Reign of Gaming

Tweaks to the Honor Initiative

When we launched the Honor Initiative we saw an immediate drop in reports. We’re happy with the results to date, and want to talk about some changes we’ve made that affect the way players earn crests. Before we get into these, we’d like to share some pretty interesting stats that helped influence our decisions:

• 73% of Honor is awarded to players you’ve never played with before
• Currently, about 1% of players have unlocked the Great Teamwork, Leader or Mentor crests, and 16% of players have unlocked the Honorable Opponent crest
• There’s a 71% chance that a game will be positive if you meet a player who’s unlocked the Great Teammate crest and an 80% chance if they’ve unlocked the Great Leader crest

Based on these observations, we believe it’s a bit too difficult to earn a “Great” crest and too easy to earn Honorable Opponent. Although players who earn Honorable Opponent are generally more positive than those without crests, the numbers suggest that they haven’t been as positively impactful as those players with the other crests. With that in mind, we’ve made a few changes to the Honor Initiative:

• We’ve reduced the threshold for unlocking the Great Teamwork, Leader and Mentor crests, making them slightly easier to obtain
  • Some players will be granted new crests, since they’ll now qualify under the new criteria
• We’ve increased the threshold required to unlock the Honorable Opponent crest, making it more difficult to obtain
  • Players who don’t meet the new criteria for this crest will have to regain it

We’ll continue to iterate on the Honor Initiative as necessary, and then start working on the next phase of the feature.

New Updates to the Tribunal

Since we released The Justice Review a few months ago, we’ve been monitoring The Tribunal and would like to share a few changes we’ve made that are subtle, but critical. These changes are aimed at improving accuracy, providing voters more relevant information, and helping to identify the most toxic players even faster. I want to quickly highlight these changes and explain the logic behind them:

• Previously, our Player Support team reviewed every Tribunal case before approving punishments and pardons. After reviewing the past 12 months of data, we’ve seen that cases that are unanimously pardoned or punished tend to be extremely accurate—Player Support rarely overturns these decisions. In light of this data, we’re now going to focus our attention on the cases that are more ambiguous—the slight majorities and split decisions. This should improve The Tribunal’s accuracy by focusing on the cases that need the most attention.

• We’ve tweaked how The Tribunal builds cases: Instead of 100% random sampling, The Tribunal will attempt to find the games that ultimately sent each player to The Tribunal. By selecting recent games that had a big impact in sending the player to The Tribunal, you should see the average number of reports against a player increase per game.

• The Tribunal system is a constant balance between speed and accuracy. In the past, we were concerned that forcing The Tribunal to find more games per case meant that toxic players were left unchecked for longer than necessary. However, we realize that more information for our judges means more accurate verdicts, and recent optimizations have allowed us to change The Tribunal so that every case will have at least two games to review instead of one.

• Players have often said they forget how many cases they’ve done on a given day—we’ve added a progress bar that will outline how close you are to maxing out your daily case limit.

These changes are already live, so go check them out and let us know what you think! We’re confident that these changes will improve the accuracy of The Tribunal system without compromising on speed. 

Visit the Tribunal today and help us make a difference in the community!