I think you might be focusing on the wrong site, Slin. If your intention is to make a better platform than tf2c/stadium/faceit then mixchamp should be exactly that. It would even do what you want it to already: balance the teams without captains.

I don't think it 'starting immediately' is an issue. Sure it's not perfect but 12 people shouldn't have to wait for another 12 people to add up just to have a bigger pool to balance with. That's impractical. I'd rather play 2 games with teams that might be less balanced than 1 game in that same timespan because there was no 4th medic for half an hour.

The only reason I haven't switched to mixchamp yet is because it's not alive when I want to play. This can be solved by promoting mixchamp more. The new map cups helped with that but it needs even more. Organize viewer mixes with the site when you stream, promote it in your videos, ask other streamers to do the same.
The newbie mixes could promote it as well. There's plenty of people there who'd be down to play more without having to rely on tf2center and aren't confident enough to add up to pugchamp

Mixchamp might not be 100% what you're looking for but it's still the best thing you can use by far to achieve it

Why not try to get to a middle ground between having captains or not ?

For example, we could have the algorithms to make the teams and then give the option to the captain to replace X players (I was thinking 1 or 2) with another player from the queue that didn't got picked by the computer. That way, we could keep the benefits of a fast picking process while avoiding problems between players that don't want to play with each other.

I like your ideas about improving pugchamp system very much.
The thing with me is I have 0 comp experience with 6s or Higlander(excluding 1 lobby that went horribly 'cos I didn't know how to play Medic) and I can't really tell what would the optimal system be based on experience-just logic and assumption.
What I can address is the in-game matchmaking system that is crucial for the wider audience(comp and casual).
Current casual servers are filled with a variety of people - friendlies, trolls, tryhards, fresh gibuses and people who want balanced matches(of course some of the groups overlap).
If Valve made several queues (and a good training mode explaining all the important stuff properly) to filter out the skill groups/ preferences it would appeal to more people.
1st queue: PUBS
Bring back the good old pubs. Mostly recommended for new players to learn the maps mechanics etc without the pressure of doing badly. Also more skilled players would troll around and experiment with different loadouts without their inner pressure of doing bad.(or is it just me?)
2nd queue: CASUAL
Pretty much the current Casual system, but that tries to pair players with similar skill level and focuses on the objective. Very close to the CS:GO casual- gaining xp but not impacting the comp mode. No weapon bans or class limits
3rd queue: COMPETITIVE
"Pure" comp matchmaking with an ELO ranking system and weapon bans/class limits and possibly several types(6s,HL, prolander, 4v4, ultiduo etc. -also available in casual for practice) for different preferences. It would serve as a genuine and reliable rank/evidence for the leagues and possibly implement them in its core.

I hope that this didn't go off-topic too far away (barely mentioned the pughchamp so waaay off-topic) as I feel this is a very important issue for the TF2 community and TF2 in general.

Best regarts,
-wannaBe

P. S. Keep up the good work with the vids <3

Am I wrong in assuming that mixchamp actually does try to balance teams?

I always assumed that players had a hidden elo (like how elo is now hidden on pugchamp), and that's what was used to calculate the "dominance score"

http://i.imgur.com/sSvOyrA.png

Is it actually a completely random spread of players to teams?

https://www.reddit.com/r/truetf2/comments/4qc208/want_to_pug_6s_try_mixchamp/

I think several of the situations you pose would be solved by an improved picking system. Keep it simple based off skill level for balanced pugs. IMO the ELO system “looks” great but is unreliable and inaccurate.

What would the problem be with adding an additional feature to the picking process where you make players pick their skill level in addition to picking the classes they want to play?

After that, Captains can still pick the players they want but would be restricted in choices for each phase so invite stacking would not happen.

This would be a simple example of the captain picking system based off skill level.
Obviously there would be more to it, but just to keep the idea basic:

Picks 1, 2 and 3 (Invite and IM Level) – Open Restricted
Picks 4 and 5 (IM and Open Level) – Invite Restricted
Pick 6 (Open Level) – Invite and IM Restricted

As someone who's been in the PugChamp slack since nearly day one, I've already seen multiple plans to make the service better, but the number 1 reason issues still exist is the development process is slow. There's only 2 developers and they can only really work on things in their free time due to school/jobs outside of TF2. You can have the best ideas in the world, but it won't matter if it can't really be implemented quickly, so developmental help would be the first thing that would have to happen if anyone wants to help make PugChamp better.

