284 replies to this topic

[Zergling]

Duke Nedo, on 25 November 2016 - 05:55 AM, said:

Aye, I am not saying it's succeeds, I just believe it tries... In prime time that works quite OK in a way, there are enough high tier players in every match to average the huge skill disparity in T1.

Problem is when there are fewer players on and it starts to wildly mix tiers to fill up the matches and (if it tries to match PSR between teams) you are matched only against one or a few potentially very good T1 players on the other team... then you'd better carry your load and carry hard... at least that's what I tell myself when I get a loss streak every time I try to basic a new mediocre mech and am in need of being carried.

I dunno, even when I'm having a loss streak I really don't see that many of the same players. One or two sometimes, but not all the time.

Stefka Kerensky, on 25 November 2016 - 05:57 AM, said:

Thanks.
I'm also interest in understanding if we really have a MM
If you have time, could you do the same evaluation, but taking into account w/l and k/d only?

I'm asking this because, imo, it's a very bad designed MM, if it is based on PSR (always up, never down) which is calculated in a bad match score system (which is, imo, irrilevant)

Sure, here it is with overall W/L in Season 5 per player:


That gives an average W/L of 1.0575 for my team, versus 1.0633 for the enemy team.


And here's another assessment of each player's skill level, this time based on their Season 5 leaderboard stats in the weight class they were using that battle, with a much stronger emphasis on W/L (which dropped my rating of myself... and tbh it should probably be lower; I don't think I'm a great player, I just got lucky with lights this Season):



My team:
2x upper 10%
1x upper 25% to 10%
1x upper 25%
2x above average
2x average
2x below average
1x potato


Enemy team:
2x upper 5%
1x upper 10%
2x above average
3x average
2x below average
1x potato


Removing those players of similar skill level that cancel each other, it ends up with this for my team:
1x upper 10%
1x upper 25% to 10%
1x upper 25%

Versus enemy team:
2x upper 5%
1x average

I'm not as sure, but I still suspect the advantage was for the enemy team.

What probably won the game for my team was eliminating the high skill player in the Viper early, when he made the unusually bad mistake of going off alone to cap a point that my team had several mechs nearby.

With him removed from the battle, my team had the skill advantage.

Edited by Zergling, 25 November 2016 - 06:55 AM.

[STEF_]

Duke Nedo, on 25 November 2016 - 06:23 AM, said:

May I ask how much you play in the group queue? I am curious how good w/l ration you can have in the long run as a 100% solo player.

I play about 95% in the solo queue mixing wildly between good mechs and underdogs, and when we do group up it's mostly just to yolo around in mixed builds drinking whiskey without much drop commanding going on. I rank like 12500 in w/l ratio but ranked 2994 for average score so it's a bit of a disconnect there.

Curious what makes that disconnect. Lack of pug-bossing? Too few drops in organized groups? My best meta-mechs have 1.5-2.0 w/l ratio but the sample is so small that doesn't have to mean anything...

Can you maintain a 2.0+ w/l ratio only dropping solo?

It depends on periods: if I want to level mechs I prefer solo ('cause pugs can feed me ), if I want to have fun with friends (then group que), If I have a bad mood (so soloing).
It's impossible for me to have real data about solo and group.
Anyway, I really don't like to drop with large group, my max group is 4-5.

From my experience, there is no MM whatsoever in group que, and I'm not so brilliant pilot. So things are up and down depending on who I found as enemies. Totally random. And you can face uber teams while you have potatoes in your team, OR also you can drop with your lance along good pilots, and face potatoes only.
You can rely on teamwork...BUT ENEMIES TOO!


I don't have true datas but there are mechs I used in solo que only.
But it depends on the mech: my "solo que only" phoenix hawks have around 1.0.
My solo que only Grasshopper 5P has 1.64
But the Banshee 3M has 2.40 w/l : it was a beast back in the days.... all those wubs!!!

[meteorol]

Duke Nedo, on 25 November 2016 - 06:23 AM, said:

Can you maintain a 2.0+ w/l ratio only dropping solo?

Kaffe got a ~ 4 w/l on a ~ 10 k/d on ~ 1k average dmg by using nothing but the KDK-3. Only soloqueue (by his own claims).

[STEF_]

Zergling, on 25 November 2016 - 06:54 AM, said:

I dunno, even when I'm having a loss streak I really don't see that many of the same players. One or two sometimes, but not all the time.




