Ladders in 1.0

Why not get some experience yourself before taking on others in the community?

Why not play with some friends?

Why not meet some of those people and team up with them and learn?

Our data is that only a small fraction of the player base actual wants to play ranked. In addition you need a large pool of people in a ranked game to actually make the math work. So again, the argument here is do we need it for 1.0 launch? Remember we are going to do competitive features, it's a question of relative priority is it not?

Does PAStats actually match people based on skill? When implemented it was first come first serve and I don't know that the number of simultaneous users has ever been high enough to justify changing.

And what about integrating matchmaking with user generated games? As long as the game is unmodded, nothing should speak against showing ranked games in server browser (with a notice that this game isn't managed by any player!), as well as possibly matching ladder players to regular games.

Or for the beginning: Just show the player rank (whatever ranking system you are using) and average rank in lobby.

The issue without ladder isn't so much the lack of such, but the lack of clear distinction between skill levels when selecting your match, resulting in rather frustating/boring matches.

It's rather logical that modded games are excluded from ladder by default. Same goes for eco changes or games with all vs AI-setups. It only gets tricky when you take custom systems into consideration since some of them are clearly to be classified as "fun map".

Anyway: Showing ranked, Uber initiated games in the Lobby would surely help boosting ranked games without immediately splitting the player base between two systems.

Auto-matching every player entering the lobby is just the logical next step. That means always showing at least 1-2 rated and properly ranked matches in the most prominent position.

This isn't suited for a officially ranked ladder since it's open to match manipulations, but that doesn't concern much. It's more to be thought of like the "level" system you have in various coop enabled RPGs.

It kinda goes both ways, to us the chain of logic saying there will still be lots of/more players a "couple of months" after release isn't obvious. Many here feel that players will peak at release and without long term hooks to hold on to players there will only be fewer once Ladders/Matchmaking hits later on.

Basically we're both seeing the same type of issues with each other's viewpoints.

Mike

Hey Guys,

Just so you all know there are non-competitive players out there. I am most certainly one of them. In fact all my mates are the same. We prefer playing friends or AI instead of random people. Always been that way and don't see us changing any time soon. I don't enjoy playing against random people in the slightest but I recognise that not everyone else is like me. @neutrino, thank you for listening to the community but also recognising not all players are the same and for building a game that will cater for all types of players. Even if all the features that I would like to see won't be in 1.0 I have faith and trust that they will be in due course.

Cheers,

-Todd

What percentage of players do you think are in this camp? Our estimate is that it's low. Perhaps yours is that it's high?

Also you can still play multiplayer on day one. If the surge of players is huge chances are you are going to be playing with a newb anyway.

I still think most of the arguments are assuming that *most* players want to play competitive multiplayer. I haven't seen any evidence from an RTS game that I've worked on that this is the case. Most players seem to want to play against the AI. Another big chunk want to play with their friends.

I don't use pastats ladder and I do not plan to use any 1v1 ladder. I've only played PA with a bunch of friends BUT I'm still thinking that ladder at is release is important (but less than a polished game).

For two reason :

- all people (except the ones, I play PA with) I know that play a game for more than few months, are on games with ladders, online stats, success, etc. Games focused on multiplayer or social aspects.

- the game will be DRM free (which is a great). When I was young, I always said in my mind: "I download the game and if I enjoy it I'll consider to buy it." In practice, I did it very few times (and half of the time it was to have some specific multplayer features).
A good way to keep the liberty of DRM-free game and push people to effectively buy the game is to give them a good reason to log on your server... ladder can be one of these reasons.

I was actually just referring to how PAStats ranks people by awarding/deducting points based on wins/losses. If your goal is to match people with similar skills, you could use a points system like that as a basis for comparison.

Maybe create a matchmaking filter in the lobby that could add matchmade games along with 'regular' ones?

I haven't seen much evidence that launch is going to be the peak player count. Games that continue to update tend to continue to grow. If this is your argument then it's not a good one.

Again remember we are talking about timing here. Not whether we are going to have the feature but when.

You don't think a big announcement about a competitive focused PA update would bring lots of players to the table to try out the new features? The idea that launch is all or nothing is simply not the way modern games work.

What kind of concurrency did Dota2 and LoL have at launch compared to now?

I know I am breaking basically both types of ways people like to matchmake, but what if ladders exist, and the matchmaking system was more like chess online?

By which I mean, you get ladder numbers, but you don't get auto'd, but when you press the matchmake button you simply get a list of other people and their rank number who clicked the matchmake button and anyone can click on anyone to start one. For ease of access, make another button that instantly tries to match with the closest ranked player in list.

Worse case scenario, and this happens with matchmaking anyway, is that you click on a few people who already clicked on a few other people so it doesn't go through the first time, basically you attempt to match with one person, you may have to attempt to match with another and another. With the faster "match with nearest rank" button it may still happen but you just click the button 2-3 times until you match with someone.

Obviously if a match was sent, the message should be accept or decline. Possibly doable from the bottom right social bar when game invites pop up.

The HARDEST part about this, is making it autogenerate fair maps and rules and such for 1v1 instant matchups. PAStats does this already so I think Uber can also follow suit and just have a bunch of templates that quickly push through to make a game while the players get the generic matchup over-screen hiding all the auto-field-generation.

The main reason I suggest all of this, is because matchmaking the player can't see, ends up possibly being really poor matches, and/or questionable in player population. 12 matchable players would make this function, I have seen it function in other games with 12 people.

Yah I agree that it's not necessarily the highest priority.

Perhaps a simple mechanism like showing an icon next to player names based on how many hours played would help? e.g.. bronze for less than 10hrs, silver for less than 100hrs and gold for over that.

Then at least in the lobby you have an idea of what you're getting into. And in a FFA you can team up on the most experienced player

In all honesty, I'm like this too. I have no interest at all in playing with randoms online, and know nobody at all who actually is. Heavy comp RTS has always seemed super niche. Honestly based solely on who I know, and some experience from looking at other games, I'd be impressed if more than at most 10% of the player base ever played a multiplayer vs humans match with random people online.

Not necessarily competitive. But often enough I don't want to play alone. And it's just as frustrating to get matched against a competitive player as it is boring to get matched against a casual player when playing competitive.

It's usually the most fun when you get matched with equaly competent players. Even when playing against the AI!

A rank system which testifies some experience surely helps a lot at matching players with comparable skill levels - even when not tied to an officially ranked ladder.

Basically this is pretty similar to just showing ranking in the server browser (which isn't the worst idea I've ever heard).

10% is the ballpark I would guess as well based on previous experience and looking at PA.

Yes, it's true that showing a ranking can help players figure this out. I'm somewhat liking the hours in game metric someone else talked about although they all have massive downsides.

To reiterate my previous suggestion...

Could we get a simple indication of player experience (in terms of in game time) added to the game?

Something that shows how long someone has played at a glance.

The reason for this is simple- you can easily identify new and vet players alike. Then everyone can make a decision of who to or not to play against, and you can manually balance teams.

Without some kind of metric to establish base skill it is impossible to manually balance a loby.

This method isn't perfect, but it is enough to avoid total stomp games- Spring has implemented this since the start and it's something that is so simple yet so effective.

Having never seen any stats.... I agree! At least for me personally and everyone I know in terms of RTS games it's always single player campaign or skirmish mode if available first. Then play with friends either online or LANs. Never touch matchmaking at all.

Cheers,

-Todd

You beat me to it!