83 replies to this topic

[Captain Caveman DE]

consider something like this (which is btw something 'zero sum'); just spitballed and an example.
it still pronounces a win over just matchscore-farming, but at least the people carrying the losing side still get something of it,
while the last 3 of the winning team still go down a little.

oh and: people have to get rid of the idea that going-down is a punishment; elo/tier/whatever is just a number to sort performance, so a matchmaker can build a 'fair and balanced' match; that's all.
having a third of all games be stomps is an alarming number and it needs to go down;

[Kurdain]

What about letting people choose their tiers?

The higher the tier the better the reward...knowing that all the good players will be there.

It'd take some good balancing to ensure it couldn't be abused but it'd remove all this auto-balancing.

5 tiers.
5 checkboxes.
Let's say I am a average player so I tick the boxes to be Q'd for tiers 2,3, and 4.
Or if I am a good player I tick 1 and 2.
Or if I am brand new i would tick 4 or 5.

Sure I guess you could have a new player constantly checking tier 1 but they'd die fast for no rewards.
Same with the expert buy 'farming' bottom tiers...lots of time spent for lower rewards.
It would certainly take some balancing.

OR keep the match system in combination with the above and allow people to select their tiers within 1 tier of their current skill.
I.e. Tier 4 pilot could select 3, 4, and 5. A tier 1 could select 1 and 2, tier 5 could select only 4 and 5.
Might allow for people that play ok get matched with better people easier and would avoid bottom feeding or newbies in #1.

[Nesutizale]

Kurdain, on 07 June 2020 - 07:41 AM, said:

Stuff

I agree that next thing on the list should be a rewiev of what gives a high matchscore.

[Brauer]

Kurdain, on 07 June 2020 - 07:41 AM, said:

Well I hope there is a rethinking on how score is determined as it now stands it seems damage is king. Sure you have to damage the other team to win but 50 pinpoint damage in the right spot is arguably more important that 200 splash damage.

Also as I have pointed out, and maybe I am just bad at playing, that when I run my NARC/Spotter mech I almost always get sub 200 match scores because I am not doing a lot of damage. Instead I am putting up UAV's in the middle of the enemies lines, narc/mark targets for 75% faster lock and 100% stronger missile tracking strength, and tagging capture points when appropriate.

I do know that when I have a good round of spotting/narc'ing that the ATM/LRM mechs...mostly LRM to be fair, usually get well over 1000 damage and often results in a win.

Also I feel that you shouldn't be penalized because you had bad luck. I crossed between 2 buildings on Grim Plexis at 150 kph and caught dual heavy guass to the chest...popped my raven in 1 hit to the core. I am sure the shot was intended for someone else and I was unlucky...but I received a penalty. Not sure how you would mitigate this but something to ponder against the guy that stand still, doing team damage...heck on HPG 3 of us was killed by the same team mate yesterday and he got a higher score than I did.

Doing damage yourself and getting kills is by far the best way to drive wins, so it is appropriate for the match score formula to highly reward doing damage.
Outside of serving as the NARCer for a group running an lrm strat you generally are not driving wins that well with a narc light when compared to lights that can actually fight and put up big damage numbers.

Also, re: getting popped by DHG, people do intentionally hit lights with DHG, especially those going in a straight line. So while that may have been an accidental kill when someone was aiming for a bigger mech it wasn't necessarily the case.

[Marikhen]

I'd just like to chime in with an agreement that any system whereby low match scores can result in tier demotion is absolutely fine, just so long as match score gets reworked too.

Punishing good players for doing anything but "moar deeps" is not good, especially when you have game modes that, at least in theory, revolve around something other than two teams just slaughtering one another. If you have a guy on your team who manages to capture the enemy's "flag" in Assault(?) while everyone else is slap-fighting then he shouldn't get the absolute lowest match score of the game and be demoted because he did 0 damage.

Edit:

Kurdain, on 07 June 2020 - 07:41 AM, said:

Also I feel that you shouldn't be penalized because you had bad luck. I crossed between 2 buildings on Grim Plexis at 150 kph and caught dual heavy guass to the chest...popped my raven in 1 hit to the core. I am sure the shot was intended for someone else and I was unlucky...but I received a penalty. Not sure how you would mitigate this but something to ponder against the guy that stand still, doing team damage...heck on HPG 3 of us was killed by the same team mate yesterday and he got a higher score than I did.

Two minor points.

#1 Pilot scoring actually should factor "luck." It sucks, and I've been on the receiving end of bad "luck" a number of times myself. It's also something I still feel should be included, at the very least because the guy who shot you still gets his reward for being lucky enough to have a lightly armored light wander into his line of fire like that.

Frankly "luck" is always a component of combat, and even the best pilot in the world can be the unluckiest SoB to the point where even though he blows the roof off of simulators nobody wants him on their real teams because he's a jinx. I don't see an issue with that being a factor of match score.

