42 replies to this topic

[Mercules]

Bhael Fire, on 21 July 2014 - 12:28 PM, said:

What is better; "holding up" 4 mechs or killing 4 mechs so your team could win? In the solo queue, killing those 4 mechs is much more valuable.

Again, I need to reiterate; teamwork is a good thing. Nobody is arguing that. But when it comes to measuring player skill on what makes for a valuable player in a PUG match, the players with the higher KDRs are going to carry the rest of the team.

It's not black and white, especially since you can easily rack up 4 kills by sitting back and waiting till your teammates have all but killed a mech and then putting the final shot into it. I'd rather have the team player with low KDR than the 4 kill guy. Ideally I'd rather have the team where everyone on it had 1 kill and 8-11 assists at the end.

KDR is NOT a good measure of skill. Part of that is because it can be so easily gamed. The other part is there is no correlation between getting the last point of damage on a mech and being a skilled player. You can pretend there is, but there isn't.

[Roadkill]

Bhael Fire, on 21 July 2014 - 12:28 PM, said:

But when it comes to measuring player skill on what makes for a valuable player in a PUG match, the players with the higher KDRs are going to carry the rest of the team.

ROFLMAO.

You do understand that "solo queue" doesn't mean "solo combat," right?

I can just as easily argue that the players with the higher KDRs are the ones sponging off of their teammates, using them as meat shields, and stealing their kills.

KDR means very little other than you got the last shot more often than someone else. But the last shot is hardly meaningful in a team game where probably 5-6 people damaged each Mech before it was destroyed.

Bhael Fire said:

If you want to win, yes. I suppose it comes down to what's important to you.

That right there boggles the mind.

What's the point of a team combat game if you're not trying to win?

KDR is little more than an e-peen measuring stick.

Which of these two solo-only players would you rather have on your side in a Solo queue match:

1) KDR 3.5, W/L 0.5, avg damage 650
2) KDR 0.9, W/L 3.5, avg damage 150

If you answer #1, you don't understand what those stats mean.

[Bhael Fire]

Roadkill, on 21 July 2014 - 12:47 PM, said:

KDR means very little other than you got the last shot more often than someone else.

And looking at a small sample, that's indeed all it would mean. However, over the course of 1000+ matches, if you still maintained a consistently high KDR, that means you're either freakishly lucky or you have talent for killing. There's an odd misconception that it's easy to "game" KDR...or that even if you could consistently game your KDR that that's a bad thing in a game that is essentially team death match.

Roadkill, on 21 July 2014 - 12:47 PM, said:

Which of these two solo-only players would you rather have on your side in a Solo queue match:

1) KDR 3.5, W/L 0.5, avg damage 650
2) KDR 0.9, W/L 3.5, avg damage 150

In the solo queue? I could care less which one of those players were on my team if I were in the solo queue — mainly because the one with the high KDR has an unbelievably low W/L which kinda makes it a wash. But I'd be inclined to prefer the player with the higher KDR, if not for the simple fact that I could expect them to kill around 3 or 4 enemies before dying. In a solo match, that's a more reliable asset than hoping the other players work with player 2's obvious talent for teamwork and command.

However, if I were in a group, I'd take the one with the 3.5 W/L.

[Roadkill]

Bhael Fire, on 21 July 2014 - 01:37 PM, said:

There's an odd misconception that it's easy to "game" KDR... or that even if you could consistently game your KDR that that's a bad thing in a game that is essentially team death match.

It's not a misconception at all, because I used to do it all the time. It's easy to do... I just don't enjoy playing that way so I don't do it anymore.

And doing it is bad for your team's results. In previous versions of MechWarrior, though, wins and losses didn't really matter in casual play so gaming my KDR was the best way to climb the ladders that were in place. I was more than happy to help my team lose as long as I was able to bolster my KDR. And I wasn't at all the only person doing it.

