Matchmaking Algorithm for Normal Warzones

[L-RANDLE]

Indeed. The troll (or is it stupid) is strong in that one.

 

I vote.... Short Bus ridah...

[Chrisweaver]

/bump

[UncelSam]

And you are wrong, on multiple levels.

Go troll the premade thread....

 

So u dont dare to fight other premades, coward teenager premade runner?

[Tezeretz]

Nice post Op.

 

Sadly you do not work for bioware...and there coders are good at two things.

reskinning old items with new colors for cartel market....and breaking currently working features.

 

The current match making system is far less complicated. So they likely wont wanna change it.

Its as follows:

 

People who click qeue PVP= get dumped into a warzone.

People who do not click qeue PVP= dont get put into a warzone.

 

They USED to have a variable.

 

Republic get put together against Imps. and vice versa.

But it was later changed to:

Same faction can be dumped together, but can fight same faction or opposing faction.

 

 

Thats as complicated as bioware can handle at the moment.

[L-RANDLE]

So u dont dare to fight other premades, coward teenager premade runner?

 

Age doesn't determine wisdom, and whether or not you have a pair....

 

Grow one please.....

[L-RANDLE]

So u dont dare to fight other premades, coward teenager premade runner?

 

Age doesn't determine wisdom, and whether or not you have a pair....

 

Grow one please.....

[Zunayson]

Age doesn't determine wisdom, and whether or not you have a pair....

 

Grow one please.....

 

Why do people always say to grow ********s? I mean they're flimsy. Grow a ******. Man those things can take a pounding!

[Khan-doh-hh]

How to measure algorithm success or end goals are important but I will mostly ignore this for now (assumed that most players share a general view on what the point of this is, and post will be too long otherwise). Other important #'s for algorithm:

 

Total keystrokes or actions/second. Good players usually do WAY more actions per second than weaker players. Keeping track of this over time (maybe last 10 wz's or some other interval that regularly resets?) would be a great way to very quickly sort the whole population into three groups: Low/medium/high actions per second players. Clickers and turners would likely populate the lower pool. This would be a pretty reliable way of sorting players into three skill pools even if no other matchmaking was used. Node guards would have a disadvantage. This kind of idea could be used in many other ways, but even if it wasn't practical as part of matchmaking it would be awesome to see this on the scoreboards.

 

An extension of the last idea would be total unique abilities used. This would be a different ratio for each skill tree. I'm not sure how many abilities each AC has but the idea is that the greater % of the available abilities used is better. I don't think items/consumables should be counted toward this. I know that certain specs have no use for all their abilities and this would make some people use abilities they didn't actually want to. Not sure if this is a good suggestion or not.

 

Each AC tree could also have milestones for 1-3 of the abilities they are supposed to use most in that spec. Ex. Jugg gets player rating if he hits 20 smashes in a match. Your rating goes up if you are meeting your marks each match. Adds metagame but could force certain abilities.

 

I do think that your player rating should be based both on individual skill and team performance. At first it is more important to sort by individual skill for balance, but ultimately the game is about team synergy and group objectives. Your team can both carry you and be the reason you lose. The majority of player rating should be decided by individual performance but a not insignificant contribution should come from team results. Further discussion on this is necessary.

 

The following kind of idea could be used to boost individual ratings based on "team" performance: Top healer on the enemy team also took the most damage. Your team gets bonus for this. Things like this.

 

PvP legacy achievements are also way too underdeveloped. There needs to be more obscure achievements that promote objectives and premades. For example: Enter wz as a 4 man premade and all 4 players get unbeatable before any of you die. All premade players must stay within 60m radius of each other at all times. Instead we have complete 20 solo kills on every map 50 times (or whatever it is) . How are you realistically going to accomplish this in so many matches without purposely ignoring objectives and breaking off from the group? This is a showing of brutal redundancy and lack of creativity.

 

Whatever happens needs to start happening now. In my view PvP players tend to be much more likely than PvE players to switch games. PvP simply exists in too many other games across many genres. It isn't hard to get your fix elsewhere. People want to love SWTOR PvP but the effort just isn't there right now. If BW can't make significant changes to the current PvP "matchmaking" it will inevitably be swallowed up by other games. An amazing PvE game has the potential to eventually go down in the "Video Games Hall of Fame" but a great PvP game will do it much quicker. I know this is a progression game but take advantage of E Sports or new try new models (like Huttball). Payoffs will be big! Sorry for saying game so much.

 

One way to measure player satisfaction? Self reports. Put two vote boxes at the end of each wz. Did you enjoy this wz? Was this wz balanced? Etc..

 

/post

[Tamanous]

Ever been an update on this?

 

Just left a warzone where the other team was at least partially from one guild and every single player on their team over level 50. Our team was 100% pug and averaged in the 30's level range with NO healers.

 

This is a complete waste of our time. Bolster cannot even come close to balancing this. Bolster is a lie.