#2 Team damage and moreso team kills should be more harshly punished. In point of fact I'd suggest that team damage's match score penalty go up in an increasing damage:point curve so that 1 damage might be 1 point but 50 damage nets you a 100 point penalty and 100 damage nets you a 250 point penalty.

Similarly team kills should be more harshly penalized in light of team damage done.

I team-killed a guy in one match with only 2 team damage meaning he was literally one-shot. It was one of those "Oh, wow" moments where properly used LRMs accidentally hit a teammate and he popped from just one or two missiles. In that case I don't mind a 10 or 25 point penalty for being unlucky or because I could have waited another half second for the guy to clear/die so that my missiles' flight path was clear.

At the same time if I'd had to hit him repeatedly and cored him out with ATMs, whether for an intentional TK or because I just didn't care he was in the way, and did 50+ damage in the process of killing him then my penalty ought to be closer to 200-250 points.

Edited by Marikhen, 07 June 2020 - 08:15 AM.

[Brauer]

Kurdain, on 07 June 2020 - 07:50 AM, said:

What about letting people choose their tiers?

The higher the tier the better the reward...knowing that all the good players will be there.

It'd take some good balancing to ensure it couldn't be abused but it'd remove all this auto-balancing.

5 tiers.
5 checkboxes.
Let's say I am a average player so I tick the boxes to be Q'd for tiers 2,3, and 4.
Or if I am a good player I tick 1 and 2.
Or if I am brand new i would tick 4 or 5.

Sure I guess you could have a new player constantly checking tier 1 but they'd die fast for no rewards.
Same with the expert buy 'farming' bottom tiers...lots of time spent for lower rewards.
It would certainly take some balancing.

OR keep the match system in combination with the above and allow people to select their tiers within 1 tier of their current skill.
I.e. Tier 4 pilot could select 3, 4, and 5. A tier 1 could select 1 and 2, tier 5 could select only 4 and 5.
Might allow for people that play ok get matched with better people easier and would avoid bottom feeding or newbies in #1.

We don't need more queues, the game has too many as it is.

[Brauer]

Marikhen, on 07 June 2020 - 08:03 AM, said:

I'd just like to chime in with an agreement that any system whereby low match scores can result in tier demotion is absolutely fine, just so long as match score gets reworked too.

Punishing good players for doing anything but "moar deeps" is not good, especially when you have game modes that, at least in theory, revolve around something other than two teams just slaughtering one another. If you have a guy on your team who manages to capture the enemy's "flag" in Assault(?) while everyone else is slap-fighting then he shouldn't get the absolute lowest match score of the game and be demoted because he did 0 damage.

1. Sitting on a cap point is not challenging, so why reward it?
2. An individual running off to cap one or more points, or base rush, is not consistently driving wins. Kills and damage drive wins.
3. Doing damage and getting kills is the most consistent way to drive wins and requires more skill than sitting on a cap point. it is appropriate to reward this type of action.


Sub-points for point 2, think about it this way: by killing and/or damaging enemy mechs you are both degrading their ability to damage or kill your force and degrading their ability to cap points or damage your base. For this reason even top level competitive matches are generally ultimately decided based on who does a better job of killing the opposing force, rather than by sitting on caps. In fact, true cap-strats as used in comp are more about maintaining cap parity or a slight lead so that you can wear the enemy down and potentially force them to engage you in a way that works in your favor, rather than relying primarily on hitting the magic number to win on caps.

Running to cap bases or shoot inanimate buildings is usually not actually playing the game mode better than players out to get kills.

EDIT: Somehow this got messed up and initially was self-quoting, fixed that.

Edited by Brauer, 07 June 2020 - 08:09 AM.

[Tarl Cabot]

After going over the last 50 games I played, (hit Print Screen to save overall End results in Screenshot folder), I calculated that for myself, damage (50% to MS) makes up 65% of my MS, overall 301 MS avg for those games while by piloting primarily medium mechs, Energy/ballastics and SRMs combos, no LRMS. Out of those 50 games, 2 assault and 3 heavies, one light. But boy, did that one light tear things up. (Energy)

Quote

by killing and/or damaging enemy mechs you are both degrading their ability to damage or kill your force and degrading their ability to cap points or damage your base.

quoted for double like. For the light pilots not applying any real damage, you may be enhancing your team's ability to do apply damage, WHEN it is utilized, but any additional damage you are able to apply speeds up the process of removing the target from the board.

Remember, Tiers are not the goals, they are only there to be utilized by PGI MM to form up players who are closer to performance then players who are on the opposite end of the Tier system. And having been narc, being an experienced player I and others know how to counter its effect. The opposing team may know where I am at, and as long as I am able to put terrain between myself and the incoming LRMS I will have very little damage. And if you are not firing on me yourself, nothing changes in that aspect.

