145 replies to this topic

[SamizdatCowboy]

For all you TL;DR folks the gist of this post is:

• K/D is a bad public leaderboard metric, but it can be a useful private stat
  
• The devs do NOT plan to use K/D for public leaderboards, they plan to use overall damage done and XP earned which are good/better leaderboard stats
  
• To successfully pursue these stats you have to be a team player, hence stat hound = team player
  
• Care bears produce rainbow colored slag (proved this through experimentation)

Now the OP:

While browsing the forums, I've noticed that there seems to be a significant number of MWO players who fear public stats, leaderboards, ladders, etc. They seem to think stat hound and team player are mutually exclusive concepts... that by introducing public stats there is no longer any incentive for team play. There also seems to be this care bear mentality that players shouldn't worry about their personal stats period.

I am a brawler, and the thrill of Mechwarrior for me is honing superior piloting skills: The ability to maximize DPS without overheating (or smartly overheating to shutdown to get in a final kill shot). Riding my throttle to maintain optimal position (staying behind a lumbering atlas... using a well timed engine stop to throw off an opponent's aim). And of course having precise aim myself, taking into account any lead a particular weapon needs.

And I haven't even mentioned the thought necessary to configure an optimal chasis/loadout...

K/D ratio is the essential metric that indicates where I'm at as a pilot. High K/D demonstrates my piloting skills are at a high level, low K/D tells me I need to make adjustments. As a brawler having high K/D is (surprise surprise) pretty challenging.

Considering the time, effort, and ability needed to be a superior pilot you're damn right I want to be able to compare myself to others. I am competitive, and I'm sorry but chasing the top of a public leaderboard motivates me to play and get better (and spend MC.. converting Mech XP to General XP is where I spend most of it).

Now, while K/D is an excellent metric for gauging personal ability, it's a terrible metric for leaderboards given things like kill stealing. Damage done on the other hand is a great leaderboard stat as it truly reflects your impact on the battlefield, while negating kill stealing.

If I have a high amount of damage done per match I am implicitly helping my team in a significant way. Also, I HAVE to be a team player and stick with the group to have both high K/D and high damage done: if I go rambo and run around in the open or into the middle of an enemy lance by myself... hello insta-dead.

Fortunately, this is exactly the direction the stats system in MWO is going:

Paul Inouye said:

The K/D ratio always steers the way people play. MWO is not your typical FPS (CoD, BF etc) where you are ranked based on your personal kill count.

In order to push gameplay to a more team oriented, strategic game style, XP and overall damage will be the ranking system for our leaderboards. Padding your personal Kill count will do nothing for you in the long run. Padding your XP gains will 1) rank you higher on the leaderboards and 2) greatly benefit your team. This game is not about personal gain but gains of the well rounded and thought out combatants which makes up for valued MechWarriors.

So for all you care bears who think it's "bad" to chase stats 1) tough, they're coming and 2) I look forward to turning you into rainbow colored slag, climbing a public leaderboard, and helping my team all at the same time.


Postscript:
All that talk about team play aside, I would LOVE to see Solaris VII style competition.. so I could compete in say the 1v1 hunchback ladder or some such. In that case it would be purely about individual play, but if it were siloed in such a way it would satisfy the cravings of tourney players such as myself without polluting the 'standard' team play mode.

Edited by SamizdatCowboy, 20 November 2012 - 11:54 AM.

[Obadiah333]

I'm not so sure it's the k/d ratio or having their (possibly) low stats publicized. It's the fear that the elitist jerks will take over and casual solo or casual group players won't have anyone to play. This fear is not irrational. It stems from the many other mmo games out there that have been damaged or ruined by this group of people. Unfortunately, MWO really caters to this type of peoples and I suspect that there will be a lot of merc. companies, clans, etc. that will end up with a k/d ratio to get in or play. If you want some examples of how this works, look up what gear check did to WoW, or go play rift.

"Your gear sucks! You can't run this dungeon with us, go F off noob."
"What's your dps? 1200? NOOB. You need at least 1500 to even THINK about entering this dungeon."