[cashogy_reborn]

Oh hey, this is my thread

 

 

Supposedly we are getting some kind of matchmaking based on your RWZ Rating, coming in 2.4. Its unknown how well it works yet; we'll know within a week tho.

[Tamanous]

Oh hey, this is my thread

 

 

Supposedly we are getting some kind of matchmaking based on your RWZ Rating, coming in 2.4. Its unknown how well it works yet; we'll know within a week tho.

 

 

Ranked isn't the issue. Ranked pvp must have match making or rankings don't work. What players have been asking for since LAUNCH is to have some sort or match making in normal warzones too.

 

In the last few days I've seen (including my previous example) situations where there are 5-6 of the same class on one team, level ranges way out of whack and teams that get 3 healers while the other gets none. This is very common and we all know it and have experienced it.

 

It goes beyond class, level and spec balance though. There are players who constantly top charts in their role (such as myself) because they are experienced players. Many others aren't. They especially get frustrated as lower level warzones must also facilitate learning curves. Slapping these people into matches against pre-mades of level 50+ experienced players is ridiculous. Each match offers stats and metrics Bioware can use for match making. Players who constantly do low damage/healing and bad kill/death ratios can be matched with similar skilled players until they learn the game. Give them a healer or 2 and let them have fun. Pre-mades with players who always top charts can be put up against similar players.

 

This isn't rocket science.

 

Many of us have been in experienced guilds and we all know why many experienced players enter these games in groups is because they can laugh about how bad they destroy players. I've heard the discussions many many times in G chat and on voice. This isn't the intention of Bioware yet all this time ignoring this issue puts the blame directly on them.

 

This isn't about hand holding. It is about offering equal challenge to all. The change to both win and lose by player/team skill and not being dictated by mechanics. I used to also gloat about killing someone in 3 globals but now I just see it as bad game mechanics.

[cashogy_reborn]

Ranked isn't the issue. Ranked pvp must have match making or rankings don't work. What players have been asking for since LAUNCH is to have some sort or match making in normal warzones too.

 

In the last few days I've seen (including my previous example) situations where there are 5-6 of the same class on one team, level ranges way out of whack and teams that get 3 healers while the other gets none. This is very common and we all know it and have experienced it.

 

It goes beyond class, level and spec balance though. There are players who constantly top charts in their role (such as myself) because they are experienced players. Many others aren't. They especially get frustrated as lower level warzones must also facilitate learning curves. Slapping these people into matches against pre-mades of level 50+ experienced players is ridiculous. Each match offers stats and metrics Bioware can use for match making. Players who constantly do low damage/healing and bad kill/death ratios can be matched with similar skilled players until they learn the game. Give them a healer or 2 and let them have fun. Pre-mades with players who always top charts can be put up against similar players.

 

This isn't rocket science.

 

Many of us have been in experienced guilds and we all know why many experienced players enter these games in groups is because they can laugh about how bad they destroy players. I've heard the discussions many many times in G chat and on voice. This isn't the intention of Bioware yet all this time ignoring this issue puts the blame directly on them.

 

This isn't about hand holding. It is about offering equal challenge to all. The change to both win and lose by player/team skill and not being dictated by mechanics. I used to also gloat about killing someone in 3 globals but now I just see it as bad game mechanics.

 

Easy partner.... what did I write in the OP?

 

 

For starters, if you are whining about lowbie PvP balance, the door is that way --->

Lowbie PvP has absolutely nothing to do with balance in any way, shape or form.

 

And there is equal opportunity for everyone to succeed, even currently. People have a hard time understanding that equal opportunity =/= equal result. If the average player took personal responsibility for their performance (rather than blaming it on some aspect of the other team, that they had ZERO control over), and worked to improve their game a little bit each day, we wouldnt even need matchmaking for regs.

 

But the average player is a window licking moron. So we need matchmaking so they arent brutalized for attempting to distinguish their monitor's flavor in the middle of a warzone.

 

The point of this post was to show how easy it would be to develop an algorithm that was mostly dependent on personal performance. Thats the kind of matchmaking that would be useful for Regs. We will likely *never* be getting matchmaking in regs.

[ManiacDavis]

I really enjoy your algorithm. The only problem I see is that you can make clutch plays or help your team win which is not accurately reported in your end game statistics.

 

For example, stunning an enemy opponent so that your team can capture a node is not reflected in statistics. Cleansing a friendly target is not represented in the statistics. Slowing an enemy team is not represented (think voidstar). All of this time helping your team win the game in turn lowers your DPS/Heal numbers.

 

This is why if your team wins, you should not lose rating because you could have been an all-star support leading your team to victory, but it is not reflected in the end game statistics. Other than that, I like your algorithm and think it would be highly beneficial in normal warzones.

