1911 replies to this topic

[Mawai]

Interesting ... a topic where a developer actually appears to read and respond

So ... I will post my major concern at the moment and see what happens

----------

I'd like to mention one topic I find of particular concern. Proper testing of the new matchmaking algorithm.

Match formation will be limited by the least common mech weight or class. With a 3/3/3/3 distribution (unless the mech disttribution entering the queue is absolutely perfect) there will be queues for certain mech weights and classes. With even a small discrepancy in mech classes launching there will be huge queues forming very quickly. Less than an hour of uptime will generate queues of 30 minutes or more in some cases for some weight classes.

Long queues would be extremely undesirable from a player or corporate perspective in my opinion. No one wants to have to wait more than a couple of minutes for a match that only takes 10 minutes on average.

Here is a link to a more detailed post on the subject: Hopefully someone is listening ...

http://mwomercs.com/...82#entry3228282

[Karl Berg]

Veranova, on 14 April 2014 - 09:52 AM, said:

That actually brings up another issue, of bigger 'mechs ELO being worth the same as a smaller 'mech's ELO, which I don't think it necessarily is. If you looked at my ELO in periods of running with assaults in my group, compared to periods of mainly running lights regularly, you'd probably find the modifier you need to balance ELO against the weight class.

We actually separate your Elo's based on weight class right now. You have a separate Elo when running lights, mediums, heavies, and assaults. This has both benefits and drawbacks, which have been covered to some extent in this thread already.

As for the multitude of the suggestions on how to handle the coordination between groups and Elo better, there are some highly insightful ideas in this thread. At this point, I really want to test several of these suggestions out against the data; run a bunch of samples in order to determine which work out best for MWO's dynamics.

[Karl Berg]

IronChance, on 14 April 2014 - 05:41 PM, said:

Seconded on the color-blindness issue. It hasn't been that big of an issue, but it does feel like functionality non-colorblind people are getting that I'm missing out on. And, yeah, the gauss charger sound is too quiet (which is a sound I rely on since I can't see the readiness indicator on the cursor).

I'll definitely raise the issue of colour-blindness with the UI team. I know it's been brought up as a major concern before.

Mawai, on 14 April 2014 - 06:38 PM, said:

Interesting ... a topic where a developer actually appears to read and respond

So ... I will post my major concern at the moment and see what happens

Yup, we're well aware; but thanks for raising your concerns. Absolute worst case scenario, we've built in the ability to toggle the 3/3/3/3 ruleset on/off. We will definitely be monitoring the launch of 3/3/3/3 very closely to ensure the impact on wait times is minimal.

[Dimento Graven]

Another question:

Has there been any thoughts on the post match summary screen, posting the Red and Blue Team's average ELO (available only in non-12 man queues), and each side's Total Weight?

Probably after the implementation of 3/3/3/3 (a truly bad solution in my opinion) the later won't be that critical, but the former... That might still be worth seeing.

No one would know their own score, just what the team's ELO score was.

The lack of any sort of disclosure at all is what makes a lot of the things MM does feel 'arbitrary' at best...

Edited by Dimento Graven, 14 April 2014 - 09:24 PM.

[Karl Berg]

Dimento Graven, on 14 April 2014 - 09:22 PM, said:

Has there been any thoughts on the post match summary screen, posting the Red and Blue Team's average ELO (available only in non-12 man queues), and each side's Total Weight?

If we were asked to, we could do this. The intent is not to though. We really want whatever skill-based matching we use to be an invisible system, it's purely there to produce better games. Even showing just an Elo mismatch would probably be counterproductive. There is future design coming down the pipes that will make that statement make even more sense; it just didn't make it in time for first release of launch module.

[Tekadept]

Karl Berg, on 14 April 2014 - 09:52 PM, said:

If we were asked to, we could do this.

On this point, obviously design do their thing, and you do yours, do you /have you ever had any kind of input that has been pushed up the chain? ie you say to design hey guys you know we can do x and it would be really cool and help the gameplay etc, and they go oh yeh your right full speed ahead. Or are you only there to do designs bidding

Edited by Tekadept, 14 April 2014 - 11:03 PM.