Marikhen and Kurdain, another way of looking at it, the higher the avg MS, the more said players are relying on damage output w/team work with everything else usually being secondary. Then the question you have to ask yourself, do you really want to be grouped up and against players w/300 MS+, or would you rather in drops with players farther down the ladder, closer to your output and potential playstyle?

[Cluster Fox]

Thanks for listing to the feedback Paul.

Responding to some earlier posts, sorting players and giving a simple rank doesn't reflect how they did relative to others. It's better, but not ideal. A well-matched team would get the same spread than one with a very large skill spread.

For this to be done correctly, the PSR system needs one more variable:

Average team score. (let's call it avgms)

Then the vector brackets (--, ++) can be based off this relative match score. Ideally 5 brackets, but 4 like the current system could work too.

This is almost the same as Paul suggested, except matchscore is variable. Reason at the bottom of this post. Table format cause it's easier to visualize.


Disclaimer: I don't care what values Win / Loss should be like. That's not my point and it's been discussed to hell and back on the other thread. It'll be the same over time anyways.

Reason to do this: If match score is the PSR metric, the match is a closed system and we should rank the players against the other players within this closed system, not against a fixed outside number. Each match becomes a sorting event, kind of like a giant bucket-sort algorithm.

Also, this is not necessarily zero-sum for a single match. But it's much closer than anything using a fixed value. It also converges to zero sum over time.

Edited by Cluster Fox, 07 June 2020 - 10:47 AM.

[MODOK69]

First off, thanks for looking at the feedback. none of this is going to be easy or acceptable to everyone but the closer we get to getting it right the better off we will be. I'm a bit lost on how to start a new forum thread so I'll ask my off topic questions here, hoping they get to the right ears:

1. Is there any truth to the rumor that MWO will be moving to MW5 and become an updated/re-worked better version?

2. Gonna catch a lot of heat for this question. But if player base and funds are an issue, maybe this shouldn't be free to play (it really isn't now but...) I would be willing to give a monthly fee to play the game, as long as I go an honest hard effort to improve/market the game. Its my belief that improving the game is what will bring back the player base. Or maybe MWO just might not appeal to the younger generation that doesn't understand the lore etc...

I've said it before but one more time won't hurt. This is a gem of a game and it can be saved!

Edited by MODOK69, 07 June 2020 - 10:31 AM.

[Brom96]

Cluster Fox, on 07 June 2020 - 09:27 AM, said:

Sorting players and giving a simple rank doesn't reflect how they did relative to others. It's better, but not ideal. A well-matched team would get the same spread than one with a very large skill spread.

For this to be done correctly, the PSR system needs one more variable:

Average team score. (let's call it avgms)

Then the vector brackets (--, ++) can be based off this relative match score. Ideally 5 brackets, but 4 like the current system could work too.

This is almost the same as Paul suggested, except matchscore is variable. Reason at the bottom of this post. Table format cause it's easier to visualize.


Disclaimer: I don't care what values Win / Loss should be like. That's not my point and it's been discussed to hell and back on the other thread. It'll be the same over time anyways.

Reason to do this: If match score is the PSR metric, the match is a closed system and we should rank the players against the other players within this closed system, not against a fixed outside number. Each match becomes a sorting event, kind of like a giant bucket-sort algorithm.

Also, this is not necessarily zero-sum for a single match. But it's much closer than anything using a fixed value. It also converges to zero sum over time.

Yes, parameters within single match, whit all participants and particularly the ones in the same time, under the same condition, would be more just way to rate a single pilot.

Also, lets for a moment think of a winning side. If you kill an enemy by alpha striking his back, you score a kill, remove enemy from a combat and do it so in an expedient manner. You basically can quickly erode enemy team's combat ability. Yet, I guess you will get less points for that then me who will do it shooting the enemy mech all over, removing one torso, then the other one, and taking four times more time to do so, and taking more damage then you, so my ability to fight after that kill is proportionally lesser then yours.

So who's contributing more to the team? According to proposed system (and the current one, of course) - me. Which is laughable.

There is no way to satisfy everyone, but I do think the system you proposed is more just then the alternatives.

[OneTeamPlayer]

Well while we're making requests if you're switching up the system would you mind putting a significant matchscore penalty on team damage above a very small number, say 5 to 10.

No reason to be lighting up teammates.

Edited by OneTeamPlayer, 07 June 2020 - 12:58 PM.

[RRAMIREZ]

Brauer, on 07 June 2020 - 08:08 AM, said:

Running to cap bases or shoot inanimate buildings is usually not actually playing the game mode better than players out to get kills.

Even if I don't radically desagree with that,
It seams that today playing objectives is even counter productive… which seams to me that the way game is played is more or less always the same, whatever the game mode.
And I think that is goes along with "max damage and kills" over strategy.
That's what I found interesting in having WL > MS
Sort of changing the focus on players mind.
But I don't feel like this change going to happen.

