Players Are Still Failing Ara-Kara and Halls in Season 3 Week 3

The third week of Mythic+ this season has gone well. Once again, Nightstalker314 is taking a look at the weekly data provided by Raider.IO. The total drop-off this week (6.75%) was rather small.

Once again, Chinese server data is the main factor. On those servers, it was a dip of just 2.58% while in the remaining four regions, it decreased by 10.09%. In addition, any major trend on Chinese servers continues to shape the global data. The shift toward higher key levels is as pronounced as the dislike of running the Ara-kara dungeon. So, let’s look at the details.

Weekly Total Run Performance Compared To Other Seasons

Since this is a comparison to old data, we exclude current Chinese data from these metrics. Given the 10.09% drop-off, TWW S3 is currently on track with DF S3 and TWW S1. In week 4, TWW S3 might fall behind because during DF S3 and TWW S2, we had a dungeon event. The same goes for TWW S1 with major tuning efforts in week 4 and a dungeon event in week 5. For TWW S3, such an event happens in week 6. This comes right before the Turbo Boost for this season in Patch 11.2.5.

When looking at the weekly global numbers just for this season, we can clearly see the dominance of Chinese data. Activity on those servers is declining much more slowly. And within the next few weeks, their global share might surpass 50%. Especially during dungeon event weeks and the Turbo Boost start in TWW S2, their numbers shot up a lot higher. The drop-off in total numbers in general is rather small. While the other regions lost around 197,000 runs compared to week 2, that value was just above 40,000 for China. 

Total Runs Across All Dungeons

The order of dungeons, as well as their distribution, is much more dynamic this season. While the top two dungeons haven’t changed, Dawnbreaker dropped from #3 to #5, and Streets of Wonder climbed from #7 to #4. Priory switched positions with Gambit, and Ara-Kara is still dead last. The global share for this dungeon has now dropped from 11.26% to just 10.01%. On Chinese servers, it is sitting at 9.02%. This should be a clear signal for the developers to consider more tuning.

The in-time percentage for all dungeons (81.44%) went up slightly compared to last week (80.71%). For most dungeons, the changes are rather small, but Halls of Atonement stands out. Eco-Dome is still leading with 88.80% while Streets of Wonder is now comfortably at 85.28%. And this during a Tyrannical first week, when in week 1, also Tyrannical first, it struggled at just 55%. Gambit is following at 84.20% with Priory (83.15%) close behind. 

The Dawnbreaker is leading the lower half at 81.86%. This dungeon has made a decent return so far, but it’s plagued by Skyriding bugs between the ships. Floodgate (79.89%) did rather well last season and seems a bit overtuned in comparison. Halls of Atonement (76.98%) improved quite a lot, and it might just be the impact of having Tyrannical active first instead of Fortified. If the value for week 4 is once again dropping off, this might be a hint for required trash tuning. 

Ara-Kara Needs More Tuning

The overall in-time percentage (68.32%) is still stuck below 70%, regardless of the affix combination. Nerfing the first boss helped a bit during the last two weeks. But the major issues are within several trash pulls, especially the first one. When looking at the heatmap provided by keystone.guru, we can clearly see that the first big pull is leading to the most wipes within the entire dungeon. The issue comes from groups feeling forced into the Season 1 meta-strategy that can fail far too easily.

The number of casters and the amount of area denial in this pull are simply too much for most groups to handle. A similar, but less severe scenario happens during the second pull after crossing the bridge behind the first boss. 

Split Across All Key Levels

The shift towards higher keys continues with 63.39% of all keys already being completed at key level 10 or higher. 33.24% were done on just level 10 alone. The bump around key level 7 is still noticeable, but all of this is once again shaped by Chinese data. 71.20% of their keys were done at 10 or higher, while basically half of those (35.64%) were just key level 10. All levels below 9 see less relative activity over there, going as low as 60% at level 2 compared to the other four regions.

On the other hand, 59.31% of all level 11 keys were completed on Chinese servers alone. For total, global numbers, the activity at key level 13 and above is between 2 to 13 times higher than last week. TWW S2 and S1 topped out at key level 17, respectively 14 in week 3, while we already saw level 19 timed in S3.

Success rates have slightly improved on key levels 7 and 8, which is rather odd for a Tyrannical first week. The step up from 6 to 7 resulted in a drop of 10.56 points last week. For week 3, this value reduced to 8.46 points. And the drop-off from level 9 to 10 improved from 5.26 points to 2.24 points this week. These numbers are also major improvements compared to the same week in S2 and S1. But these are just the metrics across all dungeons. Within the next few weeks, we’ll be looking at key level breakdowns for all dungeons while examining the existing tuning issues. 

Conclusion and Week 4 Forecast

With just three weeks into the season, the overall data is looking good. But the tuning problems for Halls of Atonement and Ara-Kara should be addressed sooner rather than later. The terrible in-time ratio in Streets of Wonder saw immediate reaction from the developers. Meanwhile, the most annoying trash mob in Halls, right before the final boss, stays untouched. Unlike the two big pulls to the left in the first section, which caused many wipes, this mob just wastes your time. And for the suicide pull at the start of Ara-Kara, the devs might need to lean into supporting this strategy with major trash nerfs. Increasing the timer would be another option.

As of this writing, 120 hours after US reset, hourly numbers for the US, EU, and CN are 14.2% to 16.6% lower than the previous week. This can be due to the rapid shift towards higher keys, resulting in fewer required runs for the crest cap. Or the current tuning may be discouraging some players. But we will take a look at those numbers next weekend.