The more I think about it, the more I realize DbD's balance mathematically doesn't make sense.

BHVR just added a bunch of pallets to every map. Survivors have a lot of tools to speed up their repair speed. Hens set the SWF speedrun record for escape in under 2.5 minutes last month due to One-Two-Three-Four!. And while coordination certainly helps, the solo speedrun record is still under 3 minutes set a year ago.

The pace of the game is simply fast. Killers getting stuck in a long chase is a death sentence which is why you often see people playing the best killers who can handle the tools survivors have in chase.

While yes, these sound scary, these assume the first survivor can last 90 seconds in a chase.

Not necessarily since toolboxes and to a lesser extent gen perks & great skill checks exist. Plus you have to account for player travel times/spawns, pickup/carry time when the first survivor is downed, travel time/player positioning after the first hook, and so on. Lots of guaranteed free time on gens. True that it's not necessarily time on the gens that should be done, but a smart team can play around that really easily and avoid getting 3-genned.

Mathematically the game really is in the survivor side's favor in most cases. The problem is the general survivor playerbase doesn't even attempt to play efficiently a lot of the time and the non-functional MMR system doesn't help one bit.

Theres alot more factors than just the chase time, positioning / where you bring the killer are equally important, its why sometimes the optimal play when chased is literally running into a corner on the opposite side of the map, farthest away from a hook cause that forces the killer to waste 10-15 seconds just walking to a hook, and cause they are also in the farthest corner away from a gen, thats even more time to walk back, which can make even a 30 second chase give 60 seconds of time or more.

3 Gens are strong, almost necessary sometimes, but its mostly up to the Survivors where the last 3 Gens are, as long as you fix even 1 of the gens in that position early, it will end up split across the map, and if the Killer only sits around that position it allows you to just run far away if they initiate a chase, and you can infinitely reset if they don't commit, so either you slowly whittle the gens down, or they commit for the hook and lose their 3 gen.
(Not Counting Outlier 3 Gens like currently that TWD Map with 3 within like 10 meters, they need to fix that, its goofy)

And this is just speaking basekit, with Perks you can force alot of lose-lose situations for the Killer, theres a reason Survivor perks are VERY Limited in Comp, cause if not 90% of Killers would have 0 viability in that space, the only thing keeping this 'balanced' is the fact people dont commonly face these calliber of teams in normal games.

Most of this doesnt really matter against Solo Queue uncoordinated squads, they will 3 gen themselves or get chased near important gens, i think Solo Queue really needs Coordination to close the gap slightly, a Chat-Wheel would be sick, cause BHVR doesnt wanna do VOIP.

Yeah, game is supposedly balanced around 30-45 second chases. If 3 gens are popping around the time of first down it means the chase took extra long.

I also get a deep sense of trepidation when I see a math post, so let's dive in.

Let's highlight something at the end:

And before someone comments it, yes, again, I know. This is not common place, but it's methematically possible

Yes, this is true for any competitive activity. It's possible for one side to dominate the other. Example: in the semi-finals of the 2014 World Cup Germany was up 5 to nothing 30 minutes into the match. It happens even at extremely high level of play.

If they just decide to crank a gen, and the next chased survivor can survive 90 seconds (or less depending on perks and time to travel) the other 2 gens can be popped and it's already end game.

If the survivors do really well in chase, especially two 90 second chases with the killer doing nothing about those on gens, yes, the survivors are going to win.

And if they only last 10 seconds in chase, they are going to lose.

Basically: if the survivors do really well, they win, if the killer does really well, he wins. Like I'm not sure what the issue is here.

fact it's possible even with the killer making no mistakes (and even if the survivors make a few minor ones) makes me realize the balance is an absolute joke.

Wait, woah, when did we get to this? The killer was in two 90s chases. What exactly was happening that they made no mistakes?

a team dedicated to this, with at least 2 good chasers, can totally pull this off if they tried or get so close that they're still ahead by a mile.

Not sure how the survivors force the killer to chase these two people, especially being they are cranking gens which should make them relatively easy to find.

It also explains why gen regression is king, despite how pitiful it is, because no matter how bad it is, other perks cannot save even near that amount of time, with many of them even having penalties that make their bonuses pointless. Camping is a thing because killers simply cannot pressure gens in many cases in any effective manner, meaning their hook is currently the most important defence as it's the only time they can force other survivors into shorter chases.

