43 replies to this topic

[hammerreborn]

Thirdstar, on 01 February 2013 - 09:50 PM, said:

Despite our recent disagreements I still find your zest to gather data very heartening.

I expect to continue to see Awesomes in some quantity for the entire life of MWO, due to currently being the cheapest assault, frequency as a trial choice and the 9M variant.

The shift away from Ravens is probably due to the stripping (in part) of the lagshield.

The shift TOWARDS Fatlasses is also probably because the lagshield fixes.

I wonder if a reconnect option will significantly effect metrics, my gut says yes. I also expect 12 vs 12 to reduce the effect of 1-2 DCs/AFKs.

Your summary of course is spot on.

Doubtful for the reconnect, which is why I truely believe they aren't focused on the reconnect option currently. Technically one of the AFKs arrived after the match was already 7-0 and only sprung to life when I shot him in the face (possible bot?). I mean theres nothing he could have done by then. Any assault that loses a minute or so from disconnect, restarting the client and then redropping into the match is going to be too far behind to do anything worthwhile (except in premades where the team members might wait around for him to come back), especially when the larger maps are released.

Also I firmly believe the awesomes are out in force lately because of the hero mech. I know at least 3 guys in my merc group that are running them right now in preparation. Until they get away from the class based matching I truely believe Awesomes will be always dragging behind the Stalker and Atlas, who can do pretty much everything better than the Awesome except for speed.'

I worry about 12v12. At the current going rate....it's going to be 6-7 assault mechs a match on each team, and without R&R there's no downside to doing so. So overtime I actually expect more and more assaults to appear except for those running their own personal favorites.

Edited by hammerreborn, 01 February 2013 - 10:19 PM.

[Thirdstar]

hammerreborn, on 01 February 2013 - 10:18 PM, said:

Doubtful for the reconnect, which is why I truely believe they aren't focused on the reconnect option currently. Technically one of the AFKs arrived after the match was already 7-0 and only sprung to life when I shot him in the face (possible bot?). I mean theres nothing he could have done by then. Any assault that loses a minute or so from disconnect, restarting the client and then redropping into the match is going to be too far behind to do anything worthwhile (except in premades where the team members might wait around for him to come back), especially when the larger maps are released.

I pretty much pulled my hypothesis from my reconnect experiences in WoT, though now that you point it out I remember that the last time I reconnected, my tank had driven off of a cliff and into the ocean. Mind you WoT loads the client up very fast. Still, I believe having the option is better than not.

hammerreborn, on 01 February 2013 - 10:18 PM, said:

Also I firmly believe the awesomes are out in force lately because of the hero mech. I know at least 3 guys in my merc group that are running them right now in preparation. Until they get away from the class based matching I truely believe Awesomes will be always dragging behind the Stalker and Atlas, who can do pretty much everything better than the Awesome except for speed.'

Completely agree. As long as weight class based matchmaking exists success will swing towards players who bring the heaviest mechs in each of those classes (given equivalent skill and teams). I hope upcoming PPC features and fixes will help somewhat but anything a Awesome can do, with respect to PPCs, a Stalker can usually do better.

hammerreborn, on 01 February 2013 - 10:18 PM, said:

I worry about 12v12. At the current going rate....it's going to be 6-7 assault mechs a match on each team, and without R&R there's no downside to doing so. So overtime I actually expect more and more assaults to appear except for those running their own personal favorites.

It's certainly yet another issue that needs to be looked at. Limiting assaults per drop would be a solution, but would come with it's own issues. It also occurs to me that larger maps will make bringing more assaults less attractive, specially Atlases and Stalkers, but would require a significant size increase to truely matter. Will have to see what the upcoming map looks like. A 50% increase (number pulled out of my butt, sort of) won't be enough.

[hammerreborn]

Thirdstar, on 01 February 2013 - 10:36 PM, said:

It's certainly yet another issue that needs to be looked at. Limiting assaults per drop would be a solution, but would come with it's own issues. It also occurs to me that larger maps will make bringing more assaults less attractive, specially Atlases and Stalkers, but would require a significant size increase to truely matter. Will have to see what the upcoming map looks like. A 50% increase (number pulled out of my butt, sort of) won't be enough.

Except you could argue mobility is the whole point of conquest but what you're seeing is still all out brawls at theta. People are still going to lean towards big and heavy no matter the map size, the loadouts whats going to chance (more guass/ppc and less streaks depending on cover).

Then those same people will complain non stop on the boards about getting capped. Or will end up just sitting on their base and letting the other team come to them, which is a tactic seen in most 8 mans right now from my limited experience.

[Kamikaze Viking]

Thirdstar, on 01 February 2013 - 09:50 PM, said:

I wonder if a reconnect option will significantly effect metrics, my gut says yes. I also expect 12 vs 12 to reduce the effect of 1-2 DCs/AFKs.

A reconnect option for those who have legit client issues (black screen, hud or fps problems) and can re-load fast enough to get back into the game would be great.
And it may also smooth out the data you are getting about DC/afk games.

