What is skill or a win? The Lead Game Designer knows it the best

Whats lazy about it? if they tried doing it the hard way and it didnt work out, whats wrong with a more simple way?

If you run into a brick wall, do you just keep smashing into it until it breaks?

Yeah. If it's as simple as escape equals a win, death equals a loss they should really make it more like 0 escapes = 100 percent loss, 1 escape each each survivor gets 25 percent of a win, 2 escapes each survivor gets 50 percent of a win, and so on. It would promote more team play (needed at the solo level) and it would be a more accurate level of skill as well. Sometimes you can just be the guy the killer wants out of the game, if he spends 3 gens to kill you, this making it easier for the others to get out, you absolutely should get a share in that victory. You contributed to the team's success. Just as those that pounded gens, tried to unhook, tanked, took agro, etc should have some share as well.

Are there better ways to determine skill. Obviously. But if we're really going to simplify it to escapes/deaths than it should be based on how the team does over the individual.

And that's why we go for averages. Just because the simple solution is smart and easy, does not make you lazy for picking it.

Sounds like you are really low MMR, i have not seen a player idd consider inactive like that in ages. Never the less, I bet they are trying their best anyway, its very unlikely anyone makes their way through the game like that, even past the bloodpoint incentive. If you think its possible, please demonstrate if you can make it from low to high mmr by just hiding.

I dont think BHVR is being lazy they tried to find the solution, but it just was not meant to be. We gonna have to move on.

So they tell you that anyone who crosses that exit line and escapes raises their MMR. But you still think they have low MMR. Gotcha.

I mean it's not like people on here don't post about how MMR is not working as intented every single day, because if a survivor has high MMR, they should not be doing these things (hiding all the time, not being efficient etc). THAT is the argument, not their "averages".

Like if they want to base it on escapes, sure go for it. Just stop calling it "skill" when we (the people who carried them through the match) know it only took them holding W towards the gate to get there.

Absolutely correct. That's why I've been trying to argue that James Edwards (avg 3.5 pts per game) was just as skilled as Michael Jordan. They were both on the same team and had the same amount of wins!

LOL

I think all the arguments are interesting and have some merit, but that Dev's metaphor was pretty bad.

In general though, it's just hard for me to justify such a simple way to determine skill/MMR. Are these players solo? SWF? Are they running the strongest perks? Meme builds? No perks? Are the spending time in matches focusing on archive challenges? Did someone disconnect early in the match? Was someone AFK? Map size? Which killer?

These are just a few things that can really have an impact the likelihood of any survivor escaping.

Also in a game that is rich on several different side objectives and play styles (totem/boon hunter, altruistic, looping, gens, hide and seek, etc). It seems like players can carry their skill in certain play styles, so just like in Hockey sure you have players that fill different roles on the team. Nicklas Lidstrom was arguably the best player in the league for many years, because he had some offensive skill, but mainly because he was the best defensive player in the league for about ten years. Much different than Sidney Crosby or much different from the impact any of the best goalies have had.

I understand their reasoning to an extent, but I just think it's way too watered down for this game

I hate devil's advocate, but I'd be willing to bet if you ran plunderers, left behind, iron will, and call spirit with a key and you played like an absolute ghost you could probably escape in at least half of this games if not closer to 75 percent. You'd be playing scummy and would be bored, but it's certainly doable.

But in general you're probably right.

Its just a matchmaking value. The game is not judging you. #########?

its just checking your averages and matching you fairly with people who has the same.

Nobody is expecting a perfect system, but at least a decent one would be nice. Basing it only a single criteria was a massive miss.

The tweet is portraying a flaw of thought that many other people also fall victim to. They take a ridiculous scenario, an exception, and make it something that should be systematically accounted for, a rule.