The next thing is money related. PugChamp does not operate at a profit or break even, and therefore most of our servers are donated. In NA we were at 4 servers on PugChamp and 2 servers on MixChamp for a long while, with all 4 PugChamp servers being used only a handful of times a month. We had to downscale to 3 and 1 purely from someone who was donating servers to us deciding they didn't want to anymore, and had to scramble for some replacements. Once we're able to support a larger amount of players at once on the site from the additional dev support, then we'd need the money, or even temporarily donated scrim servers if we're able to get to that point (loaning out your personal scrim server for a specific window of time for the site to use), to play on.

There's a couple things on your google doc that I don't really agree with, but there's no point in even getting deep/specific into that at the moment because nothing can really change at a faster rate until assistance is given via the two points above.

who actually does this

Best suggestion in the thread

One thing that really shouldn't be allowed imo is the constant killing of pugs when people miss the drafting, sometimes that can be acceptable if it's an inactive 12 player pug but if the idea of having 100 ppl added up I really think that pugkilling should come with a penalty since it just slows down the process even more.

I can assure you it was either never ready to be used, or not a single invite player knew it was ready for use.

I can assure you you're either remembering incorrectly or only speaking for yourself.

I'm not sure how you could expect those to be run when not a single one of my teammates, besides maybe Vand (since he was an admin), was invited to that though. I have one of the most pugs on that website, I asked at least once a week after the domain was up but servers weren't connected to it, but tsc and erynn just replied that they were busy, which is fine.

I seriously don't believe the website was ever up with servers connected past the beta/alpha phase. I checked the website and I asked tsc/erynn frequently. It's definitely not the players' fault if I had to put in that much effort to not even see an official opening, not beta.

? you think i care elo? idc? and? for what?

Hey Jarrett,

I understand that PugChamp hasn't seen much development since August 2016, as I talked to Erynn about this at length. I'm willing to look at starting something new if that's what it takes. At this point in time Erynn says she's still working on the site and still trying to develop it so I'm not necessarily trying to hijack her project. This whole thread started out for me as I was thinking of ways to expand PugChamp and grow the top end of the pugging scene.

I agree with a lot of you guys in this thread that maybe what we need is an improved MixChamp instead of an improved PugChamp. Take the auto-picking but adapt the system so that it can handle more than 12 players at once to create balanced teams. Assuming we use a good skill rating system, I think it would work then.

My motivation for expanding PugChamp is as follows:
1. I think pugging is the lifeblood of comp TF2 and I don't think there is a good option for new players to play 6s and get into the game. We begged Valve to create this experience for us but it didn't go very well, and we should look into taking matters into our own hands through 3rd party MM services. By creating a centralized pugging experience you have a path for players to play against people of the same skill level while also working towards playing against higher level players.

2. PugChamp is already pretty successful and has most of the playerbase. I'm not looking to get 100 people in the queue right away, I was simply looking for a way to expand PugChamp past the 36 person capacity problem. First we make it possible for 40 people to add up, then work towards 60, 80, 100 as we expand. As many of you mention in the thread you've never seen 40 people add up to PugChamp and it's not because nobody knows the site exists -- it's because the 40th person knows they won't get picked. Rather than waiting around to not get picked they go to other services that cater to new players instead like TF2Center and MixChamp. Newbie Mixes are great but aren't scalable since they only happen at a specific time of the week and require a lot of manpower. Automated picking can solve this.

3. If we have an idea for sustainable pugging at scale then I'd knock on every door and see if they could build it and how much time it would take. If MixChamp can't do it then I'd ask FACEIT or Valve or whoever. I've also asked some of my co-workers to see if they'd be willing to contribute lol. It's a pet project for sure, but I feel it's a good cause.

I'm not really asking you guys to do the work I'm just trying to get your help to vet the idea. If any of you are software developers that's a plus.

I didn't send out invitations so I can't comment on that (or whatever happened in EU) but what I do know for sure was that in March of 2016 there were attempts in NA to start up the beta testing of the pro league, and there was a server connected to the site because I played one of the games that erynn tried to gather people for (she asked multiple times so I'm sure there were at least a handful of games, even if they weren't 100% invite players, but that's ok 'cause its just in beta and you want to make sure shit works).

I'm pretty sure the pro league never got past the beta phase mostly because the feedback that was received from the people who did participate were mostly neutral and negative towards the idea, and partly due to it never being used naturally outside of erynn trying to gather people, despite advertising on pug.champ.gg, posting in the original PugChamp thread, and whatever other invitations erynn/tsc gave out via Steam and Discord. I tried idling in it added throughout the day for a week or two to try to get it going but the most I ever saw personally added to the site was around 4 people.

