Jump to content

Mektek's Heavy Gear Assault


275 replies to this topic

#161 Heffay

    Rum Runner

  • PipPipPipPipPipPipPipPipPipPip
  • The Referee
  • The Referee
  • 6,458 posts
  • LocationPHX

Posted 22 January 2015 - 06:23 PM

View PostRedDragon, on 22 January 2015 - 04:03 PM, said:

Bwahahaha! :D
Oh wait, you were serious? :blink:

Elo can never work in a game like MWO if it isn't combined with some kind of BV system. You can't compare it to a shooter where every player starts with the same gear. That's like saying Elo works when it matches two generic shooter players of arguably the same skill, one starts with only a knife and the other with a SMG or sniper rifle and full body armor.


If someone got to an 1800 Elo rating using just a knife, and the other person got to 1800 using a sniper rifle and full body armor, then yes, that is a well balanced match.

That is why Elo works. If you can't understand why those two players are equivalent despite radically different gear, then I don't know what to tell you. It's pure math. The math doesn't lie. And if you change the conditions, it is *self correcting*.

BV/TV will never have that feature.

#162 TrentTheWanderer

    Member

  • PipPipPipPipPipPip
  • Legendary Founder
  • Legendary Founder
  • 264 posts

Posted 22 January 2015 - 06:36 PM

@Heffay Right now in the lobbies you can look and see who is there before the match starts, and if you're worried you might be outmatched you can check on the leaderboards ( http://www.heavygear...ort=kd_adjusted ) and use the (placeholder) scores listed there to evaluate skill level. Heavy Gear Assault is going away from the "server authored battle instances" that we've seen in games like Hawken and MWO where the server makes all the decisions about who gets to play against who, and now the players are getting to make that decision!

I'm glad you guys have been talking a bit about the financial model, since it's been such a hotbutton issue especially in the wake of free2play robot games like Hawken and MWO. Of course, the most up to date info can be found on the Heavy Gear Assault web page and on the Heavy Gear Assault forums (making an account and posting/browsing the forums is free, and the Devs post and reply there every day). If you check the FAQ you'll find the answers to all of your questions Heffay! Here's a link in case it doesn't show up right away: http://www.heavygear.com/game/faq

A really big difference in the business model and the product delivery between what you might be used to with MWO is that Heavy Gear Assault is going to be provided through a modular release system, a lot like Star Citizen. On completion players who have purchased the game (with a one-time transaction) will own the game, and like game releases of old as long as they have a computer that can play Heavy Gear Assault they will be able to play it forever! The server code and everything needed to play will go to the customer, who won't ever have to pay again. This also means that if the main servers ever go down the players can still play Heavy Gear Assault and host their own servers, as opposed to situations where if the MWO servers turned off no one could play anymore.

The economy is also taking a new approach that i'm really interested to see, and it reminds me a lot of the Firefall economy. Players gain recipes for unlocking equipment production by playing or purchasing a game package. Once the recipe is unlocked they can fabricate the item and put it on the in-game marketplace where it's bought and sold for in-game currency. This is a pretty progressive approach to managing in-game items and progression, as it lets the free market manage things like supply and demand and generates a lot of metadata that can also be used to help keep the game balanced and fun.

More than anything else though, MekTek are the old guard who have played and loved mech games as long as any of us, and they've always supported the mechwarrior community (as long as they were legally permitted) with their time, skill, and effort to bring mech games out to more people. Even if it isn't a "replacement" for MWO (they're really so different already!) it's an entry in a series that i've always loved done by the guys that have been giving me great experiences to enjoy for years. I'd recommend anyone take a look at what they're doing with Heavy Gear Assault, it really is something special.

#163 Heffay

    Rum Runner

  • PipPipPipPipPipPipPipPipPipPip
  • The Referee
  • The Referee
  • 6,458 posts
  • LocationPHX

Posted 22 January 2015 - 06:58 PM

View PostTrentTheWanderer, on 22 January 2015 - 06:36 PM, said:

I'd recommend anyone take a look at what they're doing with Heavy Gear Assault, it really is something special.


