TF2 General Chat and Speculation Station V14 - The Bargaining Stage
999 replies, posted
The went into private beta, and have spend the last couple years working on the next update.
I expect lots of VIP content when they finally finish.
Would we really need autobalance that much if Valve finished bot support for all gamemodes and added nav meshes for all the maps?
Do you want to play with bots?
Have you played CS casual before?
Nope. I'm assuming that game uses bots and lets you take one over when you are dead or was that a different game?
bots take empty slots on-the-run and are immediately removed if it's filled by a real person.
It's a good system.
The idea that "bots sux xD" is also extremely archaic and is just assumption from other videogames where bots are indeed unhealthy. TFBOTs however are a marvel and remain outstanding work from mr Mike Booth due to their complex simulated awareness and ability to make players feel like they had a fair fight and weren't just killed by a walking aimbot.
Spending tons of effort to make a surperb bot system only for 3% of people to see it outside of MvM, very Valve move.
TFbots still aren't capable of rocket jumping or any other form of advanced mobility, however, which is a shame.
They are mostly finished in every core mode except CTF, which doesn't have proper Engineer AI or defensive strategies in general. The problem is lack of nav meshes for the maps and alternative gamemodes which they may as well leave bots off for those gamemodes anyway since hardly anyone queues for them.
I'm up for the experiment but only in extreme scenarios. TF2 is a lot more complex than CS I'd imagine and a lot more magical things happen such as server-wide melee battles and the like, don't think technology is at a point where it can detect that.
At the same time bots could also cause some magical things to happen; which is why im up for experimenting with it, why not.
Bots are probably more useful than most of the players I see in casual.
And then a human player plays Spy and backstabs a bot and the other bot enemies know exactly where they are and that they are a Spy, and immediately light them up, and evem the highest difficulty bots literally can't get past a Sentry in the red spawn-first point hangover on Dustbowl's third, and you realize thst they're still AI opponents and nothing like human players.
They also haven't been officially updated for years, aren't compatible with new weapons, aren't programmed to use non-stock items, and have a common tendency to get stuck even on well-done nav meshes.
If the TF team were to do an update that added ai replacement players into quickplay, it'd have to be an update completely dedicated to overhauling the system, and they'd have to spend months just giving the bots nav meshes for every new map, if they wanted them to be barebones useful.
Good god I want to go back. I wish we had something to talk about. I've done nothing but lurk for about 2 months and nothing has happened.
The recent supposed Portal 3 ARG has made these feeling all but alleviate. I wish it was still 2012. My favorite game has begun to slowly stagnate and I'm unable to do anything.
Hopefully there will be something to talk about in two weeks or so(Smissmas).
Off-topic:
There's one good thing that came out from the games sections removal. I actually started posting in other games thread, and now Facepunch is not just a TF2 forum for me, but a video games one.
I couldn't find a "specific mention" of the Match Maker creating dead games, although I did find some instances where you were inferring as such?
Although the above was really talking about the Turn-over rate and not so much dead servers caused by the match maker.
In this bit I can only assume you were inferring the Match Maker is "plagued by the same issues" you claim Quick Play had.
But this is beyond the point. I used all of that information to back up my claims with a bit of context, and you instead focused on the chunk of text as some kinda "you didn't say this so lemme prove you wrong" attack.
I explicitly say that the dead servers are being caused by the Match Maker and provided several different reasons as to such in that post and prior: The MMR prevents people from joining specific servers, the queue system itself appears to hang up at times preventing any new players from connecting, a penalty that bars people from rejoining the queue would do just the same.... etc.
"Lack of players" in quickplay is very much not just because of hour of day..? There were a bunch of times where atleast I played Quickplay at prime time on upward and the server just fizzled down to 6 people and no one would join - because if you are searching for an upward server, would you rather join the 18/24 one or the 6/24 one?
I'm just not even certain how you could believe a player base, regardless of size is not effected by "hour of day." I didn't even make the claim that "hour of day" is the one and only factor for emptying servers. As you mentioned, tendencies in human nature can lead to a bias in which server the user will pick... There is more than one factor and while I didn't list more (as I probably should have)
I by no means went and said something like "In my personal experience, at 5PM servers average 16, 9PM: 14, 12AM: 12, and its just so plainly obvious that 'hour of day' must be the only cause of this" I think you get the joke here.
And once again, let me repeat myself; It doesn't matter if you offer people the best time of their lives. They will leave for a variety of reasons unless you penalize them. By the way, I have never mentioned that I personally believe penalties are what we need to introduce to Casual; I just view this as the only "solution" there is, and it's a shitty one.
