67 replies to this topic

[CobraFive]

I don't think any one of us can make really informed decisions on matchmaking until we have an idea of the gameplay. For example if the mechs are close in power, we might net even need matchmaking at all!

That said I'd enjoy A BV-style system. not a clone of table top, but something that ranks mechs on their overall power (not necessarily just combat power, either).

That or something based on the pilot XP/Mech XP.

[Kristov Kerensky]

BV systems don't work well when player skill is the determining factor in how the combat will actually go though, so it's not something I'd personally want to see used.

Again, Panther vs Atlas, skilled player in the Panther, newb in the Atlas, who do you REALLY think will walk away from that fight alive?

Now, if PGI removes player skills from the equation, by all means, go with the BV system, I personally won't be playing the game, so I really won't care. That's not a threat or anything, just a statement of fact, I won't play if player skill is irrelevent to who wins or loses in a computer video game based on PvP.

[Nik Van Rhijn]

Kristov Kerensky, on 22 December 2011 - 11:32 AM, said:

BV systems don't work well when player skill is the determining factor in how the combat will actually go though, so it's not something I'd personally want to see used.

Again, Panther vs Atlas, skilled player in the Panther, newb in the Atlas, who do you REALLY think will walk away from that fight alive?

Now, if PGI removes player skills from the equation, by all means, go with the BV system, I personally won't be playing the game, so I really won't care. That's not a threat or anything, just a statement of fact, I won't play if player skill is irrelevent to who wins or loses in a computer video game based on PvP.

Then what system would you propose?

[Woodstock]

Nik Van Rhijn, on 22 December 2011 - 11:39 AM, said:

Then what system would you propose?

I second this question ... and raise you a:

'How do you code a system to take the players manual dexterity, game experience, having a good/bad day, played 10 other fps games in a league but only played mwo for 2 days (so far) and 'x' number of other human issues into account?'

[Ghost73]

woodstock, on 22 December 2011 - 02:47 PM, said:

I second this question ... and raise you a:

'How do you code a system to take the players manual dexterity, game experience, having a good/bad day, played 10 other fps games in a league but only played mwo for 2 days (so far) and 'x' number of other human issues into account?'

It has been done before. I know BF3 uses a player skill value to help with matchmaking that is based on a number of factors including KDR, the rate at which the player gains points, etc.

Kay Wolf, on 21 December 2011 - 06:38 AM, said:

My initial thoughts when I first read about match-making was that I don't really care for something like that in campaign mode; in that mode the forces are there, at the planet, to do the job of either defending or attacking, and I'm not understanding why a match-making system would be necessary? I need to wait for more information on this.

It would be for drop in players that aren't choosing a specific match to join. Of course, this may just be how they do all matches.

Kay Wolf, on 21 December 2011 - 06:38 AM, said:

If they do regional servers, I'm going to lose 4 people from my ranks that I know of, unless we're able to proliferate across regional servers.

For an action-based MMO, it is a must, unfortunately. See below for my explanation.

Kay Wolf, on 21 December 2011 - 06:38 AM, said:

Well, if they go with Battle Value, or BV2, as from Combat Operations, a pilot's rating (Cadet, Green, Regular, Veteran, Elite) modifies BV automatically, and is the perfect system for matching out folks.

I'm hoping that the devs make it so that the BV value of individual mechs do not vary much (as in the game is very balanced), that way it won't matter who has the most expensive/heaviest/tricked-out mech.

Kay Wolf, on 21 December 2011 - 06:38 AM, said:

Is there a way to do this? To determine someone's ISP speed up and down, and then mix them with other appropriately leveled individuals? If there is, then I will lose, again, some of my folks.

Yep its called ping. Well, technically it doesn't record your ISP speed because it doesn't matter since most online games use extremely little bandwidth. Think of ping like sonar ping: it records the time it takes for a signal to reach the destination and back. This number is your latency in milliseconds. The smaller, the better. It is also the main reason they will have regional servers because trying to play with 300ms (.3 seconds) of latency between players on different continents is ridiculous (people can actually differentiate that amount of time in their head). For an action-based FPS like MWO, 100ms would be pushing it. I believe average latency between NA west coast and east coast is about 150ms, so I am hoping they have two servers here.

Edited by Ghost73, 22 December 2011 - 04:09 PM.

[UncleKulikov]

Matchmaking should group players based on their skill level, which can most accurately be based on their win/loss ratio.

If you join with a Lance/Squad, before entering matchmaking everyone would select their mechs and thus have a BV total that is used to match you against other full lances in a given planet.