I'm probably going to get in at the $40 level, because I want to support this game company, and the more interest there is for giant stompy space robots from the future style games, the better for the genre. It's a big universe, and there is room for both companies. I really hope they can deliver on the HG IP!

I may even make some videos. Hopefully fan art is something they support through quasi-legal misappropriation of their game assets... :ph34r:

#164 Dino Might

    Member

  • PipPipPipPipPipPipPipPipPip
  • Little Helper
  • Little Helper
  • 2,030 posts

Posted 22 January 2015 - 07:15 PM

Well, I will probably get it this weekend - DANGIT. I really need to stop adding distractions to my life, but HG2 was the heat!
Anyone remember the epic battles between MERC and C7C? CTF 1k Karaq, 3 cap max, lasting for 45 minutes because each team was so darn good.

I still remember a bunch of the handles of guys I played with way back when and their unit designators. I finally stopped when there were only 2-3 servers left and maybe 20 people on per night. It was a sad day when it finally shut down.

Kodiak, LRG, FC2. Fire off all 48 shots, get about 20 kills in that span of ammo, then swap to fists and go Mike Tyson on the enemy gear bay.

Also, anyone remember the admantium base turrets? You hit one of those things going 150kph in your light and you just blow up, because any collision with them damaged you, and the collision detection warped you partway through them, causing hundreds of collisions in less than a second? Hilarious as gear parts go scattering across the enemy base. I would suicide on our own base turrets when I ran out of ammo and needed a new gear.

#165 TrentTheWanderer

    Member

  • PipPipPipPipPipPip
  • Legendary Founder
  • Legendary Founder
  • 264 posts

Posted 22 January 2015 - 07:36 PM

View PostHeffay, on 22 January 2015 - 06:58 PM, said:


I'm probably going to get in at the $40 level, because I want to support this game company, and the more interest there is for giant stompy space robots from the future style games, the better for the genre. It's a big universe, and there is room for both companies. I really hope they can deliver on the HG IP!

I may even make some videos. Hopefully fan art is something they support through quasi-legal misappropriation of their game assets... :ph34r:


I'm looking forward to hearing "What does the Gear say" lol :P

#166 Heffay

    Rum Runner

  • PipPipPipPipPipPipPipPipPipPip
  • The Referee
  • The Referee
  • 6,458 posts
  • LocationPHX

Posted 22 January 2015 - 07:37 PM

View PostTrentTheWanderer, on 22 January 2015 - 07:36 PM, said:


I'm looking forward to hearing "What does the Gear say" lol :P


Skate skate skate skate skate skate skate skate?

...


I might have to come up with something different for that. ;)

#167 Wolf486

    Member

  • PipPipPip
  • 85 posts

Posted 22 January 2015 - 08:20 PM

View PostHeffay, on 22 January 2015 - 06:58 PM, said:


I'm probably going to get in at the $40 level, because I want to support this game company, and the more interest there is for giant stompy space robots from the future style games, the better for the genre. It's a big universe, and there is room for both companies. I really hope they can deliver on the HG IP!

I may even make some videos. Hopefully fan art is something they support through quasi-legal misappropriation of their game assets... :ph34r:


That's awesome! I'm glad to hear you had a change of heart!

#168 Vassago Rain

    Member

  • PipPipPipPipPipPipPipPipPipPipPip
  • Bridesmaid
  • Bridesmaid
  • 14,396 posts
  • LocationExodus fleet, HMS Kong Circumflex accent

Posted 22 January 2015 - 08:38 PM

Posted Image

#169 Anjian

    Member

  • PipPipPipPipPipPipPipPipPip
  • FP Veteran - Beta 2
  • FP Veteran - Beta 2
  • 3,735 posts

Posted 22 January 2015 - 11:59 PM

View PostHeffay, on 22 January 2015 - 04:21 AM, said:


Battletech has a BV system that is essentially the same. And it suffers from the exact same problems. It's incredibly easy to game. The meta will shift to low TV builds that have incredible synergy. And that can't be fixed then, because if you muck with the TV of any individual component, it'll ruin ALL the other builds that use one of the components.

