• 0 Vote(s) - 0 Average
  • 1
  • 2
  • 3
  • 4
  • 5
OPEN DISCUSSION: HOW TO BETTER VALVE MATCHMAKING & BRING IN NEW PLAYERS
#21
imagine a future where you can play NA highlander at 2:00 pm on a summer weekday that would be awesome
I'd probably still be playing tf2 these days if I wasn't banned from tf2center

hlpugs is neat but it doesn't work well for the current highlander demographic. I don't wanna spam adding up to maybe get picked and hope people like me
I just wanna press a button and play some HL.

Rather than complaining about it on the forums for 3 years straight how about you just take the L and realize people don't like the platform.
  Reply
#22
Here are my two cents.

If new players are brought to the game. The in-game competitive mode would need to be updated.

I think the in-game competitive mode needs to stick to one game type, and as it stands now, it is NR6s. I'm going to make the presumption that this will not change. Highlander would be too many players for matchmaking, and there would likely be arguments about who would play what. 4s is not really a common game-mode. The other game type that could possibly be played is Prolander, but I doubt they'll just add one player.

Class Restrictions would likely need to be implemented for the higher competitive scene to give any care at all about valve matchmaking. You could keep this as simple as no more than one or two of each class, or you could implement the more complex class restrictions of traditional 6s. Most likely, if class restrictions were placed in at all, they would be no more than a certain number of a class.

The map pool of valve matchmaking is actually pretty good. It has a good mix of the most popular maps from every game mode. Just so you all don't have to look it up, here it is below:
Badlands
Badwater
Foundry
Gorge (Control Points)
Granary
Gullywash
Metalworks
Process
Snakewater
Sunshine
Swiftwater
Turbine
Vanguard
Viaduct
So not too bad, but there are some problems with it. Some maps could be swapped around a little. I think some of the 5CP maps could be removed in place for another decent KOTH map, seeing that there is only one. Swiftwater is a m a s s i v e map, almost too big for six players. It could be swapped out for Upward pretty reasonably. Now with the minor gripes dealt with, I can talk about Turbine. Capture the Flag is an unimaginable hunk of garbage of a game mode. It is certainly not in a good state for competitive play in Team Fortress 2 at any level. This map would need to be removed or CTF would have to be fully reworked.

The ranking in competitive means almost nothing. It's relatively easy for a half-decent player to get to Death Merchant. This would likely change better players queued for competitive, or if there was an occasional reset or re-placement period at the beginning of a "season".

This leads to the biggest problem of valve matchmaking, nobody cares about it. No good player will ever queue for matchmaking with the serious intention of improving. They do it because it's a way to kill time in a game they enjoy. If anybody wants to seriously improve they join PUGs, lobbies, or teams and run scrimmages. These are just more established and have better players in them.

When it comes to bringing in new players, that's a hard one to tackle. My girlfriend has recently started playing TF2 and I've been able to ask her the things that she doesn't like or has a hard time understanding. Everything below is what I've learned from her and what I personally find to be problematic.

The default settings of the game are absolutely horrible. I had to give her a version of my config and the basic launch options to make sure the game would run well and she would have the basic necessary options, like hit sounds, damage numbers, network settings, removing an FPS cap, and enabling the developer's console. These things should be stock with the game. After changing these setting I was still able to pull ~90fps on a $500 laptop with decent graphics settings.

The cheating problem in casual and competitive needs to be tackled aggressively. This is a massive annoyance for any player when they have to kick someone or change servers constantly because of cheaters. I understand that this problem will never be completely resolved, but it can be made better.

Random critical hits and random bullet spread need to be removed. Completely unnecessary random chance really has no place in TF2. I could explain this in-depth, but Uncle Dane made a good video going over the subject here: https://youtu.be/WHvwijT2ss8.

Everything under the "FIND A GAME" button should be fully playable. A decent tutorial and practice range would also be a good addition. These are standard with many modern FPS games. These would give new players the basics of TF2 and a place to practice or test weapons if they don't want to do so in a public game. The other game mode that should be fixed up a little is Mann vs Machine. As it stands now, MvM is dilapidated and glitchy. There have not been any new missions added since Two Cities. I'm not suggesting to add missions, I think that the game mode as a whole just needs some glitches patched.

After the game is cleaned up, it would need to be advertised somehow.
  Reply
#23
Before I took my one of my hiatuses I (vaguely) remember people who played b-pugs enjoying the b-pugs, not sure if that's no longer the case because everybody I knew has either quit or because they've gotten good enough to get picked in the a-pug, or if there is something else that makes them not want to add up anymore even for b pugs.

