+ Reply to Thread
Page 1 of 3 1 2 3 LastLast
Results 1 to 15 of 34
Like Tree4Likes

Thread: A way to reward skill in PvP?

  1. #1
    Sword of Telara
    Join Date
    Sep 2014
    Posts
    888

    Default A way to reward skill in PvP?

    Quote Originally Posted by Daglar View Post
    There is no "skill based reward system" for PvP in place. We have discussed this subject in the past and were not able to arrive at any sort of an acceptable system with equal or less exploitation than the current system that involves requiring player "skill" however you decide to measure it that would improve RIFT. If you have one that has not been considered over the last several years - in the context of RIFT, please post it.

    ~Daglar
    I've no way of knowing what has or hasn't been considered, but I do see what I believe to be a fairly good way to do it. It might well require more effort on Trion's part to make it possible, though: without information about how the matchmaker selects people (does it look at players and tries to assign them to matches, or the other way around?) it's pretty much impossible to come up with something that would take only a small modification.

    The basis is to reward players for winning streaks, and alter the matchmaking system so that the team skill level required to get larger streaks rapidly increases without making life too boring for those who aren't so good (for the sake of brevity I'll term them "bad" players from now on, I don't mean any disrespect/offense and all that usual stuff).

    Firstly, this should only affect the 65 PvP bracket. There's not enough people for it to work anywhere else.

    Now, on to the matchmakng. It seems (based on the Fernand incident) that the current matchmaker attempts to produce teams with identical average ELO ratings. This is flawed as it essentially turns PvP into a bit of a grindfest without particularly rewarding good play, because your ELO will increase until you are paired with "bad" enough players for you to start losing. It's also accommodating of "bad" play by the reverse mechanism. Either way, both good and "bad" players get the same rewards - which, according to Daglar, is why we can't have an equal gear rate to PvE. Whether the current state of PvP gearing is or isn't appropriate given this fact is a topic left for the other thread, I think.

    First major change: winning players should be given the option to queue for another warfront with their current team (or just be left in the party with afterwards like dungeons), even if it is forced to be the same map. If premades weren't possible, losers should always be thrown out. I'll explain the significance of this below.

    So, onto the matchmaking changes: 2 values should be stored on each player: their ELO (same idea as what we have already) and their current win streak. Suppose on a team the average win streak is 2 and the average ELO is 1000. I'd want it to pick players based on these priorities:

    1. Their win streak is 2. Doesn't matter what their ELO is for these purposes (I'll explain why later), but it should be preferred to wait for players to become available than fill the game as fast as possible.
    2. Their win streak is 1 and their ELO is close to the average of the enemy team but not too far away.
    3. Their win streak is 0 and their ELO is close to the average of the enemy team but not too far away.
    (it may be necessary to include win streak 3 if the enemy team's average streak is higher to prevent the win streak match "tiers" from slowly collapsing as the match progresses)

    Now for the explanation. For a win streak 0 match, it should pair people of similar ELOs only, which is fundamentally different to how it works currently. Suppose two games are over: one of low ELO ("bad" players) and one of higher ELO. Now both have a win streak of 1, they will be paired against each other. The result? The higher ELO should win comfortably, and be paired against another team with a winstreak of 2.

    The point is that games after the first winstreak should avoid being ELO matched as much as possible. That way, good players will tend to get larger winstreaks, and "bad" players should fail quickly. Those "bad" players should quickly be dumped back into the winstreak 0 bucket and matched against lower ELO players.

    Now, onto throwing out losers. If there was no option to group queue, it would be good as a team that got enough wins to just fell short of their reward can join again in the hope that they can get lucky. That would be unfair against random queuers as it would make PuG-ish premades: I'm on the fence about whether this should be done or not as it would be even more unfair to random players. Maybe only if the enemy team has group queuers in it...

    Now, onto the main point of this: you can attach higher rewards to winstreaks, and have "good" players recieve them at a significantly higher rate than "bad" players being carried.

    A few more changes would be pretty much essential:
    1. Normalise warfront lengths. I haven't played BA or GSE since the update, but Scion still drags out forever. They're already unequal with the current reward system, but this would make it worse.
    2. After a player's winstreak gets too high (eg above 5), apply a stacking debuff that (for example) reduces damage and healing done by 5% and increases damage taken by 3% gaining one stack per win beyond the 5th. Then the premades can't get streaks lasting forever, which will slow their reward progress a bit.
    3. Soul recall and actions which afflict you with Ascended Disgrace remove your winstreak. So DCing and rejoining halfway is safe. Harsh on those who DC, but I don't see any other way to make it work. (I guess you miss rolls if you DC just as the raid boss dies, too?)
    4. An option to manually reset your win streak, as it would greatly increase your queue time.
    Changes can go further, too. It might be necessary to shorten WFs to increase the "bad" player elimination rate (shorter WFs = higher streak required to get rewards) or decrease the team sizes (so 1 "bad" player is a larger hindrance to your team).

