Beta End
Assistant Producer Sean Meek and Balance Designer Clint Tasker discuss our beta for Dawn of War II: Retribution. Click on any of the images to view a larger version.
End of the Beta
Although the beta has come to an end, the team here is on the final push to fix a few more bugs before the game goes live to you on Tuesday. Here at Relic we consider the beta to have been a great success. We received a huge amount of great feedback from all of you and we tracked down and solved several problems that would have been issues at launch. Our matchmaking has been improved, our chat system is helping to build our community and doesn’t lag the game like it was at the start of the beta.
Over the course of the beta we’ve gathered a sizable amount of data and we thought that it might be interesting to share a small portion of it that you might find interesting. Before we jump in I want to make note of a few facts.
1. Some of the numbers in these reports are skewed by a variety of factors such as Imperial Guard being new and thus overly popular
2. The sample size for some of these statistics is relatively small (1% of users played as the Lictor Alpha or the Ravener Alpha in the beta)
3. These are not all the stats that we collect and our decisions are not based solely off of the statistics we gather.
Ok. Disclaimers are out of the way. Let’s look at some numbers.
Let’s start with population. The first wave of Beta keys were distributed on Feb 1 (with a few small problems hehe) and these stats end on Sunday Feb 20. The table below displays the number of unique player matches and player matches distributed by match type. A player match is a single player in a game mode so a 3v3 game is actually 6 3v3 player matches. This allows us to see the relative popularity of the different areas of the game.
There are a couple of interesting things we saw here. There were problems with the matchmaking service in the first week which lowered the number of automatch games, but once we deployed our first fix in Beta 3.5 (Feb 4) the matchmaking number began to spike. Things tapered off quickly and we fixed another issue in Beta 3.6 (Feb 7) the day before the beta opened up to all Steam users who owned a Dawn of War product. At this point the player base grew, matchmaking was working properly and the automatch numbers took their rightful place as the most popular game mode.
Some other numbers that were interesting to us were the hero popularity numbers. Unsurprisingly, the Imperial Guard numbers are overwhelmingly high but the breakdown of which heroes are popular is valuable information. Interestingly a lot of the other races maintained their normal levels of popularity and it seems like most of the Imperial Guard players came out of the Space Marine player base, which is normally quite a bit larger. -Sean Meek
Why Did We Change Chaos Tier 1?
The Chaos tier 1 changes in patch 3.9 took a lot of players by surprise and I wanted to take some time to clarify the reasoning behind them. When we first released Chaos Rising there were some glaring balance issues that overshadowed other parts of Chaos. We didn’t have time to really look at how the metagame developed and the style that Chaos was played. It eventually developed that a triple Heretic build became the most viable build in a variety of matchups. This build became even more powerful with the introduction of Noise Marines. Based on stats we knew that this wasn’t necessarily unbalanced, but it left little room for Chaos Space Marines and Havocs. Our changes were not to necessarily address racial balance but to repair unit relationships, and balance the economy of Chaos.
Our own internal balance designer, Braden “bC” Chan had been focusing on Chaos from the beginning of the beta. He was the first to bring the “3tic to NM” (3 Heretic and Noise Marines) build to my attention. I tried it myself and it seemed pretty clear that this was the way things were going to go. Aside from the issues with this build, Chaos also suffered from a bit of a lopsided economy.
Early in a game Chaos didn’t need to significantly invest in power. Heretics don’t require a lot of upgrades to be successful and Noise Marines were at a rather lower power cost. This meant Chaos could tech quickly to the “fast bc” (Blood Crusher), but that they would also quickly suffer from requisition drain due to Heretics not scaling as well as other units.
In come the changes. Heretics were always considered a good melee counter, but were intended to scale by supporting Chaos Space Marines. Worship was initially developed because we felt without it Chaos could not stand up in the Space Marine matchup. Regardless, Chaos didn’t quite come out the way we initially planned. Heretics were too effective for cost and CSMs were not cost effective enough. We increased the cost of heretics to dissuade users from “spamming” them, as some players put it. We want Heretics to be a vital part of Chaos but it has been a design goal that a combination of different units is always more effective that a lot of one particular one. This meant that reinforce cost for Heretics would go up and the bleed would be worse, but this would be offset by our CSM change. Chaos Space Marines were made cheaper primarily because they were not cost effective, un-upgraded. Secondarily because Heretics would be a heavier drain on the economy of Chaos and a cheaper base cost for CSM would help alleviate this.
The results of these changes were a more versatile Chaos tier 1. It took the handcuffs off the early game for Chaos and allowed for some more variation in build order. A player can still play a straight three Heretic build but it is going to be less effective. Building one or two CSM squads is now much more viable and because these units are more durable and scale well with upgrades it reduces the late game requisition drain Chaos was suffering.
Before we launched these changes I discussed them with some top level Chaos players as well as with Braden. Most everyone seemed to agree that this would allow more flexibility in Chaos and not reduce or enhance their competitive edge. These changes were never intended as a buff or nerf to Chaos. We saw a metagame trend that was not intentional and, by our and other’s standards, not very fun.
Noise Marines were also changed in cost. The changes to Noise Marines were not as inter-connected as the changes to Heretics and CSMs but they do have a lot to do with the Chaos economy. Because Chaos was a race that really didn’t require a lot of early power compared to races such as Space Marines, Noise Marines were increased in power cost. But that wasn’t the only reason. Noise Marines act as a very versatile unit and generally there are high power costs attached to these kinds of units. Comparatively a higher power cost is more in-line with Flamer upgrades in other races. After 3.9 a lot of players pointed out that the cost of the Blastmaster upgrade put Noise Marines proportionally far above similar units in cost, such as Plasma Cannon Devastators and Zoanthropes. There is an advantage to upgrading an existing squad rather than building a new one but on this point you guys were right on. Chaos’s power advantage was in the early game and the cost was frontloaded. There was no reason to have this additional power cost, so it was removed.
I hope this provides some insight into how this and some of our other changes are made. Below is some data that backs up our assumption that Chaos is no better no worse off with these changes, at least in the short term. We’ll continue to monitor these and other changes while we ship Retribution, and if changes are necessary in the future, they will be made. -Clint Tasker
Once again, thank you to everyone that participated in our beta for Dawn of War II: Retribution. We're looking forward to playing the game with you all when it is released next week.