[JP_Legatus]

so the premade v pug issue is a big one. lot of discussion/debate surrounding it. in my mind, the best way to solve it is to institute some kind of algorithm based matchmaking that looks at your individual performance, and not at whether you win or lose (if your PUGing, you obviously dont want someone else's mistakes to affect your rating).

 

(edited for space)

 

I think this is a well-intended idea, but it won't realistically help anything. Let's look at how this would play out.

 

The first problem with any individual rating in a team game is that the teams you play with impact your rating. Just being in a team of guys you play with a lot is an advantage because you know what they're doing. So let's say you play with 3 buddies a lot, and you're not the greatest players but knowing each other helps a bit. Let's say in a 1500 starting rating, it might impact your score by 10 points being with friends you know. Now multiply that times 16 since there are 16 players each with their own rating who may also be doing this.

Also basically if I play with good teams a lot, my rating will be high, then if I decide one day to solo queue, my rating will be way higher than it normally would be if I played alone. Say I queue with good teams half the time and solo queue the rest of the time. Say my leet teammates pad my rating about 50. Again multiply that by 16.

But wait, that isn't the only issue. Simply playing with 3 people who share voice comms and using real time communication is a big advantage. So if, IE, you play with a guild (as many do) you are always performing at a level above what your individual rating would suggest when you are with that guild. Say voice comms accounts for a rating pad of another 50. Again multiply that inaccuracy by 16.

Now remember that swtor isn't just a team with 16 players, it's a game where composition impacts the game dramatically. The optimal mix of players is (generally agreed as) 2 healers, a tank, a node guard, and 4 dps (2 mara and 2 sniper if possible). Say my guild likes to practice for rateds with a healer, tank, dps, nodeguard composition in our 4 man group when we queue for regs. I'd argue that's worth at least 100, but for arguments sake lets just say 50 again. Multiply it by 16. Oh wait, this means you're always serious in regs too and you're never off in a corner screwing around hunting Gudarzz with 2 deception sins. Add another 50 and multiply that by 16 as well.

 

I think I was pretty reasonable with the numbers up to this point. I never gave more than 50 points to a single "advantage", and I could argue many of those factors impact a players rating by more than 100. I could also come up with a huge list of other things, such as having maras vs having concealment op dps, or merc healers vs op healers. Maybe you're queuing against pub guilds a lot because of the time of day you play, or maybe the opposite: nobody good plays when you're on. Maybe you just finished gearing your sorc but you've had a geared sage for years and are a top sage in the game. Maybe your power goes out a lot but you're really good. The list of things that impact these ratings goes on forever.

Anyways, the total difference from just the stuff I listed? 410 x 16, or a difference of 6191.

 

So what's the point who cares? Put it in anyways! Well basically all I'm saying is this system, as well designed as you could possibly make it, would still not change warzones at all, and would probably just result in more elitist ******** (such as the kind we already see from rated teams who typically beat the exact same team every day and call themselves the best players ever) than it would in actually helping any balance in warzones.

 

I hate to be the bearer of bad news, but the reality is that rating systems don't work in small communities. You need a massive quantity of different players going against each other constantly to ever get a system like this to work, and how many pvpers really still play this game?

Honestly I'm pretty sure this game is in "milk it til its dead" mode so I doubt you'd see anything resulting from an idea like this, even if it would work to begin with.

 

I've had some experience with this kind of system in a small community before and that's why I bring it up. It didn't really do anything for the game either way, but it turned the community into a bunch of rating whores.

 

IDK man, all I can say is good luck to you and I hope something positive comes out of this, but don't hope for drastic improvements to the game.

[PloGreen]

So u dont dare to fight other premades, coward teenager premade runner?

 

There's another thread for this kind of qq, use that.

[Savej]

Assuming that we have a lot of players from all brackets q'ing it works.

 

At this moment on POT5 @55 there are 3 wzs running (all voidstars - not sure what that means). There aren't enough players to break that into brackets without adding a lot of wait times to qs.

 

X-server is the magic bullet but it's too late.

[cashogy_reborn]

I really enjoy your algorithm. The only problem I see is that you can make clutch plays or help your team win which is not accurately reported in your end game statistics.

 

For example, stunning an enemy opponent so that your team can capture a node is not reflected in statistics. Cleansing a friendly target is not represented in the statistics. Slowing an enemy team is not represented (think voidstar). All of this time helping your team win the game in turn lowers your DPS/Heal numbers.

 

This is why if your team wins, you should not lose rating because you could have been an all-star support leading your team to victory, but it is not reflected in the end game statistics. Other than that, I like your algorithm and think it would be highly beneficial in normal warzones.

 

I think that is an inherent flaw in the way its dependent on the WZ stats being collected.

 

However, I think that the players that are frequently contributing on that level are going to produce good stats anyways.

 

 

Just spit-balling an idea here, but what if you were to get more objective points for captures depending on how many enemies are in the immediate vicinity? It wouldnt solve every scenario, but would make clutch CC during cap opportunities contribute more to your objective score