If a survivor is so much better than the killer they are facing that they can "run them for 5 gens", that survivor dying in that match is absolutely a crass exception to the rule that that survivor will be surviving. Play that same match a thousand times and 900+ times that survivor will be escaping alive. Note: Playing that same match again would not mean the survivor would again lead the killer on a 5-gen chase, they might not even be chased altogether, or only for 2-3 gens. There's an entire world of things that may happen in those matches... but since we know that the survivor is so much better than the killer that they can potentially lead them on 5-gen chases, the likeliness of the matches ending in that survivor surviving is overwhelming. This is what MMR accurately measures, over large enough sample sizes, and it very much accounts for those chase skills for that reason.

Not only that, if other players in that match simply did gens and escaped, they are absolutely justified in gaining MMR. Sure, it didn't take much skill for them to do so, but in this case MMR is not about the skillfulness with which a victory is achieved, it's about achieving a victory at all, by doing the correct win-oriented things, which is playing effectively. And since it is still a victory it only makes sense for MMR to increase. Because if it increases, those players will face stronger opponents, and those easy, non-skillful ways to play will stop giving them victories. In this example it is clear that they will not keep having teammates that run the killers for all gens. Because the killers are better, because 5-gen chases are an extreme rarity to begin with even in the most lopsided matches, and because the other players might very well also just be people that do gens and aren't very skilled at chaseplay. Again, over large enough sample sizes, this will correct itself, and players that can only do gens and aren't good at much of anything else that actually takes skill will not keep climbing since they will not keep surviving.

Big text warning for the rest, which is just additional discussion for anyone interested in it.

Of course, if there are easy things that don't require skill but still consistently lead to victories even against equally or even higher-skilled opponents, those would pose problems for such a system that would actually prevent it from selecting for skill. But not only would those be problems of game design that have to be remedied by game design, rather than problems with the rating system and let alone problems that could realistically be solved by it, but I am not convinced that there is much of anything in this game that actually fits those criteria. Again, doing gens is certainly easy, but it's also certainly not capable of letting worse survivors consistently survive against better killers, or we would be seeing much higher survival rates and we would not be seeing killer players that win the vast majority of their matches, because doing gens is simple and easy and obvious and so if it alone would be capable of consistently leading to victories regardless of skill disparities, it would be a huge balance issue that would have huge impact on survival rates. But they're under 50% globally, and good killer players win 80+% of their matches. Even in tournaments where some of the best players face off and people are literally paid to do gens as efficiently as peak humanly possible, kill rates usually settle above 50%. Simply "doing gens" is not at all effective enough a strategy itself to undermine the impact of skill on the outcomes of rounds. Hiding even less so, for another thing people sometimes bring up. People that just hide will die even more frequently than people that just do gens.

The other big flawed argument in this regard is of course camping/tunnelling/slugging, and how they could undermine the impact of skill on outcomes by leading to kills without requiring skill. First of all, they still require skill, not only because you actually still have to win chases to be able to down and hook survivors in order to camp, tunnel and slug them (and those strategies are all the more effective and regularly indeed only at all effective if you do so quickly, which is obviously tied to chase skills), but there's actually a whole lot of nuance to these strategies on both sides if you actually have equal players face off. That this is true is obvious to see when you look at the fact that even in tournaments, where killers are encouraged to use every trick in the book and camp/tunnel/slug with a vengeance whenever opportune, they are not simply dominating by doing so. Leatherface, the least skill-requiring camper character in the game, does not in fact get easy 4ks in these scenarios. On the contrary, it is precisely the most skill-requiring characters (Nurse and Blight most of all) that are the ones capable of consistently 4k-ing here. So it is obvious that primitive camping alone is also not effective enough a strategy to undermine the impact of skill on the outcomes of matches.

