Jump to content

Average Match Time Statistics From Pgi - Cauldron Patch And The Future

Balance Gameplay Weapons

180 replies to this topic

#41 Teenage Mutant Ninja Urbie

    Member

  • PipPipPipPipPipPipPipPip
  • The Tip of the Spear
  • The Tip of the Spear
  • 1,678 posts

Posted 28 April 2021 - 12:04 PM

View PostKrasnopesky, on 28 April 2021 - 10:27 AM, said:

Here are two examples that will likely make it in:
UrbanMech(s): Ballistic ammo quirks (+10 AC20 shots per tonne of ammo for example)


Posted Image

Posted Image

#42 Krasnopesky

    Member

  • PipPipPipPipPipPip
  • CS 2022 Gold Champ
  • CS 2022 Gold Champ
  • 217 posts

Posted 28 April 2021 - 12:17 PM

View PostHeavy Money, on 28 April 2021 - 11:44 AM, said:


Those both sound great!
Will there be a period for the community to give feedback on the proposed quirks before they're locked in?

I know we've spoken a bit in MWO Comp Discord about the Marauder II. Is there a more permanent place than a discord channel for people to submit ideas/changes/concerns?

It'd also be nice if we knew what guidelines you are all using. For example, up till now, quirks were often very specific to stock loadouts. Is that policy changing? What rules of thumb are you following in balancing quirks? What's an example of a working mech that we're trying to bring weaker mechs up to match?

Thanks!


Quirks are a very large undertaking and greatly impacted by Weapon and Mobility changes. Ideally we would have quirks finished well before the time of implementation and publish a list of them for feedback purposes, but we would have to have a finished list first! Of course we will not rush such a process and if we have a lot of doubts we will not hesitate to delay implementation to ensure it is done correctly the first time (to the best of our abilities).

I am working with PGI now to have our own section on the official forums, hopefully that will happen sooner rather than later. Until that happens we have a channel on the official MW5/MWO discord that we are now using instead of the MWO Comp discord (The MWO Comp discord should remain competitive focused only). Here is an invitation to the feedback channel in the MW5/MWO discord: https://discord.gg/hUWc3fYDm9

I have also been collating feedback from everywhere I can for The Cauldron to consider.

As for guidelines we are establishing a particular method of identifying which 'Mechs need help and how they need it. Later on when we publicly release our information there will be an explanation of the process we undertook to arrive at the point we did. There will be some quirks that follow stock loadouts / lore / thematic considerations and there will be some that follow more 'interesting' and competitive considerations in order to bring variety and usability to as many Mechs as possible.

Soon we will publish a list of mobility changes that will have an explanation of the process that we undertook and the reasons why we decided on that method.

#43 MechTech Dragoon

    Member

  • PipPipPipPipPipPip
  • Elite Founder
  • Elite Founder
  • 308 posts

Posted 28 April 2021 - 02:41 PM

There is the "fear factor" to consider. Despite TTK decreasing significantly, match time is relatively unaffected due to your average players fear of actually making plays.

More time is spent on positioning and vying for position since mistakes are far more fatal.

#44 FupDup

    Member

  • PipPipPipPipPipPipPipPipPipPipPipPipPipPip
  • Ace Of Spades
  • Ace Of Spades
  • 26,888 posts
  • LocationThe Keeper of Memes

Posted 28 April 2021 - 03:00 PM

Match time is pretty far detached from TTK so it would be more useful if we had statistics like how long the mech survived, how much health % it had when it died, how much damage it was able to do before dying, etc. That would provide a much clearer picture of how the game as been impacted.

#45 Krasnopesky

    Member

  • PipPipPipPipPipPip
  • CS 2022 Gold Champ
  • CS 2022 Gold Champ
  • 217 posts

Posted 28 April 2021 - 03:25 PM

View PostFupDup, on 28 April 2021 - 03:00 PM, said:

Match time is pretty far detached from TTK so it would be more useful if we had statistics like how long the mech survived, how much health % it had when it died, how much damage it was able to do before dying, etc. That would provide a much clearer picture of how the game as been impacted.


I think match time and TTK are directly correlated. I can't see how they wouldn't be. Obviously there is more to the picture such as walking time on large maps, objective play etc, but I am sure if PGI decided to revert the double armour mechanic implemented extremely early in this games life we would see a large drop in match time due to the drastic reduction in TTK.