These are common arguments, but neither makes sense.

Gen defense: what perk would make a difference in this scenario? The killer isn't getting downs, they can't really activate anything useful. Even blocking the gens is going to have a minimal impact if the chases are taking this long.

Camping: even less true. In your scenario the survivors intentionally leave the target on hook. Now instead of two people doing gens, its three.

-

Basically: you're arguing that if things go really well for the survivors then, mathematically, they are going to win. Yes, obviously. Its the same for the other side. None of this is getting into balance issues or game design (such as the guessing game in the lobby or the over weighted impact of early game). But I think the ease in this scenario at which the survivors are doing gens and the survivors are getting long chases would be indicative of killer mistakes, randomness really helping them out, goated survivor play, winning lots of 50/50 guess, or some combination of all of that.

I get what you mean about it being an extreme-yet-not-viable scenario, but the top 26 on the speedrun leaderboard are all under 5 minutes, and most of them aren't by Hens or his friends. In fact, the top 10 are almost all under 3 minutes, and the top 20 for solo specifically are all under 5 minutes. These are people who were willing to record and verify their results for the leaderboards, as well as follow any rule requirements for the category, and says nothing for the millions of games not represented. Its actually quite interesting to see how the times have changed over time. Here's the swf category, but it also has others like solo on there:

https://www.speedrun.com/dead_by_daylight

It doesn't mean your average survivor group is escaping lightning fast due to no fault of the killer, but this is the level of punishing that survivors lament in regards to tunneling: Your opponent's focus on maximizing efficiency reduces your ability to make mistakes (or sometimes even play) until you have no hope of recovering your objective. When criticizing killers in these games saying "if they made no mistakes, the time doesn't add up" is a laser focused perspective on the broader issue: how many mistakes are either side allowed to make to stand a chance. If survivors get all 5 gens done lightning fast, there is no time for the killer to build pressure and/or snowball, just like a survivor who is chased off the hook has no time to defend themselves reasonably.

Its an integral part of the game's design for the survivors to have their strength be at its peak at the start, while it slides into the killer's favor as resources are used up. That means that the burst potential of the survivors at the start has to be addressed in an equal manner to the killer's ability to force resource consumption, most notably in the cases of hook states and teammates in the game.

Short game durations also affect different killers differently, a nurse or blight isn't going to care as much (and will have more opportunity to even punish) vs builds dedicated solely to speedrunning trials vs a hag or trapper who might not even start a single chase by the time gens start popping. I remember a while back there was a push for an "early game collapse" style of phase for trials where the survivors wouldn't be able to maximize their strength at the start of the match, but in actuality an idea like that should only come with the introduction of more stern (and actually well implemented and totally not just lazy attempts at appeasement) anti-tunneling design.

In the end, that leaderboard is basically just a list of failures to a smoke test. Trials should not be able to go that fast, just like people shouldn't be able to be forced out of the trial before being able to have an opportunity to participate in a fair chase.

I get what you mean about it being an extreme-yet-not-viable scenario, but the top 26 on the speedrun leaderboard are all under 5 minutes, and most of them aren't by Hens or his friends.

Actually the extremeness isn't really my point, though its related.

For discussing math/balance (not fun, which is very important but a different topic) on the gen rush the issue remains of if instead of trying to get the world record for escape, how much more or less successful would the team/player be under the same conditions using other builds? That a team can escape really fast I don't doubt, but what's the overall impact on their escape rate?

Here's the swf category, but it also has others like solo on there:

It also has the killer records, and some of them are also incredibly fast.

If survivors get all 5 gens done lightning fast, there is no time for the killer to build pressure and/or snowball, just like a survivor who is chased off the hook has no time to defend themselves reasonably.

But survivors maximizing gens have nothing else. A good chase by the killer, having the right killer/perks, and the survivors can crumble. Like lots of things that maximize efficiency, there's no backup, making a recovery extremely difficult.

I remember a while back there was a push for an "early game collapse" style of phase for trials where the survivors wouldn't be able to maximize their strength at the start of the match

I don't have any problem with the concept, I think there's a lot of merit to it, but I think the speedruns are bad at exhibiting the reasons for it.