When people argue "a Bubba getting a 2-hook 2k with facecamping is seen as more skillful by the system than a Nurse getting a 8-hook 1k", there are various flaws in the thinking. First of all, this is regularly only possible due to the matchmaker not actually pairing players of similar ratings. That is not a problem with the rating system, but with the matchmaking system which could in theory even completely ignore those ratings. The rating, as it is based on survival and kill rates, would dictate that the survivors in the match if we actually have a match where players are appropriately paired, would be able to counter the Bubba, since they play to survive and are good at it. Just like Bubbas don't get easy kills in tournaments where people are close in skills. One facet of this fact is that it is rather difficult for Bubba to get downs on equally or higher-skilled survivors to begin with. Sure, go ahead, facecamp that hook you got after multiple gens already popped, show me how you get more than a 1k out of that if the survivors actually play to survive (which, again, an appropriate rating level would dictate they do). The next flaw here is that they assume the Nurse is actually more skilled in this scenario. I guarantee that if a Nurse does not manage to get more than 1 kill against a team a Bubba could get more kills against, the Nurse is not more skilled. Over large enough sample sizes, a more skilled Nurse player would always outperform a Bubba player against like opponents (which they over large enough sample sizes have to be assumed to be facing). Another flaw is that they assume a 2k would allow Bubba to gain rating, usually that would basically just lead to their rating becoming stagnant.

And the final flaw that is worth taking a more comprehensive look at is that they assume camping/tunnelling/slugging do not require skill, as opposed to hooking often, and that hooks should be the win condition. There's many issues with this assumption.

First of all, camping/tunnelling/slugging do require skill. Like I mentioned earlier, for one obvious thing they still require players to win chases. But they also have various aspects of proactive play and counterplay themselves that are mentally and mechanically demanding, require players to make good decisions and plays. Sure, they may not be the types of skills one necessarily likes, but they are skills, and if a player is capable of winning consistently by being skilled at utilizing these strategies, that player should absolutely be gaining MMR. And that leads to another issue with the assumption, namely that in a scenario where we just base MMR on hooks, people that camp/tunnel/slug would be low MMR, get weaker opponents, yet basically kill everyone all the time and make for very lopsided and undesirable matches. The next issue is that people care about killing and surviving more than about getting hook stages or preventing hook stages. I doubt people at large would ever be satisfied with a system that tells them they lost even though they all survived, or even though they killed all survivors. The game simply is about killing and surviving, and the vast majority of players always have and always will play to that end and care most about those things.

Yet another issue is that the game is simply not balanced around the idea of winning (killing) with many hooks. In scenarios where you have equally skilled players on both sides, camping/tunnelling/slugging are absolutely necessary to be able to win, you will practically never be getting 12-hook 4ks there, as opposed to less-than-12-hook 4ks. Sure, we could argue that the game should ideally be balanced around being able to kill with many hooks instead, but the assumption does not acknowledge that the game in fact just is not balanced for that. There are arguments for and against "many hooks" yielding a more desirable gameplay than the gameplay revolving around playing for kills by any means, including camping/tunnelling/slugging, and while I do agree that the game would likely stand to benefit from at least being shifted somewhat to revolve more around getting many hooks than merely kills, there could also be arguments that the game should be balanced for... being able to compete even without using perks, being AFK for 30 seconds, or cleansing all totems and looting all chests before touching generators. If the game were balanced such that that is possible it might be better, or it of course might very well not be, but the reality simply is that it is not balanced for such things, and winning with less-than-12 hooks is a possibility innate to its balancing, due to the existence of those aforementioned strategies, and as such has to be accounted for in the MMR. That is the name of the game, and while a player that plays to get and does get many hooks may certainly be more skilled at chasing than a player that plays to get as many kills as possible by any means, the latter player absolutely is more skilled at winning the game and just factually playing the game to win as it is designed to be played, and therefore entirely naturally gaining MMR. If you go out of your way to get as many hooks as possible instead of getting kills by any means possible, you are deliberately hampering your performance in terms of the actual win condition the game is actually balanced around. That doesn't mean you are a bad player, and if you still kill most survivors playing that way you will still climb MMR, but if you don't, you shouldn't climb MMR, because you are factually not good enough to be able to get enough hooks to kill survivors consistently playing this way. Indeed you wouldn't even want to climb MMR with such a playstyle, since you will get tougher and tougher opponents against which you get fewer and fewer hooks.