The difference between MWO and MW4 is that in MWO you need a group to farm pugs. In MW4 I didn't need a group, and I could farm the pugs on both teams at the same time.

Bhael Fire, on 21 July 2014 - 01:37 PM, said:

But I'd be inclined to prefer the player with the higher KDR, if not for the simple fact that I could expect them to kill around 3 or 4 enemies before dying. In a solo match, that's a more reliable asset than hoping the other players work with player 2's obvious talent for teamwork and command.

As expected, you really don't understand what those stats mean. I used crazy numbers for a reason.

#1 is a loser. Sure, he gets a lot of kills, but it's clearly to the detriment of his team. He's no good to you if he gets 3.5 kills on average if that means your team is going to lose 2/3 of its games. His unbelievably low W/L ratio is the point. He's terribad for the team.

#2 is a winner. No, he doesn't get very many kills, but his teams win. Constantly. He's either an amazing scout - and simultaneously good at communicating his findings - or he's an amazing commander who's able to consistently and reliably get a random bunch of pugs to follow his lead. A 3.5 W/L ratio is insanely high... I'd wager that it's virtually impossible given the way the game's matchmaker is set up. The guy wins 78% of his games... in the solo queue. That's crazy high talented.

There's no question at all that you should want #2 on your team in the solo queue. Stats clearly can't capture this guy's skill very well, but you absolutely want him on your team.

The fact that you'd be inclined to prefer #1 explains why you think KDR is a leading indicator of player skill - you simply don't understand what the various stats actually represent.

[Joseph Mallan]

Bhael Fire, on 21 July 2014 - 12:28 PM, said:

What is better; "holding up" 4 mechs or killing 4 mechs so your team could win? In the solo queue, killing those 4 mechs is much more valuable.

Again, I need to reiterate; teamwork is a good thing. Nobody is arguing that. But when it comes to measuring player skill on what makes for a valuable player in a PUG match, the players with the higher KDRs are going to carry the rest of the team.

Or they will have 3-4 kills and still lose the game. Don't get me wrong, in most PUG matches the group doesn't act like a team and dies accordingly, when the group becomes a team then you see a good game... maybe not close, but a good game anyway.

Road... You really should have used A and B instead of 1 and 2. Cause I don't want #2 on my team.

Edited by Joseph Mallan, 21 July 2014 - 03:08 PM.

[Roadkill]

LOL Joe you crack me up.

[Bhael Fire]

Roadkill, on 21 July 2014 - 02:58 PM, said:

The fact that you'd be inclined to prefer #1 explains why you think KDR is a leading indicator of player skill - you simply don't understand what the various stats actually represent.

I absolutely understand what those stats mean and the implications they have in solo and group queue.

Player 2 has an insanely high W/L ratio but that's going to be practically useless in the solo queue if nobody listens to him. I can guarantee that he didn't get that high of a ratio playing in PUGs. This player would fare much better in the group queue because that environment would work in tandem with his skill set, unlike the solo queue where player willingness to work as a team is not always reliable.

Player 1 has a very good KDR and can be expected to get plenty of kills, so even with his extremely (almost supernaturally low) W/L ratio, he's probably going to do better in the solo queue than Player 2 will.

[SolasTau]

I think this is an interesting discussion. As a new player, I'm currently just trying to get damage, and it's been my primary benchmark of if I'm doing alright or not. But there are ways to game the system. I was watching one of TheB33f's videos (The Dapper Jager) which basically waits for its friends to strip armor off the mech then goes in for all the crits. That would greatly artificially inflate the xp values per match (B33f even says as much in the video, claiming it to be a great way to let the xp and cbills roll in).

It's interesting. Another Mech Warrior and I have been having an extended private chat about what 'skill' even IS in this game this week.

[Bhael Fire]

Joseph Mallan, on 21 July 2014 - 03:05 PM, said:

Or they will have 3-4 kills and still lose the game. Don't get me wrong, in most PUG matches the group doesn't act like a team and dies accordingly, when the group becomes a team then you see a good game... maybe not close, but a good game anyway.