[p4r4g0n]

Karl, I posted some questions in the Feedback to the Regarding the Launch Module and Group Sizes here. Based on past experience, it is unlikely I will get any meaningful responses there but I'm genuinely curious about the matters I raised

I wonder if you would be able to answer any of those questions, as long as it doesn't step on anyone's toes.

Also, if it is possible for solo players to be given the option to join either the public matches (1 group & solos) and the group public matches (groups & 1-3 solos (per team)), could this option not also be extended to 2-4 man groups?

Would this complicate or facilitate matchmaking?

[Kmieciu]

Quote

In addition to this we found a further issue where the explosion code in CryEngine was limiting the number of explosions processed on a given frame and then deferring explosions to subsequent frames to effectively load balance the work for the server

Why do you even use the explosion code for SRMs? A year ago Paul set the explosion radius to 0.05 meters. Yep. That's 5 centimeters. Why do you waste server resources to simulate explosions with a radius of 5 centimeters? The missile itself is bigger than that.

Why not using ballistic projectile code for SRMs? These are anti-armor missiles designed to punch through armor. If a missile hits a hitbox, it should apply full damage to that hitbox. No fancy equations. Just like LBX, but a lot slower (300 meters/second vs 1100) and with shorter range (270 vs 1620 meters).

[Karl Berg]

Tekadept, on 14 April 2014 - 11:03 PM, said:

On this point, obviously design do their thing, and you do yours, do you /have you ever had any kind of input that has been pushed up the chain? ie you say to design hey guys you know we can do x and it would be really cool and help the gameplay etc, and they go oh yeh your right full speed ahead. Or are you only there to do designs bidding

Oh my... I shouldn't. All right everyone, pay close attention. This is the very definition of a ban-able offence.

When we started MWO, design made a confluence page (internal wiki for those not in the know) for developer suggestions where we could write up all our amazing ideas for the project. Paul's very big on narrative design, of course, so we had to frame all the design suggestions in a very specific format for him to review. He then took the time, despite his extreme candy crush addition, AND his obsession with cutting out articles from 'Indie developers weekly' to make an enormous collage in the shape of Kazuki Morishita's left arm, to visit each and every single suggestion, and overlaid them with an image very similar to this:

And that was my experience with offering suggestions to design.


* I kid of course. The above did happen for mostly as I described, but they will certainly listen to feedback. In fact, technical review is a hugely iterative process between the design and engineering teams. Design writes up documents, we tear them apart in a big meeting, they write a new version addressing our concerns, and we repeat.

[Karl Berg]

p4r4g0n, on 14 April 2014 - 11:06 PM, said:

Karl, I posted some questions in the Feedback to the Regarding the Launch Module and Group Sizes here. Based on past experience, it is unlikely I will get any meaningful responses there but I'm genuinely curious about the matters I raised

I wonder if you would be able to answer any of those questions, as long as it doesn't step on anyone's toes.

Also, if it is possible for solo players to be given the option to join either the public matches (1 group & solos) and the group public matches (groups & 1-3 solos (per team)), could this option not also be extended to 2-4 man groups?

Would this complicate or facilitate matchmaking?

Some of what you ask would step on toes. We haven't started the technical review process for this yet, so I don't know that I understand all the edge cases in the design.

I believe the answer to this particular question of yours:

> Allows 2-10 man groups.
> • 11 and 12 man groups; OR
> • 11 man groups only?

I think this means only groups of size 2 to 10, so that there will always be a minimum of 2 groups per team; but I haven't clarified this, so I could easily be mistaken.

[Karl Berg]

Kmieciu, on 14 April 2014 - 11:40 PM, said:

Why do you even use the explosion code for SRMs? A year ago Paul set the explosion radius to 0.05 meters. Yep. That's 5 centimeters. Why do you waste server resources to simulate explosions with a radius of 5 centimeters? The missile itself is bigger than that.

Why not using ballistic projectile code for SRMs? These are anti-armor missiles designed to punch through armor. If a missile hits a hitbox, it should apply full damage to that hitbox. No fancy equations. Just like LBX, but a lot slower (300 meters/second vs 1100) and with shorter range (270 vs 1620 meters).

I haven't looked at this in detail, so unfortunately I can't give you an accurate answer. Sorry

[p4r4g0n]

Karl Berg, on 14 April 2014 - 11:53 PM, said:

Some of what you ask would step on toes. We haven't started the technical review process for this yet, so I don't know that I understand all the edge cases in the design.

