Jump to content

Applied Some Math In Business


76 replies to this topic

#21 Dogstar

    Member

  • PipPipPipPipPipPipPipPip
  • Bad Company
  • Bad Company
  • 1,725 posts
  • Twitch: Link
  • LocationLondon

Posted 15 April 2021 - 03:58 AM

Well done Nightbird, this is really good.

Could I ask people to ask Daeron if we can get more stats released please. In a similar manner to the Cauldron I believe that the community can provide an insight into player numbers and match making that PGI, with limited time and effort, can't.

#22 Nightbird

    Member

  • PipPipPipPipPipPipPipPipPipPip
  • The God of Death
  • The God of Death
  • 7,518 posts

Posted 16 April 2021 - 11:43 AM

A minor prediction about the Cauldron patch, made before the patch comes out.

I will personally enjoy the patch, but I predict the overall effect will be negative on retention after 1-2 months. To explain my reasoning, I point towards LRMs. LRMs are quite weak in high level play, but there is no attempt to make them more viable because they are already too effective in lower level play. Around the middle skill level where most players are, LRMs are just right, not too effective or ineffective. My impression of the Cauldron patch is that it will make a lot of weapons viable at a high skill level, which is great for players that play at a high level, but it will disrupt the current excellent balance in the middle skill levels. As a result, my prediction is 2 months of above 60% retention rate (grey dashed line) for April and May, then subsequent months below 60% until the patch after.

No need to explain to me how wrong I am, let's re-visit after 3 months.

#23 Escef

    Member

  • PipPipPipPipPipPipPipPipPipPip
  • Overlord
  • Overlord
  • 8,529 posts
  • Twitter: Link
  • Twitch: Link
  • LocationNew England

Posted 18 April 2021 - 11:14 AM

View PostMechaGnome, on 11 April 2021 - 10:30 PM, said:


Yes they do lie.. And companies can do their own meta data.. it still fails miserably most of the time, also half of the data can be faked and regularly is with games companies..


Are you accusing Nightbird of faking their data?

#24 Nightbird

    Member

  • PipPipPipPipPipPipPipPipPipPip
  • The God of Death
  • The God of Death
  • 7,518 posts

Posted 01 May 2021 - 03:03 PM

Too early to make anything of the Cauldron patch, keep in mind the April numbers only tell us how much of March's players stayed. We'll need to wait for the end of May to see first month full data for the Cauldron changes.

Posted Image

Edited by Nightbird, 01 May 2021 - 03:04 PM.


#25 Nightbird

    Member

  • PipPipPipPipPipPipPipPipPipPip
  • The God of Death
  • The God of Death
  • 7,518 posts

Posted 02 June 2021 - 08:58 AM

Posted Image

The first month of Cauldron data is in. As can be seen on Jarl's list, we did get more players which is good. Bringing people into the game via hype/excitement is half the equation. Retaining them however is more difficult, and from that point of view the lowest hanging fruit I still think is the broken Matchmaker.

#26 Grus

    Member

  • PipPipPipPipPipPipPipPipPip
  • Little Devil
  • Little Devil
  • 4,155 posts

Posted 02 June 2021 - 11:03 AM

Square root of 9 is both 3 and -3...

Mind blown...

#27 Nightbird

    Member

  • PipPipPipPipPipPipPipPipPipPip
  • The God of Death
  • The God of Death
  • 7,518 posts

Posted 02 June 2021 - 06:25 PM

View PostGrus, on 02 June 2021 - 11:03 AM, said:

Square root of 9 is both 3 and -3...

Mind blown...


???

#28 Dogstar

    Member

  • PipPipPipPipPipPipPipPip
  • Bad Company
  • Bad Company
  • 1,725 posts
  • Twitch: Link
  • LocationLondon

Posted 03 June 2021 - 04:43 AM

View PostNightbird, on 02 June 2021 - 08:58 AM, said:

The first month of Cauldron data is in. As can be seen on Jarl's list, we did get more players which is good. Bringing people into the game via hype/excitement is half the equation. Retaining them however is more difficult, and from that point of view the lowest hanging fruit I still think is the broken Matchmaker.


Nice work! There's a definite uptick and especially amongst the least performing players - I would guess that the Cauldron patch, which has improved weapons in general is allowing poorly performing players to do a bit better and thus stick around a bit more

#29 Nightbird

    Member

  • PipPipPipPipPipPipPipPipPipPip
  • The God of Death
  • The God of Death
  • 7,518 posts

Posted 03 June 2021 - 06:46 AM

View PostDogstar, on 03 June 2021 - 04:43 AM, said:

I would guess that the Cauldron patch, which has improved weapons in general is allowing poorly performing players to do a bit better and thus stick around a bit more


Not really, every line is within the margin of error from previous patches. There is no uptick at all. I would want to see something like a 5% improvement before declaring an improvement in retention has occurred.

#30 bilagaana

    Member

  • PipPipPipPipPip
  • 131 posts

Posted 03 June 2021 - 07:28 AM

View PostNightbird, on 03 June 2021 - 06:46 AM, said:

Not really, every line is within the margin of error from previous patches. There is no uptick at all. I would want to see something like a 5% improvement before declaring an improvement in retention has occurred.