If the only changes we are making are directly related to a 'Mechs capability to survive and deliver damage (which is the case) I would say that comparing match time before and after the patch is an excellent way to ascertain the impact said changes have had on TTK. Of course more statistics would be excellent, but PGI would need to devote the time and resources to gather and process said data first and secondly be willing to share it with us.

I can only see positives if match time has stayed the same or increased after the next few planned patches.

#46 Nightbird

    Member

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

Posted 28 April 2021 - 03:38 PM

Funnily I expect increased mobility to cause TTK to go down rather than up. "How, when you can spread damage better?" You might ask. Mobility does increase defense capabilities, but it also increases offense by allowing you get around a corner faster, poke more aggressively, track a target more accurately. My bet is on the latter being a larger impact than the former.

#47 Butter Boy

    Rookie

  • Ace Of Spades
  • Ace Of Spades
  • 5 posts

Posted 28 April 2021 - 03:42 PM

Excellent post Kras, thanks for putting this together and all of your hard (&unpaid) work.

I think the decrease in match time, our surrogate for TTK, was inevitable given the buffs in weapons. We can't rationally assume that TTK will go anywhere but down when the only changes have been across the board increases in alpha and/or DPS with minimal decreases in damages for certain weapons. Overall a small drop given the multi-step approach, the others which will presumably increase TTK. I too have seen matches with teams that are timid and afraid to commit, in FP as well. Not convinced its any more than it was pre-patch but we will see. I think with the planned future changes, +/- a few tweaks for unforeseen issues, that the trend will reverse.

I agree that a month-month comparison is a great idea going forward. These first few weeks are going to be 'statistically variable' as people find their new preferred builds. Also- anything making LB20's and snub nose ppcs a viable build is OK in my book.

#48 Firewuff

    Member

  • PipPipPipPipPipPipPipPip
  • Overlord
  • Overlord
  • 1,204 posts
  • LocationMelbourne

Posted 28 April 2021 - 03:59 PM

View PostKrasnopesky, on 28 April 2021 - 06:41 AM, said:

The second set of data shows an increase in match time during the past weekend compared to the initial 48 hour timeframe.

Of course as I described in my post, this is a rather small sample size. In the coming months we will see the results for average matchtime over a month long time period (provided PGI continues to give us this) and player numbers.



I've posted something similar on FB but overall Match time is not a good indicator of TTK. 1 shadowcat sniper can drag a match time out significantly even if the 20 out of the other 24 are dead already. Same goes for the post earlier today on this group where 2 mechs were left no weapons and had to cap. This also does not account for the characteristics of different game modes. This is further complicated by the different game modes.


What would be a much more useful is TTK for the 1st, second, 3rd mechs etc to fall and look at the overall distribution of when mechs are dying over the duration of the game. This also needs to be done across different tiers individually to see if any change good or bad is affecting any particular player segment differently. Mode (most common) rather than Media (average) is also a better measure as it tends to ignore outliers. This is why Mode is a better economic indicator of income than Media, its not biased but a few extremely large incomes (think Besos etc).

#49 Big Tin Man

    Member

  • PipPipPipPipPipPipPipPip
  • Rage
  • Rage
  • 1,957 posts

Posted 28 April 2021 - 04:00 PM

FWIW, when the cauldron is looking at those reductions in average time percentages, please consider the average time per match to come into contact with the OP.

I don't particularly enjoy the 3+ minute mech walking simulator on Polar, but TTK has decreased far more than what the high level metrics show. If that data can be broken down to average match time per map, I think that would be more telling as to actual effects.

Edited by Big Tin Man, 28 April 2021 - 04:00 PM.


#50 Butter Boy

    Rookie

  • Ace Of Spades
  • Ace Of Spades
  • 5 posts

Posted 28 April 2021 - 04:04 PM

View PostFirewuff, on 28 April 2021 - 03:59 PM, said:



I've posted something similar on FB but overall Match time is not a good indicator of TTK. 1 shadowcat sniper can drag a match time out significantly even if the 20 out of the other 24 are dead already. Same goes for the post earlier today on this group where 2 mechs were left no weapons and had to cap. This also does not account for the characteristics of different game modes. This is further complicated by the different game modes.


What would be a much more useful is TTK for the 1st, second, 3rd mechs etc to fall and look at the overall distribution of when mechs are dying over the duration of the game. This also needs to be done across different tiers individually to see if any change good or bad is affecting any particular player segment differently. Mode (most common) rather than Media (average) is also a better measure as it tends to ignore outliers. This is why Mode is a better economic indicator of income than Media, its not biased but a few extremely large incomes (think Besos etc).



