Allow us to pick rating brackets for TMM
-
I agree with giving higher rated players priority because it's easy enough for lower rated players to get games anyway. This would also promote more higher level games, instead of TMM spamming 1300-1900 games and leaving out all the 2k+
Atm it's only possible for me to get games when i queue with people >500 rating lower than me, and even then it's honestly really hard.And then I'd like to add it's awful that I can't have any other games open while waiting, because that immediately crashes the TMM launch. Really makes the waiting experience that much worse when I can't A. walk away from desktop for prolonged period of time and B. can't do anything while I wait
Don't know if it can be fixed but would be really nice if it could (custom games don't have this issue do they?) -
@bulliednoob said in Allow us to pick rating brackets for TMM:
And then I'd like to add it's awful that I can't have any other games open while waiting, because that immediately crashes the TMM launch. Really makes the waiting experience that much worse when I can't A. walk away from desktop for prolonged period of time and B. can't do anything while I wait
Don't know if it can be fixed but would be really nice if it could (custom games don't have this issue do they?)On a related note I was doing some cleaning while in the queue and I felt like I had to glance at the screen every now and then in the small chance that I actually get a game. It would be nice if we could enable some very noticable sound notification (preferably several seconds long) when we get a game. Then I could at least leave the room safely while I have speakers on.
-
Speaking as a 1400-1500 rated player, I wouldn't be able to tell the difference between an 1800 and a 2300 teammate/enemy whilst playing the game myself (unless there was 1 of each to compare). I'm perfectly happy to have players like Blodir in my game, as long as the match is balanced and they can forgive my mistakes, but I do enjoy games more when everyone is below 1600.
A rating bracket system sounds alright, but not if the brackets are forced. Like if I have to play in the 1500+ bracket whenever I reach over that rating, I'm just always going to be the worst there and have no fun. With an optional bracket system it allows people who don't want to play with noobier players to opt out of those games, while still allowing those of us who don't mind playing with noobs as much to continue as we are. Those players who opt in will just have to accept the fact that imposing their own brackets will increase wait time.
Also, as a side note, whenever I get into a game where everyone is 1400/1500+, the map always seems to be 20x20, often with lots of navy, whereas the lower games are at least 75% land maps. Why is this, and can it be changed?
-
Higher rated game -> more difficult maps. A 10x10 is extremely simplistic gameplay in a 4v4 generally. Bases are compact, mexes are all taken and controlled by min 5, ACUs walk mid, all aggression is over. If that was made more pervasive at the top level of tmm, I’d definitely just stop queueing entirely.
In terms of sea based 20x20, they’re just easier to make than a decent land 15x15 or 20x20 so there are more of them on the vault.
-
I'll also throw my two cents in, sitting around 1600 tmm rating it's incredibly difficult for me to get games as well and it's causing me to play less and less tmm. I usually have to queue with lower rated friends if I want to get semi-consistent games. It's pretty annoying being one of the first in queue, then seeing the queue grow to 32+ players and then not getting a game. And that happens over and over. The last few weeks I've basically given up solo-queueing and just went back to hosting again. I imagine it's a similar-but-worse situation for even higher rated players as Blodir describes.
-
@ftxcommando That's understandable, however it leads to a, perhaps unintended, increase in noobiness. For a newer player who predominantly plays TMM, and who hasn't had years of experience in custom games, we learn how to do well in those low level TMM land maps and practice over and over at them. Then we rank up into more experienced games and the whole dynamic shifts - our prior experience is largely thrown out of the window. I think this is the primary reason that 1800+ players consider many 1300-1500s nooby - we just don't know what we're doing and haven't practiced 20x20 naval maps. This means its very easy for us to make "simple" mistakes which garners BM from better players.
I don't have a solution to this, but do other people feel the same way?
-
@brokenprincess well that's the same as other "noobiness".
If your 800 rated and play other 800 rated your opponent will use bad strategies (as he's only 800), so you will learn how to counter those strategies and incorporate them in you gameplay. But your counter strategy is also bad and only works, because your opponents strategy was even worse and will cease to work once you meet an opponent of a certain level.
But you still shouldn't play 2k rated players as an 800 just to prevent learning noob strategies/counters from playing against bad players. that's just how you learn and improve at the game.
The only problem here would be the higher rated players being mad at the lower rated players for making mistakes.
They may be mad at the system for matching them with worse players, but being mad at lower rated players for making mistakes is just stupid. -
It isn’t like you randomly unlock 20x20 maps at the highest bracket. You get more and more of a percentage of your pool consisting of them while they also increase in complexity in turn.
It also isn’t like all current high level players sprouted with the knowledge to enjoy those games either. Lots of people gained rating playing gap, isis, wonder, canis, and so on. Eventually you just get tired of it and, if you had interest in getting better, you started joining the more complex maps with the better players.
Is there a growing pain period? Yes, but that’s just what improving is. You aren’t going to improve at anything if you just stay comfortable all the time.
-
@blodir said in Allow us to pick rating brackets for TMM:
It would be nice if we could enable some very noticable sound notification (preferably several seconds long) when we get a game. Then I could at least leave the room safely while I have speakers on.
FAF phone app, when?
-
@Nex I concede, you make a good point - the only problem is 800s are expected to play poorly and 1500s are expected to be somewhat adept, so it's a shock to my team mates when they witness my utter lack of skill xD
@FtXCommando I agree that growing pain periods are required, but the issue is that these periods affect team mates too. As I'm sure you're well aware, having a sub-par team mate lose a whole side of the map is very frustrating. This leads to less desire to play TMM for the top rated folks, which is the point of this thread.
-
Honestly it really depends on the situation. I honestly don’t even enjoy kicking like dual gap 1800 dudes from games as they at least want to try to play different stuff.
