30 replies to this topic

[Bishop Steiner]

Ghogiel, on 23 September 2015 - 07:16 AM, said:

W/L imo is the most important metric at the end of the day. More so in solo queue than group. It is the best stat at measuring the individual pilots ability to influence a matches final outcome. Players who do this consistantly are the good players in solo queue in my eyes. That ability translates to high dam and kills anyway.


This is where to start.

............. so wrong, on so many levels, reasons already stated in earlier posts.

When I can lay out 1000+ damage, 6 kills, do every thing possible....and the other 11 scrubs don't break 250 dmg, and have 1 kill between them, that W/L proves NOTHING about my individual performance or skill level.

[Ghogiel]

Bishop Steiner, on 23 September 2015 - 07:19 AM, said:

............. so wrong, on so many levels, reasons already stated in earlier posts.

When I can lay out 1000+ damage, 6 kills, do every thing possible....and the other 11 scrubs don't break 250 dmg, and have 1 kill between them, that W/L proves NOTHING about my individual performance or skill level.

So you are saying you should go up in PSR with that sort of performance? Well good news, you would!

[Lily from animove]

Bishop Steiner, on 23 September 2015 - 07:19 AM, said:

............. so wrong, on so many levels, reasons already stated in earlier posts.

When I can lay out 1000+ damage, 6 kills, do every thing possible....and the other 11 scrubs don't break 250 dmg, and have 1 kill between them, that W/L proves NOTHING about my individual performance or skill level.

funfact is, with the current system they all just get an up as well because it's enough for the "winning" condition to trigger "ups" for all of them.

Ghogiel, on 23 September 2015 - 07:21 AM, said:

So you are saying you should go up in PSR with that sort of performance? Well good news, you would!

yes but both gamers the 250 scorer and the 1000+ damage guy will reach the same T1 filled bar and I doubt they are equal in skill if they both regulary make those scores.

@Bishop

that is how I would rate it, with existing higher ups, normal up,s stays, down, and heavier downs



but since this was an assault scenario, there still needs to be a lot more score granted for ticking on the base to take into account the effort of a player going for the mission goal.

those 2 guys at the 26 scoring are a bit a borderline case, same score one ahs more assist but less damage. Hard to judge the circumstances of what was less bad than the other.

Edited by Lily from animove, 23 September 2015 - 07:54 AM.

[Revis Volek]

Bishop Steiner, on 23 September 2015 - 06:13 AM, said:

match score
accuracy
some mix of damage and KDr
things like solo kills should heavily impact it compared to kill steals

what should not, is W/L which in the end is not in your control in solo queue. While a Tier 1 should not have to worry about getting dropped with a team of tier 5s and facing an enemy team of tier 3s, which results in the tier1 and his team getting stomped, the outcome of a win or loss really is not an individual stat.

If there was anyway to determine kill efficiency, it would be great to add it.... as the guy who Gausses the eye out of your Atlas is probably more skilled than the guy who uses 1000 LRMs to kill one.

Of course, the great unwashed would cry because their laservomit would actually hurt them. The skilled laser users would not be impacted much...but actual skill in this game is few and far between.

W/L seems to be weighted very little, i had matches last night that i lost and still went UP in PSR or stayed the same. Only 2 out of the 25 or so i played that i lost did my PSR actually go down.

2 kills like 400 some dmg and a few assisits made my PSR go UP in a loss....i can post the screen shot when I get home. and i am a very secure (2/3rds full bar) Tier 3 player.

Edited by DarthRevis, 23 September 2015 - 07:53 AM.

[ColourfulConfetti]

It should revolve around kills first, damage second, whether your mech dies or not shouldn't affect matchscore at all and whether you win or lose shouldn't have such a large impact either, it's PILOT skill rating after all. It's rewarding the comp group droppers and really hurting the guys who solo drop 90% of the time. Though it does work out quite well for lrm and streak boats. It's completly stupid how the game punishes you so severely for dying, especially for heavies and assualts, it's in the job description to tank damage, this kind of scoring encourages poor play.