Almost all modern competitive games who rely on matchmaking based around skill, not gear. LoL, DOTA, WoW, SC... I haven't seen anything about HGA having some sort of player ranking system. They really need one.


MWO uses ELO, and yet there is plenty of stomps.

Hawken also uses an ELO system, now you can see where that went.

On the other hand, World of Tanks, probably one of the most competitive PvP games out there, uses a Tier system (matchmaking based on gear), and I encounter much less stomps on that game than MWO. Every tank therefore has a certain tier or rank, and constantly balanced to fit that rank. The matches though have a tendency to see more of the lower tier tanks get wiped out early, then let a small elite group of higher tier tanks carry the rest of the game to the end.

War Thunder uses a BR or Battle Rating system, which is also gear based, but BRs also go up and down, based on server statistics on K/D and survivability (length of time it stays in the game). While highly controversial, its the one game that I engage that has the least stomps, and the situation is so fluid, you can have the tides of war reverse and counter reverse, and a near victory turn into defeat all of sudden.

Under the BR system, if one plane or tank is a bit too strong, or has overly strong synergy to its tier, the BR would rise correspondingly with its K/D ratio and game survivability. When the BR rises, it will face increasingly stronger and more advanced planes or tanks. If a plane or tank proves too weak for its tier, then its BR falls where it will be matched with increasingly weaker competition. The advantage of the BR system is that the attributes of the plane or tank is preserved, and the only thing War Thunder has to do is to adjust the said vehicles to be more authentic to their historical counterparts. These adjustments are made when they are able to find more historical documents and data, then adjusts said plane or tank to the new data. Thus in War Thunder they don't boost or nerf for "balance" but to adjust in terms of improving historical accuracy, then let the floating BR adjust the vehicle to a rightful tier. The system can be rage inducing, as its not perfect, and many things seem too powerful for one group, too weak on the next as the BR doesn't respond quickly enough, though War Thunder always tends to have microupdates and patches very frequently. But overall, War Thunder doesn't have that predictability in MWO in that sense that you are going to win or lose.

Back to Heavy Gear 2, I remember the matches, but I dont' remember the game having the ugly stomps that MWO has. However, it has a problem with stacking just like Mechwarrior 4, which happens when players are allowed to choose their teams and they often choose the teams that they know have good players. And that problem occurs in all these earlier PvP games and this created the matchmaking ELO system or forced randomization of each team. But I also the fact when I was playing HG2, good players would voluntarily switch to one team or another just to balance the teams out.

I remember the 1K, 5K and unlimited battles. I also remember that at a certain time, due to the way the Japanese internet priced themselves (free unlimited at certain hours, usually midnight to 3AM Tokyo time), the Japanese players would all come in, and there were lots of them. They usually set up 3K TV games but some would also participate in the "west" unlimited TV battles.

#170 RedDragon

    Member

  • PipPipPipPipPipPipPipPip
  • 1,942 posts
  • LocationKurpfalz, Germany

Posted 23 January 2015 - 03:50 AM

View PostHeffay, on 22 January 2015 - 06:23 PM, said:


If someone got to an 1800 Elo rating using just a knife, and the other person got to 1800 using a sniper rifle and full body armor, then yes, that is a well balanced match.

That is why Elo works. If you can't understand why those two players are equivalent despite radically different gear, then I don't know what to tell you. It's pure math. The math doesn't lie. And if you change the conditions, it is *self correcting*.

BV/TV will never have that feature.

The (pretty obvious IMO) point is not that both players earned their Elo with different gear but that those players have the same skill level and yet are thrown into the game with highly unbalanced gear.
In MWO, if you are put in two identical games, the first time with an Awesome with nothing but a TAG and the second time with a Meta-Direwolf, don't you see that Elo just doesn't work? You have the same skill and Elo value in both matches, but in one of them you are an effective player, in the other a liability to your team.
Sure, this example is a hyperbole, but the mechanics behind it work their dirty magic every day in MWO. The best player becomes a lot less useful while grinding bad variants for example. And yet he is forced against players of the same Elo value with Meta-Timberwolfes.

