Hi Folks,
It's been a while and I'm now back in the office catching up on what's been happening. I'm just doing a brief update for you all since you should know by now that there are some really cool features coming down the pipe.
UI 2.0
Sean and I are going through your feedback from the last PTS build of UI 2.0. We will be addressing what we can but keep in mind that we are up against a hard date of February 4th. There are a lot of good ideas in the feedback and we will implement them over the next little while as UI 2.0 matures and stablizes. One of the key things that you noticed during the PTS testing period is the lack of an overall view such as that on the "smurfys" website. This is something that we've determined as a necessity but will be added post UI 2.0 launch. There are numerous other features like this that will be coming out post launch and we'll keep you updated as they come on-line.
Match Maker
Some of you may have noticed a change made to the Match Maker. Previously, the match maker would find the Elo values of a team/player and try to find opponents in a narrow tolerance +/- of the given Elo. After waiting for 3 minutes, the Match Maker would start increasing this tolerance to a maximum of +/- 1000. This worked fine for a long period of time until the curve you've seen on graphs, settled down into the bell curve that Elo creates. But what has happened is that there are players who are at the extreme ends of the bell curve. Very high Elo rated players and very low Elo rated players were not able to match make properly because there were not enough players in their bracket.
For example, a player had an Elo rating of 2790/2800. That person was trying to match make with players with a score of 1790 or higher. The majority of players in MWO do not fit that criteria so the match maker would fail consistently. To remedy this, the max tolerance was set to 2800. This means after 3 minutes, the match maker would start widening the search tolerance around a player's Elo score to +/- 2800. Theoretically, this means a 2800 player could be matched with a 0 rated player. But that is purely theoretical.
We've seen numerous reports of players saying they're being mixed with players far outside of their Elo rating. This is odd because 1) no player knows their Elo rating and 2) metrics is proving this to be a misconception. As of yesterday (Jan 20), the average deviation between team Elo ratings was approximately 175 points. The extreme maximum was around 750. What this means is that match maker is finding people of relative skill compared to your personal Elo with a maximum tolerance of 750.
If you believe that you are being matched with very low Elo players, 250 for example, that means that your current Elo is at the MOST, 1000. This is below the average threshold and you are being properly matched. If you believe that you are a very high Elo player, 2100+ for example, the EXTREME lowest Elo rated team you're going to be going up against is 1350 which is just above the median player of the game.
Now I was inundated with screenshots and PMs concerning the issue of "very high Elo players" being matched with "very low Elo players" so I decided to look directly into the issues themselves in a direct manner. The screen shots showed complete stomps and high discrepancies between damage done on each team. I used 4 screenshots that were submitted (and keep in mind, these were complete wipes/stomps) and compiled the team Elo rating for each team involved. In screenshot 1, the Elo difference between the two teams was approximately 10. In screenshot 2, the Elo difference was 47. In screenshot 3 the difference was 101. In screenshot 4, the Elo difference was 65. What means is, that even if a game ends in a "stomp", there is a high chance that the opposing team was just outplayed via communication and on field maneuvering.
Another thing to keep in mind is that there was a tonnage difference between the teams. There is a general misconception out there that I'd like to clear up. Elo does NOT balance tonnage per team. The only relationship between Elo and a BattleMech is that the player has 4 WEIGHT CLASS Elo scores. A light score, a medium score, a heavy score and an assault score. This allows the match maker to adjust your match base you being great in a light, but not so great in a heavy. Again, this has no relation on team tonnage.
In short, the MM Elo system is kicking off matches between teams with a fault tolerance on average of approximately 175. While this is not 100% perfect (you will never get this number to 0), it is VERY effective.
Clan Tech/BattleMechs
We are still looking at numbers for weapons and technology and are working on testing some of those values out internally with the design team. It is still too early to comment on this stuff at the moment and as soon as it becomes clear on how we'll be addressing specific weapon systems, I will let you know in a Clan specific update.
Dave Bradley (the guy with the epic beard) will be finishing up a write-up on how Clan BattleMechs are assembled and customized very soon. As soon as that write-up is ready, I'll have him post it here in the Command Chair.
Mech Modules
Tier 1 weapon modules are now in development builds. We will be looking at exactly how these modules will be affecting gameplay and the first set of modules affect weapon RANGES. Expect to see these in game fairly soon!
Community Warfare
Engineering is currently working on the new Lobby system required for faction matches. This is no small feat as a Lobby is a completely new state of the game in which players are not just sitting in the UI/Front End and they are not connected to a dedicated game server. It is essentially a game state in which players can communicate together and prepare before dropping into a match. This game state will eventually allow us to have players reconnect to lobbies from which they have accidentally been disconnected from. This is also the game state in which players will be able to vote on game modes/maps prior to the match starting (pending final design). Lobbies will eventually be used for private matches that you've hear of from Bryan's earlier posts. More details to come as development is moving forward but I thought I'd update you with this exciting new feature in the works right now.
Please leave your feedback about this update in the Official Feedback Thread. Thank you!
1
Project Update - Jan 21/2014
Started by Paul Inouye, Jan 21 2014 02:16 PM
No replies to this topic
2 user(s) are reading this topic
0 members, 2 guests, 0 anonymous users