Hlpugs will probably never be good for the really low levels because captains do sort of need to be able to know the players they're picking (league pages sort of ameliorates this issue but not entirely, infinite roster riders as one example) but if I was currently a main/advanced player I think I would lose my mind playing tf2center with people who are literally fresh out of casual.  Unfortunately you do need a critical mass of players to actually get the b-pug rolling, but sort of the sacrifice you make for a more ""quality"" pugging experience.
  Reply
#24
(05-22-2020, 07:46 PM)Ruins Wrote: The map pool of valve matchmaking is actually pretty good. It has a good mix of the most popular maps from every game mode. Just so you all don't have to look it up, here it is below:
Badlands
Badwater
Foundry
Gorge (Control Points)
Granary
Gullywash
Metalworks
Process
Snakewater
Sunshine
Swiftwater
Turbine
Vanguard
Viaduct

bro what
  Reply
#25
(05-23-2020, 01:33 PM)pajaro Wrote:
(05-22-2020, 07:46 PM)Ruins Wrote: The map pool of valve matchmaking is actually pretty good. It has a good mix of the most popular maps from every game mode. Just so you all don't have to look it up, here it is below:
Badlands
Badwater
Foundry
Gorge (Control Points)
Granary
Gullywash
Metalworks
Process
Snakewater
Sunshine
Swiftwater
Turbine
Vanguard
Viaduct

bro what

I get that this is an RGL discussion, but if we want to practically bring in new players then the base game needs work. This is the map pool for valve matchmaking. I think that that map pool is pretty good, considering that Valve would want to show off all of their major game modes in TF2. Below that, in my first post, I put down my critiques for it. The biggest problem is Turbine, and while there may be some other problematic maps, like the choke points of Gorge and the Concrete area of Viaduct, it really isn't that bad and would not need a massive overhaul to fix.

The thread title is "HOW TO BETTER VALVE MATCHMAKING & BRING IN NEW PLAYERS." I saw a lot of reply posts talking about HLPugs, and I don't have a good answer or opinion about HLPugs.
  Reply
#26
Only a select few remain that played CTF and turbine pro. I can say from first hand experience the game mode and map failed to be any good or entertaining. Which was forever dropped after s8.
I will always be a pyro main <3
  Reply
#27
I prefer the "choose your own slot" system that TF2Center has as opposed to HL Pugs' picking system. It feels more casual and inviting that way.
  Reply
#28
(05-23-2020, 05:26 PM)Ruins Wrote:
(05-23-2020, 01:33 PM)pajaro Wrote:
(05-22-2020, 07:46 PM)Ruins Wrote: The map pool of valve matchmaking is actually pretty good. It has a good mix of the most popular maps from every game mode. Just so you all don't have to look it up, here it is below:
Badlands
Badwater
Foundry
Gorge (Control Points)
Granary
Gullywash
Metalworks
Process
Snakewater
Sunshine
Swiftwater
Turbine
Vanguard
Viaduct

bro what

I get that this is an RGL discussion, but if we want to practically bring in new players then the base game needs work. This is the map pool for valve matchmaking. I think that that map pool is pretty good, considering that Valve would want to show off all of their major game modes in TF2. Below that, in my first post, I put down my critiques for it. The biggest problem is Turbine, and while there may be some other problematic maps, like the choke points of Gorge and the Concrete area of Viaduct, it really isn't that bad and would not need a massive overhaul to fix.

The thread title is "HOW TO BETTER VALVE MATCHMAKING & BRING IN NEW PLAYERS." I saw a lot of reply posts talking about HLPugs, and I don't have a good answer or opinion about HLPugs.

objectively this map pool is terrible and if u wanna bring in new players, have maps that aren't dogshit for 6s. half of these maps are awful and are not reflective of a true competitive experience
  Reply
#29
(05-23-2020, 12:15 AM)spring rolls Wrote: Before I took my one of my hiatuses I (vaguely) remember people who played b-pugs enjoying the b-pugs, not sure if that's no longer the case because everybody I knew has either quit or because they've gotten good enough to get picked in the a-pug, or if there is something else that makes them not want to add up anymore even for b pugs.