#171 Heffay

    Rum Runner

  • PipPipPipPipPipPipPipPipPipPip
  • The Referee
  • The Referee
  • 6,458 posts
  • LocationPHX

Posted 23 January 2015 - 05:01 AM

View PostRedDragon, on 23 January 2015 - 03:50 AM, said:

The (pretty obvious IMO) point is not that both players earned their Elo with different gear but that those players have the same skill level and yet are thrown into the game with highly unbalanced gear.


You have that completely wrong. The players have vastly different skill levels, but are both rated 1800 based on the gear they play with. You take the guy with the knife and give him decent gear, and he'll rise out of the 1800 bracket, because the beauty of Elo is that it is self correcting. Same thing with your Awesome vs Dire Wolf example: Elo will correct itself after a few matches.

That is what makes it a good match. The guy with the knife playing at at an 1800 Elo isn't a hindrance to his team, because he *got* to 1800 using just that knife. That's skill right there.

Edited by Heffay, 23 January 2015 - 05:03 AM.


#172 Dino Might

    Member

  • PipPipPipPipPipPipPipPipPip
  • Little Helper
  • Little Helper
  • 2,030 posts

Posted 23 January 2015 - 11:34 AM

I think Heffay has it right in terms of understanding ELO and how to use it for a matchmaker. I think it's crucial that they allow games to be formed with limited TVs, and I would be ok if they use ELO for balancing teams or if they don't.

Playing every game as unlimited will become really annoying, if HG2 is anything to be considered. The other problem is how to proportion ELO and TV if you try to use them both to balance teams. What if a low ELO player makes such a high TV gear that his rating is considered "powerful," but in reality, he can't even figure out how to move and shoot? I think TV should not be part of the matchmaker to balance things, but rather it should be an inherent restriction in the match itself. The host can choose the TV limit, or maybe there are lobbies dedicated to certain limits.

#173 Heffay

    Rum Runner

  • PipPipPipPipPipPipPipPipPipPip
  • The Referee
  • The Referee
  • 6,458 posts
  • LocationPHX

Posted 23 January 2015 - 11:44 AM

View PostDino Might, on 23 January 2015 - 11:34 AM, said:

What if a low ELO player makes such a high TV gear that his rating is considered "powerful," but in reality, he can't even figure out how to move and shoot?


That's the beauty of Elo (It's a guy's name, not an acronym btw): After a few games, it'll drop even further to take into account he can't walk and shoot at the same time. It's self correcting. Once he figures out the basics and can start doing things, it'll rise as he gets better, gets better gear, etc.

Having TV determine the maximum amount of power you could bring to a match isn't a bad idea, but once you're in the lobby the game has to figure out how to split up the players. Can't have all the veterans on one side and noobs on the other. Well, I mean you can but that's a horrible matchmaking system. If you set up a 4 person lobby with your 4 man team and invite pugs to play against... yikes.

#174 TrentTheWanderer

    Member

  • PipPipPipPipPipPip
  • Legendary Founder
  • Legendary Founder
  • 264 posts

Posted 23 January 2015 - 11:54 AM

Something that's interesting about how the stats are being managed and tracked now in the alpha is that kills, deaths, and matches are tracked- and they are hybridized both into k/d (which reflects raw kill death ratio regardless of experience and circumstance) and k/d-a which awards a higher score to players who consistently score higher over larger match counts as opposed to a player who has two matches and a very high k/d.

With the (placeholder) stat system on the leaderboards now a player with two matches, 8 kills, and 0 deaths (8 k/d) is actually registered on the k/d-a scale much lower than a player with 100 matches, 50 kills, and 25 deaths (2 k/d). While this isn't the "finished" system that will be used for any kind of bracketing, it does show that the Heavy Gear Assault devs are thinking hard about how best to use the stats (and what kind of stats to track) to make sure everyone is having fun and challenging games at their own skill level with similarly strengthed opponents.

Edited by TrentTheWanderer, 23 January 2015 - 11:58 AM.


#175 Vassago Rain

    Member

  • PipPipPipPipPipPipPipPipPipPipPip
  • Bridesmaid
  • Bridesmaid
  • 14,396 posts
  • LocationExodus fleet, HMS Kong Circumflex accent

Posted 23 January 2015 - 07:51 PM

View PostHeffay, on 23 January 2015 - 11:44 AM, said:


That's the beauty of Elo (It's a guy's name, not an acronym btw): After a few games, it'll drop even further to take into account he can't walk and shoot at the same time. It's self correcting. Once he figures out the basics and can start doing things, it'll rise as he gets better, gets better gear, etc.

Having TV determine the maximum amount of power you could bring to a match isn't a bad idea, but once you're in the lobby the game has to figure out how to split up the players. Can't have all the veterans on one side and noobs on the other. Well, I mean you can but that's a horrible matchmaking system. If you set up a 4 person lobby with your 4 man team and invite pugs to play against... yikes.


Elo is strictly for 1v1 stuff.

I like that you described all the best parts of MWO's matchmaker in this post. Was that on purpose?

#176 RedDragon

    Member

  • PipPipPipPipPipPipPipPip
  • 1,942 posts
  • LocationKurpfalz, Germany

Posted 24 January 2015 - 04:04 AM

View PostHeffay, on 23 January 2015 - 05:01 AM, said:


You have that completely wrong. The players have vastly different skill levels, but are both rated 1800 based on the gear they play with. You take the guy with the knife and give him decent gear, and he'll rise out of the 1800 bracket, because the beauty of Elo is that it is self correcting. Same thing with your Awesome vs Dire Wolf example: Elo will correct itself after a few matches.

That is what makes it a good match. The guy with the knife playing at at an 1800 Elo isn't a hindrance to his team, because he *got* to 1800 using just that knife. That's skill right there.

That is when you assume that every player strictly sticks to his gear, what is neither right for the average shooter nor for MWO. In a game where changing gear is not only possible but also encouraged (grindign mechs) El would have to adjust every 2 or 3 matches, and not only for a few points but for a whole new bracket.

And you saying "Elo will correct itself after a few matches" is exactly what is wrong with it in a game like MWO. You have a high Elo but want to make a quick match in an inferior mech? Suck it up, because you will be matched with pro gamers for the next few matches and be a hindrance for your team, and finally, when your Elo has adjusted and you want to go back to your better mechs, you will be paired with some new players that can't even turn their torso.

#177 Heffay

    Rum Runner

  • PipPipPipPipPipPipPipPipPipPip
  • The Referee
  • The Referee
  • 6,458 posts
  • LocationPHX

Posted 24 January 2015 - 05:31 AM

View PostRedDragon, on 24 January 2015 - 04:04 AM, said:

That is when you assume that every player strictly sticks to his gear, what is neither right for the average shooter nor for MWO. In a game where changing gear is not only possible but also encouraged (grindign mechs) El would have to adjust every 2 or 3 matches, and not only for a few points but for a whole new bracket.

And you saying "Elo will correct itself after a few matches" is exactly what is wrong with it in a game like MWO. You have a high Elo but want to make a quick match in an inferior mech? Suck it up, because you will be matched with pro gamers for the next few matches and be a hindrance for your team, and finally, when your Elo has adjusted and you want to go back to your better mechs, you will be paired with some new players that can't even turn their torso.


You vastly overinflate how much Elo will change. It's a team game, and your contributions is just 1/12th of your team. Even switching between the most decked out meta mech in your deck and a brand new trial mech in stock configuration, it'll probably only cause it to fluctuate a couple hundred points at most.

No other matchmaking system out there is self correcting. Elo is a low maintenance and incredibly *fair* method of creating matches, which is (again) why almost all modern gaming systems use it in some fashion, whether it's an outright score or a ladder.

As just posted in the HGA thread, they too will be running a ladder. Guess how you calculate what part of a ladder someone belongs?