More issues with making MMR revolve around hooks is that it would necessitate major redesign. Since survivor MMR would also have to be based on hooks lest we get a scenario where both sides can win simultaneously and MMR is rendered pointless altogether and/or where campers kill everyone all the time due to remaining at low MMR, we would need to get rid of hook timers and the ability to kill oneself on hook, since otherwise survivors could simply play to die on their first hook and not unhook each other, denying the killer hooks and thereby guaranteeing a "win". Then we would also need to resolve scenarios where every survivor is slugged or hooked, allowing the round to recover. And then even in general it is often impossible for killers to play for many hooks because they simply cannot find the survivors that haven't been hooked yet, or because those that have interact with them beforehand. There are tons and tons of scenarios that are largely out of the killer's hands that see games end prematurely without 12 hooks, even if the survivors are not actively killing themselves to prevent them from being able to get many hooks.

Finally, MMR being based on kills rather than hooks is not even an issue for players that care more about hooks than kills, because they can simply keep playing for hooks, and if that doesn't allow them to climb the MMR "ladder"... so what? There's no reward up there, the top MMR realm contrary to what seems to be another common misconception is not supposed to be the valhalla where the best, most desirable gameplay happens and everyone should want to strive to be at. We aren't even being shown whether we are there. That realm is simply about having players that care the most about winning and are the best at winning, ideally being paired against one another, to create a very competitive gameplay environment. It's precisely about "tryharding", "sweating", playing as efficiently as possible without any regard for fun or fancy play, only concentrating on fulfilling game objectives and victory conditions. And while winning in such a highly competitive environment absolutely does involve a ton of skill and "doing gens" and "camping" is not a simple win-for-free card by any stretch, and while for a lot of players playing to win in tough competition is actually desirable and fun in itself, including the utilization of things such as camping, it also doesn't prevent anyone from simply playing in other ways that they consider to be more desirable, fun, skillful, and so on. On the contrary, such a system rewards you for playing in those ways by keeping your MMR at a level where you more often face more like and like-minded opponents that are not playing only to fullfil win criteria either, but that also care more about other stuff, are less objective/win-oriented, and as such yield matches where everyone has more fun due to having more time/opportunities to get many hooks, loot chests, try weird builds and plays, and whatnot.

To quickly touch on another point: While they could and probably should base MMR on the group performance (how many survivors survived rather than did you yourself survive) since that would account for teamplay and the effects a player's impact has on the overall survival chances in a team better (and since that way rating adjustments on both sides could reasonably be based on the average MMR of the group, rather than individual ratings which can skew adjustments unreasonably sometimes), again, over large enough sample sizes even the per-survivor rating adjustments will come out to more than accurately enough measure average performance levels, and thereby skill, experience, "sweatiness" - in short, all of the things that contribute to survival and kill rates. Accurately enough because ultimately this is not about actually measuring skill or showing anyone how skilled anyone is, it's just about making matches happen that are more fair in terms of survival and kill chances, that create more actually-competitive gameplay, since that is objectively more healthy. If you dislike camping and tunnelling or genrushing, you should embrace the MMR system because that way campers and tunnellers and genrushers face tough opposition that can actually compete with that, rather than beat up on casuals like they have been for years. And if you yourself fall more on that casual spectrum of not wanting to play to win so much, you will face less campers and tunnellers and genrushers as a result of not being high MMR.

The issue with "SBMM" that we still see lopsided matches happen is not due to flaws of how the rating system is meant to work - selecting for kill and survival rates is a great way to do MMR, any competitive format bases MMR on win conditions, and you may not like it, but the win conditions in DbD are killing and surviving. The issues could be part-systemic, in as much as the rating adjustments might not be harsh enough, allowing players to climb MMR merely by winning against lower-rated opponents, or despite not consistently winning against equal and higher-rated opponents. The MMR cap could also be much too low, leading to a pooling-up of players at that rating that aren't actually close in kill and survival rates. But the major issue is most likely simply the matchmaking system, and it regularly throwing players of wildly differing ratings into matches.