[Dimento Graven]

Quote

Nearly 10% of 3L pilots have disappeared.

I find this interesting and it more or less lends evidence towards confirming something I stated in the anti-3L thread that was raging earlier, namely that since the netcode changes, all the lesser 3L pilots have moved on to other 'mechs (in my view, based on anectdotal evidence, no recorded stats) such as the Hunchback, Centurian, and Cicada.

The remaining 3L pilots are the ones that are actually "good" at piloting the 3L, and we may yet see even more decline in the numbers of 3L's as we may have a few 'average' or 'below average' pilots who are stubbornly sticking to the build, or are slowly saving the money to purchase an alternate 'mech.

In short, if you die to Raven 3L now, don't feel bad, he's actually a pretty good pilot, and if you kill one, you should feel good, because he was actually a pretty good pilot.

Edited by Dimento Graven, 02 February 2013 - 08:30 AM.

[hammerreborn]

Dimento Graven, on 02 February 2013 - 08:28 AM, said:

I find this interesting and it more or less lends evidence towards confirming something I stated in the anti-3L thread that was raging earlier, namely that since the netcode changes, all the lesser 3L pilots have moved on to other 'mechs (in my view, based on anectdotal evidence, no recorded stats) such as the Hunchback, Centurian, and Cicada.

The remaining 3L pilots are the ones that are actually "good" at piloting the 3L, and we may yet see even more decline in the numbers of 3L's as we may have a few 'average' or 'below average' pilots who are stubbornly sticking to the build, or are slowly saving the money to purchase an alternate 'mech.

In short, if you die to Raven 3L now, don't feel bad, he's actually a pretty good pilot, and if you kill one, you should feel good, because he was actually a pretty good pilot.

The problem is that while yes, it's great that the ravens are no longer 60% of all lights with the 3L being 30% of all lights fielded (as it was in december with the trial raven), we're losing diversity rapidly between classes. Lights and Mediums were already on a steady path to extinction once R&R was removed, and now with netcode fixes (which I'm not complaining about in anyway), the "light" advantage that was keeping the class alive is gone, and with the brawl like nature of the game due to map size and ECM I wouldn't be surprised if lights and mediums are below 15% atm, and assaults above 40%.

I'm going to do another mech diversity run after the Pretty Boy is released to confirm.

[Taizan]

Not really surprised at the amount of premades in this dataset. Teamplay in a group is the premise of the game, though you can lone wolf if you want to.

[Jman5]

The unreliable way of determining premades in the game means that this is at best a conservative estimate. I did an experiment like this a few months ago and I found there were premades out there that flat out refused to admit they were in a group, or became hostile when I asked. A lot of groups either don't talk or think you're just going to yell at them if they admit it.

At this point in the game, I wouldn't be surprised if 90% of matches had a premade in it.

[Yokaiko]

Jman5, on 02 February 2013 - 12:13 PM, said:

The unreliable way of determining premades in the game means that this is at best a conservative estimate. I did an experiment like this a few months ago and I found there were premades out there that flat out refused to admit they were in a group, or became hostile when I asked. A lot of groups either don't talk or think you're just going to yell at them if they admit it.

At this point in the game, I wouldn't be surprised if 90% of matches had a premade in it.

Or you have people like me that tell you its a 4 man if someone asks me if I grouped.

whether I am or not.

[hammerreborn]

Taizan, on 02 February 2013 - 12:08 PM, said:

Not really surprised at the amount of premades in this dataset. Teamplay in a group is the premise of the game, though you can lone wolf if you want to.

I'm more surprised at the number who admitted to being premades than the number of them. Though it seems a number of them knew who I was from my earlier studies and were glad to help.

It definitely showed I was wrong that one time I said that premades were matched against others though. And that my results from the new player experience were low (though that could have been because I was in a trial and actually dragging a friendly premade team down rather than rolling in my Jenner which I can get at least a kill a match) when I found that premades only won around 60-70% of the time..

I think I might try in my next diversity to track premades, team mech loadouts, and then enemy ECM. It won't be pretty but it might yield results that can be compounded to either data results.

[hammerreborn]

Jman5, on 02 February 2013 - 12:13 PM, said:

The unreliable way of determining premades in the game means that this is at best a conservative estimate. I did an experiment like this a few months ago and I found there were premades out there that flat out refused to admit they were in a group, or became hostile when I asked. A lot of groups either don't talk or think you're just going to yell at them if they admit it.

At this point in the game, I wouldn't be surprised if 90% of matches had a premade in it.

Nearly everytime a premade was on my team they would either speak it aloud or do team chat only, so the amount of premades on my own team is probably accurately represented.

For the enemy team, I mean there was still plenty shown, and some of them were actually filled in after the fact because similiar to the two games where i noted it was the same premade from earlier, there were 2 other premades I caught by playing a second round with them shortly after with the same makeup (i.e. 3 games later and the same 4 people are there you're a premade liar).

