So imagine that Smurf widow maybe not a Smurf but testing out the latest hacks. Gets crazy amount of kills in a game. Oh well gg go next.
Next game - relief! A normal game. Now that Smurf last game makes me want to try widow, how hard can it be! Now your game you just did, win or lose, is being measured against that Smurf/ hacker stats a la pbsr.
Now smurfs in one person game sucks (great if on your team). But it maybe rare(? That’s a different debate).
But that cluster of stats of ALL smurfs / hackers/ around all the games “create a cluster” of statistics that yes maybe all outliers but still and has shown - to move the average.
Since we don’t know how many bronze games are played, how many games have smurfs - we cannot say definitely they don’t affect the average.
Thoughts?
Disclaimer, I do not seek to rank up only to improve the quality of “skill rating” overall.
Its worth noting that not only does PBSR split things up by rank (a bronze will never be directly compared to a GM) but even if there are a couple thousand cheaters in that rank, we’re still probably talking in the ballpark of like 10% of the population at most. Realistically, if I had to guess I’d say they just raise the average somewhere around 2 elims a game. Sure, definitely an impact, but far outweighed by the number of players around them.
And that’s still assuming the cheater has a static rank, which they don’t. PBSR is specifically made to see this stuff and bump them out of said SR, minimizing further impact. Sure, they can throw back down, but now you’re taking a 50+ elim game followed by a 0, still averaging to 25 over time.
TLDR, does it effect things? Yes, but not much to care about.
Genuine smurfs struggle to stay low. This would imply that whatever the system is doing, it recognises anomalies. If a player has stats that are the average of the rank 1000sr above (let’s say a 1200 bronze player has the stats of a 2200 good player) then why would they compare average bronze players to such stats?
At the end of the day though, you said it yourself.
So I’m really not sure what you’re trying to get out of this thread .
If at all. 2 more elims on average would be like a 10% increase of the average. I think that’s already more than it actually is.
Let’s assume that the average elims for DPS are 20 kills per 10 minutes. Now let’s assume that there is a Smurf who pops off and stomps the enemy team and finishes the game with 40 kills after a 10 minute games. Those 40 kills is obviously way above average but due to the stomp, the enemy DPS both got maybe 10 kills each. In the end, it evens out.
Your doing a dis service to yourself with these quotes and it not answer your own questions.
To answer your where, imagine a 2d point graph I imagine hackers and smurfs create a point cluster above and away from average.
I’ve been accused multiple multiple times of only coming up with suggestions simply to increase my rank… so I can see a different pixels being used for my rank icon. I assure you, I make these suggestions and raise these points so that my games in my lowly ranks result in more “ggs” than “ez”.
Lastly I’m glad you all think - andreassuring this great system. But that 40 elims example can just as be skewed by analyzing killing blows vs elims.
The other dos on the team maybe only got 3 each… maybe the tanks got the killing blows. Making numbers to a hypothetical situation might help you sleep at night but real analytics have to done.
I don’t know what that means either. You really have to be more precise with your words. Are you implying that the full quote answers my question? Because they do not.
They sure are but their impact is negligible to the average, as i have explained above. Besides, you imagine this to be this way is not the same as “it has been shown…”.
I am not sure who asked or how this is relevant to my question, but okay I guess? I literally don’t know what “improving the quality of skill rating” means though.
Well, my thinking is reinforced by concrete and thought-through examples. What about yours?
All i see is you claiming something and not providing any convincing argument as to why that is the case. You think this to be the case and dismiss everything that contradicts it while not contributing anything of value.
Well, if you actually thought about this at all you would have come to the conclusion yourself that killing blows are a subset of elims. It doesn’t make a difference whatsoever.
As i said, all killing blows are elims (but not all elims are killing blows) so I am not sure what you are getting at. They are also a separate stat and if you substitute “elims” with “killing blows” in my above example, it will literally have the same outcome.
True, but if you can’t even give me a hypothetical example convincing me of your claim then it’s probably not true.
Not really, others have agreed it exist. Now we’re trying to figure how by how much.
This came from another thread.
I can’t reall argue the number of an example of a hypothetical.
You said a Smurf has 40, everyone else has 10 so it averages out. All I’m saying if a Smurf has 40, I can too come up with convenient numbers and say all other dps has 5 because in that situation the tanks took some elims away…. I mean who cares. If you can come up with numbers and so can I, that means we have to see the real numbers which we won’t so we have to go back and just acknowledge the possibility.
All due respect it is a good point smurf stats may average down, but someone making up numbers for one scenario while someone else can make up another set of numbers means we need more concrete evidence.
Until we get a report I’d just like to add to the “Smurf” / hacker issue in that I was not awarded 1-5 sr a game cause of some smurfs in another game some long time ago. Over time causes sr erosion / across multiple accounts.
Dont you see that the exact numbers don’t matter? The harder a Smurf stomps a team, the lower the enemy stats will be. If the Smurf only plays a little above average, it will give room for the other DPS to also play close to average.
Completely true. Unfortunately we don’t have anything to go off of withou knowing rough totals of the number of smurfs/cheaters and how that compares to the actual active playerbase. If we had that, we could calculate the rest, but unfortunately for now all we’re left with is baseless conjecture and logic problems.
I do believe they try to keep PBSR relative to the current playerbase to try to account for the constantly changing environment. Of course, I have no evidence, it just makes sense considering we already know they look at activity for SR calculations (if you’re new or haven’t played in years, your SR will swing much more wildly).
This is true, but then you fall into a range where you’ll be slightly better than the boosted average and go back to gaining 1-5 SR a match. Over time, you’ll level out to a point that’s likely " close enough" to your “true rating.” (Like say +/- 50)
It’s also worth noting that SR erosion is a thing anyways, even without smurfs. Plat players from 4 or 5 years ago would place in the likes of Silver now, just because Plat players, and the community as a whole have all gotten much better over the years. If you can’t keep up with the constant trend of skill improvement, then you’ll find yourself slowly falling anyways, even if you play the exact same way you played back then.
Okay and what makes you think this 4th DPS is going to have above average stats?
Well i don’t think they will, but for the sake of argument let’s assume the 4th dps is being enabled by his Smurf DPS and gets 30 kills instead of the average 20.
50+30+5+5 = 90/4 = 22.5 avg kills per player
This is just above a 1% increase and this is assuming that 25% of all DPS players are smurfs which pop off every game.