It's this mentality and the fear of being shunned and unable to play that players fear, not having their poor stats on display. Belive me, if stats tracking is introduced into MWO, you can bet that it will negatively impact a fair portion of the players in the community.

Quotes in the near future of this game:
"What weapons are you running on that Hunchie? Are you F'ing kidding me? That build sucks. You definately can't drop with us. Go L2P!"
"Your K/D ratio is what? 2:1. Hah, I can get that in a weaponless commando! Get lost, noob."

It will be the downfall of this game, rest assured. I say that because PGI is obviously catering more and more to the casual crowd, and once the players that love this game have left because of the sweeping changes to accomodate the casuals, all you will have left is the casuals. Once they go (and go they will, like locusts, to ruin other games), you will have no one left to play. It will be elitist jerks vs elitist jerks. I don't forsee that being a very long or profitable scenario.

[Weeble]

Wow, so much wrong in a single post. Postscript explains it, though. I'll leave it at that- the flaming will commence shortly.

[dario03]

I would say stats aren't bad for most people but the problem comes from the people that put to much concern into it.
Some annoying stuff:
-It's annoying if I go to help a ally just to have that ally run off at his first chance and leave me alone with 2 enemies (so he doesn't die and drop his K/D).
-A teammate really wants to raise his K/D so he keeps shooting a weaponless one-legged enemy when theres other enemies blasting the rest of the team.
-When teammates hold fire even when they have a clear shot so that they can time the kill shot (though that one is hard to tell)
-Teammates running off and powering down if the team is down by 2 or 3

Stats can't always tell a whole story. Yeah someone with a high K/D is probably pretty good but it can be hard to compare. Is the guy always pugging, always in pug+voice, always in a established clan with planned out mechs, a mix of those? Does he always do the same role, run the same mech, same loadout? Does the guy fight to the end, fight till the odds are really bad, or run off at the first sign of trouble?
Even if they do add more stats (which I would like) theres still always going to be problems.
-Average damage should definitely be in there but that doesn't tell us if your being effecient with your shots or just blasting every spot on the mech.
-Accuracy on each component could help with the last point but is someone with 70% CT/5% cockpit hits really more accurate than the guy with 60% CT/5% cockpit hits or is the 60% guy taking more weapons off of the enemy when the situation makes that the best option.
-Average XP is good but I think they will need to change how XP is earned. Why does the person that does the capping get more capping xp than the guys that are holding off the enemy team? I know you can get a lot of points for assists but I think killshots will generally get you more (I could be wrong on that though). How about more points for helping the team with things like Narc/tag. There was some other stuff but I can't remember right now.

I think they also need to seperate the stats per mech. I do way more damage and kills with my hunch but I can distract the enemy, cap, and cover more ground to help teammates in my commando.

[One Medic Army]

K:D ratio is meaningless in this game. Good players are typically those who stick with their team and focus fire, and when focusing fire who gets the kill is fairly random. I've lobbed a AC/20 shell 400m+ at an atlas and gotten a kill from taking just one shot at them, and I've stripped an atlas of both side torsos and left him a walking weaponless twig while I dealt with someone else who was still a threat. There's so much more to this game than who got the last shot.

Win:Loss means more, but it's skewed depending if someone PUGs or plays via team.

Personally I prefer average XP/match as a performance metric, but since there's a 200XP discrepancy between winning and loosing it's also slightly skewed.

[Teralitha]

Obadiah333, on 17 November 2012 - 11:24 PM, said:

I'm not so sure it's the k/d ratio or having their (possibly) low stats publicized. It's the fear that the elitist jerks will take over and casual solo or casual group players won't have anyone to play. This fear is not irrational. It stems from the many other mmo games out there that have been damaged or ruined by this group of people. Unfortunately, MWO really caters to this type of peoples and I suspect that there will be a lot of merc. companies, clans, etc. that will end up with a k/d ratio to get in or play. If you want some examples of how this works, look up what gear check did to WoW, or go play rift.