Unless you mean "ready to be used" as in fully released with invite pugs ready to happen daily, which yeah it wasn't, mostly because of what happened when it was ready to be used at all for beta testing.

I wasn't the one doing the gathering so that'd be on erynn/tsc to clarify that. The good news at least is if people want it I'm sure it could try to come back purely based on how much tsc wanted it to happen back then.

pugchamp is for pugs and removing captains is a terrible idea that would leave us with 0 pug websites

it would be nice to have a way to play low level pugs there instead of getting fatkidding for the same 12 people every time their game ends, add like aids.pug.champ.gg or something that IM+ players or players over a certain ELO can't queue on (if the latter you'd have to make it not change your elo) EDIT: but you shouldn't ruin the possibility of having high level pugs by doing this, that would kill the site

the drafting process is fun and important and I'd argue it's good for newer players to learn, plus that way we could play mixchamp level games without having to play gran/reckoner/(insert map you hate here) all the time

I just asked a bunch of top invite players and they all said they never played in it nor were they asked, almost all of whom still want an invite only pug system. Something's not adding up.

i knew about it

it was a joke and a bunch of mid im players had access cuz they farmed low open pugs for elo so no one added that knew about it or even talked about it

Something something, xkcd competing standards, something something.

Whatever happened to in house groups? That would solve like 90% of the issue right there...

i wanna say that mix champ helped me improve an insane amount at the game, and improving it would be really cool to see. honestly, I don't really see your complaints with pug champ that much: the time wait has never bothered me, and i'd like to see stats where you're saying players who know they won't be picked go to other websites. honestly, before i started playing mix champ, i just practiced my dm, jumping, and watched some Marxist videos for about a month, then started playing, and I did really well. imo, if players put a little bit of time and effort into practicing they would be so much better, the barrier of entry isn't that large, the only problem is that mixes rarely happen.

eu pugchamp pro never got off the ground because the requirements to play were much too strict and never changed, seemingly due to tsc/erynn not having the time to work on making the changes. looking at pickup2/eu pugme there has always been interest for an invite only pug system in europe, perhaps less so now due to the playerbase being unaccustomed to having such a service available.

if anyone does choose to take on the work for an invite only pug site, it really doesn't matter if it's automated or captains pick - just don't make the initial requirements to play so strict and inflexible that the player pool you draw from is too thin to assure activity. also don't give sideshow admin and allow him to unban himself for the duration of your sites existence.

True - but if no one uses the system, it's not likely to ever get fixed, as the Devs will just assume it's not a popular system and any changes to it could just be a waste of time. Using it frequently and submitting feedback daily is more likely to convince them that it is a valued feature, and worth the time and effort to improve.

TF2 Center actually has a balancing system (I think they refer to them as Advanced lobbies), which resorts both teams when both are full. I'm not sure how good it is at actually balancing teams, but it's not very popular. A lot of people like to play lobbies with friends, and that obviously doesn't guarantee you get the team or players you want.

After talking with someone I'm changing my comment on PugChamp. I thought the limit of three pugs at a time is due to the fact there are only 3 servers (which is the case), but, there's also an issue that not many people want to captain or play medic. One of the problems with the captain system, that I didn't see mentioned is that when there's a really good player captaining, it means any other potential captains also have to be really good - otherwise their team is at a disadvantage from the start. For instance, if a high level player like b4nny is a captain, and I wanted to captain against that person (keep in mind I'm considerably lower level), no matter how good a player I pick as my first choice, his team gets him automatically - in addition to whichever good players he picks. So I drag my own team down if I choose to captain against a better player, which deters me from captaining in this type of scenario.

As for MixChamp, another problem is that it's hard to find folks willing to play Medic at that level (the less cohesion a team has, the more the medic will suffer for it), and there's also only one server for MixChamp, so only one game can be active at a time. :(

I agree that the wait time in between when pugs are finished drafting and when the next pug is able to start is too long and if possible it should definitely be eliminated (i.e., the next pug can start instantly after the 12 people are chosen/maps are picked instead of leaving that screen up forever and preventing the next one from starting). This gets exacerbated by people deciding to kill pugs, which often results in even more delay if nobody steps up to captain since the person who killed the pug gets temporarily restricted.