#178 Strum Wealh

    Member

  • PipPipPipPipPipPipPipPipPipPip
  • The Raider
  • The Raider
  • 5,025 posts
  • LocationPittsburgh, PA

Posted 24 January 2015 - 06:53 AM

View PostHeffay, on 24 January 2015 - 05:31 AM, said:


You vastly overinflate how much Elo will change. It's a team game, and your contributions is just 1/12th of your team. Even switching between the most decked out meta mech in your deck and a brand new trial mech in stock configuration, it'll probably only cause it to fluctuate a couple hundred points at most.

No other matchmaking system out there is self correcting. Elo is a low maintenance and incredibly *fair* method of creating matches, which is (again) why almost all modern gaming systems use it in some fashion, whether it's an outright score or a ladder.

As just posted in the HGA thread, they too will be running a ladder. Guess how you calculate what part of a ladder someone belongs?

Bayesian skill rating systems, like "TrueSkill" (see also, here), are self-correcting, as are unique systems like the "ECF grading system" (see also, here) and the Glicko & Glicko-2 rating systems (see also, here & here).

Quote

Although the Elo model will get you far, there are a few notable things it doesn’t handle well:
  • Newbies - In the Elo system, you’re typically assigned a “provisional” rating for the first 20 games. These games tend to have a higher K-factor associated with them in order to let the algorithm determine your skill faster before it’s slowed down by a non-provisional (and smaller) K-factor. We would like an algorithm that converges quickly onto a player’s true skill (get it?) to not waste their time having unbalanced matches. This means the algorithm should start giving reasonable approximations of skill within 5-10 games.
  • Teams - Elo was explicitly designed for two players. Efforts to adapt it to work for multiple people on multiple teams have primarily been unsophisticated hacks. One such approach is to treat teams as individual players that duel against the other players on the opposing teams and then apply the average of the duels. This is the “duelling heuristic” mentioned in the TrueSkill paper. I implemented it in the accompanying project. It’s ok, but seems a bit too hackish and doesn’t converge well.
  • Draws - Elo treats draws as a half win and half loss. This doesn’t seem fair because draws can tell you a lot. Draws imply you were evenly paired whereas a win indicates you’re better, but unsure how much better. Likewise, a loss indicates you did worse, but you don’t really know how much worse. So it seems that a draw is important to explicitly model.
The TrueSkill algorithm generalizes Elo by keeping track of two variables: your average (mean) skill and the system’s uncertainty about that estimate (your standard deviation). It does this instead of relying on a something like a fixed K-factor. Essentially, this gives the algorithm a dynamic k-factor. This addresses the newbie problem because it removes the need to have “provisional” games. In addition, it addresses the other problems in a nice statistical manner. Tracking these two values are so fundamental to the algorithm that Microsoft researchers informally referred to it as the μσ (mu-sigma) system until the marketing guys gave it the name TrueSkill.

(source)

While the Elo system is widely used, it is not the only self-correcting option... nor is it necessarily always the best option among the competition.

Additonally, the HGA team has already indicated (see here for the query, and here & here for the responses) that they are aware of the potential issues & are exploring/devising methods to address/avoid/mitigate said issues.

#179 Colonel Fubar

    Member

  • PipPipPipPipPipPipPip
  • Survivor
  • 650 posts
  • LocationPlanet Agoge in the Mitera System

Posted 24 January 2015 - 01:57 PM

Until That Day!

#180 Aerik Lornes

    Member

  • PipPipPipPipPip
  • 183 posts
  • LocationAlshain , December 31st, 3078

Posted 24 January 2015 - 03:14 PM

I'm likely grabbing the 39.95 pack for Heavy Gear and the $40 Urbanmech package here come Friday the 30th, thus eliminating my gaming budget for the next month.


Always liked the setting and Mecha for Heavy Gear but never got a chance to play it,* so at least in this one I WON"T be a TT grognard. :-)


*Tribe 8 is as close a I got to playing anything from DP9.





5 user(s) are reading this topic

0 members, 5 guests, 0 anonymous users