Sure, here it is with overall W/L in Season 5 per player:


That gives an average W/L of 1.0575 for my team, versus 1.0633 for the enemy team.


And here's another assessment of each player's skill level, this time based on their Season 5 leaderboard stats in the weight class they were using that battle, with a much stronger emphasis on W/L (which dropped my rating of myself... and tbh it should probably be lower; I don't think I'm a great player, I just got lucky with lights this Season):



My team:
2x upper 10%
1x upper 25% to 10%
1x upper 25%
2x above average
2x average
2x below average
1x potato


Enemy team:
2x upper 5%
1x upper 10%
2x above average
3x average
2x below average
1x potato


Removing those players of similar skill level that cancel each other, it ends up with this for my team:
1x upper 10%
1x upper 25% to 10%
1x upper 25%

Versus enemy team:
2x upper 5%
1x average

I'm not as sure, but I still suspect the advantage was for the enemy team.

What probably won the game for my team was eliminating the high skill player in the Viper early, when he made the unusually bad mistake of going off alone to cap a point that my team had several mechs nearby.

With him removed from the battle, my team had the skill advantage.

Thanks again.
Well, from what I can see the values were more or less better than I was thinking.
But, of course, I'm seeing the MM doesn't take into account classes and who is piloting them.
With this, things are not so balanced: take a look about who is piloting good mechs and who bad ones, relative on thier statistic.
For instance, charlie lance on good mechs with good tonnage was piloted by bad players; the upper % was piloting a viper.
The winning team has bravo lance with good players and also good mechs.
No idea how it could be possible to balance this, tho.....

[Wil McCullough]

NoiseCrypt, on 25 November 2016 - 06:18 AM, said:

The beautiful paradox of a guy wading in fallacies inorder to win his own version of the discussion. You are a true case study

I hope that you one day can move on from this methode

the irony in this post is so delicious. if only you could see it.

[MW222]

NoiseCrypt, on 24 November 2016 - 01:13 AM, said:

Unless you post all of your mechstats i call poppycock

Respects but we do not call "poppycock" in MWO the accepted phase is "Shenanigans!" or so I have been told.

[Davegt27]

I got zero damage in this match lol

[Snazzy Dragon]

am I good yet Dad
probably not, need 220 more damage than that

Edited by Snazzy Dragon, 26 November 2016 - 01:24 AM.

[meteorol]

Snazzy Dragon, on 26 November 2016 - 01:22 AM, said:

am I good yet Dad
probably not, need 220 more damage than that

Dude... Posting screenshots of single matches still doesn't veil your average stats over 1500 matches everyone can look up. You could post 20 screenshots of 1k+ damage here, people could still look up your average stats and guess how much damage you deal on average, seeing those high scoring matches are either seldom outliers for you or that you are having plenty of terrible matches to offset them aswell.

Just stop it.

[Appogee]

Zergling, on 24 November 2016 - 07:51 PM, said:

Yes, I expect to still produce decent stats even when the matchmaker is giving me bad teams.
If I'm doing badly when the matchmaker is giving me bad teams, or if I'm unable to produce a good W/L when I'm playing at a decent level over a large number of battles, then it is likely my own fault.

And yes: If I'm producing good damage and K/D along with correctly supporting my teams, then I expect to produce high W/L. And yes, I can do that even with 'crap' mechs.

In overview, you believe your personal performance was the key determinant in your high win:loss ratios. Or put another way, you single-handedly turn the tide of battle when you are dropped with poor teams, and cause the poor team to win. Even when you're not in a levelled meta Mech.

Okay.

Do you have a video channel and any recorded matches I can watch where you demonstrate this prowess? I will be very interested to see you in action and learn how you do it.



On another point, I noticed something interesting about our respective stats...

Your average match score is 285, mine is only 235.
Your kill:death ratio is 1.75, mine is 1.9.
Your win:loss ratio is 1.3, mine is 1.1

I'm losing more matches than you. That almost always means I've been killed. Despite that, my kill:death ratio is higher.

You are winning more, and therefore dying less. But you're achieving fewer kills than me per death. Taking into account your fewer deaths, you must be achieving far fewer kills per average match than me, for me to have a higher kill:death ratio than you.

It seems anomalous that you, who are personally carrying bad teams to victory, are achieving far fewer kills per match than me, who isn't able to personally carry bad teams to victory.

What are your thoughts on why this could be...?

Edited by Appogee, 26 November 2016 - 02:43 AM.