We are on the same page, Joe. I agree, teamwork is essential and teams that use team work will tend win more often.

I was just saying that players with high KDRs will tend to do better in the solo queue and that players with high W/L ratios tend to do better in the group queue...but that both stats are needed to give an idea of the player's overall skill.

[Roadkill]

Bhael Fire, on 21 July 2014 - 03:44 PM, said:

I can guarantee that he didn't get that high of a ratio playing in PUGs.

No, you can't, because I specifically stated that both players were solo only players.

In other words, player #2 is apparently so good at leading that he's capable of doing it even in the solo queue.

And player #1 is apparently such a maverick that despite his great KDR, he manages to help his team lose in 2/3 of his games.

Translation: KDR isn't a good indicator of skill. You might be able to argue that it is usually an attribute of a skilled player, but it isn't an indicator. There's no causality in that direction, as demonstrated by my example. You're confusing correlation and causation.

[Asmudius Heng]

There really is no way to accurately gauge the quality of a player using statistics.

There are so many things a player does that cannot be linked to a stat that can lead to better performance.

Win/Loss ratio is helpful but wen you team mates are so variable its hard to pin that one down too - then throw in ELO trying to balance your W/L by pitting you against better opponents or making you try to carry a team full of bads.

Anything used is an extremely rough measure

[DAYLEET]

Joseph Mallan, on 21 July 2014 - 11:21 AM, said:

As rarely as I get to Cap in an Atlas I had no idea there was no XP for Capping. It seems rather silly that there is no XP reward for performing the main objective for the conquest mission.

You get a 1time cap xp IF you cap it from zero to full and IF you win the game, but if you cap more you dont get more xp and i think its 75xp.

[Prezimonto]

I will add that the XP measure can be very heavily skewed by the 2x win of the day.

Casual players, or old timers who don't play for hours, will often go through and win 2x matches and that's all.

It's still probably a better stat than damage dealt, but it's still not all that effective.

What irritates me is the "match score" in the game that directly linked to damage score.

[Artgathan]

OP here with a quick point: I didn't mean to seem as though I was suggesting that Avg XP/Match is the best indicator of skill. My point was that it is better than using Avg Damage/Match.

I can have a high avg damage/match and lose all the time, but my avg xp/match will suffer.

[Pyrrho]

I don't think someone with a high K/DR but low W/LR can be said to be "helping his team lose the game" -- you've given us nothing but the end result, and nothing about what the other 11 members of the team were doing. They could have all been Disconnects in every game he has played -- which would make his high K/DR thoroughly impressive.

But, also, I don't think KD/R and W/LR are important when thinking about what makes a player "skilled" or a valuable asset to his team.

[Varik Ronain]

So much about this topic can be very subjective. All this talk about what measures skill and if XP is a gauge is at least a different conversation. I can link so many screenshots of final match stats but I will go with my best match yesterday and you can look at the XP, c-bills and base stats. I know none of my stats are amazing on their own but the ability to do a bit of everything and contribute to my teams victory at least makes me feel good.







For me the measure of this game is having fun, being unique in builds while still being able to contribute and the community of players you spend your time with. I do not consider myself an elite pilot by any stretch of the imagination as I drop with several who are much better than myself and I have been on the receiving end of a pummeling by the top tier units out there.


Bonus points for anyone who can guess the mech and build

I can get a copy of the end screen for a teammate but it was a 12-9 match that came down to the wire.

Edited by Varik Ronain, 21 July 2014 - 05:25 PM.

[Joseph Mallan]

Bhael Fire, on 21 July 2014 - 03:54 PM, said:

We are on the same page, Joe. I agree, teamwork is essential and teams that use team work will tend win more often.

I was just saying that players with high KDRs will tend to do better in the solo queue and that players with high W/L ratios tend to do better in the group queue...but that both stats are needed to give an idea of the player's overall skill.