[Damocles]

Ghost73, on 22 December 2011 - 04:06 PM, said:

Yep its called ping. Well, technically it doesn't record your ISP speed because it doesn't matter since most online games use extremely little bandwidth. Think of ping like sonar ping: it records the time it takes for a signal to reach the destination and back. This number is your latency in milliseconds. The smaller, the better. It is also the main reason they will have regional servers because trying to play with 300ms (.3 seconds) of latency between players on different continents is ridiculous (people can actually differentiate that amount of time in their head). For an action-based FPS like MWO, 100ms would be pushing it. I believe average latency between NA west coast and east coast is about 150ms, so I am hoping they have two servers here.

Growing up playing Jane's WW2 Fighters on a 56K modem and MW3 the same way (which was fast back then) I always laugh so hard when people have such issues with lagshooting.

It is nice that these things have mostly gone away but I just remember when it was part of the games and still is in some cases (Red Orchestra).

Cheers
/Damo

[Threat Doc]

Ghost73, on 22 December 2011 - 04:06 PM, said:

Yep its called ping.

I understand latency, but I was talking about a deeper program, something that could act more accurately. I know it's possible, so what about a means of averaging power across systems, to perhaps allow for more power to alleviate latency issues?

UncleKulikov, on 22 December 2011 - 04:20 PM, said:

Matchmaking should group players based on their skill level, which can most accurately be based on their win/loss ratio.

If you join with a Lance/Squad, before entering matchmaking everyone would select their mechs and thus have a BV total that is used to match you against other full lances in a given planet.

I would agree with this, but what happens if you have a Company sized battle preparing to take place, each unit actually having more than one Company of personnel, and one side can't match a full Company because their people are ranked to skill and/or ping levels lower than the average allowed by the other Company. No, I don't think either of those methods will work; even if the lower-ranked Company will lose to the full Company, the battle should be allowed to move forward, allowing all those who want to play to do so.

[Ghost73]

Kay Wolf, on 22 December 2011 - 04:30 PM, said:

I understand latency, but I was talking about a deeper program, something that could act more accurately. I know it's possible, so what about a means of averaging power across systems, to perhaps allow for more power to alleviate latency issues?

Do you mean like delaying information sent to the server so that everyone has the same ping? AFAIK, it is impossible to reduce latency through programs, as it is entirely dependent on the quality of the cables that the information goes through. Maybe once everyone uses fiberoptic cables, we can have a global server for action-based MMOs.

Kay Wolf, on 22 December 2011 - 04:30 PM, said:

I would agree with this, but what happens if you have a Company sized battle preparing to take place, each unit actually having more than one Company of personnel, and one side can't match a full Company because their people are ranked to skill and/or ping levels lower than the average allowed by the other Company. No, I don't think either of those methods will work; even if the lower-ranked Company will lose to the full Company, the battle should be allowed to move forward, allowing all those who want to play to do so.

Matchmaking is for people who just want to jump in a game and play. For planned matches like these ones, matchmaking is not considered because it is more about teamwork than individual skill.

[Nik Van Rhijn]

The trouble is I'm not sure that there is any existing system designed to deal with the complexities of MW both pilots and mechs, particularly in the Campaign game. As for Woodstocks question about peoples previous experience. PGI can only go on the stats they have in game so all people will start off at zero. Which means many times we will have total mismatches as there are some extremely experienced FPS players around, with far more of us at a more mundane level. To put it bluntly - a lot of people will get stomped to start with until levels get established. In fact I think it would perhaps be a good idea if the campaign game did not start for a few weeks after the game goes on line to allow stats to be established. No one wants to bring their House or Merc Co into disrepute because they had the bad luck to start out against groups that later proved to have some of the top players in.

[Hellen Wheels]

I envision something along the lines of MPBT:3025 and would hope that there is a similar type of communication interface that broadcasts updated newsfeeds about ongoing or planned battles (that a lone wolf might drop in to, just to shake things up, or maybe even the odds, or fulfill a contract).

Example: some House declares for an unclaimed planet or sector and sends a couple of lances there, orbiting in their Dropship. If no one takes them on w/in 15 minutes, then they claim the planet (and the c-bills) without a fight. Or, another Merc group or House sees this intrusion, and sends their Dropship. Now it's 8 versus 6 in orbit, the battle will start in 7 minutes, and the six needs a few mercs or at least a lone wolf.

Seven minutes later, the dropships drop their loads, whether it's 8 v 6 or 9 v 10. And it would be nice to be able to "hot drop" into an ongoing battle. You could have hot drops landing all the time, contesting that one planet all night long.