[Zergling]

Appogee, on 26 November 2016 - 02:23 AM, said:

In overview, you believe your personal performance was the key determinant in your high win:loss ratios. Or put another way, you single-handedly turn the tide of battle when you are dropped with poor teams, and cause the poor team to win. Even when you're not in a levelled meta Mech.

Okay.

Do you have a video channel and any recorded matches I can watch where you demonstrate this prowess? I will be very interested to see you in action and learn how you do it.



On another point, I noticed something interesting about our respective stats...

Your average match score is 285, mine is only 235.
Your kill:death ratio is 1.75, mine is 1.9.
Your win:loss ratio is 1.3, mine is 1.1

I'm losing more matches than you. That almost always means I've been killed. Despite that, my kill:death ratio is higher.

You are winning more, and therefore dying less. But you're achieving fewer kills than me per death. Taking into account your fewer deaths, you must be achieving far fewer kills per average match than me, for me to have a higher kill:death ratio than you.

It seems anomalous that you, who are personally carrying bad teams to victory, are achieving far fewer kills per match than me, who isn't able to personally carry bad teams to victory.

What are your thoughts on why this could be...?

While I don't think I am single-handedly responsible for turning the tide of battle in all of my wins, it is likely I have made made enough positive influence to cause a decent number of wins where my team would have lost without my contribution.


As for videos/recordings of my play, I don't have any. There are far better players than me you are better off watching anyway.


As for stats, note that I've played 241 battles in Season 5, and have scored 265 kills; that is 1.099 kills per battle.
You have played 172 battles in Season 5 for 173 kills, which is 1.006 kills per battle.

Combined with my higher average match score, that means I am scoring more kills and doing more damage per battle, which results in a higher influence on each battle.


Also note that in Season 4, I had a 1.30 W/L, 1.77 K/D and 265 average MS.
And in Season 3, I had 1.28 W/L, 1.24 K/D and 233 average MS.
I don't have any stats for Season 1 or 2 however, as I wasn't playing then.

Combined across those 3 months of play, that is 842 battles, 474 wins, 368 losses; 1.29 W/L. That is a quite significant number of battles, so it appears I am able to consistently maintain a decently positive W/L ratio.


And my stats in the Season 5 assault leaderboard have gone up even higher since I played another 7 battles in my Nanuq:


Will those stats remain at that level or increase further? I don't know; they might very well decrease.

But I will be surprised if I drop below 1.50 W/L and 500 damage/battle in my next 10 battles; that would require me to win less than 4 of those 10 battles and do less than 394 damage/battle for those 10 battles.

The losses would certainly be possible if I get a bad losing streak in the matchmaker (everyone gets shafted by the matchmaker here and there), but it is unlikely my damage/battle would drop that low.

Edited by Zergling, 26 November 2016 - 04:01 AM.

[STEF_]

Duke Nedo, on 25 November 2016 - 06:23 AM, said:

Can you maintain a 2.0+ w/l ratio only dropping solo?

I've just remembered that I have an alt account which I used to drop solo que only, testing trial mechs only.

I've checked its stats and tt has k/d ratio 2.12 and w/l ratio 1.78 (tier 2)

So maybe with good leveled mechs I should do w/l 2+

Edited by Stefka Kerensky, 26 November 2016 - 04:31 AM.

[STEF_]

Zergling, on 26 November 2016 - 03:57 AM, said:

As for videos/recordings of my play, I don't have any. There are far better players than me you are better off watching anyway.


As for stats, note that I've played 241 battles in Season 5, and have scored 265 kills; that is 1.099 kills per battle.
You have played 172 battles in Season 5 for 173 kills, which is 1.006 kills per battle.

Combined with my higher average match score, that means I am scoring more kills and doing more damage per battle, which results in a higher influence on each battle.


Also note that in Season 4, I had a 1.30 W/L, 1.77 K/D and 265 average MS.
And in Season 3, I had 1.28 W/L, 1.24 K/D and 233 average MS.

I would like to understand how average MS works, and if it's really worth to take into account for the PSR values and tiers.

For instance, we have the same average MS: 266, right now. But I have 2.46 w/l ratio and 2.30 k/d ratio.
Something makes no sense, or I cannot get it.


With the alt account I use to test trial mechs in solo que only, pizzafly, I have MS=2.49 and w/l=2.20 with k/d=1.89 (season 4)

Imo, MS is very bad implemented, and that's why MM is truly wonky

Edited by Stefka Kerensky, 26 November 2016 - 04:45 AM.

