LOW Field Of View BREAKS 5v5 Matchmaker, "ELO Hell" is REAL!

Bug Report
1

How FOV breaks 5 VS 5 Competitive Matchmaker. STAY WITH ME HERE:

  1. Field Of View affects how fast you move the mouse and acquire targets

  1. Overwatch 2 INCORPORATES how fast you move the mouse and acquire targets in calculating MECHANICAL SKILL and WIN PROBABILITY in matchmaking

  2. When I lower my FOV to 97, my ACCURACY goes up but obviously response and acquisition speed go down SLIGHTLY.

  3. HERE’S THE PROBLEM: I am a PLAT (sometimes low Diamond) DPS player when I grind, otherwise casually Gold. I MAIN Soldier 76 and Bastion. I got downranked to SILVER when I set my FOV to 97 from the default to 103. The capability of my teammates plummeted, meaning the game “read” me as way less capable.

  4. CAUSE: The matchmaker PROFILES ME (not racial profiling but ability profiling) AS WAY LESS IMPACTFUL THAN I AM with VERY SMALL reductions in FOV (from 103 to 97). It then gives me correspondingly NOVICE teammates that LOSE. I can hear the skeptics “SO SmartyPants if you were impactful you’d WIN and your MMR would rise.” FALSE-- THIS difference is SO DRASTIC and ABSURD there’s NO WAY my game impact goes down as as much as Overwatch 2 believes from 103 to 97 FOV.

  5. HYPOTHESIS: The PREDICTIVE MODEL for WIN PROBABILITY is NORMALIZED to 103 FOV where more than 99% of the player base is. Overwatch’s programmers never bothered to normalize the predictive model to 102, 101, 100, 99, 98, 97 FOV because the population sample size is too small. THIS CAUSES IMPLICIT BIAS because OF COURSE these players READ as having slower response time and movement due to low FOV with increased accuracy. This is a MAJOR FAIRNESS oversight for the small population of players reducing FOV.

  6. I GUARANTEE my SLIGHTLY slower response time and movement speed from 103 to 97 FOV (for increased accuracy) does NOT reduce my VICTORY IMPACT (and slightly raises it), but I am PENALIZED by an inaccurate, GENERAL predictive model. This model DOWNRANKS ME ARBITRARILY because OTHER PLAYERS AT 103 FOV who generate the same telemetry are its “standard.”

  7. HYPOTHESIS 2: Overwatch 2 MIGHT enact HEURISTIC aimbot detection. For aimbots that have NOT been discovered and banned yet. It MIGHT do this by taking player telemetry, including target acquisition time and accuracy, into consideration. HIGHLY VARIABLE playing at LOWER FOV (which makes aim-botting easier) could TRIGGER the matchmaker to discretely “poison pill” behavior that LOOKS like cheating. Blizzard would PREFER To RETAIN cheaters as CUSTOMERS while NOT allowing them win games unfairly.

  8. A GREAT way to retain cheaters BUT discourage cheating would be to use certain cheaters as a “gift” to create fair games with VERY novice teammates with lower win probability, preventing them from “ranking up” (presumably their reason for cheating).

  9. It’s POSSIBLE that I accidentally triggered an undisclosed heuristic “anti-aimbot” detector in the matchmaker that is OF COURSE undisclosed. It has been programmed to be TOO SENSITIVE and I am just UNLUCKY that my combination of settings, attributes and account history triggered it.

  10. I don’t know how to add screenshots here, but this matchmaking change caused by 103 to 97 FOV is SO DRASTIC that I go from 80% WIN RATE (starting in Gold 3) to 90% LOSE RATE (down to god-knows-where, I stopped at Silver 5) by SIMPLY changing FOV and it shows in match history with games I can link to.