"Your gear sucks! You can't run this dungeon with us, go F off noob."
"What's your dps? 1200? NOOB. You need at least 1500 to even THINK about entering this dungeon."

It's this mentality and the fear of being shunned and unable to play that players fear, not having their poor stats on display. Belive me, if stats tracking is introduced into MWO, you can bet that it will negatively impact a fair portion of the players in the community.

Quotes in the near future of this game:
"What weapons are you running on that Hunchie? Are you F'ing kidding me? That build sucks. You definately can't drop with us. Go L2P!"
"Your K/D ratio is what? 2:1. Hah, I can get that in a weaponless commando! Get lost, noob."

It will be the downfall of this game, rest assured. I say that because PGI is obviously catering more and more to the casual crowd, and once the players that love this game have left because of the sweeping changes to accomodate the casuals, all you will have left is the casuals. Once they go (and go they will, like locusts, to ruin other games), you will have no one left to play. It will be elitist jerks vs elitist jerks. I don't forsee that being a very long or profitable scenario.

This scenario is not possible with MWO. For one... any weapon build can be made to work with skills. PUGs are randomly set, and you cant view teamates weapons.

As for unit requirements, thats entirely up to them and will not affect what happens to MWO.

One Medic Army, on 18 November 2012 - 12:50 AM, said:

K:D ratio is meaningless in this game. Good players are typically those who stick with their team and focus fire, and when focusing fire who gets the kill is fairly random. I've lobbed a AC/20 shell 400m+ at an atlas and gotten a kill from taking just one shot at them, and I've stripped an atlas of both side torsos and left him a walking weaponless twig while I dealt with someone else who was still a threat. There's so much more to this game than who got the last shot.

Win:Loss means more, but it's skewed depending if someone PUGs or plays via team.

Personally I prefer average XP/match as a performance metric, but since there's a 200XP discrepancy between winning and loosing it's also slightly skewed.

Then the only pure metric is damage dealt? As long as all weapons are equally balanced....

[One Medic Army]

Teralitha, on 18 November 2012 - 01:08 AM, said:

Then the only pure metric is damage dealt? As long as all weapons are equally balanced....

Like I said, I prefer XP/match, since it rewards those behaviors that the devs feel should be rewarded.
It'd be nice if the XP rewards got a bit more involved though.
Blow off a centurions left arm? 15xp. Blow off an Atlas-RS's arm with 2 large lasers? 15xp
Spot an enemy for LRM fire of one volley that doesn't even hit since you lost lock? 25xp
Spot an enemy for LRM fire that kills him? 25xp

Lots of the XP reward stuff is still in an early stage. It's not perfect, but no metric ever is.

As to the damage dealt, the higher your XP:damage the better you're generally doing at focusing your damage and being efficient.

Edited by One Medic Army, 18 November 2012 - 01:12 AM.

[Asmosis]

SamizdatCowboy, on 17 November 2012 - 11:06 PM, said:

While browsing the forums, I've noticed that there seems to be a significant number of MWO players who fear public stats, leaderboards, ladders, etc. They seem to think stat hound and team player are mutually exclusive concepts... that by introducing public stats there is no longer any incentive for team play. There also seems to be this care bear mentality that players shouldn't worry about their personal stats period.

I've been cored, had my arms blown off or leg blown out by "friendly" mechs after getting a kill due to kill stealing (wut?) so yes i do think this is a concern if MORE people start to care rather than just the die-hard clan players protecting their 5-1 ratios and such ( i know thats the reason because they said as much after doing it).

I've also had people quit seconds before getting killed to avoid that nasty "death count" since that means they have to kill another 5+ mechs without death otherwise. I can only imagine what it would be like if these stats actually meant something.

Edited by Asmosis, 18 November 2012 - 01:17 AM.

[Kobold]

Leaderboards keep players interested. Personally I am waiting for the Merc Corp leaderboards that will allow us to let units waggle their e-peens about (especially the smaller merc corps that may not be taking over planets, but just taking lots of contracts for houses).