[Zergling]

Stefka Kerensky, on 26 November 2016 - 04:40 AM, said:

I would like to understand how average MS works, and if it's really worth to take into account for the PSR values and tiers.

For instance, we have the same average MS: 266, right now. But I have 2.46 w/l ratio and 2.30 k/d ratio.
Something makes no sense, or I cannot get it.


With the alt account I use to test trial mechs in solo que only, pizzafly, I have MS=2.49 and w/l=2.20 with k/d=1.89 (season 4)

Imo, MS is very bad implemented, and that's why MM is truly wonky

It's weird and I don't know why, but your average MS is bizarrely low for such a high W/L and K/D. Your killrate isn't even that high; 627 kills to 629 battles.

For that W/L and K/D, you should be up around 300-400 average MS.

There must be something unusual with how you play to produce such a high W/L and K/D with such a low average damage, but I have no idea what it could be.

Edited by Zergling, 26 November 2016 - 04:59 AM.

[Duke Nedo]

Zergling, no offense but you are still in tier 2 so you will on average face less good opponents and be expected to carry less. I am sure you're a good player still but your job is easier. Was true for me as well, while climbing I was well over 2 in k/d but after maxing the XP bar I have dropped to about 1.85 or so.

Stefka, was that in t4? I wish there were flags to filter data for queue, dates etc.... I bet all data is there, we must need a better stat browser...

[johnwolf]

recently i've been having a lot of decent matches in my Marauder sense decided to swap out the PPC's for 4 mediums and, because i'm mastering the Mech, one has dual LBX 10's instead of dual UAC 5's. I have a lot of fun in these builds and have broken 300-500 on average. that being said, i sometimes have poop to the brain and get too aggressive and pop fast. Or i don't pay attention to the map and i got to far from my teammates.


seriously, what's with people not looking at their map, IT'S SO USEFUL!

[Zergling]

Duke Nedo, on 26 November 2016 - 04:57 AM, said:

Zergling, no offense but you are still in tier 2 so you will on average face less good opponents and be expected to carry less. I am sure you're a good player still but your job is easier. Was true for me as well, while climbing I was well over 2 in k/d but after maxing the XP bar I have dropped to about 1.85 or so.

My stats actually went up when I increased from Tier 3 to 2 (Season 3 I was Tier 3, hitting Tier 2 at the end of the month). There are potatoes at every tier, so I doubt my stats will drop much when I hit Tier 1 in another month or two.

[Duke Nedo]

Yes, but on average, less potatoes. You can still drop in a tier 4 game, soon you can't.

[STEF_]

Zergling, on 26 November 2016 - 04:54 AM, said:

It's weird and I don't know why, but your average MS is bizarrely low for such a high W/L and K/D.

For that W/L and K/D, you should be up around 300-400 average MS.

Imo, the issues with MS, PSR (and, so, MM) is that PGI gave too much importance to stuff that really doesn't matter
For instance, I never use AMS, rarely I use ECM, and I try to do the less dmg possible to kill a mech (that is I don't spread dmg, I aim at only one hitbox in the enemy mech....and of course I don't use lrm)

This because it's the killing that matters if you want to win, in any game mode.

I don't really know how MS is calculated, tho.

Edited by Stefka Kerensky, 26 November 2016 - 05:15 AM.

[Zergling]

Duke Nedo, on 26 November 2016 - 05:07 AM, said:

Yes, but on average, less potatoes. You can still drop in a tier 4 game, soon you can't.

Hmm, we'll see. If I find things too tough at Tier 1, I'll just run more meta builds to compensate.

I started Tier 2 around the end of September, and since then I'm up to 60% of the way to Tier 1. I figure another 300-400 battles should get me there.

Stefka Kerensky, on 26 November 2016 - 05:14 AM, said:

Imo, the issues with MS, PSR (and, so, MM) is that PGI gave too much importance to stuff that really doesn't matter

For instance, I never use AMS, rarely I use ECM, and I try to do the less dmg possible to kill a mech (that is I don't spread dmg, I aim at only one hitbox in the enemy mech....and of course I don't use lrm)

This because it's the killing that matters if you want to win, in any game mode.

I don't really know how MS is calculated, tho.

Before PSR Tiers, MWO used an Elo system based entirely off W/L... and I'm kinda doubting the PSR system is better.

I generally don't try to spread damage; I go for the kill, but if I don't have a shot at a near dead torso I'll just fire at whatever I can hit.

Edited by Zergling, 26 November 2016 - 05:24 AM.