But looking at the first mechs TTK would have the same drawbacks (albeit the opposite direction) as would looking at overall match time. Plus, if they were able to give us the TTK of the first three or so mechs like you ask for, why wouldnt they just give us the TTK for each mech, or the average for each match? As much as I would love all these numbers like alot of people posted, doing stats is incredibly time consuming work.

Edit- Also I think the only way to actually record the TTK for a mech would be to directly observe. And how many matches would we need to observe to achieve statistical significance? More than id have time to volunteer. People can feel free to correct me but unfortunately match time is probably all we've got that would be quick enough and cover enough matches easily to be of any relevance.

Edited by Butter Boy, 28 April 2021 - 04:08 PM.


#51 Anomalocaris

    Member

  • PipPipPipPipPipPipPip
  • Bad Company
  • 671 posts

Posted 28 April 2021 - 04:10 PM

View PostBig Tin Man, on 28 April 2021 - 04:00 PM, said:

FWIW, when the cauldron is looking at those reductions in average time percentages, please consider the average time per match to come into contact with the OP.

I don't particularly enjoy the 3+ minute mech walking simulator on Polar, but TTK has decreased far more than what the high level metrics show. If that data can be broken down to average match time per map, I think that would be more telling as to actual effects.


That's a good point. It would be better to be able to break down match time from when the shooting begins. Figure that's about a minute in on most maps which means the reduction in actual fight time is a higher percentage than just looking at gross match time.

But who knows if we'll get that. And for now, we can still use the relative changes in match time as an analog for TTK. It's most useful as a directional indicator rather than looking actual magnitude of change. TTK was expected to drop after the weapon patch, let's see if it goes back up after mobility. I don't know if it will (interesting theories for why it would go down or up being posited here), but if it does, then the mobility pass is working in the right direction.

#52 Krasnopesky

    Member

  • PipPipPipPipPipPip
  • CS 2022 Gold Champ
  • CS 2022 Gold Champ
  • 217 posts

Posted 28 April 2021 - 04:59 PM

View PostFirewuff, on 28 April 2021 - 03:59 PM, said:



I've posted something similar on FB but overall Match time is not a good indicator of TTK. 1 shadowcat sniper can drag a match time out significantly even if the 20 out of the other 24 are dead already. Same goes for the post earlier today on this group where 2 mechs were left no weapons and had to cap. This also does not account for the characteristics of different game modes. This is further complicated by the different game modes.


What would be a much more useful is TTK for the 1st, second, 3rd mechs etc to fall and look at the overall distribution of when mechs are dying over the duration of the game. This also needs to be done across different tiers individually to see if any change good or bad is affecting any particular player segment differently. Mode (most common) rather than Media (average) is also a better measure as it tends to ignore outliers. This is why Mode is a better economic indicator of income than Media, its not biased but a few extremely large incomes (think Besos etc).

View PostBig Tin Man, on 28 April 2021 - 04:00 PM, said:

FWIW, when the cauldron is looking at those reductions in average time percentages, please consider the average time per match to come into contact with the OP.

I don't particularly enjoy the 3+ minute mech walking simulator on Polar, but TTK has decreased far more than what the high level metrics show. If that data can be broken down to average match time per map, I think that would be more telling as to actual effects.

View PostAnomalocaris, on 28 April 2021 - 04:10 PM, said:


That's a good point. It would be better to be able to break down match time from when the shooting begins. Figure that's about a minute in on most maps which means the reduction in actual fight time is a higher percentage than just looking at gross match time.

But who knows if we'll get that. And for now, we can still use the relative changes in match time as an analog for TTK. It's most useful as a directional indicator rather than looking actual magnitude of change. TTK was expected to drop after the weapon patch, let's see if it goes back up after mobility. I don't know if it will (interesting theories for why it would go down or up being posited here), but if it does, then the mobility pass is working in the right direction.


Of course more data looking at multiple different variables and considerations would be better, but I don't have that. I shared literally all the data I currently have from PGI.

Matches are complex and have huge variables at play. The data can be skewed by many occurrences that may happen.

We could look at when the shooting starts (presuming PGI has the ability to do so) but perhaps that Locust ran straight for the enemy team, dropped a UAV and immediately died? Maybe that 4 man team of comp players all running full meta push builds just rolled the entire enemy team in a matter of minutes? Sure Mechs like the Shadow Cat can draw out matches, but how often does that happen over hundreds of thousands of matches? I don't know and I doubt PGI would know without spending a lot of time processing data to find out.

