, a few tweaks here or there in the first few test rounds until you come up with the "right" config. Now since you cannot reset or save any, the stats that you get for the games 61+ (which will probably be a lot better than the first few ones) will be added to the stats of your games 0-60. So, after a hundredandsomething rounds with this mech, you better kept a close track of its performance all the time or the K/D as well as any other stats are nothing but a mishmash of not-telling-you-anything. Even worse, if the Mechlab is your second home (which i doubt to be very likely since UI2.0, but anyways) and there is not a night without a different config.

Same problem with Wins and Losses. I assume that on average, your overall performance in a game that is won will be at least a bit better than in a game that is lost, especially if it´s one of those 0:12 games. There is no option to split up your K/D between games won and games lost.

It is good that there are stats, but they seldom tell the real story. They are in my opinion right now at the same level of usefulness that UI 2.0 is.

Some time in the past, someone promised to add a lot more features to the whole stats system, on a monthly basis if i recall that right - unfortunately this promise went the same way the "no 3rd pv" and "cw coming soon" (and spoiler alert: loads and loads more) promises took...

[mogs01gt]

MWO needs to revamp the profile system. It really should be K/D/A ratio since you can go 0(kills)-0(deaths)-12(assists) in any given game.

[Wildstreak]

For me, K/D only measure how much I have improved myself.
I used to have horrible K/D when I started. My base stat averaging all my play now archived was originally ~0.25. That improved to 0.35 about Winter 2013-14 when I started using certain designs after a break.

Now the few games I played since the stat wipe, I am 0.75. Helluva improvement but definitely not into high ELO based on other stats I have seen posted, maybe middle or between middle-low.

[T0ADKILLERD0G]

mogs01gt, on 28 March 2014 - 09:59 AM, said:

MWO needs to revamp the profile system. It really should be K/D/A ratio since you can go 0(kills)-0(deaths)-12(assists) in any given game.

Good idea

[Josef Koba]

As far as I know, the KDR of one player isn't accessible to anyone unless that player makes it so, so really it's more of a matter of personal pleasure. For me, a positive KDR means, simply, that I'm killing more than I'm dying. This is a good thing. It's a very rough measure of how I'm helping my team (I also keep an absurd and pointless spreadsheet with all sorts of stats so I can adjust my own tactics as necessary). Now, I'd rather have a positive KDR AND substantial amounts of damage and assists. But, as with anything, it depends on the role a pilot is trying to play. Support mechs probably won't have positive KDRs, though it's certainly possible.

I'm a little bummed about the stats resent for one reason: I was an abysmal player for six months with a deep red (negative) KRD. I clawed and clawed my way back to being able to respect my own abilities as a pilot. Eventually I crested 1.0, then 1.5, then 2.0, before finally settling in at around 2.11. At that point there wasn't a ton of movement. It seems that I was about a 2.11 pilot. Some matches I'd kill four or five or even six mechs, then I'd go a bit without doing more than 2.0. Now, well, my KDR is higher, no doubt reflecting my better skills and ignoring that six month's worth of terrible piloting but man...I worked hard for that 2.11. I'm glad they at least kept it archived.

[Dunning Kruger Effect]

KDR is not the only (or even the best) measure of player skill. Your overall match score is a better indicator. A "global" KDR stat is far from meaningful and kills within a single game is also completely irrelevant and not indicative of a "trend".

Like most stats, there is little context. However there is still SOME context (type of mech) and if there are many games played, "abormalities" in gameplay stats tend to level out.

What it does measure "usefully" is player kills vs deaths, in a particular mech, over a sustained period. I.e. the ability to deal killing blows whilst preserving ones own mech. That is not "meaningless" or "worthless" stat and can be useful for adjusting your game play style.

For example, I have an Awesome with a .12 KDR (!) over 20 games. Since my average is .75 with my next-worst assault mech, I know I need to radically change my loadout or playstyle. I could be very valuable to my team in other ways, but it's safe to say I probably need to look at my playstyle if my KDR is that low.

FACT #1: Killing the enemy team is how you win. This is good.

FACT #2: Losing your own mechs is how you lose. When you are dead you cannot help your team. This is bad.

FACT #3: Since killing > dying, a high KDR is "better" than a low one.

FACT #4: A person who plays 100 games and kills 1 mech and loses 1000 (0.01) IS NOT a better player than one who kills 1000 and loses 1. (100:1). That's an extreme example, but according to those who say "KDR is worthless" then a 0.01 KDR can be the sign of an excellent player. I think that is a silly assumption. All we know is for sure that the .01 guy is 100 times more likely to die than he is to get a kill in any given game.

Since the average player (83% of the player base) plays PUGS, we can presume the low (0.01) guy is not persistently "carrying" his team to victory with amazing and obscurely unquantifable acts of teamwork. In fact, statistically he is most likely to be "dead" spectating his team mates who are -1 mech down.

At bit off topic: There also seems to be this fallacy that a high KDR and teamwork is mutually exclusive. I remember in Planetside 2 people used to say "I'm a medic and a good team player, therefore I have a very low KDR" ignoring the fact other medics did just as much teamplay, while maintaining a very high KDR. The second medic actually killed all the nearby enemies before reviving, meaning you could revive in safety, and spent less time on the spawn screen and more time actively being a "team player." A low KDR does not automatically make you a good team player. It simply indicates you die a lot more often than you land a killing blow. And remember folks, killing enemies and not dying is how you win!

