51 replies to this topic

[Ted Wayz]

OP, you are over a year late on this suggestion. Maybe read up on some of those suggestions before dropping the mic?

[adamts01]

Ted Wayz, on 25 July 2016 - 02:36 PM, said:

OP, you are over a year late on this suggestion. Maybe read up on some of those suggestions before dropping the mic?

We didn't have leaderboards a year ago...... Did you have your coffee this morning?

[White Bear 84]

Yeah right, so you would place a mech pilot that gets carried by their team in an upper tier just cause they have a stupidly high W/L ratio... ...sounds more broken than ELO

lol.

[Ted Wayz]

adamts01, on 25 July 2016 - 04:26 PM, said:

We didn't have leaderboards a year ago...... Did you have your coffee this morning?

We have had leaderboards for a looong time. By chassis in weekend comps. You play this game or just post on the forums?

[adamts01]

White Bear 84, on 25 July 2016 - 05:45 PM, said:

Yeah right, so you would place a mech pilot that gets carried by their team in an upper tier just cause they have a stupidly high W/L ratio... ...sounds more broken than ELO

lol.

I'm a decent player and I have an overall 1.35 W/L. You're consistently better than me, Which reflects in your 1.45 W/L. With a large enough sample of games, I think W/L is the most accurate stat we can use. Damage is easy to farm from the back of your team, even if you don't really contribute. Better players ignore sticks and focus on the greatest threat, while bads get tunnel vision trying to increase their K/D. Kmdd is good, but half the important jobs that are out there don't rack up damage (being the tip of the spear, providing ECM instead of getting in a good firing position, running back to stop a cap, suiciding to stall a cap....). When it comes down to it, W/L is all that matters, and I don't believe a program will ever be able to dissect a match and determine who pulled their weight and who didn't.

Ted Wayz, on 25 July 2016 - 06:15 PM, said:

We have had leaderboards for a looong time. By chassis in weekend comps. You play this game or just post on the forums?

Mostly talk **** on the forum. Really though, this is the first I've seen of the leaderboards. Was it just a weekend comp thing? Or was it recently brought to the limelight because now it's public? I don't do comp stuff so if that's where it was then I wouldn't have had a clue about it.

[White Bear 84]

adamts01, on 25 July 2016 - 08:19 PM, said:

I'm a decent player and I have an overall 1.35 W/L. You're consistently better than me, Which reflects in your 1.45 W/L.

TY for the compliment there,

What I am thinking though, is say I have a W/L of 1.45 playing pugs and another player has a W/L of 35 as a group.

Who is the better player & who would be ranked higher?

I mean that W/L of 35 at most suggests that the players TEAM is good, but not necessarily the pilot (although one would assume with the support of other stats the other pilot is good). At this point you need further indicators to identify how a player has performed in a group, but then the same would/should apply to pug pilots too.

Of course, damage can be farmed.. ..so what else could be reliably used? Do we need new stat sets for specific roles e.g. ratio of damage tanked/given as assault, number of spots if a light, flanks if in a medium etc etc..

Edited by White Bear 84, 25 July 2016 - 08:52 PM.

[InspectorG]

EZ-est solution:



JUST MAKE IT RANDOM AND LET NUFFLE DECIDE!!!!

[adamts01]

White Bear 84, on 25 July 2016 - 08:51 PM, said:

TY for the compliment there,

What I am thinking though, is say I have a W/L of 1.45 playing pugs and another player has a W/L of 35 as a group.

Who is the better player & who would be ranked higher?

I mean that W/L of 35 at most suggests that the players TEAM is good, but not necessarily the pilot (although one would assume with the support of other stats the other pilot is good). At this point you need further indicators to identify how a player has performed in a group, but then the same would/should apply to pug pilots too.

Of course, damage can be farmed.. ..so what else could be reliably used? Do we need new stat sets for specific roles e.g. ratio of damage tanked/given as assault, number of spots if a light, flanks if in a medium etc etc..

Maybe only base your rank off of solo que, it's not like there's much match making going on in group que anyway. And there's no way to keep people from being carried and having their stats inflated from a good group. Really, there's no perfection to be found, and I think using W/L is the lesser of the evils. There's all these little formulas people think will save the world, but the end result is winning, so just skip to that.

The other thing that I think is critical is having the community split between 3 tiers, everyone plays their own tier, and in order to move up you have to be better than someone else and kick him down. None of this XP bar nonsense. I really think 3 tiers is the magic number. We already have 3 different kinds of games, 1-3, 2-4 and 3-5. They're such a mess, why mix everyone up so badly? This way middle of the road players won't have to fight super comp players or absolute noobs. I don't think wait times would be any longer, but even if they were, I think the closer skill levels would make it worth it.

[NoiseCrypt]

kf envy, on 25 July 2016 - 12:00 PM, said:

just

Don't be such a *****(cat)

Edited by NoiseCrypt, 25 July 2016 - 10:19 PM.

[NoiseCrypt]

Ted Wayz, on 25 July 2016 - 02:36 PM, said:

OP, you are over a year late on this suggestion. Maybe read up on some of those suggestions before dropping the mic?

Maybe read the entire tread instead of just responding to the thread title with a one liner... ?

Viktor Drake, on 25 July 2016 - 10:16 AM, said:

There is a fatal flaw in using the leaderboards for MM purposes. The leaderboard scoring doesn't mean your the best player in the game, rather it means that your the best player in your Tier.

Keep in mind that all the players are already broken up into tiers for MM purposes so all I have to do at Tier 3, is to outperform all the other players how are also at Tier 3 to get a monster leaderboard score. Hell you could theoretically have a bottom Tier player win a leaderboard challenge just because is just good enough to dominate the lowest tier but not good enough to advance a tier.

Also keep in mind that the higher you go in tiers, the more and more skilled players become and as players become more skilled, the ability to stand out gets harder and harder. I wouldn't be surprised if very few of the top leaderboard players are actually playing at Tier 1.

What part of "remove PSR/TIERS" is it that you don't understand ?

[Dago Red]

InspectorG, on 25 July 2016 - 09:04 PM, said:

EZ-est solution:



JUST MAKE IT RANDOM AND LET NUFFLE DECIDE!!!!

If sync dropping could be completely eliminated I'd actually be fine with that in quick play solo.

Already got an altar set up and everything.

[Lukoi Banacek]

If winnowing down PSR to one statistic was your goal, I could see it being WLR (although a less damage dependent match score would be more holistic), but that over simplification is exactly the problem.

The benefit of the leaderboard, and in fact of expanding the leaderboard, is that more data allows for better understanding of a player's skill level which in turn should impact MM.

You cannot farm every facet of this game to look better if there are enough legitimate metrics. Which means those people who are truly better players will represent across multiple facets or stats and be tiered appropriately. Right now PSR is closer to an xp bar (although some players are entrenched enough in their playstyle to stay rooted in a lower tier), and match score is heavily based on damage.

Weighted differently, with broader considerations, the matchscore could more accurately predict player performance for MM than simple WLR ever could. But more importantly, factoring MORE granular examples of performce paints the best picture.

Instead of arguing over which single stat to base MM on, we debate which should hold the most weight in a reasonable formula.