But it won't prevent me from playing and hoping maximum players investment over salt.

[Nesutizale]

I think to make playing the objectives more interesting there needs a lot to be done. From reviewing how people play the modes vs how they should be played over to matchscore and maybe even if the maps are right for the mode. Sometimes I think even respawning could be an option but all of that is a different topic. Lets try the new PSR first and then see how it works.

[Brauer]

Nesutizale, on 07 June 2020 - 01:27 PM, said:

I think to make playing the objectives more interesting there needs a lot to be done. From reviewing how people play the modes vs how they should be played over to matchscore and maybe even if the maps are right for the mode. Sometimes I think even respawning could be an option but all of that is a different topic. Lets try the new PSR first and then see how it works.

How people play vs how the game should be played? I'm pretty sure trying to force your playerbase to play a game in one very specific way is a great way to see players flee.

Even in comp you see a divide. Sometimes teams focus on controlling the map, and by extension a few caps, other times you see a team go all in on killing their opposition quickly enough to cap after. It just so happens that in a QP environment teams aren't going to be coordinated enough to reliably carry out map control or cap strats.

Edited by Brauer, 07 June 2020 - 02:30 PM.

[Wraith of Shadow]

Brauer, on 07 June 2020 - 08:08 AM, said:

.... 1. Sitting on a cap point is not challenging, so why reward it?....

... because that's the actual goal of the mission you're on?

Honestly, I'd suggest changing what grants Match Score depending on the scenario of the match in question. For a match where killing the enemy team isn't actually a goal, maybe damage should give little to no points at all... but that might be too complicated.

Kurdain, on 07 June 2020 - 07:41 AM, said:

...
I do know that when I have a good round of spotting/narc'ing that the ATM/LRM mechs...mostly LRM to be fair, usually get well over 1000 damage and often results in a win....

Maybe the 'Spotting' reward should be modified so that if a unit is NARC'ed then a % of whatever damage missiles do to him is given as points to the one who NARC'ed him (while it's active)?

Edited by Wraith of Shadow, 07 June 2020 - 02:40 PM.

[Tarl Cabot]

PGI current proposal does not have a static PSR range, and the middle two ranges covers too much ground. It should be broken up into 100 ranges, as well as adding another PSR level.

In the graph below, the first one is broken up into 100 MS ranges, the 2nd one into 125 ranges but it may be tougher to swallow.

0, 1, 3, 5, 7 and -7, -5, -3, -1, 0.



With that said, this is what the current setup looks like, then comparing it to the upcoming PGI update. I do not expect PGI to do a complete rewrite to obtain a zero sum for every drop. It would be easier to add another level and equal out those levels.

[Cluster Fox]

Tarl Cabot, on 07 June 2020 - 03:17 PM, said:

...

I do not expect PGI to do a complete rewrite to obtain a zero sum for every drop. It would be easier to add another level and equal out those levels.

You nailed it Tarl. I don't expect PGI to re-write their whole system either. Trying to build/tweak as close as possible to what has been proposed by Paul is the most likely way to have something that works. Balancing change vs results.

I still strongly hope PGI goes relative to team / match for a much more effective PSR sort, but getting that extra variable might be buried deep in the code-of-worms. In the absence of which, a fifth bracket like you suggested would be an improvement... as long as the global average match score is the median value, otherwise the distribution skews.

Also, I don't think this should be limited to same-size brackets. Average match score distribution is not linear. So brackets shouldn't need to be.

Ref: https://leaderboard.isengrim.org/stats

[Wraith of Shadow]

Tarl Cabot, on 07 June 2020 - 07:33 AM, said:

.... Below would add at least one more threshold breaking apart the 101-250 and the 251-400

0-100
101-200
201-300
301-400
401+

PSR can still preform the reset and apply the new values and review the result for the month to make a decision on what minor changes needs to be made, be it adding an additional threshold to tweaking the values themselves.....

Honestly, to me at least but I'm pretty much a layman about this sort of thing, the easiest thing to do would be to take the above table and match it to the players average match score for his last bunch of games (10, 30, 100, whatever is fair). That way people would get placed in whatever Tier they usually end up performing in consistently.

Tier 5 = 0-100
Tier 4 = 101-200
Tier 3 = 201-300
Tier 2 = 301-400
Tier 1 = 401+

Nice and simple.

Mind you you'd still probably have to tweak how much certain actions contribute to Match Score. Damage being too important, non-combat contributions being undervalued, etc.

Edited by Wraith of Shadow, 07 June 2020 - 04:55 PM.

[Bowelhacker]

I said it in the other thread and I'll say it agin: fix how match score gets calculated before you start ******* around with this other stuff.