Hlpugs will probably never be good for the really low levels because captains do sort of need to be able to know the players they're picking (league pages sort of ameliorates this issue but not entirely, infinite roster riders as one example) but if I was currently a main/advanced player I think I would lose my mind playing tf2center with people who are literally fresh out of casual.  Unfortunately you do need a critical mass of players to actually get the b-pug rolling, but sort of the sacrifice you make for a more ""quality"" pugging experience.

can more advanced players add up Smile
im tired of being the only "bad" player to add
  Reply
#30
I think the reason nobody wants to add up to HL pugs and that no one want to say is that if a lot of lower elo players add up, some idiot captain will end up picking that one invite sniper who didnt get in A pug and he will ruin the pug experience for everyone else. At least in in-house pugs you can restrict invite players.

On a TF2c system, I mean it can go better or worst because then they can add up to whatever they want, but like I said at least a TF2c system you could directly restrict some lobbies to be lower elo only without having to worry about a funny captain wanting to win the pug so he picks the fatkids from pug A
what the hell am i doing here?
  Reply
#31
(05-25-2020, 02:24 PM)Gobitoe Wrote: I think the reason nobody wants to add up to HL pugs and that no one want to say is that if a lot of lower elo players add up, some idiot captain will end up picking that one invite sniper who didnt get in A pug and he will ruin the pug experience for everyone else. At least in in-house pugs you can restrict invite players.

On a TF2c system, I mean it can go better or worst because then they can add up to whatever they want, but like I said at least a TF2c system you could directly restrict some lobbies to be lower elo only without having to worry about a funny captain wanting to win the pug so he picks the fatkids from pug A

Playing with and against good players makes you better 
Who cares if you eat shit it’s a pug I eat shit every other time I play hlpugs and I still have fun and add up
  Reply
#32
(05-23-2020, 07:48 PM)pajaro Wrote:
(05-23-2020, 05:26 PM)Ruins Wrote:
(05-23-2020, 01:33 PM)pajaro Wrote: bro what

I get that this is an RGL discussion, but if we want to practically bring in new players then the base game needs work. This is the map pool for valve matchmaking. I think that that map pool is pretty good, considering that Valve would want to show off all of their major game modes in TF2. Below that, in my first post, I put down my critiques for it. The biggest problem is Turbine, and while there may be some other problematic maps, like the choke points of Gorge and the Concrete area of Viaduct, it really isn't that bad and would not need a massive overhaul to fix.

The thread title is "HOW TO BETTER VALVE MATCHMAKING & BRING IN NEW PLAYERS." I saw a lot of reply posts talking about HLPugs, and I don't have a good answer or opinion about HLPugs.

objectively this map pool is terrible and if u wanna bring in new players, have maps that aren't dogshit for 6s. half of these maps are awful and are not reflective of a true competitive experience

So here's the deal, Valve didn't make the matchmaking system to appease the hardcore sixes community. While the payload maps may not be what traditional sixes has played on, they're a legitimate form of competitive TF2. The design philosophy of matchmaking would arguably have to be changed to remove them, which is an option. So here, let's go through every single map.

Badlands
Pretty odd second and a very tight last point, but has one of the best middle points. The flow of the map isn't fantastic and I could see a good argument for this one being taken out. Still a decent map overall and has its place in competitive TF2.

Badwater
While this isn't played anymore, it's still a solid map. A good amount of space in all the points and the points all feel unique and fun to fight in. A solid payload map for the team size and game mode.

Foundry
Here's the first map I'll say is a legitimate problem. Foundry isn't an awful map, but all the points have issues of their own. Lighting is a problem in some areas, especially the space between mid and second. The spawn on last can be choked out easily. It would need a sizeable change to be considered for modern competitive play.

Gorge
Way too many tight corridors. The mid feels awful to fight in, as it's horribly claustrophobic. Second is a little better, but all the doorways are just too restricting. Same story for last. This map is a problem and would need to be removed.

Granary
Granary is a solid map. It suffers a little from the same problems that Gorge does, but to a much lesser degree. A pro version of this is played currently, and while it's not the most popular map, it is well respected and deserves its place in the map pool.

Gullywash/Metalworks/Process/Snakewater/Sunshine
These are the big sixes maps. I'm going to assume that this is what you mean by a true competitive experience. I could go into the great points about the maps, but I'm just going to mark that all of these are good and deserve their place.

Swiftwater
Swiftwater is a staple highlander payload map, but with six players, the map is too big. A good example of this would be the second point, where there are three decent doors. Teams can just attack the least defended one and bomb up or walk to the high ground. I'll mark this is a problem map and say it needs removed.

Turbine
No, I've already explained this one.

