70 replies to this topic

[MischiefSC]

Slambot, on 30 December 2017 - 12:25 PM, said:

How about Average match score x WLR? This might help a bit given the proponents of both arguments and the validity of their points. Would also like to see match score revised to emphasize KMDD over kill shots.

The only relation of average match score to w/l is that winning more raises your match score. So you're double dipping that value.

Otherwise it doesn't matter what your match score is if you're consistently winning drops -

Because the matchmaker isn't trying to predict who's going to get the best score, it's trying to predict who's going to win.

[Spr1ggan]

Iain Sumner, on 13 December 2017 - 09:09 AM, said:

It's important to some people...I don't even know my stats nor do I care about them. But sometimes you get the occasional "I have 1.2 moron" in game chat. People focussing their competence based on vague stats and brag about their K/D or W/L, stats that were "invented" in games like Doom or Quake back then and had a real and important value behind them. In MWO, it's just an old and outdated metric.

Why would anyone brag about a 1.2 ratio? That's pretty poor in this game.

[Tarogato]

Spr1ggan, on 30 December 2017 - 12:59 PM, said:

Why would anyone brag about a 1.2 ratio? That's pretty poor in this game.

I was really proud about my 1.2 for a long time. I played 900 games in a Jenner with Narc when I first started this game. Once I grew out of that and started playing real builds, I took note at my WLR and KDR, which were always identical. 1.11. Then 1.12. Then 1.13.... eventually 1.15.... and the month I reached 1.20 I was like "WOW I MADE IT". It was a lot of hard work. Even now that I'm somewhere around the top 2% or 3% of players, I still occasionally get seasons around 1.2 WLR when I'm leveling up particularly lousy mechs that can't carry.

Only the top 25% of players achieve 1.20 WLR. Similar goes for KDR, the top 25% is something like 1.23. So it's actually better than you think. Or ... people as a whole are worse at this game than you think.

[Spr1ggan]

Tarogato, on 30 December 2017 - 02:59 PM, said:

I was really proud about my 1.2 for a long time. I played 900 games in a Jenner with Narc when I first started this game. Once I grew out of that and started playing real builds, I took note at my WLR and KDR, which were always identical. 1.11. Then 1.12. Then 1.13.... eventually 1.15.... and the month I reached 1.20 I was like "WOW I MADE IT". It was a lot of hard work. Even now that I'm somewhere around the top 2% or 3% of players, I still occasionally get seasons around 1.2 WLR when I'm leveling up particularly lousy mechs that can't carry.

Only the top 25% of players achieve 1.20 WLR. Similar goes for KDR, the top 25% is something like 1.23. So it's actually better than you think. Or ... people as a whole are worse at this game than you think.

I was the opposite, i always hated my stats being that low. I didn't even think of my stats even being decent until i got my account k/d to 3.

[Nightbird]

GQ or SQ?

That's my one complaint with stats, it should be separated so as to be comparable. SQ is your personal skill, GQ is your team's skill.

[Tarogato]

Spr1ggan, on 30 December 2017 - 03:02 PM, said:

I was the opposite, i always hated my stats being that low. I didn't even think of my stats even being decent until i got my account k/d to 3.

I'm still below 1.7 KDR. (edit, err, well, overall. If I just take my past 17 months, I'm over 2.0 pretty consistently)

The problem is... group queue inflates your WLR and KDR by stupid amounts. I'm like 90% solo queue. I even asked PGI to look up in their database and tell me what percentage of my games are solo queue and group queue, they refused to tell me. =[

Edited by Tarogato, 30 December 2017 - 03:10 PM.

[Spr1ggan]

Tarogato, on 30 December 2017 - 03:08 PM, said:

I'm still below 1.7 KDR.

The problem is... group queue inflates your WLR and KDR by stupid amounts. I'm like 90% solo queue. I even asked PGI to look up in their database and tell me what percentage of my games are solo queue and group queue, they refused to tell me. =[

Yeah i pretty much play solo queue exclusively. group queue is pretty much hold w and crush all enemies in 2 mins most of the time. Was one of the reasons why old Lords started only doing group queue in 4 mans.

[Tarogato]

Spr1ggan, on 30 December 2017 - 03:12 PM, said:

Yeah i pretty much play solo queue exclusively. group queue is pretty much hold w and crush all enemies in 2 mins most of the time. Was one of the reasons why old Lords started only doing group queue in 4 mans.

Wait, so this is mostly or all solo queue? Which seasons in particular might have been 100% solo? If any of them are, then the gap between me and you is substantially larger than I thought... as if it wasn't obvious enough from the Cougar leaderboard. =3

[Spr1ggan]

Tarogato, on 30 December 2017 - 03:14 PM, said:

Wait, so this is mostly or all solo queue? Which seasons in particular might have been 100% solo? If any of them are, then the gap between me and you is substantially larger than I thought... as if it wasn't obvious enough from the Cougar leaderboard. =3

The only time i played a lot in groups was probs when in was in AS which was before leaderboards. When i was in Lords old+new and Osiriz i was probably 80% solo queue, possibly higher. I wouldn't say i'm even that good, i do well in solo queue for the most part but in comp i always felt out of my depth (In Lords+Oz mostly). I think that's cos i'm not really a numbers guy or super cerebral player, i just go on instinct. That being said i was on AS's RHoD EU winning team.

*Edit* One way to tell if someone's doing a lot of group is the matchscore in relation to their w/l+k/d and weight class. If i'd been doing lots of group in my lights, i would probably have a matchscore around 300 instead of around 500.

Edited by Spr1ggan, 31 December 2017 - 03:28 PM.

[Brain Cancer]

Tarogato, on 30 December 2017 - 03:08 PM, said:

I'm still below 1.7 KDR. (edit, err, well, overall. If I just take my past 17 months, I'm over 2.0 pretty consistently)

The problem is... group queue inflates your WLR and KDR by stupid amounts. I'm like 90% solo queue. I even asked PGI to look up in their database and tell me what percentage of my games are solo queue and group queue, they refused to tell me. =[

This. Group queue with a good group will end up giving you better stats, simply because an organized group is greater than the sum of it's parts.

The two queues should never share stats.

[Shred Jesse]

Would it be possible to get an API where we can query statistics over time on our own?

Would enable people to setup all of these possible combinations and potentially make them available to others.