Quote

Or you have people like me that tell you its a 4 man if someone asks me if I grouped.

whether I am or not.

Yes, there were a lot of you people out there. Like the guy on my team that said he was an 8 man sync drop. Of course I'm not an ***** and can figure it out.

[Wraith05]

Thanks for taking the time to put this together

[Yokaiko]

Yeah like the guy that thought that a goup of players with the same ping MUST be a premade.
...or the number of founders is in anyway indicative of a premade
....or the presence of ECM.


You don't know, until they let people put unit logos, and mark teams no one has any idea.

None.

[hammerreborn]

Yokaiko, on 02 February 2013 - 12:34 PM, said:

Yeah like the guy that thought that a goup of players with the same ping MUST be a premade.
...or the number of founders is in anyway indicative of a premade
....or the presence of ECM.


You don't know, until they let people put unit logos, and mark teams no one has any idea.

None.

You're so cool and hip. Can you teach me to be a giant tool like you?

[Yokaiko]

hammerreborn, on 02 February 2013 - 12:42 PM, said:

You're so cool and hip. Can you teach me to be a giant tool like you?

Are someone's panties bunched?

I'm sorry, there is no verifiable means to tell the teams, so you data dump is useless. Just like the guys that say it wrong because there is a premade in every group I was pointing out that the error could be in the other direction.

If I don't know you the odds of me giving you a correct answer on my group status in game is about NIL, if that makes me a tool, fine I've been called worse.

But you may want to look in a mirror before you start calling names.

[DodgerH2O]

No hard numbers to back it up but given all my jumping around between TS server LFG channels and my time PUGging I feel like teamplay is fairly common. About 3/4 of the time I'm on during weekday evenings if I'm in a group, the other side has several things suggesting they also have a group.

Commonly, i'll be matched against one or more folk I've grouped with in the past via TS, so I find it reasonable to assume they're in a group on TS, occasionally I'll be on their same server and thus verify that indeed, they're in a group. Other times I'll see the whole thing where 4 of the enemy team are sticking close and watching each other's backs (not a guarantee of a group, I try to do this for my team when PUG playing, but often a good indicator.)

I've lately wondered if there isn't part of the matchmaking system that attempts to match groups and PGI just hasn't said anything, letting us come to our own erroneous conclusions. When I play solo I feel like I have a much lower chance to be matched up against folk who I know regularly play on teams.

[Lefty Lucy]

The appropriate way to incentivize the use of non-assault mechs is not, and has never been, R&R costs, but rather better game balance and missions that give players a reason to bring lighter mechs.

[hammerreborn]

It's also the time I played at. When I found a smaller percentage of premades per game I was also doing a weekday run during the holidays from like 2-8, which is not likely to find premades in those first 4 or so hours before everyone gets home.

As this was completely done during prime time the incidence rate increased.

Quote

The appropriate way to incentivize the use of non-assault mechs is not, and has never been, R&R costs, but rather better game balance and missions that give players a reason to bring lighter mechs.

We'll have to see how the new maps pan out with their size. I'm usually pretty optimistic about this game (I really think the map lineup is nice and that PGI has been doing a great job with this game, if maybe a bit slower than I'd prefer). But I sincerely believe that even in conquest mode these maps are still going to end up being a mass of assaults finding and killing each other, just at range rather than in brawl.

R&R wasn't great, but it was a deterrent to running an Atlas 24/7, and it's removal didn't have anything to replace it.

You can also look at the 8 man scene and see how heavies and assaults are just the dominant force used, and dragons are being used as scouts because they won't be cripped by a single guass salvo while still being fast enough to **** of danger if necessary.

And the community as a whole are a bunch of crybabies. I've seen teams open fire on friendly lights trying to cap, and one player even commented that he was TKing all 3Ls on his team while I was playing. You touch the base and you get both teams threatening to kill you as if you've just stabbed a baby and are about to kill it's sister.

Edited by hammerreborn, 02 February 2013 - 01:00 PM.

[Yokaiko]

Lefty Lucy, on 02 February 2013 - 12:54 PM, said:

The appropriate way to incentivize the use of non-assault mechs is not, and has never been, R&R costs, but rather better game balance and missions that give players a reason to bring lighter mechs.

They better come up with maps that require movement then, even with a Awesomes its rare that you can't get into engagement range on these tiny little maps.

really though, heavies seem to be the sweet spot between armor/damage/mobility, I see more heavies than any other weight class per game.

[Corvus Antaka]

Ran an LRM boat for a few days now, still nasty and effective. Why? Because people think ECM makes them immune.

We still need more options like BAP & NARC to counter and deal with ECM, however, as I get used to the ECM I think SSRMS are probably a much bigger problem than ECM right now - and how poorly TAG works vs ECM ie no targetting box, odd behaviour, inability to break through ECM even with a TAG if multiple ECMS are present, but these are more bugs.