9 replies to this topic

[Binkus]

One thing that i think could do with a slight rework if it is on anyones radar and can be bothered

The pilot rating system - carrot and stick needs to support wider team

Example:

You have played your heart out on Incursion game as a light

you fed that radar like a hungry child and supported the wider team

you had little chance to get much pew pew in

your team get rolled

you get rolled as 8 assaults bear down on you

you are a negative pilot rating - you bad bunny you

infact if you are a light generally you are going for a negative on any lost game

it cannot just be based on damage and kills it only supports larger mechs, there has to be other factors added to it included game contribution of some kind

the game to support all players and incentives to better their class needs to feed through what they achieve

infact much of the experience needs to be reworked in a similiar fashion damage and kills does not reflect the variety of mechs and roles

just thought i would throw that out there...

[El Bandito]

Accurate objective based rewards have been requested by players ever since the game was in beta. PGI hasn't delivered it yet.

[Angel of Annihilation]

The Tier system is a joke but honestly I don't know any system that is going to be accurate. The problem is tracking the things that make a good pilot good. For example you bring up light mechs. A good light mech pilot is doing much more than damage. He is spotting, off on the flank acting as a distraction, running through the middle of the enemy team acting as a distraction, scouting the map, etc. How do you quantify those actions?

Here is a prime example. A light mech running off to one side of the map and finding no enemy mechs can be as valuable as it coring out a KDK-3 from behind because when he does that the team can know that their flanks are secure. However, how does the computer track doing something useful like scouting the flanks and assign it a positive numerical score and differentiate between that and just running off and hiding to preserve his K/D which would result in a negative score.??

[El Bandito]

Viktor Drake, on 31 May 2017 - 10:33 AM, said:

The Tier system is a joke but honestly I don't know any system that is going to be accurate. The problem is tracking the things that make a good pilot good. For example you bring up light mechs. A good light mech pilot is doing much more than damage. He is spotting, off on the flank acting as a distraction, running through the middle of the enemy team acting as a distraction, scouting the map, etc. How do you quantify those actions?

That can't be quantified. Simplest thing to do is to give bonus rewards for playing Lights and Mediums.

[Angel of Annihilation]

El Bandito, on 31 May 2017 - 11:11 AM, said:

That can't be quantified. Simplest thing to do is to give bonus rewards for playing Lights and Mediums.

But I already generally get higher rewards when I am playing my lights and mediums than I do Assaults so it seems a bit unbalanced that they should qualify for a better reward just based on their weight class and not on what they do.

Honestly lights are already getting rewarded for their playstyle because the reason I general get more reward from a light comes from 3 sources.

1 - Flanking - Generally when I am playing a Light, I am getting this bonus right and left (no pun intended by the way). I don't see this near as much on my heavy or assault mechs because those are usually facing off against the enemy rather than doing much flanking.

2 - Protected Light - You get this whenever your are near a mech two weight classes above you so when you consider how many heavy and assault mechs there usually are each match, you can really rack up the reward as well.

3 - Hit and Run - Kind of goes along with Flanking as in a light your often going to be poking the enemy then running away before you got shot back at.

Then on top that it isn't uncommon for me to knock out 400+ damage in a light or even get several kills. Heck solo kills and KMDDs aren't out of the realm of possibilities either and your likely racking up 8+ assists a game even if you don't get a kill.

So for me at least, it isn't that uncommon for me to make between 50-150k more per match with my Pirates Bane, Arctic Cheetch or hell even a Myst Lynx than I do in my Marauder IICs.

Of course we are taking about PSR not Rewards so this is a bit off topic.

Edited by Viktor Drake, 31 May 2017 - 11:34 AM.

[Exilyth]

Viktor Drake, on 31 May 2017 - 10:33 AM, said:

The problem is tracking the things that make a good pilot good.