Repeating yourself doesn't make your opinion more right than you might think it was the first time you said it. I often refer back to things I've said prior so as to keep a flow of contextual information connected to the additional justifications I provide to help better explain my opinions. I don't just repeat things in the hopes it rings better the 3rd time.
What you said was also contradictory but I'm sure you're aware of it this time... right? Peddling the idea that something like this is the "only solution" (which is very one minded) on a board that may or may not influence the decisions Valve makes for future implementations while simultaneously "disliking the idea" is the crux of stupidity, especially considering you state you'd rather not see something like this implemented.
I disagree with this "solution" and believe there are far better ways of handling this. This is the entire purpose of my walls of text. To provide a counter argument backed by needless amounts of information that point out "Hmm, people do tend to stay in servers longer on average if they're enjoying something" as would be true of any experience they are happy about.
That huge ass wall of text with the 5 different scenarios that are now impossible due to the nature of Match Makers could have been reduced to 1-2 sentences where I say "In my experience, the auto-balance prior to the Match Maker update was so much better and this experience of mine warrants justification because its just so stupidly obvious, facts aren't needed to back anything up."
I could just do that with the rest of my opinions... Condense them into 1-2 sentence that are emotionally based criticisms but I'd rather not reduce my brain to mush.
No, and for the sake of this lengthy response I'll keep this one short. Your inconsistent use of the terms "Quick Play" "Casual" and "Match Maker" drive me up the fucking wall. It makes it hard to read just what exactly it is you are arguing for the sake of whatever it might be. I could've sworn you typed:
That reminds me. Today we lost two matches on Suijin in part because the other team's experienced kunai spy was getting endless free 300 max health from a 0 points heavy on our team. And he was wearing the dead-ringer, of course, so he was virtually impossible to kill for real.
the problem with tf2 updates in the current year is that no matter what the size is, it's still not pleasing
for me for instance, doesn't matter how many community maps they add, I keep tabs on tf2maps, I've seen loads of good maps. I'm also not interested in seeing any new weapons because I know they'll be annoying and rarely will they be both fun to play with/against and balanced, and also add something to the game. these are things a lot of people are looking for, new maps and weapons. I can safely say 90% of the community is pleased when maps/weapons arrive, especially valve made maps
the number 1 thing I care about is getting ranked play off the ground & 2ndly improvements to perf & vac, and there's still many people looking for MVM, people looking for EPIC taunts and skins, people wanting improvements to casual and people looking for all types of stuff
I don't see any update coming actually quelling the problems loads of people have with the game.
I kind of miss when I was just looking forward to the next weapon pack and like one map. nowadays i'm both super jaded, and the scope of the game has increased to the point where I genuinely don't think I'll ever be happy with the game again and I don't think the community will be either
It wasn't specifically mentioned, but more as heavily implied, so I suppose I shouldve made myself clearer.
And I said "*just*" the hour of the day. Of course it's defined by the hour of the day and I have never said that isn't a deciding factor. I'm saying that despite Quickplay having a more open auto-balance and general more open team switching options and the like servers still died or fizzled out. Reasons for that are that players simply didn't rejoin because there was no system that assured people would actually join and players did it manually. You don't have to explain the night and day cycle of videogames to me, this wasn't supposed to be the focus.
And I can see what was confusing about this, so sorry for not making myself clear; What I meant with "Quickplay" in the context of Overwatch was it's variant of TF2's "Casual" mode. It's "Quickplay"(Automated Matchmaking system alike the one TF2 has) is what I'd consider one of the best on the market. The idea I got from you was that if "Quickplay"(Here: TF2's old Quickplay) were to be reintroduced into TF2, we would not have any/way less dead servers and the overall player enjoyment would be higher due to things such as more open autobalance/team switching etc.
I simply disagree with the first statement given dead servers always happen in any game, no matter the system in a "casual" mode of a game for reasons already explained and Quickplay(TF2) not doing anything to ensure servers are kept alive/it being entirely depending on player-input rather than an automated system which (sadly in tf2's case doesn't do it's job properly - this however can be improved without going back to player-input) keeps a constant stream of players to servers and systematically detects and fills dead/declining servers, not to get started on unbalanced matches due to no nature of MMR or a rating system.
I've been lacking sleep as of late, so sorry if this discussion is a fucking zodiac puzzle. The initial point of this was about auto-balance, and I once again agree with your point about autobalance and want Valve to fully experiment with it.
And if it helps condense my position further about the rest; I simply want valve to step down and put a foot down on what they want casual to be. Competitive lite? Slap a penalty on leaving there which ensures games are less likely to lose players and make rank actually matter a bit. Mega Casual? Reintroduce Team switching like you suggested alongside of tuning autobalance to be a bit more free, reintroduce the team scramble vote etc. - I am relatively neutral towards both options when it comes to what to *do* out of casual.