I assume there is still no data made available as to how many of these supposed new accounts are verified new players and not just alts? I can testify certain knowledge as to a number of higher tier players, including some of the most vocal Cauldron members, indulging in seal-clubbing. So much for their concerns about improving the new player experience.

Edited by bilagaana, 03 June 2021 - 08:03 AM.


#31 Vlad Ward

    Member

  • PipPipPipPipPipPipPipPipPip
  • The Merciless
  • The Merciless
  • 3,097 posts

Posted 03 June 2021 - 08:54 AM

Considering that there is no verification process, it tracks that there would be no verification data.

Edited by Vlad Ward, 03 June 2021 - 08:54 AM.


#32 Nightbird

    Member

  • PipPipPipPipPipPipPipPipPipPip
  • The God of Death
  • The God of Death
  • 7,518 posts

Posted 03 June 2021 - 10:10 AM

View Postbilagaana, on 03 June 2021 - 07:28 AM, said:

I assume there is still no data made available as to how many of these supposed new accounts are verified new players and not just alts? I can testify certain knowledge as to a number of higher tier players, including some of the most vocal Cauldron members, indulging in seal-clubbing. So much for their concerns about improving the new player experience.


Seal clubbing with alt accounts is a problem, but it is has been a constant problem throughout the length of the graph. I don't think it has gotten better or worse.

As for solutions, a half decent Matchmaker would be able to detect a good player's alt and put them in Tier 1 before they finish skilling their first mech, removing any incentive to create alts.

#33 Vlad Ward

    Member

  • PipPipPipPipPipPipPipPipPip
  • The Merciless
  • The Merciless
  • 3,097 posts

Posted 03 June 2021 - 10:46 AM

Placement Matches? Posted Image Posted Image Posted Image Posted Image Posted Image Posted Image Posted Image Posted Image Posted Image Posted Image Posted Image

#34 Nightbird

    Member

  • PipPipPipPipPipPipPipPipPipPip
  • The God of Death
  • The God of Death
  • 7,518 posts

Posted 04 June 2021 - 07:32 AM

View PostVlad Ward, on 03 June 2021 - 10:46 AM, said:

Placement Matches? Posted Image Posted Image Posted Image Posted Image Posted Image Posted Image Posted Image Posted Image Posted Image Posted Image Posted Image


Nan, those are a bad (easily exploitable) mechanic.

#35 Nightbird

    Member

  • PipPipPipPipPipPipPipPipPipPip
  • The God of Death
  • The God of Death
  • 7,518 posts

Posted 04 June 2021 - 08:23 AM

Complaints like this thread are a sad sight fore sure: https://mwomercs.com...80453-win-loss/

That been said, the normal healthy response when you play this game enough, get familiar with and become disappointed by the bad matchmaker is to stop playing. That is what my graph shows.

#36 Vlad Ward

    Member

  • PipPipPipPipPipPipPipPipPip
  • The Merciless
  • The Merciless
  • 3,097 posts

Posted 04 June 2021 - 08:37 AM

View PostNightbird, on 04 June 2021 - 07:32 AM, said:

Nan, those are a bad (easily exploitable) mechanic.


What's the alternative?

#37 Nightbird

    Member

  • PipPipPipPipPipPipPipPipPipPip
  • The God of Death
  • The God of Death
  • 7,518 posts

Posted 04 June 2021 - 08:44 AM

View PostVlad Ward, on 04 June 2021 - 08:37 AM, said:


What's the alternative?


Adaptive decision making stats model.

#38 Vlad Ward

    Member

  • PipPipPipPipPipPipPipPipPip
  • The Merciless
  • The Merciless
  • 3,097 posts

Posted 04 June 2021 - 08:49 AM

View PostNightbird, on 04 June 2021 - 08:44 AM, said:

Adaptive decision making stats model.


I'm not aware of any video game MM that does this. Do you have any examples?

#39 Nightbird

    Member

  • PipPipPipPipPipPipPipPipPipPip
  • The God of Death
  • The God of Death
  • 7,518 posts

Posted 04 June 2021 - 08:52 AM

View PostVlad Ward, on 04 June 2021 - 08:49 AM, said:

I'm not aware of any video game MM that does this. Do you have any examples?


It's not something successful companies would freely share with their competitors, plus only about 50k people in North America (1 in 7k people) have the qualifications to do it. It's niche.

#40 Vlad Ward

    Member

  • PipPipPipPipPipPipPipPipPip
  • The Merciless
  • The Merciless
  • 3,097 posts

Posted 04 June 2021 - 09:09 AM

View PostNightbird, on 04 June 2021 - 08:52 AM, said:

It's not something successful companies would freely share with their competitors, plus only about 50k people in North America (1 in 7k people) have the qualifications to do it. It's niche.


What would be next on your list if PGI is unable to hire one of those people?





9 user(s) are reading this topic

0 members, 9 guests, 0 anonymous users