I can dream, can't I? :-)
H

[UncleKulikov]

Kristov Kerensky, on 22 December 2011 - 11:32 AM, said:

BV systems don't work well when player skill is the determining factor in how the combat will actually go though, so it's not something I'd personally want to see used.

Again, Panther vs Atlas, skilled player in the Panther, newb in the Atlas, who do you REALLY think will walk away from that fight alive?

Now, if PGI removes player skills from the equation, by all means, go with the BV system, I personally won't be playing the game, so I really won't care. That's not a threat or anything, just a statement of fact, I won't play if player skill is irrelevent to who wins or loses in a computer video game based on PvP.

You have to use both Battle Value ( a measure of a player's on paper capability) in addition to their win/loss record.

Why Battlevalue:
You need to include the value of the equipment that you bring into battle, regardless of your skill. If I bring a PPC to the field, that is a PPC that can either be the most effective PPC ever, or the worst PPC ever. That variability still contains both extremes, and so the constant (the weapon) needs to be included in the calculation.


Why win/loss:
Win/loss ratio is a well rounded view of a player's performance. If a player consistently wins against people, they go up because they are reliably performing better than their adversaries. If a player is consistently losing, then that means they are reliably performing worse than their adversaries. When a player is winning and losing at about the same rate, they are roughly in the skill slot that is appropriate to them.

Why not Kill/Death ratio:
players can provide utility to the lance without bagging kills, like spotting for LRM artillery or snooping enemy movements. Especially in Objective matches, the win should be the deciding factor for progression. KD ratio reinforces selfish play.

[Kristov Kerensky]

How would I balance out the matches if skill is what determines who wins and who loses?

Simple..PGI is trying for company vs company ranked matches, 12v12. Give each lance, 4 mechs, X tonnage total to drop with. Each lance decides what they'll use, and once each company is ready, they all drop.

That's it..no looking at who's been playing for X minutes, who's got Y kills, who's got Z losses, each side gets X tonnage per lance, that's it, let the PLAYERS make it happen.

Why I don't believe in using stats for matchmaking? Simple, too easy to pad those, just look at every game out that tracks stats and you can see it. HUGE issue in Battlefield 2, something EA never has gotten a handle on. Same with the console BadCompany games, Modern Warfare games, and BF3 itself. I know people who did this, entire clans of guys would work together to pad their own stats with smurf accounts as the targets. They are doing it today in the newest ModWar3 and BF3, they'll do it in every game where stats are tracked, it's an ego thing for people without skill to prove they are better then those with skill. Sucks to be them doesn't it.

I'm sorry if this makes some of you feel that you'll always be losers, none of us STARTED out as expert killing machines in online games, it took time, practice, dedication and the skills to build on. If you can't put in the time and don't have the skills, why should you expect anything else? Online games are a sport like any other, you have to have the basic skills required in the first place, then you have to put in the time and practice to become better, and you never stop working on getting better or you find yourself getting your head handed to you all too soon. I don't play PvP games just to kill time, I've got MMORPGs for that, cause that's what those are designed for, killing time. PvP games, that's about testing yourself against someone else, to win against someone else, because it's fun, it's enjoyable, we are hardwired for it. Don't want to, or can't, spend the time to be the best, then accept that, don't try and make ME be less because you can't be MORE, that's bs. Competition is about who's better, it's not about who's more equal.

Best way to become better..play against people who are better then you. Rookies in sports don't play against other rookies, they get tossed right into the thick of things, so they can learn and become better..or they fail. Do or do not, there is no try, as a wise hand once said.

[John Frye]

Kristov Kerensky, on 22 December 2011 - 06:31 PM, said:

-snip-


Don't want to, or can't, spend the time to be the best, then accept that, don't try and make ME be less because you can't be MORE, that's bs. Competition is about who's better, it's not about who's more equal.


-snip-

Kristov, I feel like I am seeing a theme in how you use the term "skill" across a number threads. How about giving us a better definition of what you are calling skill, since it seems like it's a more specific (or maybe its broader, either way, we can call it different) one than alot of us seem to be thinking of when we read the word. That way we won't be misinterpreting the concepts you are trying to share with us.

Edited to be less leading

Edited by John Frye, 22 December 2011 - 08:37 PM.

[Mavek]

Kay Wolf, on 22 December 2011 - 04:30 PM, said:

I understand latency, but I was talking about a deeper program, something that could act more accurately. I know it's possible, so what about a means of averaging power across systems, to perhaps allow for more power to alleviate latency issues?