This will always be a perspective variable regarding matchups, and kinda misplaced since few of the matches in Hens' video for example were just 1 outs.

Okay, this why I mentioned streaming earlier as well. I don't watch a ton of live streams, but I've caught some of Hens videos were he's trying ideas, and the Youtube video that emerged didn't cover the times the idea crashed and burned. I don't think he's being deceptive, I think he is trying to focus the video on entertaining material, and so the matches that don't accomplish that get tossed. That's always the problem with using edited videos as evidence, we have no idea what is excluded.

I can't side with dismissing a very clear issue based solely on either frequency or viability downplaying.

Not dismissing, though weighing in context, but I'll get to that a little more later.

That said, notice how its not a very contested category, the fastest time is still 2 minutes, and it has no rules around which killer or any other types of curation to the data?

If I'm looking at the site correctly, it is actually split up by each individual killer, probably the reason for less results because its divided across all the killers.

Its also a category that can be easily thrown off by things like ragequits/sandbagging/etc, which would not be able to be recorded or observed by the killer's perspective. As survivor (especially in an SWF) its much more possible to document everything in a way that allows verification of proper play.

More possible, but you still have a few issues. Whether the killer gets a a slow start or not, and what constitutes 'serious' play, which if I remember correctly Hens worried was going to get their record video tossed out.

This is also why I compare it to tunneling, as it puts optimizing one's own chances of success well before the opponent's chance to have a fair match. 

Sure, this is a good comparison

The argument that ultimately won BHVR on tunneling was fun, not balance. As a balance issue, tunneling was all or nothing. If the killer's strategy was to tunnel the first person*, and that person happened to have OTR and DS, killer probably lost (or another survivor had Reassurance, etc.). On that sense, gen rush is in a similar space, you're really hoping the killer doesn't have Corrupt and Dead Man's, and I suspect some of the killers might be intrinsically harder to run this against (the S tiers obviously, but also killers who can damage grouped up survivors would likely be very dangerous).

*we can then get into degrees of tunneling, such as proxy camping, just as we could get into degrees of gen rush, but that's going to explode into lots of variables.

The question is where is the line on fun. Having advantages based on what you selected in the lobby with no knowledge of what the opponent has, is pretty baked into the game. I thought tunneling went too far into the lobby guessing game. Does extreme gen rush reach that stage? I don't know, when I've had it run against me I don't find it particularly annoying because its at least quick and a rush, which I enjoy, but I can't speak for anyone else.

Tunneling is an excuse because of gen rushing, gen rushing is an excuse because of tunneling

Agreed.

Speedruns are like the kill streaks everyone loves focusing on hyperbolic examples of. 

For the record, I don't think that's a good argument either even if I get drawn into it on 'But what about this SWF that…' style.

I think talking about the upper end of the game is fine, even if that's the 1%, but once we start discussing the 0.000001 (the 1 in a million), I don't think its productive. That's why I've always thought it is better to focus on how many killer players have gotten things like a 50 win streak instead of the Blight (kind of like your comment on barely break 20, if a 20 win streak is seen as no big deal I think that's a bigger indication of balance).

But yes, people absolutely do over focus on Momo's Blight run.

People go to these examples because they represent extremes that they find unacceptable, regardless of whether they apply to their specific cases. 

This is where I'm going to kick back to the idea of dismissing an idea vs weighing it in context.

Single videos are very evocative, they really get people to pay attention. Data tends to be pretty boring. That's why I'm saying its not fair to evaluate without all of the data and a comparison. So if this group played with a different, more traditional build, and got an escape rate of 80%, and with gen rush they got 80% or less, well its not an issue on balance (again, on fun, if people really dislike it that's a totally different story).

This is were we are all at a loss in comparison to BHVR's decision making because we don't have the larger numbers they have access to, and even if we focus it on a specific player its going to be a smaller sample size (Otz and Hens have both done some larger data collections, but even then we're still talking relatively small footprints).

I'm going try and shorten this up because I feel like it will quickly become a disagreement over the weighting of certain examples (note: I failed at shortening).