Vanguard
Interesting map. I feel that most of the community didn't hate this one but knows that there are better maps. I'm going to go ahead and say this is fine, the flow of play on it is good enough to keep in this map pool

Viaduct
This is the only KOTH map in the matchmaking rotation. I think that if this was removed, another KOTH map would need to be added to represent the game mode. KOTH is played in almost every form of competitive TF2. I don't think stock Viaduct is very great, and I don't think any of the other KOTH maps already in the game are ready for sixes play. This is an impasse for me when thinking about this response, so I'll mark it as a problem map. I think that if a version of Clearcut or Product could be added, then it should be in this rotation.

So there, five maps are a serious issue in the map pool out of fourteen, not half. This is still a significant percentage of the maps, and that's why I brought it up in my initial post. This matchmaking, even if it receives work, will not be an in-game version of TF2C. Valve will want to represent all of the major game types of TF2 and that is likely something that we'll have to accept.

Here's what I would do with the map pool, under the assumption that I had to include every major game mode (except Capture the Flag) in TF2.

CP
Gullywash
Snakewater
Metalworks
Sunshine
Process

PL
Badwater
Upward
Borneo

KOTH
Viaduct (Hopefully Product)
Highpass
  Reply
#33


its unrealistic (and unproductive, frankly) to have a discussion about changes that "should" be made to tf2 with regards to the game itself. literally any real progress tf2 ever made competitively was brought about by the community's collective effort. even balance changes to an extent. the scene (at least hl) was at its peak between 2012 and mid 2015. idk how long you have been around but for the past couple years competitive tf2's main goal has been sustenance rather than growth. tf2 loses players quicker than it gains them. its a dated game with mechanics that don't appeal to mainstream audiences, even in a casual setting.

but to get back on-topic, the main reason you dont see growth in the comp scene is literally just lack of interest. most of the people who play tf2 are strictly casual players who either play tf2 a couple hours a week, are newbies who installed the game 20 minutes ago, or are just pubstars/traders who maybe dipped their toes in the scene or just decided to never play it. thats something you have absolute no control over. tf2 is not a competitive game inherently, so trying to make it appear that way to casual audiences is a lost cause.
normal human
  Reply
#34
(05-22-2020, 03:30 PM)Jimbob Doohicky Wrote: imagine a future where you can play NA highlander at 2:00 pm on a summer weekday that would be awesome
I'd probably still be playing tf2 these days if I wasn't banned from tf2center

hlpugs is neat but it doesn't work well for the current highlander demographic. I don't wanna spam adding up to maybe get picked and hope people like me
I just wanna press a button and play some HL.

Rather than complaining about it on the forums for 3 years straight how about you just take the L and realize people don't like the platform.

This. Anyone can scream "add up" repeatedly and act like they're helping in some capacity, but the opinion of 10 people on the forums begging others to like the system is not changing the fact that hlpugs isn't a popular format for pugs for the majority of players in the highlander community. It's the equivalent of the pug groups that happen through discord and then always die after one month, except with hlpugs you have a website to do back end things for you.

If you can't "fix" the system or otherwise change the community's opinion on it, it does not matter how or why it doesn't work for the majority of players. The problem is that you clearly aren't convincing more than ~25 people to add up by just saying "add up" on the forums or by pinging 20 discords telling @everyone to add up. So maybe try something else.
  Reply
#35
3 more for bpugs add up

if you are not good and add up, other not good players will add up
all of a sudden we have bpugs
  Reply
#36
The HLpugs discussion has happened many times over the last couple of years, tl;dr clearly in-house pugs will always be a better experience because they're curated by hand, but the failure of arguably the best current automated pug system for HL comes down to the unwillingness of players to widely adopt it in a chicken-and-egg problem. A leap of faith is needed, whether it will happen is a different argument.

For the short period of time that it was around, TF2PL was fantastic. Despite the technical limitations of Faceit's platform, new players could get together with experienced players consistently and games were happening at various times in the day. Its class picking system had the best middle-ground between pugchamp's automated system and in-house groups' total indifference to classes: you displayed the classes you're willing to play, then a game starts with 12 players and you had to organize your classes correctly no matter what. Without the need to wait for the fulfillment of often crippling class requirements, games just start as soon as you have enough people, and more often than not they still get classes they're alright playing. A rudimentary MMR system gave you an idea of what kind of experience they had. Late in its lifespan, three skill divisions for the Total Newbies, the Uncompromising Elite, and Literally Whatever The Fuck. You were always covered.

RGL not having the greatest resources for introducing new players to competitive play is well known, and also not really unique to RGL. Face it, we've always kinda sucked at bringing new people in. I've been looking into changing that.