I would agree with this, but what happens if you have a Company sized battle preparing to take place, each unit actually having more than one Company of personnel, and one side can't match a full Company because their people are ranked to skill and/or ping levels lower than the average allowed by the other Company. No, I don't think either of those methods will work; even if the lower-ranked Company will lose to the full Company, the battle should be allowed to move forward, allowing all those who want to play to do so.

I think the whole MM question may stem from a game some of us beta tested many years ago called 3025. This was a lance based game of territory control. I was on board with Davion when we roflstomped Kurita. I believe that there were shenanigans involved where people were loggin in as Kurita and dropping with alts and intentionally losing to themselves, or some such nonsense. The Kuritans only answer then was to hack into the game and rofl-stomp the Davs with aim-bots and other such nonsense. EA bought and canned the whole thing in the middle of this nonsense, dooming all of us to more than a decade of waiting for what for many of us is the end game of end games...MWO.

The problem with just letting a team of new guys go up against a team of highly coordinated, hardcore BT guys, is that Luthien will be under Davion control within days or even hours.

Hence, the whole MM conundrum.

I like a BVxPlayer ability/XP system. Each mech should be assigned a relative BV based on a 4/5(average) player, but then adjusted for the xperience or ability level of each pilot, maybe with a multiplier, just like TT, allowing newer players to field more heavies (firepower), to perhaps help to balance out the descrepancy between two companies of combatants.

This means two things I can think of:
1) Each of us will have some kind of rating/ranking/multiplier attached to us, so that we know where we stand in this world.
2) Each of us will need a stable of different mechs and/or access to a lot of different chassis at the beginning of each match, so that we can tweak the numbers to allow for a relatively balanced drop, at least at the bigging of the match.

How each match plays out will of course be up to the real skill and teamwork of each side, and a bunch of green players will still probably end up getting beaten by the more organized/experienced team. But I think that all players will appreciate that something is being done to make it so the outcome is always in question, as opposed to it being a blowout. More fun for everyone. More likely that everyone sticks around and plays for longer.

PS...This even makes more sense after reading Kristovs comments. I dont believe he is being condescending at all. Every game I've ever played had 'fightclubbers' out there padding their stats, so they could get onto the forums and tell everyone how great they are...which of course screams: POSER. But I digress...padding your stats...in the system I just advocated...would be counterproductive. If someone did try to pad their stats with smurf accounts, all that would do is artificially drive up their own BV and give their opponents a tonnage/equipment advantage, with which it would be assumed that the fight-clubber, epeen type guy would get his *** kicked relentlessly, because he has gimped himself and exposed his own lack of ability to win against real players.

Edited by Mavek, 22 December 2011 - 08:45 PM.

[Threat Doc]

Kristov Kerensky, on 22 December 2011 - 06:31 PM, said:

How would I balance out the matches if skill is what determines who wins and who loses?

Simple..PGI is trying for company vs company ranked matches, 12v12. Give each lance, 4 mechs, X tonnage total to drop with. Each lance decides what they'll use, and once each company is ready, they all drop.

That's it..no looking at who's been playing for X minutes, who's got Y kills, who's got Z losses, each side gets X tonnage per lance, that's it, let the PLAYERS make it happen.

Well, okay, but with Battle Value in play, player skill in accordance with their past performance acts as a playing field leveler. Tonnage will make it so that each certain type of game-play, the same static, boring chassis and loadouts as have gone into all previous games of type, will be used. That leads to strategic and tactical complacency once in-game, because everyone knows the same BS they did last time is the BS that's called for, now, depending on the type of map.

Battle Value helps remove that. Everyone on a team knows what the max Battle Value is, some players will take the same 'Mechs time-and-time again, while other players, with a much broader selection to choose from, even if it's not the 'Mech they have the most MXP in, will take different configurations, different 'Mechs, and their fire team and Lance mates won't know exactly how to act with their own team mates, let alone the surprises that are most likely to come from the other team choosing different 'Mechs, as well. Now you have zero complacency, you have new tactical advantages to exploit and disadvantages to either turn into advantages or protect from, and you have different and useful 'Mechs showing up all the time.

THAT is what I call fun, especially since there's only going to be a single fight, at present, for each planet .

[Mavek]

Look at it like handicapping in golf. If you have a 'ZERO' handicap, that means you are as good as it gets, and dont need any help to win. But if you just started and have a 20 handicap...now you have a shot to make it interesting for the guy that hits 95% of the fairways with each drive.