Bots can play Control Points and Payload well enough but they don't know how to play:
Payload Race
Player/Robot Destruction
PASS Time
And there are some maps where the in-game nav updater doesn't work properly and causes bots to think that they're stuck (maps where doors open and close based on payload moving rather than Control Point capture ie. Snowycoast, maps like these would need a nav_interface in the level to RecomputeBlockers?)
But for the most part, bots tend to do an okay enough job. If only you could kinda mess with them more to fix those cases where an Engineer on your team builds a Sentry nest facing walls or your own team's Engineer not moving up at all and just building in a single location through-out the entire game.
Putting in bots is quality of life improvement as long as they aren't too good.
I agree if a team has less players you should be able to switch. I'd rather switch earlier than later, but the autobalencer can't know if that is a good idea, so it is conservative until later. It would feel less of a pain if there were a few more safeguards, like don't autobalence a person while they (themselves) are capping last.
To allow spectating in matchmaking without it being used to restack the teams (for shame), maybe make spectating a mode instead of a team. Have it as a limited use thing and either only allow players to switching into it when dead, or turn you into a bot like how Skial turns you into a bot when you are afk. If STV was available for all games, some of the usages of spectating could be fulfilled out of game as some of the time consuming tasks could be done later.
For leaver penalties, I wonder how well those would work if they were visible to the transgressor, as then some would go friendly or afk instead. Overall if you aren't interested in playing the game on your assigned or autobalanced team, or chosen team (if the team switching to lower player team was implemented), then I probably don't want you in my game because you are actively seeking or at least making actions towards a worse balanced match. There's probably a big enough pool of players you can put those players together so the people that consistently leave when they are losing get stuck in a shithole with all the other people that leave when they are losing. They wouldn't necessarily be stuck there forever, but with better behavior they'd be matched in more games with people that don't leave as often when they are loosing. Similar to what Rockstar did with the cheaters in Max Payne 3. Again, I don't think this would be much help if it told players that it was happening to them. Once you get to a certain level farther down in this pool, I wouldn't expect many to crawl out there, but maybe they would be happy there as they'd be with a bunch of other people that don't take games very seriously, so there shouldn't be as much conflict.
After explosivedeer pointed out that the casual MM is visible, I'm curious what other people here have and any observations they have noticed on changes to theirs. What is the default MMR? https://steamcommunity.com/my/gcpd/440/?tab=playermatchhistory
My MMR is at 1479 right now and I'm at Tier 3 level 113. Looking back to October my MMR was about the same there, but there was a period in-between where my MMR was in the mid 1500s. My guess is playing Halloween skewed it as a lot of non-regular players came out for it and I very heavily played Cursed Cove putting me at a huge advantage over the players that launched into that map for the contract towards the end of the event. If something basic like that is actually what happened, it makes me thing the concept of one casual MMR for all of TF2 absolutely screwed. At Halloween there's definitely some players that due much better with crumpkins, or spells, or underworld and players that due worse when one of those items are out. It seems like if you usually play your best classes on your most played maps and then queue up for your least played maps and play your least played or worst classes, it could be rough as the MMR expects you to perform the same. Does it mean if you don't play the generalists very often, even though you are playing 12v12 you will have lower MMR because statically playing the generalists wins you more games (forgetting sniper is broken especially in high player environments) even though you could be as skilled as a person that plays mostly generalists? What if you "play" CTF or even Mannpower? Generating enough data for making a perfect MMR for each situation is impossible, but I think trying to condense it to one number is too simplified to be useful in enough cases.
Tier 3 Level 61 and it says my MMR is 1505. Whatever that means. It probably doesn't matter.
Bots "performance" should be depending on how badly enemy team is losing, bots sound amasing for balance but too much work for tf2 team
My favorite bot behaviour is them taunting if you look at them, allowing for easy kills. Does anybody have that video of a sniper killing a bunch of bots by looking at them at the start of a round?
I admit is nice to have engie bots build teleporters first.
All the more reason there should be a proper engineer tutorial that shows the importance of the function of your buildings, and which buildings you should prioritize to support your team.
I'n all honestly, anyone interested in improving their engie gameplay will just watch a youtube video.
I'd post my List Of The TF2 Team's Dead Children here, but apparently I don't have access to my own posts anymore. RIP. Something something competitive matchmaking.
it's ok this thread has enough of that kind of content already
id hold onto the next big boy update (maybe smissmas according to that leddit post) before closing the coffin
is that a reason not to have better tutorials? most people don't give nearly enough of a shit to even want to improve in a game, let alone actively look it up
Sorry, you need to Log In to post a reply to this thread.