I agree that K/D is not a good stat in a game like MWO. Someone who plays fire support for a good team may never get a death while racking up 1 kill per game, for a massively inflated ratio. Meanwhile the medium brawler on his team might go 3/1 every single round. Who is the better player, the 3 kills then dead HBK, or the 1/0 every round LRM boat?

I think MWO needs a stat kind of like QB Rating in the NFL, where they combine a number of factors to create a single number. It won't be perfect (just like passer rating inflates for QBs with high completion percentages on easy passes, punishes QBs who have low completion percentages but longer YPA, and doesn't take into account other skills like rushing), but it would be a good start.

This is similar to my unwavering desire for a BV analog.

The math doesn't need to be perfect, but having SOMETHING would be better than nothing.


VIVA LA BV.

[Asmosis]

On a positive note, they are rebalancing the rewards to be more team focused so if that works, and the KD ratio gets removed as a result it might promote better teamplay. It'll probably factor into phase 3 for matchmaking as well.

[JebusGeist]

Now, I'm not sure what I think I'm seeing is accurate as I've not bothered to record a sample group of 100 or so match results to figure out my XP average but, I think that XP average per match is taking in to account the XP boosts I get from premium time. (Can anyone accurately confirm or deny this?)
If this happens to be the case I have one thing to say: Inb4 the inevitable "premium time is p2w the ladder system" posts.
As long as I feel like I'm doing well and kicking *** I don't care if I end up in the top 10, top 100, top 1000 or even top 10000, and I do intend to play competitively.

Edited by JebusGeist, 18 November 2012 - 01:23 AM.

[Kobold]

Asmosis, on 18 November 2012 - 01:19 AM, said:

On a positive note, they are rebalancing the rewards to be more team focused so if that works, and the KD ratio gets removed as a result it might promote better teamplay. It'll probably factor into phase 3 for matchmaking as well.

I am against removing stats. Instead, we should be adding MOAR STATS.

I'd like to see the following, clearly listed:

• Damage per match (separated out by weapon, possibly normalized by "per shot" with the weapon)
  
• Accuracy (separated out by weapon)
  
• XP per match breakdown (average amount of spot XP, damage XP, kill XP, etc)
  
• Stats when using specific mechs/chassis
  
• Stats able to be separated by when pugging, when in a group of 2-4, or when in a "raid group"

MOAR STATS.

[Commodus]

Please! No official stats. They ruined so much games i played in past.
This e-peen comparision is so annoying and unnecessary.

I´m playing with the 12th Donegal Guards. We won 100% of the matches with 8 - premades against pugs and even 95% of the matches in clanwars.
And now with the smaller 4men/women groups we still win more than 75% of the matches (btw. a lot more fun and excitement).
How should we compare this to pug- players? And why should we do that?

Solaris arenas with 1 vs 1 or groupfight competitions are the place for ladders and e- sport. But that is future.

Edited by Commodus, 18 November 2012 - 01:26 AM.

[SamizdatCowboy]

One Medic Army, on 18 November 2012 - 01:12 AM, said:

Like I said, I prefer XP/match, since it rewards those behaviors that the devs feel should be rewarded.
It'd be nice if the XP rewards got a bit more involved though.
Blow off a centurions left arm? 15xp. Blow off an Atlas-RS's arm with 2 large lasers? 15xp
Spot an enemy for LRM fire of one volley that doesn't even hit since you lost lock? 25xp
Spot an enemy for LRM fire that kills him? 25xp

Lots of the XP reward stuff is still in an early stage. It's not perfect, but no metric ever is.

As to the damage dealt, the higher your XP:damage the better you're generally doing at focusing your damage and being efficient.

I think the XP stats should certainly be in there (and they will be), but they don't necessarily indicate pilot ability since XP encompasses a lot of non-combat stuff like spotting and capping. Important, interesting, reflects teamwork... but doesn't necessarily reflect how good a pilot someone is (sorry but sitting on a cap requires zero pilot ability).

A more interesting stat would be ratio of damage done per death. Again that's per death, not per match, which I think better reflects a player's efficiency... it retains the spirit and intention of the K/D stat without suffering its problems.

