Hacker News new | past | comments | ask | show | jobs | submit login

I don’t think I have a warped view of SR. Yes the idea that you win more than 50%, you’ll eventually get there is flawed. Especially in lower ranks where there’s an overcrowding of salty folks who leave, throw, smurf, pick heroes not based on need and just straight up don’t play the objective. I agree that the logic works at higher ranks, but it simply doesn’t at lower ranks. Performance based SR is important for lower ranks and that’s something that straight up gets ignored. It’s especially brutal because there’s no way to reset your SR (it carries over from season to season) except creating a new account and losing all the progress from thousands of hours of playing.

I find the competitive in OW especially bad considering other games like Apex really do it well. There is simply no excuse to ignore all of these problems. I’ve literally had games where 5 of my teammates quit, and I sit through it because I’ll lose SR to only find that I lost more SR by sticking around. It’s inexcusable.




Even in lower ELOs each team should have, on average, the same amount of trolls, leavers, etc. Actually, if you can guarantee that you aren't one of those people, then there are X slots on the enemy team that might be a joker and X-1 slots on your team, so the odds should favor you.


I'm unsure of the intricacies of Overwatch's rating system, but the problem I've noticed in most rating systems is that players don't play enough games for the rankings to stabilize. You end up with a situation where a very large amount of players end up close in ratings, where only a few wins or losses can propel them way higher or way lower.

This might not be a big deal, but the game isn't static. It changes, which means that the skills someone picks up might not be as useful in the next patch anymore. This keeps the rankings somewhat unstable.

Team games inherently require more games to stabilize a rating too. You can get lucky or unlucky with your teammates. On top of that, a game where you can end up with a character you're not strong with probably increases the number of games to stabilize a rating even further.

On average it might work out, but are enough games played to actually reach that? My guess is that for most players the answer is 'no'.


This is anecdotal, but when I host my buddy at my place (he is high diamond on all roles) sometimes he'll hop on my account (low silver on all roles) and steamroll. I'd say 3-5 wins per loss, and he tends to stay off his mains. Yes bad games happen, but he consistently will leave my account at a higher SR than when he arrived (please don't ban me Blizzard). I am pretty confident a diamond level player could boost a bronze account in any role without much trouble.


I recall ster making a new account, sandbagging all of his placement matches to land in mid Bronze, and then reaching Diamond in 6 hours. [1] I think it took him another week to make it back to Grandmaster. Granted, this was in Season 2 and I quit around Season 4, so I have no clue if they made any changes that would affect rank gain much.

[1]: https://www.youtube.com/watch?v=NaORwv8dkt0


there used to be different mechanics like a win streak modifier, which are gone now

but if you have a high win rate you climb more or less immediately (once you've unstuck the MMR confidence value)




Join us for AI Startup School this June 16-17 in San Francisco!

Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: