227 replies to this topic

[Sandpit]

VanillaG, on 07 March 2014 - 03:40 PM, said:

The percents are given as number of launches, not how many players only PUG or play in groups. .

Exactly

The fact that they don't seem to understand this is part of why the launch module isn't going to fix any of the issues they think it will while just causing some to lose interest or lose enough interest they stop spending money

[VanillaG]

Sandpit, on 07 March 2014 - 03:48 PM, said:

Exactly

The fact that they don't seem to understand this is part of why the launch module isn't going to fix any of the issues they think it will while just causing some to lose interest or lose enough interest they stop spending money

Their proposal is based on the fact that at one time on average 84% of the current pending matches are PUGs. Depending on how granular their raw data is you could argue that the PUG number might be lower but it would still make up a large majority of the pending launches. Their initial design lowers the possibility of more than 1 four man team on a side for PUGs (pre made boogie man), allows slight customization and rudimentary private lobbies for 12 man teams, and allows 2 teams each with one Premium member to have highly customized matches that include smaller and unbalanced teams. It is a step in right direction with the last two items while the PUG queue is slightly tweaked.

The only way that they are going to change how the match maker work is get a large numbers of players dropping as groups. The reality is that while groups are fun, they are also less efficient if you are just trying to grind out cbills or xp. Right now it to hard to form groups, but that is not a match maker issue that is UI issue. We will have to see when they start rolling out CW if that includes some sort of lobby system so it makes easier to form groups.

[Sandpit]

VanillaG, on 07 March 2014 - 04:07 PM, said:

Their proposal is based on the fact that at one time on average 84% of the current pending matches are PUGs.

Exactly, which I've pointed out is simply not accurate a few times for various reasons.

VanillaG, on 07 March 2014 - 04:07 PM, said:

Their proposal is based on the fact that at one time on average 84% of the current pending matches are PUGs. Depending on how granular their raw data is you could argue that the PUG number might be lower but it would still make up a large majority of the pending launches. Their initial design lowers the possibility of more than 1 four man team on a side for PUGs (pre made boogie man), allows slight customization and rudimentary private lobbies for 12 man teams, and allows 2 teams each with one Premium member to have highly customized matches that include smaller and unbalanced teams. It is a step in right direction with the last two items while the PUG queue is slightly tweaked.

The only way that they are going to change how the match maker work is get a large numbers of players dropping as groups. The reality is that while groups are fun, they are also less efficient if you are just trying to grind out cbills or xp. Right now it to hard to form groups, but that is not a match maker issue that is UI issue. We will have to see when they start rolling out CW if that includes some sort of lobby system so it makes easier to form groups.

Their numbers, drawing an assumption based on previous statements and data over the last 2 years, is that they had more people dropping in groups before the group limitations. Then they had less after the group limitations. That COULD be a spurious relationship but I sincerely doubt it. It shows a direct correlation which tells me one of two things

Either they don't understand stats
or
They don't want groups in CW

Either way that's not a good sign for a good portion of the population. I have no doubts the majority (not a large majority mind you) drops solo. There's reasons for that though.
How many new players immediately know about TS servers?
How many new players visit the forums? (By PGI's own admissions it's a VERY small % of players that visit forums)
How many players are able to easily seek out groups?
How many players drop solo because they can't group with 4 others that are currently on?
How many players drop solo sometimes because they can't or don't want to group at that moment?

All of those questions? THAT is how you begin to interpret data sets like that. You don't look at it and say "Yup, 84% drop solo so now we're going to scrap the whole any group size thing because it's no longer the majority"

It's because of this and PGI's lack of understanding this that I think they've just made a huge mistake.

[Alex Warden]

Kyynele, on 03 March 2014 - 07:28 AM, said:

I'd claim that premades by themselves make a far smaller difference than the sheer difference in the amount(s) of good players in the opposing teams. Of course, this means that when really good players group up, they may have that advantage often, but in usual cases when regular players group up, you might hardly see a difference.

i tend to believe the same

[Sandpit]

basically if I can look at the data they provided, read what they wrote, and almost immediately point out that many questions, holes, reasons, etc. that their interpretation is wrong without even having to dig then it SHOULD show that there are some serious issues with whoever is interpreting that data. Whoever it is simply doesn't understand how to collect and review that kind of data and they REALLY need someone who does to do it and make business decisions based on that instead of "Yup, that's 84% alright" which is exactly what it seems like they're doing now.

This is the only time (that I can recall) where I'm flat out stating PGI has done a horrible job and I could do better. Stats are a lot more complicated than just looking at a few numbers and making a decisive action based on a surface glance or thinking you understand something that you don't.

One of my biggest criticisms of PGI has always been that they don't operate like a professional company sometimes. This is a prime example. They're game developers, not statisticians. I don't expect them to fully understand how stats work. I DO, however, expect them to either hire someone to do that or at the very least get someone who DOES understand how to break down data like that to review it and THEN make an educated decision.

I could be completely wrong but based on what I've seen, read, been told, conversations, etc. I really don't think I am. I'd be more than happy to let them buy me a copy of SPSS and review the data for them.

P.S.

If you have no idea what SPSS is without googling you have no business reviewing and interpreting statistical data. Not that you HAVE to use SPSS but anyone who's actually taken stat courses and such will know what that is

[Roland]

Sandpit, on 07 March 2014 - 04:33 PM, said:

They're game developers, not statisticians. I don't expect them to fully understand how stats work.

I think most companies that attempt an online competitive game like this actually would have a statistician on staff.

