MR_SLIN
Account Details
SteamID64 76561197980196963
SteamID3 [U:1:19931235]
SteamID32 STEAM_0:1:9965617
Country United States
Signed Up July 21, 2012
Last Posted February 15, 2022 at 3:31 AM
Posts 3982 (0.9 per day)
Game Settings
In-game Sensitivity 2.133
Windows Sensitivity 6/11
Raw Input 1
DPI
1600
Resolution
1920x1080
Refresh Rate
144hz
Hardware Peripherals
Mouse Zowie FK1
Keyboard Filco Ninja Majestouch-2 (Cherry MX Reds)
Mousepad Steelseries QcK+
Headphones Sennheiser IE 80
Monitor BenQ XL2730Z
1 2 3 4 5 6 7 8 ⋅⋅ 260
#79 How to make PugChamp better in TF2 General Discussion
GentlemanJonMR_SLINThe code is all publicly available right now but it's polite to ask. You should have asked your co-worker about the legalities, it's not polite to ask, it's legally required. The code is unlicensed and subject to all normal copyright laws. Just because you can see it, it doesn't mean you can take it or make derivative works from it.

You don't have to lecture me, my dude. Pugchamp is fully open source.

posted about 6 years ago
#69 How to make PugChamp better in TF2 General Discussion

Okay, if that's the case, then by definition PugChamp is a pug and MixChamp is matchmaking.

So what I'm hoping for is to expand MixChamp's capacity, team balancing capabilities, and just overall fun factor for players of all skill levels by allowing more people to add at once. By starting pugs in waves, you can sync everyone up and ensure that games are balanced as well as possible. No matter what skill level you're at you should hopefully be able to find games with players of a similar skill level.

This differs from FACEIT in that you can add on a per-class level and get a skill rating on a per-class level, something that isn't offered by FACEIT at this time.

I suppose PugChamp could be left alone so nothing gets messed up there (nothing lost). You could bring skill ratings back for PugChamp but you could also just leave it off and then only bring back a public-facing skill rating for MixChamp.

posted about 6 years ago
#67 How to make PugChamp better in TF2 General Discussion
Tino_Wait I am confused Slin. Are you making a pug system or a matchmaking system? Because what you suggested was not a pug system, it was a matchmaking system.

What's the difference, a pug is a group of 12 random people coming together and matchmaking them together is kind of the same thing. Maybe I'm getting my terms confused but let me know if i'm using the wrong word.

Does matchmaking presume that you match people by skill rating?

posted about 6 years ago
#63 How to make PugChamp better in TF2 General Discussion
TobYou're getting ahead of yourself. First you need a playerbase... Once you have that you can start thinking about monetization. You can start advertising this very second to players on your stream, on youtube that you want to do this. Tell them to start playing mixchamp.

I can't get players to use a system that doesn't work at scale. This is the reason why people don't add up to FACEIT, why high level players don't play TF2Center, or why only users of a specific skill level play MixChamp.

The issue with MixChamp specifically is that once users fill the 12 required roles, it immediately boots up and begins the game. Because it doesn't wait for the entire pool of available players to add up, it struggles to balance games like a proper matchmaking system would.

Example:

  • 25 users are available to play. 10 of them are Invite players, 5 of them are Open, and 10 of them are UGC.
  • If they don't add up to MixChamp in exactly the right order (all Invite players first, then High Open, then Low Open, then UGC) the system will match it as first come first served.
  • Thus the pugs would likely be a mix of Invite, Open, and UGC players, so instead of two good pugs you get two bad ones.

This is why we suggest a system that queues up the games in waves. All 25 players are available for the computer to choose from at the time that you start the picking. This is simply not available in MixChamp right now. I agree that MixChamp is closer to what we're looking for than PugChamp, though.

//

Also Option A is a way to begin monetizing PugChamp without having to change anything else. Once PugChamp gets some money coming in they can purchase servers and the like.

posted about 6 years ago
#60 How to make PugChamp better in TF2 General Discussion
Tino_So who's going to do all of the backend work required to spin up a new thing?

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.

posted about 6 years ago
#56 How to make PugChamp better in TF2 General Discussion
bearodactylTrying 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.

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.

posted about 6 years ago
#45 How to make PugChamp better in TF2 General Discussion

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.

posted about 6 years ago
#21 How to make PugChamp better in TF2 General Discussion
damneasyYou can still have captain based picking, we just need more servers. If theres like 100 people added it will start like around 8 pugs, theres no way the first pug will be over before all the other ones have finished picking.

I don't know why but it always feels like picking takes 5-10 minutes, but I think it's worth sitting down one night and timing an average team drafting session. I'm pretty sure the pick timer is like 20 seconds but I'm trying to be fair in the calculation so let's say on average it takes 10 seconds to pick.