If I had faith in PGI's programming abilities I'd vouch for a more advanced way of calculating matchsore, something critical like how much damage you did to the CT or ST (if they have an IS XL)legs or cockpit of a mech. So the program when then see how much damage you contributed to that mechs death. Example, if the mech died via destroyed CT and you contributed 75% of the damage to the CT you get a reward from the reward system ("Hotshot" or something) as well as a +20-30 to the match score. The program's job is to reward who did the most damage to the destroyed component that caused the mechs death. Get's a bit tricky when you kill a mech via legs. Best way around that is to calculate who did the most damage on the leg that went down first and do the the same for the second leg but only have that damage contribution count for half. Add both of those values when the mech dies via legging and reward who gets the highest damage score on the legs. Maybe there could be a distinction base on the mech's speed and on class. So if it was a light mech and it's speed is over 120 khp, the first leg to go rule applies, if it is an assault, damage for both legs is calculated normally.

It'd also be nice to have a reward and matchscore boost towards removing a componenet with a lot of weapons and/or heatsinks in it. Give some matchscore based on how many weapons and heatsinks were destroyed from it. So if you pop a torso with 20 tons of weapon on it lets say you get 25% of that tonnage in matchsore, so 5 matchscore for killing off a gauss, 3 heatsinks, and a medium pulse that are in the hunchback's right torso and right arm assuming you were the one who did the most damage to the right torso.

But again, this is IF I had faith in PGI programming abilities or even common sense. On a similar note damage should be dropped down to 6 damage for 1 matchscore and killing blow should be about +5 matchsore instead of one, eliminating a red CT, ST or red legged mech should still be rewarded decently, you still dropped a threat from the battlefield and used your targeting info to your advantage. Using a more advanced program you can have the reward check how many weapons the mech had left and add a reduction of matchscore bonus if it had less than half or even nullify the reward completely if it had no weapons left at all.

In the end it's not about pure damage, it's about where that damage went.

Edited by PalmaRoma, 23 September 2015 - 08:02 AM.

[Greyhart]

I'd say killing blow is not a good indicator of skill as that could just mean you caused 1 point of damage (or your LRM hit at the right time) whilst the other guy stripped the CT.

On the other hand if you consistently lock mechs and hit the most damaged parts destroying arms etc then that is skilful play.

[Raggedyman]

Greyhart, on 23 September 2015 - 05:57 AM, said:

Using only things that give numbers in game and are recordable what should PSR be based on?

Something like game score divided by tier would work for me, on the assumption that game score is calculated in a reasonable manner.

[El Bandito]

Lily from animove, on 23 September 2015 - 07:17 AM, said:

as certain as PGI, becaue they said that.
They don't have a reason to lie with this, so the only chance this is not the case is when its when something was not working as they intended.

you should probably read when PGi makes posts and announcments, and try to remember those informations.

Oh yeah, then explain this.

Measly 158 score in a tie but my PSR still went up. I'm in T2, BTW.

Edited by El Bandito, 23 September 2015 - 08:22 AM.

[nehebkau]

IMHO A PSR that ignores the mech that you are in and the potential of that mech are flawed and inflate the bad mechanic of having to play meta mechs.

I guess it is another argument that pushes the need for battle-value-esque for mechs in game.

To my mind, a 300 damage, 1 kill 5 assist loss game in a MG ERLL locust 1-V is the same as a 700 damage 3 kill 9 assist win in a TBR-S Gauss x 3ERLL build when looking at pilot skill.


The Locust-5v pilot had to be just as good of a pilot as the Timberwolf -s to rack up that damage, assists and kills. Now think about ammo dependent mechs like the huginn or oxide -- if they score high damage it means they had to be very good with the limited amount of ammo they had on board in some very easy to kill mechs.

Sadly, I doubt anyone at PGI will even read this -- forget about passing it on to those in the know.

Edited by nehebkau, 23 September 2015 - 08:24 AM.

[ColourfulConfetti]

The match score just need to factor in who did the most damage to the component that caused the mech to die, put simply.

[Ghogiel]

El Bandito, on 23 September 2015 - 08:10 AM, said:

Oh yeah, then explain this.

Measly 158 score in a tie but my PSR still went up. I'm in T2, BTW.

Probably for PSR purposes ties just act like wins. Because PGI.