I believe the answer to this particular question of yours:

> Allows 2-10 man groups.
> • 11 and 12 man groups; OR
> • 11 man groups only?

I think this means only groups of size 2 to 10, so that there will always be a minimum of 2 groups per team; but I haven't clarified this, so I could easily be mistaken.

Thanks for taking a look at it and appreciate that saying anything would be inappropriate at this time..

However, got to say that it would be a little disappointing if it does work out that 12 man groups would be forced to use private matches only to play.

I can understand if the 11 man groups are excluded based on the public group matches as currently envisioned but unless there are technical reasons against it, 12 man groups should continue to be able to play in the public group matches the way they currently do. Otherwise, the current issues with people asking for 5-11 groups will almost certainly just be repeated.

It is inevitable that people will be annoyed when something they could always do is taken away.

Edited by p4r4g0n, 15 April 2014 - 12:19 AM.

[Kmieciu]

Karl Berg, on 14 April 2014 - 11:56 PM, said:

I haven't looked at this in detail, so unfortunately I can't give you an accurate answer. Sorry

SRM hit detection might seem like a minor detail, but once it is fixed it will drastically affect the gameplay. The SRMs will once again become the most efficient close range weapon and brawling will be restored. We will finally have a counter for LRMs, PPCs and autocannons.

Just let SRMs do the listed damage. That is all I ask for.

Edited by Kmieciu, 15 April 2014 - 12:22 AM.

[No7]

Thank you Karl.

The amount of goodwill you generate in this thread is only paralleled with the amount of bad-will from Russ.

Would just like to say two things that will generate a lot of bang for the buck
1. Lobby system
2. Statistics

If MWO had had a basic, crude, bug ridden, crappy, but at least something resembling of a lobby at the time of open beta launch then I can guarantee you that most of the players from closed beta would still be here. It is that simple, give the community something that they can work with and they will remain happy for a very long time and generate a lot of revenue.

As for stats, just play a game in World of Tanks and look at the end of the game screen there and see where MWO needs to go.

[Kmieciu]

No7, on 15 April 2014 - 12:26 AM, said:

Would just like to say two things that will generate a lot of bang for the buck
1. Lobby system

We can't have a lobby system, because it could be used to count how many people play the game, and this information could be misinterpreted, according to Russ :-)

[Modo44]

Can you troll harder? Karl already spilled the beans on the total number of accounts. The rest is easily inferred.

[MischiefSC]

Karl Berg, on 14 April 2014 - 07:27 PM, said:

We actually separate your Elo's based on weight class right now. You have a separate Elo when running lights, mediums, heavies, and assaults. This has both benefits and drawbacks, which have been covered to some extent in this thread already.

As for the multitude of the suggestions on how to handle the coordination between groups and Elo better, there are some highly insightful ideas in this thread. At this point, I really want to test several of these suggestions out against the data; run a bunch of samples in order to determine which work out best for MWO's dynamics.

I realize I wrote a novel prior and it's pretty much all theory crafting. What you say here though hits on the gist of what I was trying to say.

When looking at balancing with the matchmaker, I would strongly hope for balancing towards a players performance in that specific chassis and loadout, not a generalized weight class. By balancing towards weight class you push players directly towards having to pilot whatever meta works best in that whole weight class - poptart Phract, Victor, LRM boat, Spider, etc. Even when playing that build isn't what's fun it's what works, so a player has to choose between what's most likely to help him and his team win and what he's going to enjoy.

I would love the matchmaker to be balanced more towards putting me in a balanced match even when I'm in a Quickdraw SRM boat. While people have suggested trying to include factors like LRM vs ECM per side, poptarts, etc. this is obviously going to require a lot of extra prediction and segregation in the matchmaker, which will slow time and mean loosening something else.

Introducing a per-chassis, per weapon qualifier to Elo would promote playing what's fun more than what wins. That, to me, is a good thing.

Also we need a lobby in some fashion. The social component changes how people value how they play. You play different when someone you (sorta) know is there. You care more about the outcome. People pug because dropping in a group is slower and ties you to other peoples expectations both in and between matches. The ability to, effectively, 'pug with people you know' would increase investment (and thus satisfaction) with matches.

Thank you again for spending so much time here. We all appreciate it.