Edited by SamizdatCowboy, 18 November 2012 - 01:33 AM.

[Ghogiel]

AFK/suicide grinders will have the worst stats

DCing before death people will have the best stats.

[Kobold]

Commodus, on 18 November 2012 - 01:23 AM, said:

Please! No official stats. They ruined so much games i played in past.
This e-peen comparision is so annoying and unnecessary.

----------

I´m playing with the 12th Donegal Guards. We won 100% of the matches with 8 - premades against pugs and even 95% of the matches in clanwars.
And now with the smaller 4men/women groups we still win more than 75% of the matches (btw. a lot mor fun and excitement).
How should we compare this to pug- players? And why should we do that?

Part 1: I played Team Fortress Classic back from 1999-2005 or so, competitively in clan play. There was no stat tracking. Were people any less elitist? No, of course not. Instead they would brag about what clan they were in or what class guild they were in. You see a player with an XD or OD on the end of their name, and you knew they were a badass soldier. .dF for demomen, and so on. Stats didn't create elitism, it just gave people a different way to "prove" it.

Part 2: As I said above, I'd love to see stats broken down in more detail. Any sports fan knows how much fun it is to see stats like "Passer rating on 3rd and 7+" or "3rd down conversion percentage in the fourth quarter".

[IceSkraven]

Damage is not really the best indicator either(though I would agree it is better than K/D usually). High damage could be the result of poor aim requiring you to put more damage on target before it is eliminated. I would argue that individuals who routinely destroy mechs via headshot or coring through a single area of a mech to destroy it are far more skilled than someone who has to destroy almost every section of a mech before it goes down.

[FunkyFritter]

Once proper matchmaking is implemented ELO ratings will be a far better measure of skill than KDR. In the short term there's too much variance in games for stats to say anything significant about player skill.

[One Medic Army]

SamizdatCowboy, on 18 November 2012 - 01:27 AM, said:

I think the XP stats should certainly be in there (and they will be), but they don't necessarily indicate pilot ability since XP encompasses a lot of non-combat stuff like spotting and capping. Important, interesting, reflects teamwork... but doesn't necessarily reflect how good a pilot someone is (sorry but sitting on a cap requires zero pilot ability).

As they say in Battlefield3: "Do you PTFO?"
Playing the [expletive deleted] objective can be worthwhile, and a good player shouldn't just be good at combat. They should be good at knowing when combat is the answer and when it isn't.

Quote

A more interesting stat would be ratio of damage done per death. Note that I say per death and not per match, which I think better reflects a player's efficiency... it retains the spirit and intention of the K/D stat without suffering its problems.

That's certainly an interesting idea. I'd also like to see a Damage/kill or Damage/assist number. Or Kills assisted ratio (on average, how much of a damage to a killed target did you do). Definitely I'd like to see a kills/match or assists/match.
Of course no matter what metric we use there will be outliers. Like killing an Atlas-K through a rear side torso with only ~60 damage while someone else does a few hundred to their front armor without punching through.

I just happen to not like the K:D ratio much since actually killing an enemy mech is only one of many ways to be useful in this game.

And of course, before making stats go public: disconnecting before a death should still count as a death for you and a kill for your killer. This is essential.

[SamizdatCowboy]

IceSkraven, on 18 November 2012 - 01:32 AM, said:

Damage is not really the best indicator either(though I would agree it is better than K/D usually). High damage could be the result of poor aim requiring you to put more damage on target before it is eliminated. I would argue that individuals who routinely destroy mechs via headshot or coring through a single area of a mech to destroy it are far more skilled than someone who has to destroy almost every section of a mech before it goes down.

Well this really goes to the point that you need more, and more detailed stats, including things like number of headshots, accuracy, etc... with all those things laid out it's easier to understand/judge a player in aggregate. And there's no reason you couldn't have 1 leaderboard per stat.

But I still think damage done per death is a pretty good overall metric. Spray and pray pilots are probably also going to have more deaths, which would offset the extra damage they're doing...