Some measurable things which could have an influence on PSR:

• sharing armor: damage received as % of total health or % of total health lost by the team
  
• accuracy over range: hit % of shots fired, for a set of range brackets (only counts when an enemy is in front of a player to avoid firing into a wall to lower this stat, should also not count shots blocked by scenery directly in front of the player)
  
• reaction time: how long it takes a player to aim at a visible enemy
  
• kill efficiency: damage needed per kill

Edited by Exilyth, 31 May 2017 - 01:42 PM.

[Angel of Annihilation]

Exilyth, on 31 May 2017 - 01:14 PM, said:

Some measurable things which could have an influence on PSR:

• sharing armor: damage received as % of total health or % of total health lost by the team
  
• accuracy over range: hit % of shots fired, for a set of range brackets (only counts when an enemy is in front of a player to avoid firing into a wall to lower this stat, should also not count shots blocked by scenery directly in front of the player)
  
• reaction time: how long it takes a player to aim at a visible enemy
  
• kill efficiency: damage needed per kill

Sure but the fact remains that you still can use most of these to determine skill.

Sharing Armor - How does the computer tell the difference between you working as a team to rotate in and out of fire to "Share" your armor and you just being a Potato who happens to stand out in the open and get worked over?

Accuracy over Range - I don't know about you but I am often firing suppressing fire toward the enemy to keep them from moving or poking out and firing. Many times I am just sitting there firing at a wall where I know the enemy can see it just to allow my team to move without taking fire. I would consider that good and skillful play but the computer would be registering it as me having poor accuracy. So again how do you distinguish between the two using a mathmatical formula that could be translated to a PSR rating.

Reaction Time - Here is another one that doesn't always fit. Enemy moves out of cover just as I line up on enemy UAV that is allowing the enemy to rain tons of LRMs down on my team. Computer would read me not immediately firing at enemy mech that became visable as having a poor reaction time when in reality I felt that the enemy UAV was a priority target. The UAV is probably the right choice but how does the computer know it is the right choice?

Killing Efficiency (Damage per Kill) - Ok lets take a light mech for example. They are at their best when they are sowing confusion among the enemies which usually means taking a pot shot here, doing a drive by shooting there and so on. Lights aren't necessarily killing enemies and may be just poking and proding all over the place to distract the enemy team. Should they be penalized. How about that Heavy mech that only sees the arm of an opposing mech, should he not fire at the opposing mechs arm because that will reduce his killing efficiency, hence his PSR Rating?

So seriously it goes back to what I was saying, there is really no way to quantify a good vs bad player just by using the raw numbers available to a computer than can only number crunch, not observe, think and make decision on its own. Hell even if we had a true AI observing, it is easily demonstrated that someone observing something doesn't always understand what they are observing and can make erroneous evaluations based on incomplete information. Basically there is no REAL way to come up with PSR which is why skill based MM is such a bad idea.

Random MM is better because statistical probability will take care of balancing out matches. Sure you get a few unbalances matches here and there but more than likely each team will have a normal and equalizing distribution of both skilled and unskilled players resulting in balanced teams.

[El Bandito]

Viktor Drake, on 31 May 2017 - 11:29 AM, said:

So for me at least, it isn't that uncommon for me to make between 50-150k more per match with my Pirates Bane, Arctic Cheetch or hell even a Myst Lynx than I do in my Marauder IICs.

Of course we are taking about PSR not Rewards so this is a bit off topic.

Most players do far better in heavier mechs, so outlier like you can be made an exception. Also, PSR can also be artificially inflated for using Light and Mediums.

[Mystere]

El Bandito, on 31 May 2017 - 07:40 AM, said:

Accurate objective based rewards have been requested by players ever since the game was in beta. PGI hasn't delivered it yet.

You very well know that is just at most half the story. The other and possible larger portion is the extremely loud and whiny player base that cried incessantly about other players doing objectives instead of just mindlessly slugging it out somewhere close to the center of each and every map.

[kyfire]

IMHO, we have objectives, but everyone prefers to play skirmish then worry about the objectives. Also, for whatever reason, PGI choose to make damage the bulk of the reward. Therefore creating the whole, need Damage comes first objectives come last. What PGI needs to do is reduce how much damage counts towards PSR and increase how much completing the objectives are worth.