TF2Center is terrible. I don't mean that in the sense of the platform not working or not satisfying its players, but that there isn't actually a great transition from TF2Center to "real" competitive TF2. It's a hollow shell, an imitation of the form. But we have to accept that it's also the most accepted competitive entryway among new players, and we have to learn from it. Gobi makes a stellar, stellar fucking point: you go to TF2Center, you sign in, and you play. No fiddly menus, no articles in the front about the latest news. Newbies don't care about any of that, they want to get a taste of the competitive experience and we shouldn't be making it any harder. The problem comes afterwards, in how TF2Center doesn't actually make you mingle with experienced players, how it tacitly encourages players class-camping in lobbies, how the same four people make the same four lobbies with the same two maps over and over again, never to be exposed to the wider community.

Here's what I propose: TF2PL again, streamlined to TF2Center's laser point. Throw in some Highlander shit instead or something, I dunno. On the front page, just games. On sidebars, information resources on how to get into "real" comp if that's what they're interested in. Links to leagues and other fun stuff. A system that both newbies and experienced players can both actually get behind, automated to eliminate any path to becoming a closed community.
  Reply
#37
DISCLAIMER: I have no idea how feasible this is as I have absolutely zero background in programming/computer science

Couldn't you just make a tf2center clone that uses an algorithm based on RGL's API to automatically balance lobbies? Even if it was rudimentary I doubt it would be much worse than the manual captain system that HLPugs has, especially considering low level players are rarely picked if there are more than 18 added up. This would guarantee newer players a shot at playing in PUGs with experienced players, so nobody is alienated. Players with no team history are more likely to be placed with decorated players, and so on. If you already have a history of playing on teams it maybe looks at your division history/team records or something and then goes from there. No idea how it would work in practice but on paper it sounds pretty interesting. Also, since you'd have to make an RGL profile to play it would likely encourage new visitors to learn more about what the league has to offer if they're already at the website.
normal human
  Reply
#38
just out of curiosity, why isn't tf2stadium used? I wasn't around when it was launched and it's been dead (but up-and-running) for a long time. It's open source as well, so I imagine it wouldn't have to reinvent the wheel to implement what wax suggested, though it should be noted it's in a rather obscure language (Go).

However, if that tf2center clone didn't work, why would ours? Is such a thing even worth the headache of setting up, if it doesn't significantly eat into the market share of tf2center?
  Reply
#39
Stadium is dead because not enough people used it, it occupies the same niche as tf2center (easy to join very casual pugs) and tf2center is good enough for the majority of the people who use it, while having much more name recognition especially for people with 0 competitive experience. Tf2 playerbase is also too small to really have two active lobby sites, so over time everybody will converge toward the more popular one.

I think the hope is that if you create a lobby system through RGL the RGL name can kind of carry it a bit especially with people who currently play in the league. As to whether that will be enough, I have no idea. Think the biggest issue is still attracting the people with 0 competitive experience, for that segment tf2center and maybe even UGC have greater name recognition.
  Reply
#40
(05-27-2020, 06:13 PM)wax Wrote: DISCLAIMER: I have no idea how feasible this is as I have absolutely zero background in programming/computer science

Couldn't you just make a tf2center clone that uses an algorithm based on RGL's API to automatically balance lobbies? Even if it was rudimentary I doubt it would be much worse than the manual captain system that HLPugs has, especially considering low level players are rarely picked if there are more than 18 added up. This would guarantee newer players a shot at playing in PUGs with experienced players, so nobody is alienated. Players with no team history are more likely to be placed with decorated players, and so on. If you already have a history of playing on teams it maybe looks at your division history/team records or something and then goes from there. No idea how it would work in practice but on paper it sounds pretty interesting. Also, since you'd have to make an RGL profile to play it would likely encourage new visitors to learn more about what the league has to offer if they're already at the website.

you could do some stupid shit to just fish for the hl div + history of any one player even without an api

it would be exceptionally feasible if rgl has an api, because i didnt know they did

the hard part would be cloning tf2center

(05-27-2020, 09:01 PM)spring rolls Wrote: I think the hope is that if you create a lobby system through RGL the RGL name can kind of carry it a bit especially with people who currently play in the league. As to whether that will be enough, I have no idea. Think the biggest issue is still attracting the people with 0 competitive experience, for that segment tf2center and maybe even UGC have greater name recognition.

loads of new players queue for prolander pugs in the discord, which have been becoming more and more popular. i dont think attracting a playerbase would be too hard unless the experience itself is alienating
  Reply


Forum Jump:


Users browsing this thread: 1 Guest(s)