[Sandpit]

Roland, on 07 March 2014 - 04:36 PM, said:

I think most companies that attempt an online competitive game like this actually would have a statistician on staff.

I would think so as well.......

Yet here we are with several different players in several different threads pointing out several different reasons why the decision and assumptions based on the data presented isn't accurate. That's really what's being discussed I suppose. It's the fact that if I can look at the data they looked at and show why their interpretation is wrong with numerous examples and several others can do the same thing then something is wrong.

This isn't a subjective issue like "good" or "bad" mechs and weapons, this is factual data.

[Novakaine]

PGI's current mindset is that groups are evil.
Groups always stomp pugs.
Therefore we need no groups.
Therefore we need not support or encourage groups.
With no groups there will be no real need for CW.
Beyond a leader board and a map of the inner sphere.

[Sandpit]

Novakaine, on 07 March 2014 - 04:57 PM, said:

PGI's current mindset is that groups are evil.
Groups always stomp pugs.
Therefore we need no groups.
Therefore we need not support or encourage groups.
With no groups there will be no real need for CW.
Beyond a leader board and a map of the inner sphere.

A little more cynical that I'd have put it but yea, this seems to be the current case. When you look at long-term development it really does seem like PGI makes decisions based on today as opposed to trends and long-term data.

"Well we implemented this and mitigated that, now we have this issue that we'll fix" without looking at how previous decisions have steered and skewed data to drive the new trend they're currently trying to "fix"

[Aym]

Joseph Mallan, on 07 March 2014 - 05:55 AM, said:

As you should. I am not mad that a House has been trying to work as a cohesive unit. MM should be trying to group us by Faction already! Only when that happens the Lone Wolves (players not Merc Org) will complain cause they are being segregated.

Actually the best part of MMM is that we get put on random disparate teams. We'll miss getting to marik-bomb so many pug games if CW pans out the way it was originally written, but the gain of dropping with so many organized 'Marikans may well be worth it.

[Roadbeer]

Aym, on 10 March 2014 - 12:48 PM, said:

...but the gain of dropping with so many organized 'Marikans may well be worth it.

Unfortunately, doesn't sound like that's going to happen.

The fear of the Evil Premade Boogeyman is causing PGI to do everything they can to keep people from dropping in groups.

[Aym]

Roadbeer, on 10 March 2014 - 02:45 PM, said:

Unfortunately, doesn't sound like that's going to happen.

The fear of the Evil Premade Boogeyman is causing PGI to do everything they can to keep people from dropping in groups.

But what if we're not in groups, just happen to be in the same team speak because we're such an organized house, and the matchmaker tries to put marik flags with marik flags? Could be fun ;-)

[Deathlike]

Aym, on 10 March 2014 - 02:48 PM, said:

But what if we're not in groups, just happen to be in the same team speak because we're such an organized house, and the matchmaker tries to put marik flags with marik flags? Could be fun ;-)

You better bring bacon.

[Roadbeer]

Deathlike, on 10 March 2014 - 03:01 PM, said:

You better bring bacon.

It is a well known, and popular, fact that House Marik has the BEST space bacon in the Inner Sphere.
Don't accept those other house brands that are just repackaged Clanner bacon.
Clanner bacon is vat grown...ewww

[Sandpit]

Aym, on 10 March 2014 - 02:48 PM, said:

But what if we're not in groups, just happen to be in the same team speak because we're such an organized house, and the matchmaker tries to put marik flags with marik flags? Could be fun ;-)

Given as successful as we are in monday madness and the fact that in CW you have to match up with Marik players and mercs working for them, I'd say it's a good chance that sync dropping solo will wind up being viable.

[Bishop Steiner]

shame they haven't tracked which percent of that 16% are made of sync drops.....

Aym, on 10 March 2014 - 02:48 PM, said:

because we're such an organized house

can't tell if being ironic, or delusional..........................

[Roadbeer]

Bishop Steiner, on 10 March 2014 - 03:15 PM, said:

shame they haven't tracked which percent of that 16% are made of sync drops.....

I'm not sure how you could track that...
I'm also not sure they'd be able to interpret that data correctly. I mean, really, all they'd need to do is reach out to the communities who have sprung up around the game, like... faction houses, merc units, etc.
But that may shatter their preconceived notions about the dynamics of their community, so better not to.

Edited by Roadbeer, 10 March 2014 - 03:23 PM.

[Bishop Steiner]

Roadbeer, on 10 March 2014 - 03:18 PM, said:

I'm not sure how you could track that...
I'm also not sure they'd be able to interpret that data correctly. I mean, really, all they'd need to do is reach out to the communities who have sprung up around the game, like... faction houses, merc units, etc.
But that may shatter their preconceived notions about the dynamics of their community, so better not to.

track which...the irony or the delusion? (House Marik...organized...c'mon guys......that's like complaining about House Steiner running Light Wolfpacks.....)

[Roadbeer]

Bishop Steiner, on 10 March 2014 - 03:22 PM, said:

track which...the irony or the delusion? (House Marik...organized...c'mon guys......that's like complaining about House Steiner running Light Wolfpacks.....)

Well, since this is "OPPOSITE INNER SPHERE" Where Marik is the most organized House, and Lone Wolfs are the preferred front-line units...

[Deathlike]

Bishop Steiner, on 10 March 2014 - 03:22 PM, said:

track which...the irony or the delusion? (House Marik...organized...c'mon guys......that's like complaining about House Steiner running Light Wolfpacks.....)

Well, I have seen genuine scout lances that have the Steiner banner, but rarely. I prefer to enjoy a Steiner lance with full D-DC glory. At least, I don't have to scout anymore.