36 replies to this topic

[TercieI]

DONTOR, on 21 March 2014 - 07:21 AM, said:

Ill throw in my 2 cents on kill stealing.
If Im in an obscure corner of the map deulling a spider in my comando and i have it legged, and I myself have taken 20% damage. If a streak Kintaro comes around the corner and fires a salvo at the spder and kills it, it IS stealing.
I did 95% of the work, wasnt in any immediate danger, and the the final blow is taken from me.
(damn those kintaro streak boat pilots!)

Streaktaros have way more kills stolen than they steal,* amigo. That's just a fact of non-pinpoint high damage output mechs.

On topic: every metric we have has uses and limits, IMO. I concentrate on being better than I was yesterday (or at least last week). Numbers show some of that, but some of it's feel too.

*i don't actually believe in kill stealing in a team game.

Edited by Terciel1976, 21 March 2014 - 07:34 AM.

[Hangfire]

I pilot mostly lights as a pug, as a result my k/d ratio is in the negative whereas my win loss ratio is positive,,
If judged on k/d I'm rank rotten, but on w/l I'm doing allright.
It IS a team game (even for us puggers) so personal k/d and w/l stats are more or less meaningless, to me anyway, I've got more important things to worry about in game, like hassling that lrm boat, blocking enemy ECM and trying to say alive in the process.
You want to take the kill,,, by all means that's WHY I'm TAG-ing it

[AlmightyAeng]

Terciel1976, on 21 March 2014 - 07:33 AM, said:

Streaktaros have way more kills stolen than they steal,* amigo. That's just a fact of non-pinpoint high damage output mechs.

On topic: every metric we have has uses and limits, IMO. I concentrate on being better than I was yesterday (or at least last week). Numbers show some of that, but some of it's feel too.

*i don't actually believe in kill stealing in a team game.

I believe in 'killing the enemy as effectively as possible' in a team game. If that's a spider taking the kill in a 5-on-1 furball...fine...but that same spider better not be leaving a heavy alone in a 1-on-1. Get in there and help that situation, too.

'Killing the enemy as effectively as possible' also means 'support the mechs with more weapons so they stay alive to provide you with open spots to critseek.'

Edited by Ghost Badger, 21 March 2014 - 07:49 AM.

[TercieI]

Ghost Badger, on 21 March 2014 - 07:48 AM, said:

I believe in 'killing the enemy as effectively as possible' in a team game. If that's a spider taking the kill in a 5-on-1 furball...fine...but that same spider better not be leaving a heavy alone in a 1-on-1. Get in there and help that situation, too.

'Killing the enemy as effectively as possible' also means 'support the mechs with more weapons so they stay alive to provide you with open spots to critseek.'

I couldn't agree more. If anything in my post implied differently, I apologize...

EDIT: when I said "I don't believe in kill stealing," I meant "I don't believe it is a term with any meaning" not "I don't believe in doing it."

Edited by Terciel1976, 21 March 2014 - 07:54 AM.

[AlmightyAeng]

Terciel1976, on 21 March 2014 - 07:51 AM, said:

I couldn't agree more. If anything in my post implied differently, I apologize...

EDIT: when I said "I don't believe in kill stealing," I meant "I don't believe it is a term with any meaning" not "I don't believe in doing it."

Nope! Was agreeing with you

[Enigmos]

I personally agree that 'kill stealing' is a fiction, but it is a real complaint on the part of some and frankly sometimes I feel a little emotional frustration (okay: a 'surge' of emotional frustration). When a tough, close fight has had me struggling and I am nearly cored myself from it and some spider waltzes in untouched and delivers the coups de grace then to be honest, yeah, I fell peeved.

But I know full well it was better that the spidy did what he did because my core is crimson and it could have been either of us that lost. But because of that spider I am now clear winner even if he gets the kill count.

I don' think you can help feeling it, but I also don't think it is a kill steal. It was for the team. Now, if the spider next sprays my leg with his green laser I might get ticked.

[Wildstreak]

When building a Mech, my primary factor is role more than anything followed by good heat management. I have seen people use a lot of build by Smurfy stats with less than 30% Cooling Efficiency that leads to a lot of shutdowns. I tend to go higher thus avoiding that problem more. Building for role and playing to take advantage of that role while minimizing disadvantages seems preferable in match outcome.

K/D is only for personal growth to me relating to several factors (how well I use terrain to avoid enemy fire, how well I sometimes bring enemies to teammates to be killed instead of dying in a one on one fight, etc.)

Stats can also show which weapons someone is good at using at that time and which Mechs and roles they function in best.

[TygerLily]

Maybe they could tweak the scoring system? And on the end of round screen ONLY show a players score...on your personal score page it will show kills, damage, etc.