    TL;DR: Read it. If I suck at explaining things, ask.

    I can't think of any way to exploit it beyond premading (this change would actively encourage it!), but then again you can do that already. I've tried to cover everything, though there might be some logical issues that need addressing in it. The biggest pitfall I see is that the PvPing population is not large enough to support this, though that was something I somewhat discounted whilst writing this.

  2. #2
    Plane Touched Fusscle's Avatar
    Join Date
    Sep 2014
    Posts
    273

    Default

    I think matchmaking should mainly be in place to make WF matches as even as possible. I'm not sure how your idea would affect that.

    But I really like your idea of keeping the party together and queuing again as a pug PM. And it would be cool to reward skill in this way, but not at the expense of more uneven matches.
    Fusscle of Cheat Engine

  3. #3
    Ascendant wickede's Avatar
    Join Date
    Jun 2011
    Posts
    2,568

    Default

    good players are awarded with winning warfronts.

  4. #4
    Sword of Telara
    Join Date
    Sep 2014
    Posts
    888

    Default

    Quote Originally Posted by Fusscle View Post
    I think matchmaking should mainly be in place to make WF matches as even as possible. I'm not sure how your idea would affect that.

    But I really like your idea of keeping the party together and queuing again as a pug PM. And it would be cool to reward skill in this way, but not at the expense of more uneven matches.
    It depends how good you are. I can't see the point in having a PvP system in which good and bad play make no difference to your outcome. The idea of this is that "bad" players can win first game -> lose second as they have 1 winstreak, and good players can progress further if they're good enough. By preventing 0 winstreak matches from including players with significantly different ELOs, those should be more balanced than what we have now.

    Quote Originally Posted by wickede View Post
    good players are awarded with winning warfronts.
    I don't think so. See this bit I wrote in the OP:

    "It seems (based on the Fernand incident) that the current matchmaker attempts to produce teams with identical average ELO ratings. This is flawed as it essentially turns PvP into a bit of a grindfest without particularly rewarding good play, because your ELO will increase until you are paired with "bad" enough players for you to start losing."

  5. #5
    Ascendant
    Join Date
    Nov 2013
    Posts
    1,840

    Default

    Player made tournaments, so many pvpers yet so few showing up ;)
    PvP in RIFT is good only @forum.

  6. #6
    Ascendant
    Join Date
    Dec 2011
    Posts
    2,357

    Default

    Honestly, I'm not sure I quite understand your suggestions. However, I do tend to disagree with your general philosophy.

    The point is that games after the first winstreak should avoid being ELO matched as much as possible. That way, good players will tend to get larger winstreaks, and "bad" players should fail quickly.
    Everyone, including myself, wants to win. However, I don't care all that much about winning streaks. I care about the quality of the battle. I would much rather lose a battle 975-1000 than win one 1000-500.

    Rather than pack good players onto my team, I would rather see them evenly distributed in the hopes of producing a hotly contested match that is decided by the last minute showdowns of the better players on either team.

    Similarly, I don't want to pocket-heal the same awesome players for snore-fest win after win. I want to kick *** with them one battle, and then desperately flee them the next (while still healing those who will hopefully kill him/her before I die).

    Now, onto the main point of this: you can attach higher rewards to winstreaks, and have "good" players recieve them at a significantly higher rate than "bad" players being carried.
    My reward is the satisfaction of a warfront well played, and my name leading the scoreboard for healing done and objectives carried (basically vanity).

    If you are looking for a way to reward good players, perhaps PvPers could get a weekly gift box that produces items that differ in quality depending on ELO and participation. That way, there is no need to change the current matching system (which has been working rather well for me, according to the number of even matches I have been experiencing).
    Last edited by Nefelia; 01-08-2015 at 12:49 PM.

  7. #7
    Sword of Telara
    Join Date
    Sep 2014
    Posts
    888

    Default

    The whole point of this thread is really to provide a response to Daglar's post on that other thread: the reason PvP has inferior gearing is because you can fail your way to success basically, and the current matchmaking system does just that. This system (hopefully) would avoid that and allow people who perform better to reap more rewards than those who don't. Without such a system, Daglar's current stance is that PvP cannot have comparable gear rates to PvE.

    Basing rewards off winning streaks was the only reasonable way I could see to reward those players who perform better than others: that way those who aren't so good won't receive as much of a reward.

    The problem with a gift box that rewards based on ELO is:
    1. It's once a week. Comparable PvE content (NTEs) is repeatable.
    2. The current matchmaking system actively tries to lower your ELO once you raise it by putting you on a team with lower ELO players. So the only way to get good rewards would be to PvP until you win a few times and then stop until you get your loot.
    It certainly sounds like your experience with PvP is quite different to mine with regards to the evenness of matches. It seems that I spend quite a bit more time in matches where one team completely stomps the other.