Still, its kinda funny that a killer like Nurse has a worse record than a Trapper :D

Luck based, I presume, if the survivors happen to run into the early traps.

Unfortunately, one of the things that rubbed a lot of people the wrong way is how they're being so generous with some of these definitions they can't even remember them for their own powerpoint.

Ugh, what got said at one point, something like 'and a third one that I can't remember'? Yeah, that was bad, I don't know why BHVR tries to keep putting people into live speaking roles that are not their skill set.

Just like a survivor likely won't have fun if they get tunneled out constantly, a killer isn't going to have fun if they get 2-3 medium chases before the endgame starts.

So this is related to BHVR's bad definitions and what data they were getting from people. I don't mind being the target of the tunnel either, its far more annoying to be one of the three other survivors who are just slamming gens.

-onto the more disagreement points-

The big difference is that speedrunning escapes is actually more instructible than just winning as a killer. Were this not something easily emulatable by the average player

Hmmm. I spent a bit thinking about this (and what you say after), because I don't think you are by any means wrong, but I also feel a little bit off about it.

Thinking aloud: Would a killer going for a world record change their build over one going for a win streak? Yes, absolutely, things like Lethal Pursuer and other early game perks would become more essential. Is the build giving them the same weaknesses as survivors on a gen rush? Maybe, I could see both sides being possible.

I guess I'd have to look more into the world records by killers. Where the players actively seeking to accomplish it, or did they stumble into it by everything going their way?

A killer needs to play well in more than one facet of the game to get even 20 kills in a row, and thats a streak that could still be achieved in even an uphill balancing scenario. You don't have to play well to reach a sub 3min escape, you just need the right build and teammates that are good enough to allow it to be possible

They seem a little unalike in comparison. Could these survivor builds get to 20 wins in a row without needing to be good in more than one facet of the game. If they could, or even if they could just get a really high escape rate, then absolutely a problem. That runs into a structural design issue with DBD though that the amount of newer players who might be used for such a test don't have the necessary perks/items that a more experienced team does (if a person was actually going to test it).

Though I think the hypothesis would be moot if one had access to BHVR's data, do these builds result in a higher escape rate in comparison to other 'sweaty' builds (I presume BHVR compares sweaty levels which they might not)? Is so, problem, if not, still maybe a problem but a different one.

They're just extremes that shouldn't be possible in a balanced game, and people seem to pick and choose when thats allowed to be ok.

I'm so tempted to get into a larger discussion of variability in games, but I said I'd keep this short and I don't really disagree.

They could have data that would be good enough to make feedback from players redundant, but if they don't analyze it right that uninformed feedback can be more insightful than a hundred layered spreadsheet.

Yeah, true, though feedback is always going to tell them more about the 'fun' factor than data (hopefully). I guess if I had their data what I'd like to know is what happens outside the extremes with these builds.

-

Okay, I've failed at keeping it short, so theory time on the issue. There's only one 'type' of survivor so I think its critical to the game's health that survivors have multiple viable styles of 'character' they can build. I think these get split into either mixed (good at a few things) or focused (great at one thing) - with focused builds generally seeing less return for each perk slot invested. So survivors having the option to take a single healing perk and be stronger at healing, or spend their entire build and be great at healing but weak everywhere else, I think is healthy design.

Now ultimately this can make killers feel bad if looking at a singular trial as something outside their knowledge was, or at least felt like, the primary deciding factor of the outcome. In the above scenario if the survivor with the healing focused build is not the one chased and they go around healing everyone the killer is going to feel like the build is brokenly strong. For a fair evaluation we'd need to look at multiple games, best, worst, and middle of the road outcomes.

The going all out on one thing is the survivor risk/reward trade off, perhaps comparable to killer hex perks. If things go your way, great, but if not, devastating. I think that's fine, but recognize it walks a fine line that could lead to player unhappiness.

So where does gen rush fall on this? I think having a completely gen focused build falls into the above idea of doing something really well, but also full of many downsides, and thus an overall healthy option to have in the game. Is it too strong, especially with coordination? My gut leans yes, but I'd feel more comfortable with more data.

And a complete side note: When '1-2-3-4' was announced I said that BHVR was intentionally trying to reinvigorate the World Record chase (which was partly joking, but not fully).