Whenever I play solo survivor or killer, I see more tournament players in every single session than I had in years prior to MMR. There are still plenty of lobbies that seem just about as chaotic as those of the rank-based MM past, but it's blatantly obvious to me that the system is working to pair more closely-matched players much more often. I also find this obvious because all the streamers I watch have much more competitive games on average. Sure the best players still win most of the time, especially on the killer side, but they are actually having tough challenges sometimes rather than dominating constantly, and while in the past it was rare to even see a few survivors escape through exits total over a multiple-hour killer session, now they even have rounds where they outright lose, having 3+ survivors escape in one round. You had perkless 4k streaks of dozens of rounds in a row in the past, now those same players regularly can't even manage to string 10 4ks together even when using full loadouts.

All of this could've been avoided if they had just called it 'escape/kill based MMR' instead of calling it skill when what someone classifies as 'skill' in this game is such a contentious thing in the first place

although I do think not factoring in teamplay or altruism on the survivor side makes the system shallow. Same with killers, I'm not saying a killer who gets 0 kills but a lot of hooks is a 'win' but it should count for something imo.

With the dev's logic 1 hook kills are more intended than engaging all the survivors in multiple chases.

Context of the kill matters to us the players, but the results are all they care for and the methods matter little when falling in line with data goals.

Regardless of whether or not he escapes, there is no way you can tell me he isn't a good player. There are very little people that can do all those kinds of jukes. Yeah, maybe you could use different arguments of how such players stay in low ranks and so on, but there is no way you can tell me JRM isn't a good player. However, because he doesn't escape, Patrick still considers him far bellow the skill of a blendette that never touches a gen and escapes thanks to her team's efforts.

The problem with the statement from Patrick is his illogical explanation.

The fact he was comparing dbd with a hockey game didn't make sense as 1 is asymmetrical why the other are even teams.

Then the point saying that someone who runs the killer all game but eventually doesn't escape means it is 0 skill, i know he said it with different words but it is what he said.

I can hide the entire game without seeing either the killer or my teammates, never save, heal, getting chased or even touch a gen and still escape through the exit gates.

Does this mean i actually used skill?

I would say no as i didn't contribute anything in a team game.

All i did was hide and escape and nothing more.

The fact that there is so much missing in between the start and the end of the match is just sad tbh.

The way to adress this is to take actually skill in account for deciding the MMR score.

Let killers and survivors gain and lose mmr points during a match according to their actions.

So when a survivor does do absolute nothing, and i mean nothing, but still escape, they will leave the game with a negative MMR score.

Same for the killer, if he doesn't get a kill or just 1 kill/ 1 hook because he chased 1 survivor the entire game, they will end the match with a negative MMR score.

And I'm not saying this should be made a win/win situation so everyone wins, but there has to be more put into the MMR score besides kills and escapes.

Also a little side note.

The fact that in a previous Q&A it was said that SBMM based on only kills and escapes was not a perfect system, and then now 1 Q&A later it's being said it is a perfect system without any logic reasoning, will make the people very unbelievable.

Just to be clear, this is not an attack on anyone, this is just me speaking my thoughts on the whole terrible situation from the stream ✌️

The thing that people have issues with is that the devs are heavily encouraging a camping/tunneling playstyle while in the same vein trying to nerf said playstyle. Based on Patrick's dodging the basekit BT question, they are looking at more counters to camping in general, but they also aren't doing anything to encourage killers NOT to camp. The so called "nerf" to COH as well as there justification for the Twins nerf show that they have no understanding of the finer nuances of there ONLY ip, which will only lead to more disasterous events like this Q&A. Also on the subject of Twins, didn't Behavior say they weren't going to balance the game around high MMR? You can't have it both ways Patrick!

So what you are saying that if you are a new player that you start or could end up in a mmr bracket where it's full of do nothing claudettes?

Good luck escaping out of the exit gates to get to good teammates then.

A do nothing teammate is never going to be fun. It's better for those teammates to start facing killers that will punish that so they learn to actually do something.

Imagine if it worked like people wanted. Camping bubba's and do nothing claudettes both go down in mmr. So both are in the same bracket and get matched together constantly. I don't think i need to explain how impossible it would be to escape through the exit gate.

Might be better for high mmr but sadly the devs need to keep all players in mind.