Jump to content

[Idea] Stat Skill Based Matchmaking Will Be Flawed. Here Is A Solution.


2 replies to this topic

#1 IamSeanConnery

    Member

  • PipPipPip
  • Elite Founder
  • Elite Founder
  • 91 posts
  • LocationMichigan, USA

Posted 06 December 2012 - 09:06 AM

The devs have expressed matchmaking may be moving to skill vs similar skill based on stats, potentially.

This is fundamentally flawed and suffers from stagnantion in two paradoxical parts:

The first:
As skill progresses teams/players will be pitted against better teams/players. There will be LITTLE BENEFIT to do this outside of having a better challenge. Win loss ratios will approach closer to 50% values and cbills/XP of players for team averages will be the same.

I see this being abused, acually. Reward will be based on who you choose on your team in a bad way: Surround yourself by a team with players with less skill and you will be rewarded with more cbills and XP because you will outshine your teammates.(Because your team doesn't matter, you will likely have a 50% chance to win due to matchmaking, right?) That doesn't seem right as being the ONLY way to get more rewards. And for doing this? You are now considered a BETTER PLAYER based on STATS. Surround yourself by idiots and PROFIT!

Sure, games could start to be more balanced, and that's good. Maybe the only good thing

But it will not work this way for long. Eventually, stagation will destroy this system and make it much less useful. Which is my second point:
The better your team gets, the better stats you get and move up to fight better teams, right? Only in the short term. Why? Because as you fight betters teams, your stats suffer and knock you down to have worse stats. How can you measure someone based on things like k/d radio or win/loss ratio, or whatever stat you want to measure, when you throw them in the mix with people with the same ratios and measurements?
This becomes a destructive feedback loop which results in every team having more similar stats. (I generalize here, but this is very largely true)

In the END all that we have is a skill match system that measures how well a player makes a team and surrounds himself with relatively worse players, which we cannot even control beause he chooses his team around him much of the time.

(I won't even go into how one would use STATS to measure SKILL based on so many games factors)

Now that I've discussed the problem, here is the solution:

Matchmaking should be done by player options that THEY control and have BENEFITS they WANT:


Option 1: Skill: Beginner, Intermediate, Elite, Master. Each tier having higher cbill and XP outputs for the winning team only. and equally higher penalites for losing as tiers progress.
Option 2: Mode: No mech restriction, Trial Mech only, tonnage type limits
Option 3: The lobby. Pretty basic here. Based on option 1 and 2 you have the option to launch against specific players in the lobby, or to just launch against a random group.

Elegant, yet effective.
Better teams will not want to play lesser teams because the benefit is less. But they could if they wanted to. And new players could play trial mech only battles which would generate low cbills and xp, but even the battlefield some.

The SKILL model should be relatively simple to scale (as opposed to stats). If that team can keep a win/loss ratio of .5 or above on a skill level, they should benefit for being on that skill level selected and not below. If they choose higher, they will lose more and it will not benefit them as well and would even hurt them greatly to be there. I would imagine on Master if you lost you should definitely be going well into the negative for lost cbills, but on Beginner you may get cbills either way, just less of them. This will naturally work it's way out, having better players moving to better skills level matches keeping less skilled players out.
(Of course, this scaling would be on top of salvage and bonuses as developed now)

#2 RedMercury

    Member

  • PipPipPipPipPipPip
  • Bad Company
  • 223 posts
  • LocationChina

Posted 06 December 2012 - 12:28 PM

Good use of incentives.

One issue is that a skilled grifer could play at beginner all the time, ignoring the puny rewards, and just own newbies for the inherent pleasure. If there is no forced advancement, this will happen and people will be unhappy.

#3 IamSeanConnery

    Member

  • PipPipPip
  • Elite Founder
  • Elite Founder
  • 91 posts
  • LocationMichigan, USA

Posted 06 December 2012 - 04:41 PM

View PostRedMercury, on 06 December 2012 - 12:28 PM, said:

Good use of incentives.

One issue is that a skilled grifer could play at beginner all the time, ignoring the puny rewards, and just own newbies for the inherent pleasure. If there is no forced advancement, this will happen and people will be unhappy.


You're right. That would definitely be a problem. Two ideas come to mind:
1)The first skill level could be locked for people playing for in certain time period(say first 3 months). That covers the newbies. After that, as long as they rewards are good enough, players would play in their own tier. That still won't stop the occasional griefer, but then again, there will always be mixes of different skill levels in general. up.

2) At this point, Win/loss ratio could actually be used to lock out the bottom skill level. Let's say at 70% wins, beginner would not available. They move up to a higher tier and it would naturally go down some. At this point they could still occasionally go back to beginner, but they could never troll there too much.

Yeah it's not perfect. But I think it would make people want to compete better and choose a skill level that fitted their playstyle. If you liked hardcore and think you have what it takes, go to Master and if you lose prepare to pay the price, and win reap the spoils. Or Maybe you are stuck in trial mechs or don't believe in a lot of customization so you choose to do trial mechs only.





1 user(s) are reading this topic

0 members, 1 guests, 0 anonymous users