With 10 players to pick that's 90 seconds (10th player is automatically picked). Then map bans is another 60 seconds so that's 2.5 minutes or so. Then you've got the "ready up" time between games of like 30 seconds so let's just say the time to pick a pug is 3 minutes. So to get through 8 pugs that'd be 24 minutes, but you're assuming people are picking pretty quickly and nobody is intentionally killing pugs by forgetting to pick. I've definitely seen people kill pugs on purpose because they don't like the way picks are going or maybe they just AFK.

This also assumes that the first pug doesn't end in less than 24 minutes.

Either way I think that given this calculation you can make some minor optimizations to PugChamp to speed things up for sure! But I also think it's a bit annoying for the 100th person added to wait 24 minutes to ~maybe~ get picked for a pug. It's not a very good experience and if this happens enough times to you then you might just choose not to play anymore. Thoughts?

Examples:

  • 94th player adds and waits patiently for 24 minutes. Pug #1 ends in 15 minutes, so 94th player doesn't get to play for 50 minutes. Maybe gets a chance to play after 50 minutes? Again not sure because if pug group #1 ends that pug again in 20 minutes you're out of luck.
  • 94th player adds and waits patiently for 24 minutes. Pug #2 ends in 20 minutes, and assuming 3 minute picking then the 94th player doesn't get to play for 50 minutes. Maybe gets a chance to play after 50 minutes? Again not sure because if pug group #2 ends that pug again in 20 minutes you're out of luck.

I think more servers is a good idea but you've gotta remember that Erynn pays for the servers out of pocket too.

[Edit] I also made the assumption that people join the servers instantly so maybe there is some wiggle room there I didn't account for.

posted about 6 years ago
#17 How to make PugChamp better in TF2 General Discussion
TwiiKuuAre you not scared that building such an algorithm and infrastructure might be harder than what you're able to perceive? GentlemanJonSo this is just a wishlist from two people who have no ability to contribute or deliver beyond that right?

It's a forum where people talk about things, and this is just another discussion topic. Since we're in the ideation phase I think this discussion is the most important place to start since the community is group that is using it. Once the idea is solid then we go talk to PugChamp or FACEIT or whoever can support such a thing and figure out implementation later.

posted about 6 years ago
#9 How to make PugChamp better in TF2 General Discussion

Before this thread gets derailed too much, the problem I'm trying to address is this:

Imagine if 100 people are added to PugChamp. By the time you finish picking 3 pugs, B4nny and all of the invite players from the first pug are already added again. So at a certain point people aren't able to play as long as pugs are picked one at a time. If all pugs start at the same time you'd solve this problem and you can solve this by using a computer to pick teams based on skill rating.

If PugChamp or FACEIT are able to solve this problem, great. If not, then let's figure out next steps. Because until this issue is fixed, nobody is able to promote PugChamp. When new players ask "hey where do I play" we're forced to tell them to go to TF2Center which is not the right solution. We should figure out how to grow PugChamp and make it possible for 50+ people to play at the same time.

posted about 6 years ago
#5 How to make PugChamp better in TF2 General Discussion

haha woops... this is why we can't have nice things.

posted about 6 years ago
#3 How to make PugChamp better in TF2 General Discussion

Specifically looking for people who know how to write code and stuff cause those kinds of people are useful.
Erynn told me she could use developers, but if you guys feel like this idea needs work and don't want to mess up a good thing then let's talk about it. I'm happy to talk about it privately too.

posted about 6 years ago
#1 How to make PugChamp better in TF2 General Discussion

I know there's an existing PugChamp thread but I wanted to start a new discussion about some of the current issues with PugChamp and how the system can be made better. I think any existing pug system could be improved to get what we need but all systems right now have their flaws:

  • Comp MM doesn't have a good ruleset and doesn't create quality games
  • TF2Center/Stadium makes no attempt to balance Pugs
  • FACEIT attempts to balance pugs but the actual player experience is awful since you can't pick your class
  • PugChamp is great for high-level gameplay but doesn't allow for more than three pugs to be run at once
  • MixChamp is great for medium-level gameplay but starts immediately when all 12 roles are filled, so it doesn't really suit high-level pugs

===========
The Problem: Imagine if 100 people are added to PugChamp. By the time you finish picking 3 pugs, B4nny and all of the invite players from the first pug are already added again. So at a certain point people aren't able to play as long as pugs are picked one at a time. If all pugs start at the same time you'd solve this problem and you can solve this by using a computer to pick teams based on skill rating.

If PugChamp or FACEIT are able to solve this problem, great. If not, then let's figure out next steps. Because until this issue is fixed, nobody is able to promote PugChamp. When new players ask "hey where do I play" we're forced to tell them to go to TF2Center which is not the right solution. We should figure out how to grow PugChamp and make it possible for 50+ people to play at the same time.