[Threat Doc]

hehe, Never really understood golf. I think I get what you're talking about, however.

[Kristov Kerensky]

Well Hellen, it's a PvP game, what do you expect? Hand holding and sing alongs? Seriously, WHAT do you expect? Should players who don't have any twitch skills, tactical sense and a good understanding of strategic thinking be made equal to or better then those who do? What was the last PvP game you played? Did it do that? NONE of the PvP games I play do that, they don't give a rat's *** about what your skills are, you can either play the game well enough to beat others OR you can't, that's it. It's up to YOU, the player, to change that by making yourself better at the game..or accept the fact that you aren't that good and have fun anyway. I have friends who suck in most of the PvP games we play, they really do, but they have a blast anyway. I have fun whether I'm kicking *** and taking names OR watching yet another SOB take my dogtags...doesn't matter to me, the dance is what I'm there for, not the trophy. Not that I mind getting that trophy or that I won't work my *** off to get it mind you...

John..skills are a set of things that you need for a first person shooter style PvP game..

Twitch..reacting to the enviroment, aim and weapon control, anything really that's based on eye-hand coordination and reflexes

Tactical..situational awareness, IDing the weakness of your opponent/s and how to exploit them

Strategical..big picture stuff..where the choke points are on a map, best way to exploit the terrain/enviroment

Not a FULL list of what each set encompasses but enough to give you an idea. Twitch skills are really the biggest thing most people need to work on, doesn't matter how tactically minded you, what your strats are, if you can't HIT the enemy after all. By the same token, good twitch skills may let you kill the enemy, but if you can't read an ambush or just be aware of the area around you, doesn't do you much good when you get nailed in the back. Strategic thinking..this is really an optional thing for the majority, you really only need 1 good stratman in a group, someone who can look at the overview, see what needs being done, and the rest follow his plan and make it happen. I've been really lucky to game with a few master stratmen in the MW series over the years, Dacobra, CO of SRM is probably the BEST stratman I've ever seen..man is amazing with the big picture and seeing how to make it reform to HIS vision..just amazing. He's got the other 2 sets as well, better then most..or he did have..it's been a while since we danced together, just hope he's kept his skills up as I have over the years

Kay..how would you determine a player's BV score? Stats? Sorry, but I can make my stats go up or down, depending which gives the best advantage..and for BV, it would be having bad stats. People pad their stats in games they have to PAY for that don't give them a single advantage, what do you think would happen in a F2P where those stats WOULD give an advantage? That's just a really bad idea. Player stats just can't be used for anything but bragging rights, giving the player anything ingame for them WILL lead to massive abuse of the system, it's happened before. Sorry, but I've spent too many years dealing with the people who exploit anything they can in computer games to think anything but the worst of them when such an obvious exploit is offered up.

Mavek..see the above..give people something based on stats, they'll exploit that system, it's a bad idea.

Sorry folks, I'd LOVE to believe that every single one of you is just as morally bound as I am to NOT exploit anything in a video game, but the fact is, people will do it, they do it all the time, they're doing it right now in Modern Warfare 3 and Battlefield 3, they've been doing it since the days of Doom..I used to hunt for them in MW2 for the Registry on AOL, I've done it as an admin for servers for various games, I've been paid to do it professionally, so I know just how wide spread and common place it is. And in a F2P game..please..are you kidding me?

Fact is, if the game relies simply on the player's OWN abilities to be the determining factor in who wins and who loses, then exploiting the system is simply that much harder, people have to resort to straight up hacks, and those are a lot easier to detect and deal with then soft exploits like stat padding. Plus, it means that when you win, YOU did it, there was no artificial constraint on your opponent or boost for you..YOU did it. Don't know about you Kay, but it means more to me to win because I was simply better then to win because my opponent was better but the computer gimped him.

[Colonel Bogey]

What do you mean by matchmaking. Ranked matchmaking would have to be different than unranked matches. The system would also need a way of knowing what teams are premade so they fight other premade teams. Just the same but the reverse for a group of lone wolf players. Hopefully if implemented well even teams partially made up of premade and lone wolves will have equal numbers of premade on each team. I find this is an easy way to balance some games. If their is a ranking system in game, hopefully players of skill will be matched with others of similar skill, however, that is always hard since ranks usually don't have anything to do with skill. Usually players with more time on their hands will have a higher rank. I do not care too much which way they implement their matchmaking system as long it provides enough to allow players new and old an experience they can enjoy. The formula used in call of duty sounds good, however, I am not sure if it work so well here.