High score 1000+
Average score 500-1000
Low score 0-500

High reward actions (block of points)
Kill enemy who has 75% health or more (rewards marksmanship)
Destroy a component (average points per component, but most locations have a few components)
Savior Kill (Kill enemy who was attacking a friend with 50% or less health)
Team Kill (minus a huge chunk)
UAV affecting 6+ enemies
Tagging single target for 15 continuous seconds (keeps stacking...ie, a block of points rewarded every 15s)
Airstrike or Artillery which damages 6+ enemies
Win bonus

Average award actions (block of points)
Kill an enemy mech who has 50% to 75% health
Kill an enemy turret
Defensive Kill (Kill enemy who was gaining capture points in the past 10 seconds)
Tag/Narc/hold LOS on enemy that takes LRM damage
Airstrike or Artillery which damages 3+ enemies

Low reward actions (block of points)
Kill an enemy mech who has 50% or less health (the point reward went to those who did all the damage)
Capture assist
Capture initiated
Defense Assist (damage enemy who was gaining capture points in the past 10 seconds)
Airstrike or Artillery which damages 2 or fewer enemies
Unused Consumable bonus (good job managing that heat, guy)

Variable rewards (points over time)
Damage

1 pt awarded per pt of damage - ie, 300 damage is 300 points)

Plus 10% for Skirmish mode

Capture (Assault)

300 points total: X pts per second spent capturing. If multiple mechs are capturing, +X% per second equaling X% capture time decrease due to weight bonus. (ie, from start to finish you get 300 points regardless of how much help you get capping)

Resource Bonus

2.5 pts per Resource gathered (750 * 2.5 = 300pts)

Damage Soak

1 pt per 5 damage you receive

(No weapons? About to die? Rack up some points by rolling damage to every part you have left, keep those enemy guns off your friends for just one...more...second...)

Plus 5% for Skirmish mode

Friendly Fire

Minus 5 points per 10 damage to friendly forces.

Spotting/UAV

1 pt awarded per second that spotting method is highlighting a target

IE, UAV visible, shootable, costs CB, but can award high spotting bonuses

Something to this affect. The idea is to reward the myriad of things that people do during a match which lead to a win...the truth is, people focus on the info presented to them. If you "hide" kills and damage and your score is the only way to measure your performance, then you'll naturally care about those actions.

It's not to say K/D is bad...but in a team game there is a better way to quantify people's performance.

[Delas Ting Usee]

A lot of good points and suggestions brought up.
How I count my performance nowadays is by dividing the number of damage I did with the number of assists I've gotten.
That to me is a better indication on how well I performed.
I also depend on how much damage I did in a match depending on weigh class.
Light = 100-150pts of damage. Medium = 200pts and upward. Heavy = 250-300pts. Assaults = 350pts and upward.
Yup, I no longer count K/D or W/L scores anymore.
Not when you get thrown back into mech lab or forcibly disconnected.
That's just skewers your score so what's the point, right?

[Haji1096]

I posted a thread about how we could better measure a player's performance in the suggested features forum

http://mwomercs.com/...ck-performance/


Threre should be more stats to measure your performance. Things like

effective damage = (damage applied to enemies) / (total amount of damage fired)....This would measure how accurate you are, and are you firing at optimum range

overheat rate = times overheated / number of matches

damage per kill, damage per assist, damage per component destruction

some stat that could measure how much damage you have taken relative to how much attention you are taking up.

Now of course just like any stat, without context the numbers are meaning less. But you could measure them against what the average of the player base is or something. Or create expected values for each stat per chassis or mech type.

[Rear Admiral Tier 6]

I believe in the fun/time ratio more than k-d stats,if you have fun playing,it really does not matter if you die a lot,its just a videogame,not a knifefight.

[JediPanther]

My kdr is a made up number like zaboilion. I determind if a match was good on how high my kill assist and damage is. If I get one or two kills it was a great match. I've seen nearly fresh last mechs alive on my team suicide just because they were last mech or to save their kdr. Last time i saw it my kdr was .80. If you're worried more about a stat than having fun you might want to spend more time as a bookkeeper than as a person playing a video game.

[Saint Scarlett Johan]

I heavily disagree on high damage numbers. If I'm in my poptart, and I kill a guy in 90 damage or less (with the exception of an Atlas), that was a good kill and I shot well. If I'm doing about 100-150 damage a kill, I'm shooting well. In matches I get 800-1000 damage, and only 2-3 kills and 4-5 assists I had a bad personal match. Why? It means I hit everything BUT the important parts.

Same goes for my Wang, If I'm putting up high kills and low damage, it means I'm putting my AC20 where it counts. If I'm getting high damage and low kills/assists it means I played a poor match and my team carried me.


That's why I laugh at pilots that are all, "look at my 1500 damage and 2 kills, 3 assists. Hur-dur, I so 1337." When the enemy team only had 2 assaults. People like that are either really shitty shots or damage farming.


