Kill Rates are a poor balancing metric
I see many people citing that the kill rates are already high,and that the game currently favors killers.I believe using them as a means of balancing the game is wrong for a couple of important reasons:
1.One survivor can singlehandedly throw a game,resulting in a 3-4k if they DC,attempt to escape the hook or afk.It's almost impossible for survivors to win the game as a 1 v 3 from the get go(or really early on).Even if they don't do it intentionally a weak link can easily bring a whole team down.
2.The statistics don't differentiate between solo queue and teams.It's pretty obvious that a group of randoms not on voice will have a harder time and bring down the stats,but at a high level,where everyone is on the same page,and coordinated plays can be made,there definitely is a difference.
3.Since it only shows an average of kills,it is worth nothing that it is elevated by the fact that not killing anyone as killer is a pretty rare occurrence.If you tunnel someone from the start thats a guaranteed kill.You also have good chances of securing a kill in endgame,especialy if you have noed or an instadown.This is further enhanced by altruism,where people go out of their way to save someone even if it may not be possible.
4.How well does the mmr even work?It is common knowledge that queue times are favored over fair matches.There are many videos online explaing how lenient the system is in pairing people,and i can attest to it from personal experience.It is also a fact that if no one knows what they are doing the killer is at an advantage.
If this is still not enough to convince you that statistics alone are not sufficient,consider that the kill rates don't reflect the killers strength.Why aren't blight and nurse at the top of the chart?Why are killers like sadako doing so well?Remember these are the top 5% of players who supposedly know what they are doing,there is obviously a disconnect somewhere.
Comments
-
1, is fairly inconsequential in the grand scheme of things, because we're talking averages across the entire playerbase. Survivors are more or less just as likely to 'throw' a game as the killer. Survivors in one game can give up on their first hook, while killers in the next game could intentionally let everyone escape to lower their MMR. It's swings and roundabouts. It does limit the validity of kill rates I agree, but not massively so, because of the numbers we're dealing with. It prevents you using these statistics to predict the outcomes of any individual game, but it doesn't mean trends are meaningless.
2 is false, they're able to group data according to lobby composition, they can work out the average kill rates according to whether a lobby is 4 solos, 4 SWF 2 solos and a 2-man SWF, etc.
3, not killing anyone (and conversely, killing everyone, hence the Hatch) should be a rare occurrence. This is part of why kill rates shouldn't be expected to be 50%, because a 1K is still considered a loss for the killer, despite being a greater-than-zero kill rate.
Finally, if kill rates a a poor metric to balance the game by (they're not perfect, but I wouldn't say they were "poor") do you have a suggestion for a better metric?
4 -
1 - DCs throw the whole game out of the metric. Yes, a suicide/AFK can throw the game off, but so can killer mercies.
2 - They have their internal stats on SWF (they say SWF helps by about 15%), but also they have stats on MMR at various levels.
3 - Okay?
4 - This doesn't really do with the rest of the post on whether kill rates are a good metric.
0 -
The 39% escape rate included both Solo Q and SWF if I'm not mistaken. Furthermore, thrown games aren't taken into account for statistics, BHVR has ways of removing them.
There's also the fact that the devs have already stated in the past that they want the game to be killer sided, and it is. They're not aiming for 50/50's.
0 -
I'm of the opinion that they're balancing backwards. You end up with a feeling of contrived gameplay and outcomes when you balance with a kill rate in mind while favoring queue times above almost everything else in matchmaking. Trials (especially solo queue survivor) often feel like a 10 minute cutscene with pre-determined outcome.
You need quality matches to happen more frequently, and the results of quality matches should be the measuring stick for balance. They shouldn't be starting with an end goal in mind for balance and then trying to tune matchmaking around that. Games just feel pointless for experienced players when that happens. You tune matchmaking to actually produce quality matches, look at those results, and then adjust balance from there (preferably around the upper end of skill brackets).
It just feels like matchmaking is currently being used to lower queue times and contrive desired win rates for the devs. It feels cheap.
0 -
Weekly reminder that the matchmaking system we have was the one favoured by players in a blind trial.
They tested systems that resulted in closer matches with longer queue times, and everyone hated it. You waited longer, for sweatier games. Contrary to what everyone seems to believe, you do not want matches that are too close, because nobody enjoys them.
0 -
I mean kill rates aside the amount of time it takes to get good at killer is astronomically less than survivor. I don’t think anyone who has played a serious amount of both sides could even possibly make the argument that killer is harder than survivor. The game is meant to be killer sided because the fact is the amount of people who want to spend hours of games playing alone is less than those that want to play survivor. I am like 2k hours on survivor and 150 hours on killer and still 4K lobbies with streamers and 100 prestige etc last season during my first Iri 1 climb. Most killer complaints are because they play some suboptimal killer and are surprised they aren’t doing well. If I brought no mither and spies from the shadows every game I would also do awful. Not every perk or killer is made to get a 4K every game. Learn nurse or blight if you want to 4K every game.
0 -
What do you propose as alternative then?
0 -
*Some players had longer queue times
My queues were instant when bi-directional matchmaking was tested. It was actually the most fun I ever had in public matches. What they didn't like were longer queue times for players in the mid-MMR range, so they chose to alienate the outliers instead of trying to find a halfway point.
0 -
Hook-Rates per Killer.
0 -
No, just no. I like to play Sadako condemmed.
0 -
She gets 3 hooks per Condemn-Kill, never seen?
0