[Gasoline]

Karl Berg, on 14 April 2014 - 11:42 PM, said:

Oh my... I shouldn't. All right everyone, pay close attention. This is the very definition of a ban-able offence. When we started MWO, design made a confluence page (internal wiki for those not in the know) for developer suggestions where we could write up all our amazing ideas for the project. Paul's very big on narrative design, of course, so we had to frame all the design suggestions in a very specific format for him to review. He then took the time, despite his extreme candy crush addition, AND his obsession with cutting out articles from 'Indie developers weekly' to make an enormous collage in the shape of Kazuki Morishita's left arm, to visit each and every single suggestion, and overlaid them with an image very similar to this: And that was my experience with offering suggestions to design. * I kid of course. The above did happen for mostly as I described, but they will certainly listen to feedback. In fact, technical review is a hugely iterative process between the design and engineering teams. Design writes up documents, we tear them apart in a big meeting, they write a new version addressing our concerns, and we repeat.

To be honest... I feels like you could leave the disclaimer out and it would be totally legit.

Those posts below have been raised in an effort to get a response to what PGI's idea of community warfare is at the moment. The pillars of game design have been razed, 3/3/3/3 was invented instead of a rough tonnage match/limiting. Everything seems to lead away from the community warfare that was promised to us on the launch event.

http://mwomercs.com/...ost__p__3276379
http://mwomercs.com/...ost__p__3276822
http://mwomercs.com/...ost__p__3280826

I don't ask you to respond on this. You sort of already did in a reply to Chronojam. I think it was also Chronojam who issued the lack of communication in here. You sir are doing an amazing job and you have my chapeau for that. This is not the easiest terrain to settle down, but you see? You get a lot of positive feedback just hanging around here and doing the others jobs. Usually it should be the Community Manager doing this, so you probably deserve a second income .

Could you give us any insight on when we're going to read something about the current community warfare design philosophy? Cos that's the thing why we're holding out, why we still lurk around here... and why we feel constantly ignored by Paul. See, I don't want to know that there'll be beef in my dinner, I want to know if it's a hamburger or a steak... Heck, I don't want to know the whole menu, but you guys need to have at least an idea of what you're going to serve us.

[Veranova]

Karl Berg, on 14 April 2014 - 07:27 PM, said:

We actually separate your Elo's based on weight class right now. You have a separate Elo when running lights, mediums, heavies, and assaults. This has both benefits and drawbacks, which have been covered to some extent in this thread already.

As for the multitude of the suggestions on how to handle the coordination between groups and Elo better, there are some highly insightful ideas in this thread. At this point, I really want to test several of these suggestions out against the data; run a bunch of samples in order to determine which work out best for MWO's dynamics.

You've missed what I mean the problem is. I know ELO is separated right now ^^
That doesn't solve this problem.

If I run my Light in this lance: Light, Assault, Assault, Assault
Which I regularly do because I play with Competitive players. My ELO gets inflated.
3 Assaults can carry a team.

Then when I do light training with my Lance: Light, Light, Light, Light.
I have an inflated Light ELO, and usually so do my guys.
This leads to an entire night of horrific match-making, Where the equivalent ELO players on the enemy team are in much heavier 'mechs, and we get massacred. Until our ELO's balance back out where they should be.

Good Assault players usually have a higher ELO than good Light players, because they can carry the team more.
It's a natural balance, which is very easily disrupted by your choice of group composition.


Of course I've not seen the numbers. Everything I say above is based on my observations.
I bet if you ran the stats on Average Assault ELO vs. Average Light ELO, you'd see a difference.

Edited by Veranova, 15 April 2014 - 07:44 AM.

[Hylius]

Karl Berg, on 14 April 2014 - 07:38 PM, said:

I'll definitely raise the issue of colour-blindness with the UI team. I know it's been brought up as a major concern before.

This is much appreciated. I can't even see the crosshairs on most maps and just have to roughly judge where the center of my screen is (which makes arm weapons a pain in the butt). Weapons being ready/in range indication might as well not exist for me, I can't even see them if I concentrate. The compass is unreadable, especially on maps like Caustic where it blends in. Even the minimap can give me fits with the lightly colored lancemate positions and grid labels. It's really frustrating when the majority of the UI is unusable. I've been clamoring for a bit of color-blind friendliness since closed beta.