[Phex]

Hi there together,
im a light pilot and i only play pug`s.
My K/D is 0.8, but im about 150+ in wining to loosing, so im doing something right.
Right?
So the K/D depends on your playstyle and what kind of Mech you use.
Its a nice indicator of how you personal are developing but for me the win/los is more important

Greetings Firun

[Lukoi Banacek]

Chillbill, on 19 March 2014 - 10:15 AM, said:

It seems some view it as the gold standard, most don't.

I wonder where you get your statistical "most from."

However, more to your point. K/D is an important statistic, however it is not the most telling. PGI needs to allow us more granularity, not less, but until they do K/D will remain a fixation for many people.

If you enjoy the game in Mechs where you do not yet have a green K/D, I would say you are on the right track and there's no harm in trying to better that particular statistic for yourself as long as you continue to contribute to the successes of this very "team" oriented game.

I find that determining why I am less effective in a given mech when comparing certain stats (KPM, DPM, KDR, Damage/match, W/L etc) to other mechs, helps me learn to get better in those mechs. All to often, due to the initial feel a mech has in terms of weapons capability, movement profile, size etc....I have certain biases and ingrained expectations. Overcoming those incorrect perceptions to play those mechs more effectively can be key.

Prime examples for me....the non-3L ravens, the Kitfoxes etc....they are NOT lights in the manner that I prefer and I had to find entirely new ways to be effective in those mechs vice the lights I was more comfortable in. It makes you a better player.

So, having a goal to improve such stats, KDR included, is not a bad thing.

Fixating and playing to that statistic alone though, will likely not yield great results, regardless of how inflated your KDR might appear. I have a 34:1 Hunchback....I keep that stat preserved for that exact reason. To remind me that it's silly to play to that sort of fixation and that lucky runs (because ultimately it was a short stretch of games with some really great runs that won't soon be replicated by me) are simply streaks of randomness. Don't expect them, don't play to them. Your team will simply suffer for the abortive attempt and it's all just hubris anyway.

[Obscillesk]

k/d is and isn't important. If someone has less than 100 damage with 1/11 k/a then I'd argue they probably didn't do a lot for their team unless they had lrm boats to spot for. Damage done is a pretty good indicator of at least how much armor you've taken off the enemy team, but its a similar problem to above. Some guy with 600 damage and 2/1 k/a didn't do a whole lot. Yes, 2 mechs down is solid, but he also spent all his time doing nothing but dueling those 2. To me high damage and k/a is far more important than if you died or not. Plenty of games are won by the assault who charges and soaks up all the damage while his team pounds on the enemy from everywhere else.

[Cold Cash]

I would love to see avg match score along with assists added.
K/D/MS/A

Then it becomes quite a valuable indicator of a persons skill.
1.3/1/60/5 would in my opinion be a very valuable player as an example since although k/d isnt high the match score and assists are high.

Sumthing like this also rewards the guys who sacrifice k/d for support roles in a easily viewed tangible manner. I bet you'd also see an upsurge of people looking to add assists and be consistant in matches to fill out thier stats lol.

Edited by Cold Cash, 29 July 2014 - 06:25 PM.

[Adiuvo]

Assists don't really count for anything since you get an assist for any tick of damage.

[n r g]

Chillbill, on 19 March 2014 - 10:15 AM, said:

Just looking to get an idea on the populations view on the importance of 'good' k/d. It seems some view it as the gold standard, most don't.

I have two mechs with k/d's well into the green. But I enjoy playing others that I am quite 'challenged' in which kill my k/d. They are still much fun. Which is the whole point of playing, right? It's not like a pro athlete playing for his salary, etc...

Still..........

As a competitive player - K/D is irrelevant.

It's all about competitive match performance. But with that being said, if you merely do pub matches are aren't driven to competition, even then, K/D is no indication of skill rather the ability to work good in a team. But obviously even then, given the certain teams, you will get negative k/ds.

IMO, MWO is one of the most irrelevant k/d systems known to gaming - it's so team driven that it will merely get padded by playing with good teams (you don't necessarily have to be good thus the flaw) and vice versa, if you are amazing and get "mediocre" teams , your k/d will suffer.

conclusion - forget about k/d.

[L Y N X]

Does anyone reading this thread dispute that NeuroToxic is an excellent mech pilot? I bet he has mostly positive K/D ratios. If you examine the best players, you will find a commonality of positive K/D ratios. Does this mean everyone with a positive K/D ratio is a good pilot, not necessarily, but they have made more killing blows than been killed, so in some sense, they are at least competent. Does this mean you won't find any good pilots with K/D ratios less than 1? No, I am quite certainly their are a handful, but not that many.

K/D ratio is not the end-all be-all of pilot performance measurement, but it is a general rule of thumb first indicator of a pilot's general ability, that being able to kill other mechs.

[DaemonWulfe]

The unique part of reading threads like this is seeing a whole bunch of people post the same response with different wording.

Focus on having fun.
Focus on contributing to the team (focus fire, stay close to the group, stay alive, etc.)
Find a good group of people to play with. It'll help the first two suggestions come more naturally.

Do these simple suggestions and check your stats from time to time. You'll notice it doesn't matter what weight class you play frequently. You'll have positive stats on everything.

Just as an example, my Oxide (the hero most love to hate) is sitting at just under 3.0 KDR, right alongside my beloved heavies and assaults. This is because I play it's role as a support mech. I keep the lights and mediums off the tails of my fat boy team members. This in turn keeps me alive. If I only score 0, 1, or 2 kills but stay alive along with team mates over several matches, my stats will look better than that Direwolf who just got a hat trick in one match but gets killed over the next 10 games. I don't want to sound like 10 pages of long winded explanations, so I'll leave it at that.