===========
The Solution: The proposed solution is to replace the human picking of PugChamp by returning ELO or some kind of skill rating system to PugChamp and then allowing a computer to choose the teams. We acknowledge the fact that as the system grows it won’t always be the most fun for all players at all times. Sometimes there’s toxicity, sometimes you don’t get to play with your friends or you end up on a team with someone you don’t like, and sometimes the games aren’t balanced properly. However, I think given a large enough sample size, a computer could pick really well or even pick better than a human to create evenly balanced games.

Once you have this automated picking system in place, you can run an infinite number of pugs at the same time provided you have enough servers and enough medics added. We'd synchronize the start time of all the games in waves, such that the system would wait for every single pug to finish before starting the next wave. By synchronizing the start times of the pugs, you maximize the potential for evenly matched games with each wave of pugs (and you can't dodge high-level players by queuing at the right times).

Example: Wave 1 starts at 7:00pm with 100 players added. 8 pugs would be started with 4 fat kids. You have three minutes to get into the server, and the matches end in 30 minutes with no overtime. After 33 minutes, all pugs should have completed. Add in 5 minutes of buffer time and start the next wave of pugs at 7:38pm.

In traditional nerd essay fashion, Micspam and I have drafted a document outlining some of the issues with PugChamp and how these can be resolved. I also made a companion video in case you don't want to read the document but either way it's a long ride.

https://docs.google.com/document/d/11-p_dFw4urcZQ6Jxt2cCK_qcqDmxVyGb4d-s3x-jjDY/edit?usp=sharing
https://www.youtube.com/watch?v=iLCXro-EqrY

Would love to hear your feedback and if you're interested in working on this kind of project then hit me up.

posted about 6 years ago
#115 ESEA has outlived it's usefulness in TF2 General Discussion

Yeah so screwball my point is this:

1) The difference between these two leagues is marginal at best. There is obviously a switching cost involved here -- is there really that much to gain compared to how much we stand to lose?

2) The prize pool attracts the top talent, which in turn attracts the rest of the talent. Everyone knows B4nny only wants to play for money, and everyone else in Invite just wants to play against B4nny (insert joke about not liking playing against B4nny here). Sandbaggers are attracted to the Open prizepool, and everyone else in Open wants to play against the top teams in the division. The reason nobody plays in UGC is because winning UGC has no prestige -- everyone knows that getting 1st place in UGC Plat means nothing since the top teams are not there, and the same problem happens when achieving rank 18 in comp MM.

So sure, switching to a league where the current top talent isn't interested in playing (basically ditching the top end of TF2) is one way to deal with the problem. You can switch leagues and establish a new top-end of the scene. However, I think the main issue is that you're missing out on attracting users who do want these kinds of incentives, so why not search for a solution that can keep them around?

====

If you actually cost out the current ESEA fees (keep in mind I actually don't know the real dollar values) you'll get something like this:

49 Open teams * 6 players * $30 ($21 premium three months and $9 league fee) per player = $8,820
9 Invite teams * 6 players * $80 (Invite fee) per player = $4,320
Total fees collected: $13,140

Last season's prize pool (67 teams) = $19,000
This season's prize pool (58 teams) = idk

Do we think we can beat this? How much money do you think ESEA is really siphoning off?

posted about 6 years ago
#108 ESEA has outlived it's usefulness in TF2 General Discussion

Starry I think you're missing the point -- if we switched off of ESEA to some other F2P league, a ton of teams would die for pretty much no benefit. Would TF2 go on? Sure. But you're not clearly outlining the benefits of this new league that make it so much superior to ESEA.

ESEA:

  • They have prize pools
  • They have admins on duty and respond to support tickets within a timely manner
  • They have a structure and an anti-cheat system in place
  • It costs money to play
  • They don't care about growing the TF2 league or marketing it.

This new F2P league:

  • They have no prize pools
  • They have admins on duty and respond to support tickets within a timely manner... maybe
  • They have a structure and an anti-cheat system in place... hopefully
  • It costs no money to play
  • They care about growing the TF2 league and marketing it.

If the only difference between these two leagues is paying or not paying, playing for money or not playing for money, then just play UGC. Is marketing helpful? Sure but who is going to do it?

What's the point in killing ESEA if everyone is fine paying money. Is ESEA siphoning off that money for their own benefit? Probably, although some invite players have said (or maybe spread rumors) that ESEA operates TF2 at a loss. Is it worth the switch and potentially hurting the NA game? Probably not.

Give some concrete, real reasons as to what this new league would do that is simply not possible with ESEA. It might sound like I'm against this switch but I actually think it's an amazing idea if someone actually cares enough to execute and see it through to the finish line.

posted about 6 years ago
1 2 3 4 5 6 7 8 ⋅⋅ 260