I played several tmm games yesterday. I got mad when my 2000 air player made less air to ground than Blodir, had less eco than him, and had less asf than him. I expect a baseline of like some 1800+ player to, even if suffering in one area, to have some comparative advantage that I can take into account and play along with.
I had a game where I had a 1200 get utterly crushed by a 1900 on tabula, but there is no reason to be mad about it. He’s expected to die even though he was on the slot that is supposed to crush his opponent.
Another game I had a 1200 air player so rather than risk the dice roll of a 1200 v 1400 late air game, I just try to end the game with t2 air from my front slot.
I don’t think playing simple maps is going to make games more enjoyable for higher rated players. You’re just moving the game quality up in one way while decreasing it in another.
What a high level player wants is a game where strategies are necessitated due to dynamic gameplay rather than by where the drastically low rated dude on your team is. You can’t get that unless you allow restrictions on what games can be made. Poor teammates can make dynamic gameplay not matter but so can bad maps.
-
@CasterNumeroUno is that the game you are referring to? The lowest rated one here is 1k
ID: 18202637 -
The game you highlighted is a game where there were over 10 +1500 players in the Queue. And yeah it was also not fun, imagine knowing that there are high rated players in the Q and you get this as a game where other are still stuck in the Q... And yes they were stuck in the Q as I was the only one who got the game from my friend list. Morax, FtX, Farms, Silene, Utena, the ruskies and others didn't get a game, while I was thrown into this "game". Not to mention multiple games that didn't launch prior to it, due to single person not connecting. Best thing those games were actually decent as the lowest rated player was 1.3k+ But alas someone had fucked connection.
Anyway more games that shouldn't have imo happened. And I said it once, but the fact that strife can be played in this setup is already cancer inducing, add to that retards in team and it's already doomed.
Also, this screen is like 75% of my games on TMM. Which as you can imagine is nowhere as fun comparing how different it plays out compared to standard 1.7k+ lobbies. -
I run some tests with the matchmaker and I agree that the matching is not ideal at the moment. There are two things we want to do.
- Improve the game quality in general
- Find a way to bias the matchmaker to high rated games somehow.
The first one is relatively easy to achieve, as we can tune config values of the matchmaker. The second one is a bit harder as this requires that the matchmaker doesn't match a semi-high rated player with lower players, so he still stays in queue until more high rated people show up. This is not trivial to do and requires to find some sort of metric that consistently does this first.
I have a script to run the matchmaker with test data and plot the results. If someone wants to try to modify the matchmaker code or to just try different config settings, then I can happily share my script.
In the meantime here is a comparison of the current config values and the new ones I came up with.
Current:
New one:
As you can see there is always a tradeoff between game quality and wait time in queue.
With these particular settings I believe we strike a nice balance between quality and wait time.
This will lead to more people "floating" in queue and not getting matched immediately. This is what allows the matchmaker to pick better matches. It may have the nice side effect that you can actually gauge the queue activity from the amount of people in queue, even shortly after a queue pop. (E.g. 20 people in queue -> high activity, 10 people in queue -> rather low activity. -
I take it that the upvotes and lack of further discussion is consent to go through with these changes?
-
I upvoted because I thought it was good that this was being worked on and the info was being shared with the community.
However, I think this
looks worse than this
If you want to implement the new slower-matching style for just like 800+ or 1000+ rated players, that might be a nice improvement, but I think implementing it for new and low-rated players would probably be a mistake. I think fast matching times are especially important for retaining new players.
-
I would agree that it looks like the parameter changes you applied end up pretty significantly increasing the wait time for new players or people who may have lost their first few games. Which I think is a little bit against the idea of standing up the matchmaker in the first place since it was intended to be a play for new players to start and jump into the game without the politics of custom games.
So now changing that to give them a harder time of getting a match isn't great. Also this seems to just increase the wait times across the board so we will just get more complaints of why people aren't getting matched when there are enough people in queue.
But ultimately unless we want to build some very complicated logic to perform matching it basically just comes down to if we want to cater to allowing new players to get a game or cater to more experienced players getting the games they think are ideal.
-
Someone who wants to play witn his friendlist, can host a custom game. No need to make everybody else wait in average two times longer. Everybody else who are happy to play with random people. With big or small rating disparity. With lower or higer rated players.
-
I'm happy with random people. I'm not happy with animals in game. Isn't it that hard to understand?
Why join the TMM if all it's gonna give me is some dogshit games that I can as well get by hosting all welcome which is gonna be insanely more fun due to the fact I know what I'm getting into instead of realizing 5s into the match that I'm a fucking zoo keeper.If LOL ain't pushing me into playing with Bronze players why the fuck FAF have to do it? I'd rather play no game and be grumpy in aeolus than have to babysit an effin zoo. Simple as that. Like I really don't understand why is it so hard to understand that putting Master players into Bronze/Silver matches is a retarded idea that ruins the idea of having a curated ranked mode...
No game > unbalanced game.
And yeah @BlackYps I'm totally onboard for those changes. I'd rather wait 40 minutes and have a normal game than wait 20 minutes and get a retarded no-game cuz yet again the game devolves into who got bigger 800 retard on the other flank. Or even better I'd rather not play at all if I'm gonna get more of low rated games. I'd rather sit in lobby sim.
-
@wikingest said in Allow us to pick rating brackets for TMM:
Someone who wants to play witn his friendlist, can host a custom game. No need to make everybody else wait in average two times longer. Everybody else who are happy to play with random people. With big or small rating disparity. With lower or higer rated players.
Me playing with my friendslist in a premade is the whole solution to the problem of me not wanting a game where teamwork is irrelevant in a 4v4 and where I need to find the optimal cheese solution to carry from my slot with an all in play.