As Jarrett and others have pointed out, the number of servers limits a maximum of 36 people to be currently playing a pug at once, but it is certainly not impossible for more than that to be added up (I remember seeing ~42 added before, back when a lot of invite players used to add up). Recently there have also been issues with certain servers not working which limited it to 2 or even 1 pug from happening at once. Trying to reinvent the wheel and and make a whole new pug service from scratch is overcomplicating things, all that's needed now is more servers.

irc

Agree, Bear. We as a community can sit here and talk about the improvements PugChamp can make: they can increase the number of servers, they can make the picking more efficient to reduce wait times, and they can work to enforce a code of conduct that prevents people from doing dumb things like killing pugs. All of these are good ideas.

However, at the end of the day, we're limited by the two developers who own the service, and there have been no major updates to PugChamp since August 2016. Unless we get them some help, we may have to explore alternatives. In my brief interaction with the team I learned that there weren't many resources available to expand the service since it's currently operating as a not-for-profit -- they can't pay for more servers, hire developers, or expand at the speed that the community wants them to. It feels to me like they're just doing basic upkeep as a service to the community while not taking the steps needed to quickly grow for the good of the scene. It's also hard to expect them to do so while they have other obligations like jobs and life and stuff.

I appreciate the dedication to the community that the PugChamp team has had thus far. It's an amazing service end-to-end and it serves its current purpose well. I'm just trying to push the limits to see how we can continue to grow TF2. While it's not optimal to spin up a new service, I'm looking at all options for the good of the scene and leaving all options on the table.

So who's going to do all of the backend work required to spin up a new thing? If pugchamp is having issues getting more servers for free and doesn't have enough dev help, why would a brand new thing not have the exact same issues? Hell you work with the background TF2 shit, you should know better then anyone that we are short on people already. Where would these new people even come from and if they exist why are they not doing anything currently?

http://www.teamfortress.tv/41777/help-test-new-faceit-beta

Pugchamp is more limited by medics than it is by servers. Don't think I've seen more than 6 medics added over 10 times over the past 14 months I've used the websites. So that's 10 times out of the thousands of pugs I've seen start that could POTENTIALLY fill 4 servers.

Letto is right, newbie mixes would be the best thing to tell a new player to play. Then when the playerbase is actually big enough to introduce the "pro" league pugs, it would fix the first problem, which is lack of players, and then the next problems (such as servers, different pug skill divisions etc.) can be addressed when the time comes.

I think it's much better for this new service to be a for-profit endeavor whether it's FACEIT, TF2Center, Stadium, or ABC new service. Nothing can run for free like this, so you have to find a way to monetize.

Option A
Ask PugChamp if they're interested in pursuing a new monetization strategy. It's going to sound like a meme but here's what I'd do. I'd put ads on the website just like TFTV. Then I'd modify the PugChamp global chat so that if you subscribe at $4.99 per month you get no ads, a badge in the chat and next to your name on the picking interface, some sick emotes, and something cool in mumble tool (I'd find a way to recolor names or something).

The rest of it would come out of pocket and you can continue to accept donations.

Option B
Ask FACEIT to do it. They already have a monetization strategy and a will to develop a TF2 matchmaking system. Problem is, nothing they do for TF2 is because they love TF2. Everything they do for TF2 is really just a test run for a future copy pasta to their CS:GO offering, so if it doesn't work for CS:GO it doesn't work for TF2. I don't think queuing up for a service by class and class-specific skill ratings are something they're interested in developing.

Option C
Ask ESEA to do it. They've modified their pug experience in the past and I'd ask them to see if they're still interested in attracting TF2 users to their pug service. The downside with this option is that I guarantee they'd make users pay for ESEA Premium and I don't think it would scale as well if people had to pay to play. Also they have the same problem as FACEIT since everything they develop is for CS:GO and slightly adapted for use in TF2.

Option D
Ask TF2Center to do it. The main problem here is that the PugChamp value proposition is significantly different than their current product offering so I'm not sure they'd go for it. TF2Center is all about getting games started quickly with no balancing and you can play with your friends. PugChamp is about creating high quality games with solo queue only.

Option E
Ask Valve to do it. This one is also kind of a meme but a bunch of us know the devs and we can at least pitch them the idea of how they can improve their matchmaking service. Of course, they'd pursue their own ruleset, cater mostly to pubbers, and take two years to release a product so this is the worst option in the short term.

Option F
Rip the code from the PugChamp GitHub repo and spin up something new, with Erynn and Tsc's permission of course. I asked my coworker to see how much effort it would take to spin up a new service as a pet project and he said maybe 3-4 weeks for a good full stack developer, so I'd just need to see who is available in the community to pursue such an endeavor. Maybe I'd hit up the Stadium guys and see if they're still interested in building things.