    It's difficult to say what matches would be like with this system, for obvious reasons. Whether there are enough players to restrict matchmaking as I've describe I have no idea, either.


    The staying with your team thing would be optional, though it would be a lot harder to get into matches via the queue if your winstreak is not 0. That was why I suggested a way to manually reset it would be necessary.

  8. #8
    Ascendant Violacea's Avatar
    Join Date
    May 2011
    Posts
    12,128

    Default

    Quote Originally Posted by AlwaysHealNeverDie View Post
    Player made tournaments, so many pvpers yet so few showing up ;)
    I had immense difficulty finding 2 worthy team mates. So many PvP'ers, so few worthy.
    New round up of some high rank matches
    Chun-Li*E.Honda*Evil Ryu
    Abel*Ibuki*Dhalsim
    Gouken*Zangief*Ryu

  9. #9
    Plane Walker
    Join Date
    Nov 2013
    Posts
    468

    Default

    It's a tough bit, trying to evaluate PvP skill through mechanical metrics. Only to find them relentlessly gamed once any sort of rewards are tied to them. Requires both an initial and ongoing investment of resources to do, and still may spend more time than not in a rather skewed state. That includes coordinated groups performing kill, win, or streak trading.

    Though it's not just PvP. There's a reason that even for PvE in Rift's heyday they ended up doing away with contribution points and the rift leaderboards.

  10. #10
    Sword of Telara
    Join Date
    Sep 2014
    Posts
    888

    Default

    It's very quickly apparent that adding up damage/heals/kbs or DPS/HPS isn't the way to go for determining how good somebody is. Too many things get missed out in just the combat element, and that's all assuming that there are no objectives aside from do to the most healing/damage as possible.

    Wins seem to be the fairest means by their very nature.
    Last edited by Foolio; 01-08-2015 at 02:31 PM. Reason: spelling

  11. #11
    Ascendant
    Join Date
    Mar 2011
    Posts
    4,000

    Default

    Quote Originally Posted by wickede View Post
    good players are awarded with winning warfronts.
    Or 14 other players who you beat in damage, kills, damage taken, healing done, and objectives. Then you lose.

  12. #12
    Ascendant
    Join Date
    Mar 2011
    Posts
    4,000

    Default

    Quote Originally Posted by AlwaysHealNeverDie View Post
    Player made tournaments, so many pvpers yet so few showing up ;)
    Would be a complete joke now with gear. Ive competed in several of them in the past. Wouldnt even consider one now.

  13. #13
    Plane Walker
    Join Date
    Apr 2011
    Posts
    434

    Default

    Quote Originally Posted by Foolio View Post
    Stuff
    TL;DR
    All possible points you make become moot when you consider that Trion isn't willing to spend any developer resources to make changes to PvP.

    Daglar's quote at the beginning says it all:

    We have discussed this subject in the past and were not able to arrive at any sort of an acceptable system
    Note the use of the word acceptable carries with it the overbearing implication that reads: "Any system that requires more than a few hours of moderate developer coding activity".

  14. #14
    Ascendant
    Join Date
    Dec 2011
    Posts
    2,357

    Default

    Quote Originally Posted by Foolio View Post
    The problem with a gift box that rewards based on ELO is:
    1. It's once a week. Comparable PvE content (NTEs) is repeatable.
    2. The current matchmaking system actively tries to lower your ELO once you raise it by putting you on a team with lower ELO players. So the only way to get good rewards would be to PvP until you win a few times and then stop until you get your loot.
    It certainly sounds like your experience with PvP is quite different to mine with regards to the evenness of matches. It seems that I spend quite a bit more time in matches where one team completely stomps the other.
    As for problem 1: the quality of the gear/rewards in the weekly box can be determined by the week's average ELO score, the quantity of items found within can be determined by the frequency and duration of the week's PvP activities.

    As for problem 2: being on a losing team does not stop you from carrying objectives and racking up damage/heals. It just makes it harder to top the chart. Having a weekly gift box to work towards might actually motivate people on losing teams to try harder to keep up their ELO, and that would make recovery and underdog-victory just a little bit more likely.

  15. #15
    Ascendant
    Join Date
    Dec 2011
    Posts
    2,357

    Default

    Quote Originally Posted by Foolio View Post
    It's very quickly apparent that adding up damage/heals/kbs or DPS/HPS isn't the way to go for determining how good somebody is. Too many things get missed out in just the combat element, and that's all assuming that there are no objectives aside from do to the most healing/damage as possible.

    Wins seem to be the fairest means by their very nature.
    It is not about determining how good someone is. It is about determining how much they contributed towards the team's effort to win. Add in amount of objectives carried/capped and duration of the carry, and damage/heals/kbs/damage-taken is a useful metric to determine who was working for the win and who was not.

+ Reply to Thread
Page 1 of 3 1 2 3 LastLast

Bookmarks

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts