53 replies to this topic

[adamts01]

Navid A1, on 30 June 2016 - 07:22 PM, said:

Then there is the lack of a Battle Value (or something similar) system. In the current system, a mist lynx is equal to an oxide.
A Gargoyle is equal to a KDK-3.
A Dragon is equal to a Timber or Ebon.
That is Pure BS!

You just can't make that work in this game though. Most people in Dragons have some idea of what they're doing and can probably run their build well, same with the MLX. I'd rather have those guys on my team than a LRM/MG Timby/Ebon Jag. MM can't take all that equipment in to account, it would just be a mess.

[wanderer]

Navid A1, on 30 June 2016 - 07:22 PM, said:

There are many problems with MM at the moment.

The number 1 problem is that the player population is dangerously low... to the point where one should fear if MWO can go on till the end of the year or not.
That prevents any detailed match making system to work, without requiring the player to wait an hour for a game.

The leaderboard shows 37,000 or so different accounts are playing.

There is not even the loosest attempts at sorting players by skill at this point, only a waiting period for seals to hit the clubbing grounds of T3+ that happens to be a bit shorter for good players than bad. Not filtering these players out of the T1-T3 range just fertilizes the potato fields to grow a crop of frustration that rises higher and higher by the day.

That's all it really takes to start. Increasing requirements to go upwards, increasing requirements to avoid going downwards, increasing requirements to stay in place, all of which get bigger with each tier. Paul has produced a wonder: a system where everyone will be "the best", whether they're the guy who can pick up 10 legit kills in a match or the guy who can pick up 10 damage per game if he's lucky.

Until that's fixed, despair of getting anything more than lolrandom once you pass T3, much less T2.

[Death Proof]

adamts01, on 30 June 2016 - 07:38 PM, said:

I don't think you understand what we mean by common variable. Teams switch dumbasses every round, if you're having chronic losing streaks then you're one of them. But I do agree, the tier system is a mess. Everyone quickly makes tier 3, and tier's 1, 2 and 3 all play in the same matches, so there's zero point of it unless you're fresh off the boat.

Wake me up when you've seen the ending of the movie.

[Navid A1]

adamts01, on 30 June 2016 - 07:41 PM, said:

You just can't make that work in this game though. Most people in Dragons have some idea of what they're doing and can probably run their build well, same with the MLX. I'd rather have those guys on my team than a LRM/MG Timby/Ebon Jag. MM can't take all that equipment in to account, it would just be a mess.

Its a multi input- multi objective function.

I know that a Gargoyle in the hands of an elite pilot is worth 2 KDK-3s. This is where a multi input function should come in
If PGI insists on using the flawed PSR, at least they could modify it to be less dumb!

Example:
Overall player PSR = raw PSR * ( mech value multiplier + weight class multiplier )
where:

0.1 < mech value < 0.7 (calculated through a normalized function, taking into account JJs, weapon values, Heat management, engine rating, etc...)

and,

Weight class mult = 0.4 - 0.5 - 0.6 - 0.7 for L/M/H/A.

Edited by Navid A1, 01 July 2016 - 02:11 AM.

[adamts01]

double posting like a noob.

Edited by adamts01, 01 July 2016 - 02:31 AM.

[adamts01]

Navid A1, on 01 July 2016 - 02:08 AM, said:

Its a multi input- multi objective function.

I know that a Gargoyle in the hands of an elite pilot is worth 2 KDK-3s. This is where a multi input function should come in... something like this:

If PGI insists on using the flawed PSR, at least they could modify it to be less dumb!

Example:
Overall player PSR = raw PSR * ( mech value multiplier + weight class multiplier )
where:

0.1 < mech value < 0.7 (calculated through a normalized function, taking into account JJs, weapon values, Heat management, engine rating, etc...)

and,

Weight class mult = 0.4 - 0.5 - 0.6 - 0.7 for L/M/H/A.

That's not a bad idea, I just don't think it's possible. Take my 12sl Nova for example. It dies if you fire everything at once, it has a terrible heat value, but when firing 2 groups of 6 it's one of the coolest mechs I run. Plus some people just do well in certain builds and bad in others. I'm a decent Jenner pilot but I suck *** in Clan Oxides. I can't make the Atlas work to save my life but people are claiming it's a Kodiak slayer in a brawl. I think dialing in PSR, having only 3 tiers so that there are enough players to each play your own tier, and trusting good players to bring something they can use is the best solution.

[Navid A1]

adamts01, on 01 July 2016 - 02:30 AM, said:

That's not a bad idea, I just don't think it's possible. Take my 12sl Nova for example. It dies if you fire everything at once, it has a terrible heat value, but when firing 2 groups of 6 it's one of the coolest mechs I run. Plus some people just do well in certain builds and bad in others. I'm a decent Jenner pilot but I suck *** in Clan Oxides. I can't make the Atlas work to save my life but people are claiming it's a Kodiak slayer in a brawl. I think dialing in PSR, having only 3 tiers so that there are enough players to each play your own tier, and trusting good players to bring something they can use is the best solution.

Its better than what we have now.
Besides, heat management number assumes there is no ghost heat, so if you alpha, that is your fault.

Also, there is nothing stopping the introduction of chassis specific PSR. Then when you drop in your atlas for example, you'll have a lower PSR for MM to choose.

[adamts01]

Navid A1, on 01 July 2016 - 02:53 AM, said:

Its better than what we have now.
Besides, heat management number assumes there is no ghost heat, so if you alpha, that is your fault.

Also, there is nothing stopping the introduction of chassis specific PSR. Then when you drop in your atlas for example, you'll have a lower PSR for MM to choose.

I've always thought they should use mech stats instead of overall stats. They collect all the data already, and that eliminates the need for constantly evaluating what chassis are strong.

[Revis Volek]

adamts01, on 30 June 2016 - 07:41 PM, said:

You just can't make that work in this game though. Most people in Dragons have some idea of what they're doing and can probably run their build well, same with the MLX. I'd rather have those guys on my team than a LRM/MG Timby/Ebon Jag. MM can't take all that equipment in to account, it would just be a mess.

IF they did it right MM would have no problems.


You cant have the MM on the fly figure out what weapons you have what weapons the 11 others guys have the mechs etc. It would just take way to long and be a pain im sure.


BUT when you built your mech in the mech lab they could just assing values to weapons based on there balance. So a 3 LRM5, 2 ac 2 Wolverine would have say a 200 BV or PGIV (PGI VALUE) a KDK with 4 10 could be rated at a 400 (2x as deadly and effective vs the Wolverine). You just use these numbers like tonnage and try to meet equal parts of tonnage and PGIV/BV/Mech Grade.


Its not hard but it is work and its seems that is the major stopping point for most good ideas. Too much work for this company to handle i guess.

[Navid A1]

The funny thing is that they record every single stats with regards to your performance... every single possible thing!

Even LRM and LRM+artemis have different stats and each mech variant has its own data recorded.


But when it comes to evaluating PSR... its the dumbest way possible... a single number... which has NOTHING to do with those extensive data they have collected for more than 3 years.


If my stats shows that i have 2000 matches with gauss rifles with 89% accuracy... that should have more weight on my PSR if i'm bringing a dual gauss mech.
If my stats in my Atlas-K are Sh*t Tier... then it should lower my overall PSR if i'm dropping in one.

PGI... why are you not using the data you already have on hand?

W H Y?

[Morticia Mellian]

I Zeratul I, on 28 June 2016 - 05:28 PM, said:

It took me about a week to go from tier 4 to tier 3.

Then it took me 3-4 weeks to go from tier 3 to tier 2.

About 2 months ago I was tier 4. Worked my way up the ladder.

I am curious, did you achieve it from a brand new account? I returned to MWO a month ago, and I am still stuck at Tier 5 thanks to all those matches of mastering mechs and different builds last year. Whatever progress I see is with my stats and rankings, not the Tier system.