At the end of the day there is one extremely important number that trumps all else: the amount of players playing the game. If this rises and continues to rise that is the single best indicator of success. Currently that number is higher than it has been in quite a while and it is on an upwards trend. Time will tell if it continues to increase.

#53 The6thMessenger

    Member

  • PipPipPipPipPipPipPipPipPipPip
  • Nova Captain
  • Nova Captain
  • 8,045 posts
  • LocationFrom a distance in an Urbie with a HAG, delivering righteous fury to heretics.

Posted 28 April 2021 - 05:31 PM

View PostJediPanther, on 28 April 2021 - 06:20 AM, said:

July: RE-re-RE-scale.
Yeah. i'm expecting one of two things to happen. 1. Pgi will make up a reason not to. Eg. waiting to eg7 to take over and do it. 2. Pgi screws it up yet again and makes it so bad even the few 20ts are reduced to the ones with only ecm/stealth as viable.


Yeah, I agree that PGI would find an excuse not to work.

But if they do, all they need to do is follow the Cauldron's rescale instruction.

#54 R5D4

    Member

  • PipPipPipPipPip
  • Ace Of Spades
  • Ace Of Spades
  • 197 posts
  • LocationAlberta

Posted 28 April 2021 - 06:16 PM

Honestly appreciate the buff approach the Cauldron is taking.

I just hope that the Cauldron pays special attention to the impact on mobility changes (and scaling changes) to both the assaults and the light mechs. Both of these classes have, in the past, received the most savage and often undeserving nerf beatings by a relentless PGI but Lights have imo had it particularly bad since they lack the armor and weapons of assaults.

Edited by R5D4, 28 April 2021 - 06:16 PM.


#55 udoshi

    Member

  • PipPip
  • Bad Company
  • 21 posts

Posted 28 April 2021 - 06:44 PM

View PostKrasnopesky, on 28 April 2021 - 05:57 AM, said:

The Cauldron's overarching goal has always been to increase 'fun' and 'enjoyment' for the majority of people who play this game, two extremely subjective concepts. In order to make this goal more measurable and achievable we have decided to focus on making more weapons, play styles, and 'Mechs in general viable.

At the end of the day the best indicators for us and PGI are an increase in players and player retention. If we see an increase in players and those players keep playing then we are on the right path.


Hey, since you're a good point of contact, I would like to ask for a few simple low-hanging-fruit changes while you're in a fix the game mode:

1) Fix the map voting screen. Its kinda garbage that its blind pick and pick-once no backsies. SO...
a) Make the progress/vote sliders show up before voting as it used to be. Keep the 'you can only click once' functionality.
a) Add a 'pass' button that just automaticaly passes and gives +1 vote.
a) Increase the map select by 1. (so 6 buttons, 1 for pass+ 4 for maps)
a) Instead of voting for a map, put a button right below that votes against it. This uses up half your votes and counts against one specific map. (Ie, we're okay with any match except this one)

2) Add Hibernal Rift as a regular map mode! Its actually a really damn good map! Just park it in regular rotation (its faction play normally)

3) Make hotter maps have a C-Bill bonus. Making players able to fire less means less amounts of damage, means less progress. Compensate directly.

4) Make unpopular maps have a c-bill/xp bonus so there's a reason to go there.

Thanks

#56 Heavy Money

    Member

  • PipPipPipPipPipPipPipPip
  • The Marauder
  • 1,275 posts

Posted 28 April 2021 - 06:52 PM

View Postudoshi, on 28 April 2021 - 06:44 PM, said:


Hey, since you're a good point of contact, I would like to ask for a few simple low-hanging-fruit changes while you're in a fix the game mode:

1) Fix the map voting screen. Its kinda garbage that its blind pick and pick-once no backsies. SO...
a) Make the progress/vote sliders show up before voting as it used to be. Keep the 'you can only click once' functionality.
a) Add a 'pass' button that just automaticaly passes and gives +1 vote.
a) Increase the map select by 1. (so 6 buttons, 1 for pass+ 4 for maps)
a) Instead of voting for a map, put a button right below that votes against it. This uses up half your votes and counts against one specific map. (Ie, we're okay with any match except this one)

2) Add Hibernal Rift as a regular map mode! Its actually a really damn good map! Just park it in regular rotation (its faction play normally)

3) Make hotter maps have a C-Bill bonus. Making players able to fire less means less amounts of damage, means less progress. Compensate directly.

4) Make unpopular maps have a c-bill/xp bonus so there's a reason to go there.

