Zoom
Trash
enjoyable back to our October edition of Behavioral Systems updates. We have some supplementary topics for everyone this month, but lets believe a see at how we are currently tracking towards our goals, and whats coming through the pipeline next.Graves Build
Goals and Status Update
We started the journey a few months ago in the manner of the taking into account goals specifically targeting Intent to Lose and AFK behaviors:
Mitigate the impact of behaviors considering they occur in game.
Early Surrender (In Progress)
LP Mitigation: Providing baby book consolation in ranked games for verified AFKs and intentional feeding (In Progress)
Deterring this behavior from in the works in the first place.
Penalty and queue lockout increases for verified AFKs (Launched 10.19)
Bumped in the works AFK penalties applied by 10% globally
Penalty buildup correlates to a decline in overall AFKs region to region
Detection improvements for AFK behaviors (Launched 10.20)
Detection improvements for Intent to Lose (In increase - estimated 10.21)
Detection correctness updates for every behaviors (In expand - more below!)
In 10.21, you'll see an accrual in intentional feeding and griefing performer detection. We've taken a pass greater than our existing detection that relied on players repeatedly feeding and accumulating many deaths. Our 10.21 updates look at specific details of each death to flag those that appear to be intentional. This allows us to detect players afterward forlorn a few deaths even though still visceral nimble to sever intentional feeders from players who are just having a bad game.
Our seek later than each detection update is to hit at least one of the afterward without regressing upon the others:
Detect more offenders than before
Detecting offenders sooner
Reducing collateral damage from falsely detecting adorable players
Our current focus is yet upon penalizing brusque offenders after just one game, but some of the nuanced behaviors may say yes a few games before we have ample confidence to business a penalty. In the adjacent update, well be looking at touching those behaviors that agree to merged games to penalize in the same way as confidence into behaviors were feeling confident penalizing after just one game. Longer term, we'll be exploring ways of dealing in imitation of intentional feeders as the game is in progress.
Hot Topic: Reporting Confidence and truthfulness Updates and You
In our August update two months back, we referenced our reference book evaluation of hundreds of games and from function so, weve speculative a ton. Jumping right in, our biggest takeaways were:
Player reports are more efficient at identifying the symptoms of more subtle/nuanced game-ruining actions than us bothersome to clarify those symptoms through intellectual conversations.
Not every reports are created equally. There are players who are great at consistently recognizing disruptive actions and others who spam the financial credit button all match.
We're beautiful confident in the momentum we're making behind improvements to our single game detection systems, but we theoretical that with reference to games one at a epoch doesn't occupy the full describe of how players engage in various behaviors across gameseven if each individual game is expertly assessed
With this in mind, we cant wait to introduce y'all to our newest detection system that we have been iterating on for months. Using the dark magics of huge data, we have completed our first iteration of our other name game behavioral attentiveness system.
The system works by measuring and controlling for variations in performer behaviors higher than large periods of time. This gives us the completion to distinguish amid games that just didnt go your artifice because you were stomped in pathway hostile to games where your teammate ran circles roughly speaking your nexus for 20 minutes. By leveraging a longer artist history, we unlock the capability to improved inform decisions upon like to issue penalties and past to encourage off.
Players can expect that finalized system to do the following:
Identify players that consistently engage in disruptive actions for harsher penalties than our current single game detection system. The more frequently one engages in disruptive behavior, the harsher the penalty. This is already in place for AFK and Leaver penalties, but we'll be expanding it out to extra disruptive behaviors.
Distinguish in the middle of the players that are consistently great at recognizing other players' disruptiveness from those who just spam the checking account button every match.
Allow disruptive players the opportunity for behavioral reformation. One of the help of our additional system is we can track behavioral shifts beyond time. We desire to permit players who reform the inadvertent to come back, even if nevertheless visceral adept to effectively prevent bad actors from defilement games.
Weve tested this system on players deliberately feeding, griefing, and abandoning games and are feeling certainly optimistic. Our in the future tests indicate that we can simultaneously enlargement both the number of detections as with ease as our confidence in their quality.
In the upcoming months we'll be tuning this system to enlargement the confidence of our detections, expanding the breadth of behaviors we can track, and going live.