It depends on what you want out of the Que I guess. If you are a High KDR player its fine, But When I wrack up 7 assists 0 Kills and 450+ damage I did the team proud by making sure the enemy dies and YOU can keep killing. My KDR looks like crap, but Man all those wins look good.

[Bhael Fire]

Roadkill, on 21 July 2014 - 03:55 PM, said:

No, you can't, because I specifically stated that both players were solo only players.

Ok...let's ignore the fact that it's practically impossible to have a W/L ratio that high (or low) playing only in PUGs and pretend that Player 2 has somehow managed to get a consistent average W/L of 3.5 over the course of hundreds of matches.

Then yes, they would probably do well in the solo queue using their mind control powers to make complete strangers follow their every command and lead them onto victory.

However, If the Player 1 was in the same match on the same team, he'd probably get a higher score than Player 2 and get more kills even though his W/L ratio was so abysmal. In other words, it wouldn't matter if their team won or lost; Player 1 would still get the better score than Player 2.

Roadkill, on 21 July 2014 - 03:55 PM, said:

You're confusing correlation and causation.

I'm not confusing anything. I am merely stating a fact: If a player manages to get a consistently high KDR, that is indeed a good indicator that they are a skilled player — Unskilled players do NOT consistently maintain high KDR scores, even if they try to "game" the system. It just doesn't happen.

That said, just because a player's KDR is average, it doesn't make them a bad player...especially if they have a high W/L ratio, it just means the player with the higher KDR will tend to do better than them in solo matches.

Teamwork is the key factor in winning any match. It will allow players to essentially "combine" their killing efficiency to maximize results. However, all the leadership and command skills in the world won't help if nobody is listening — which is almost all the time in the solo queue. That's why those with high KDRs tend to do better than those with lower KDRs in the solo queue.

All this silly hypothetical bickering has made me thirsty...time to get drunk.

[Joseph Mallan]

Bhael Fire, on 21 July 2014 - 05:42 PM, said:

All this silly hypothetical bickering has made me thirsty...time to get drunk.

I'm at work, so drink 5-8 for me!

[Roadkill]

Bhael Fire, on 21 July 2014 - 05:42 PM, said:

Player 1 would still get the better score than Player 2.

But that doesn't necessarily make Player 1 a better player. "Better" is subjective... you clearly define "better" to mean a higher KDR. I define it as someone who helps our team win.

KDR doesn't indicate that. It might correlate, but it isn't indicative.

Quote

I'm not confusing anything. I am merely stating a fact: If a player manages to get a consistently high KDR, that is indeed a good indicator that they are a skilled player. Unskilled players do NOT consistently maintain high KDR scores, even if they try to game the system. It just doesn't happen.

That's not a fact, that's your opinion. And it's a non-sequitur anyway as that's not what either of us said.

You said that KDR is the best indicator of skill. It isn't, as has been demonstrated. We're talking about all-around piloting skill, remember, and there are many ways for a pilot to be skilled at MWO that would not be reflected in their KDR. That's why it's of limited usefulness for anything other than e-peen bragging rights.

Quote

That said, just because a player's KDR is average, it doesn't make them a bad player...especially if they have a high W/L ratio, it just means the player with the higher KDR will tend to do better than them in solo matches.

A nit-pick, but we don't have solo matches unless you arrange them via private games. We have a solo queue, but the end result is a team game. Which is why...

Quote

Teamwork is the key factor in winning any match. It will allow players to essentially combine their killing efficiency to maximize results.

Exactly. Teamwork is OP. And it isn't reflected at all in KDR.

A solo queue player's W/L record is a far better indicator of their ability to help their team than their KDR. Your KDR is irrelevant if your team loses. But winning in the solo queue consistently means you're skilled at MWO, regardless of your other stats. You can't win consistently unless you're skilled at something.

Quote

All this silly hypothetical bickering has made me thirsty...time to get drunk.

Waaay ahead of you there.