Thanks


None of these things are within the Cauldron's scope.

#57 justcallme A S H

    Member

  • PipPipPipPipPipPipPipPipPipPip
  • CS 2020 Referee
  • CS 2020 Referee
  • 8,987 posts
  • LocationMelbourne, AU

Posted 28 April 2021 - 07:31 PM

View PostHeavy Money, on 28 April 2021 - 06:52 PM, said:


None of these things are within the Cauldron's scope.


Nor PGI roadmap.

#58 justcallme A S H

    Member

  • PipPipPipPipPipPipPipPipPipPip
  • CS 2020 Referee
  • CS 2020 Referee
  • 8,987 posts
  • LocationMelbourne, AU

Posted 28 April 2021 - 07:34 PM

View Postudoshi, on 28 April 2021 - 06:44 PM, said:

2) Add Hibernal Rift as a regular map mode! Its actually a really damn good map! Just park it in regular rotation (its faction play normally)

3) Make hotter maps have a C-Bill bonus. Making players able to fire less means less amounts of damage, means less progress. Compensate directly.


Hibernal is not in Faction Play. It is already in Quick Play, and only Quick Play.

Hotter maps effect every mech. It is not as if a hotter or colder map means it takes more or less damage to kill mechs. That remains the same and this payouts are going to be unaffected.

Anyway as stated all your points are outside Cauldron scope and also PGI. Appreciate you might think they are low hanging fruits. Some just aren't, all of them require coding. I'm not sure as a result of that, that they will be on the cards anytime soon.

#59 justcallme A S H

    Member

  • PipPipPipPipPipPipPipPipPipPip
  • CS 2020 Referee
  • CS 2020 Referee
  • 8,987 posts
  • LocationMelbourne, AU

Posted 28 April 2021 - 07:41 PM

View PostFirewuff, on 28 April 2021 - 03:59 PM, said:

I've posted something similar on FB but overall Match time is not a good indicator of TTK. 1 shadowcat sniper can drag a match time out significantly even if the 20 out of the other 24 are dead already. Same goes for the post earlier today on this group where 2 mechs were left no weapons and had to cap. This also does not account for the characteristics of different game modes. This is further complicated by the different game modes.


Outlier situations are just that. The whole point of a large sample is to ensure those outliers don't adversely skew the results. Over 1,000s of games they won't so such a statement is of little relevance.

Map, mode - it's all taken care of via ways of average and sample size. It's all accounted for.

Of course more detailed stats would be useful. However if you expect there to be a huge shift it isn't going to be there be that map, mode or Tier.

Even if Tier 1-3 games TTK is lower / matches are faster (that might be true) that means TTK/match time for Tier 3-5 is even less impacted or potentially not at all due to the average over the sample size. Imagine T3-5 games not being affected much at all? It's possible based on these numbers.

Edited by justcallme A S H, 28 April 2021 - 07:43 PM.


#60 Scout Derek

    Member

  • PipPipPipPipPipPipPipPipPipPip
  • The Infernal
  • The Infernal
  • 8,015 posts
  • Google+: Link
  • Twitch: Link
  • LocationSomewhere where you'll probably never go to

Posted 28 April 2021 - 07:44 PM

View Postudoshi, on 28 April 2021 - 06:44 PM, said:


Hey, since you're a good point of contact, I would like to ask for a few simple low-hanging-fruit changes while you're in a fix the game mode:

1) Fix the map voting screen. Its kinda garbage that its blind pick and pick-once no backsies. SO...
a) Make the progress/vote sliders show up before voting as it used to be. Keep the 'you can only click once' functionality.
a) Add a 'pass' button that just automaticaly passes and gives +1 vote.
a) Increase the map select by 1. (so 6 buttons, 1 for pass+ 4 for maps)
a) Instead of voting for a map, put a button right below that votes against it. This uses up half your votes and counts against one specific map. (Ie, we're okay with any match except this one)

2) Add Hibernal Rift as a regular map mode! Its actually a really damn good map! Just park it in regular rotation (its faction play normally)

3) Make hotter maps have a C-Bill bonus. Making players able to fire less means less amounts of damage, means less progress. Compensate directly.

4) Make unpopular maps have a c-bill/xp bonus so there's a reason to go there.

Thanks

1) Not possible at the moment

2)Seems like a personal preference

3)Seems like another personal preference

4)Seems like yet another personal preference





1 user(s) are reading this topic

0 members, 1 guests, 0 anonymous users