My personal best match was 8 kills and 496 damage. Insta-gibbed a Griffon, Commando, and Shadowhawk with rear ST shots, blew out the side torso of two Jagers and a Catapult, and headcapped another Catapult. I only got out damaged by two other guys, 551 and 528 and both were quad AC2 Jags hitting everything but the CT.

The stats we have are almost meaningless without even the ability to see assists. If you have a high W/L, but low K/D and low assists, your team is carrying your useless arse around. But without the ability to see the number of assists, we can't determine this.

I'd like to see a breakdown of the weapons and not just raw accuracy, but the components I'm hitting often by percentage. If I have 88% accuracy with my medium lasers and I'm only doing 3.8 damage per shot fired. That means I need to get closer or have a steadier hand. And if I'm missing outright, what components am I actually hitting? Am I just uselessly scraping my lasers across both arms and all the torsos? Or are they drilling CT?



And as DONTOR said, there is kill stealing and kill securing. There is difference. If I'm getting beat on and beating on another mech and I've hit heat threshold and a buddy comes and kills the guy and I got to live, that's kill secured. If I started the initial engagement and a buddy joined in halfway through the fight and gets the killing blow, that's kill secured.

If I'm in my commando dueling an LRM stalker or an Atlas and I've been working him over, and just before I get the killing blow in, some **** in a spider zooms by and scrapes with a small laser for the kill or some AC2 boat flings rounds from over 1km away and gets the killing blow, that's a steal.


I'd also like to see something like PS2. Each kill is worth a "pool" of XP. If another guy plasters an enemy for 95% of the damage and I get a single round that wings him for the 5% kill, I only get 5% of the XP for the kill, and he gets the larger 95% for the assist.

Edited by JohanssenJr, 04 April 2014 - 02:13 PM.

[Mike McSullivan]

Even if i liked your post JohanssenJr, there is one point im not with you
If you did 1000Dmg that means that your opponent lost 1000HP, doesnt matter if they were spread or not. Which made it easier for the rest of the team to kill one or another enemy. Thats why A+AStrikes and LRMs are very useful. a "deep-orange" torso needs only one more AC20 to possibly kill the mech.

Aside from that: i count the kills per match with any single chassis for myself. Depending on the weightclass(or even mechtype) i can see in which mech i perform good or not (lights<1: doesnt matter; Assault<2 :meh...). Deaths shouldnt be counted in any way. Doesnt make sense to me in an environment where we fight in (time)restricted "instances" with a 50%chance of winning(solodrops). It could make sense in a persistant world like Planetside2 (K/D shows more the timelenght you survive)

[Saint Scarlett Johan]

Mike McSullivan, on 05 April 2014 - 12:37 AM, said:

Even if i liked your post JohanssenJr, there is one point im not with you
If you did 1000Dmg that means that your opponent lost 1000HP, doesnt matter if they were spread or not. Which made it easier for the rest of the team to kill one or another enemy. Thats why A+AStrikes and LRMs are very useful. a "deep-orange" torso needs only one more AC20 to possibly kill the mech.

And what if that shot took that damaged side torso and I didn't die? And instead I put out another AC20 round into your CT that killed you? Was that damage worth it? Or how people waste time on the arms of my Battlemaster while I drill their CTs? They are doing damage, they can take both arms for total of 150 points of damage. But they didn't take anything from me of value, but payed for it with their life.

Go ahead, shoot the shield arm on my Cents. It means you aren't shooting the legs or my CT. It means you aren't doing anything to kill my combat effectiveness, which I'm in turn using to kill you and your team.

[Jonathan Paine]

Important thing to understand: the difference with how well you did in ONE match, versus how good you are OVERALL.

If you want to know how good you are overall, the win/loss metric is unbeatable in the long run. If you are good, you will win more than you lose but thanks to the Elo system, it will GET HARDER the more you win.

For any given game, damage is probably a good indicator of how well you played - HOWEVER, all the damage in the world will not make up for forgetting about the cap points and having that dratted locust on the other side beat you on the cap. Even if that player did 0 damage, that was a better match than your 1500+ damage and your 5+ kills. Who cares about the KDR of a loser? Living and losing = terribad.

TL,DR:
W/L for overall skill, KDR only matters if you win.

[jper4]

DONTOR, on 21 March 2014 - 07:21 AM, said:

Ill throw in my 2 cents on kill stealing.
If Im in an obscure corner of the map deulling a spider in my comando and i have it legged, and I myself have taken 20% damage. If a streak Kintaro comes around the corner and fires a salvo at the spder and kills it, it IS stealing.
I did 95% of the work, wasnt in any immediate danger, and the the final blow is taken from me.
(damn those kintaro streak boat pilots!)

or if it's me doing the shooting my streaktaro fires three full alphas (assuming the target hasn;t been killed by the person who started fighting it by then) ,which takes off both arms, sends CT to red, right torso to orange before managing to take out the left torso and the xl engine while the surviving leg that is hanging on by a paperclip remains untouched the entire time when it finally dies.