  11. I am NOT in ANY WAY a “good” player but I am AT LEAST average and probably even a MINOR MONSTER at 97FOV, WAY above SILVER in performance. [Probably low Diamond if I didn’t have the social skill of a toddler]. So yes, THIS IS A BUG. It MIGHT BE A NEW BUG and UNINTEDNED CONSEQUENCE, so before you tell me “I am Top 500 Hitscan DPS with lifetime 95 FOV” I don’t know how long this bug has existed. It may also affect only average to slightly-above-average ability play like mine. Blizzard is really the only people who need to see this because they know their code and all the things it MIGHT be doing.

ONE MORE IMPORTANT DETAIL (WTF I can’t reply to my own threads here? You think I’m gonna BUMP a BUG REPORT as an attention grift?):

  1. MOUSE POSITION POLLING RATE and its relationship to REFRESH RATE at various RESOLUTIONS, RENDER SCALES and FIELDS OF VIEW-- particularly on multi-display desktop setups, MIGHT be factors in this bug under either HYPOTHESIS of mine-- inaccurate “impact” behavior profiling OR heuristic anti-cheat match protection.
  2. MY monitor, an Alienware AW2725QF, has dual-refresh settings. 4k 165hz / 1080p 330hz as display spec, 4k 180hz / 1080p 360hz when display is overclocked. I also run two other secondary displays with my GeForce RTX 4060 TI. My CPU (AMD Ryzen 7 3700x) and GPU are actually underpowered for my display, so I have changed render scale A LOT (e.g. 75% to maintain 165FPS or 180FPS in 4K) as well as toyed with various refresh rates and fields of view (since lower FOV also raises hardware performance). Troubleshooting this bug had me briefly playing at 60FPS (which lowers my mechanics) just to see whether the pox/curse the matchmaker had placed upon me lifted-- until I finally realized it was Overwatch 2’s OVERREACTION to REDUCED FOV.
  3. STRONG CIRCUMSTANTIAL EVIDENCE suggests my player behavior is a SQUARE PEG in Overwatch 2’s PREDICTIVE ROUND HOLE. I fell into an extremely unfair matchmaking user experience being penalized for niche attributes that made my predicted game impact WILDLY inaccurate (OVER-ESTIMATING it if I got downranked with horrific teammates or, as I hypothesized, punished by an ANTI-CHEAT heuristic that believed I was botting cause of course my crit shots are DING DING DING DING DING at 97 FOV which could seem like an aimbot to a heuristic that has machine-learned to identify cheats from thousands of players’ data.).

EDIT: One more POSSIBILITY

  1. I don’t know if this forum ties in with a ticket system that updates this whenever I edit it, but another thought crossed my mind-- it COULD be that when I reduced FOV to 97, the matchmaker OVER-ESTIMATED my game impact which ended up downranking me. There’s a “maladaptive performance zone” where my performance can IMPROVE enough to WORSEN my teammates in the “game balance” calculation but NOT enough to win games. EVERY predictive model WILL discriminate against SOME people unfairly and there’s no way to make it fair to EVERY outlier.

It is POSSIBLE that at 97 FOV, my performance input aligned with better players that would normally carry-- and I was just a dumb/untalented player compared to that median for other reasons (timing, positioning, the fact that I have no friends to play with and don’t use comms), my being reluctant to expand my hero pool to Widow, Ashe, Hanzo and others who derive more value from improved mechanics than Soldier/Bastion.

In that case, is it the predictive model’s fault for OVER-estimating my carry capacity or my fault for sucking? THIS WOULD STILL BE A BUG because it is counter-intuitive to believe improving one’s mechanics would create LESS WINNABLE games with lower win probability CONSISTENTLY. THAT is ABSOLUTELY NOT an intended outcome of the matchmaker.

I don’t have your product design meeting minutes but I’m quite sure that unless it’s part of heuristic bot detection, “Mechanics Improve and Winnability Decreases” is not a PLANNED end user experience.

I wish I could just see a spreadsheet of my own game telemetry for every game I’ve every played WITH macros to graph statistical trends. I would pay $50 per month if a BattlePass allowed me to do that with played games.