We got a sample size of 75 games played. I can say for myself i played these matches with mostly the same mindset and behavior within the games. This mindset being to climb the behavior score.
Somewhere early during these games i unbound my Enter key, my primary and secondary chatwheels and hero voicelines, push to talk button, to not be tempted to type/say anything.The intent here being, for all toxic chat/voice reports to be automatically discarded.
I was actually considering unbiding ALT key too, so i would not end up using muscle memory to ping my ultimate is 4seconds away from being ready, which is can be used as basis for ‘valid’ toxic chat reports. But i decided against that because i use ALT + QWE ASD for items and i am not relearning 6 hotkeys for behavior climbing.
With the rules and mindset i was abiding during these games explained and set. We can analyze the results.
=============================================================
Over these 75 games my winrate was 40 wins, 35 losses (close to expected 50/50 result) which should be representative of an average performance.
All behavior reports are periodic (after 15 games) meaning - i was not deemed by overwatch as a griefer a single time during this period (October 24th to November 8th) which would trigger a non periodic behavior report.
(takeaway here is that these reports are inconclusive or false as far as i can say, since i did not receive any warning msg in game or any low priority games)
I think the stats in each column of the individual reports are mostly comparable enough and therefore can be used to compare between each other.
| column name | achieved value range | theoretically possible values |
|---|---|---|
| positive matches | 10 to 12 | 0 to 15 |
| reported matches | 8 to 12 | 0 to 15 |
| abandoned matches | 0 | 0 to 15 |
| reports | 23 to 41 | 0 to 600* |
| reporting parties | 12 to 19 | 0 to 135** |
| comms reports | 7 to 23 | 0 to 195*** |
| comms reporting parties | 4 to 12 | 0 to 135** |
| commends | 26 to 40 | 0 to 135** |
my logic for calculating theoretically possible values of upper bounds
*15(games) x (4 teammates) x (5 different report options: text, VOICE, smurfing, griefing, cheating) + 15(games)x 5(enemies) x 4(different report options: text, smurfing, griefing, cheating)=
15x4x5 + 15x5x4 = 600
**15(games) x 9(solo que people of ‘party size 1’) = 15x9=135
***15(games) x 4(teammates) x 2(voice/chat reports) + 15(games) x 5(enemies) x 1(chat report) = 15x4x2 + 15x5x1 = 120+75 =195
=========================================================
With all of this in place i want to focus on the two reports in particular, yellow compared to the green one.
It is clear to see that by any metric conceivable, the yellow row has better stats than the green row, yet one resulted in gain of around 250 behavior score while the other deducted almost 900
The only factor that is radically different in a direction that could be considered negative is the time span over witch the 15 games have been played.
The latest report Sunday, March 8, 2499 12:39:48 PM was generated over a period of roughly a day, while the report from Tuesday, October 31, 2023 6:27:22 PM was generated over a week.
This leads me to believe the length of time over witch you play the 15 games somehow plays a role in the behavior score value change.
Conclusions:
The question that i need to raise is:
Should the length of time, that takes you to play 15 games, be a factor in your behavior score summaries?
I can only believe this is an oversight, in the same vein like how previous system of Commends and Reports used to only give you a certain amounts for a week before resetting arbitrarily, regardless of how many games you played in the particular week.
What change do i suggest?
Removing any ties of time towards the reports within the behavior system.
If i am playing 50 games in a week and i am toxic in 5 of them, I should not be punished as harsh as somebody else who plays 10 games a week and is toxic in 5 of them.
Since comparatively, one is toxic 10% of the time while the other one is toxic 50% of the time. and yet the system to me currently suggest they are both equally bad.
Way to prove his point lmao