illusorywall
patreon
Elden Ring Invasion Testing
🕑 Added 2022-03-31 04:20:53 +0000 UTCI talked about my invasion testing in the last couple posts but wanted to provide more details on what that looks like and how it's been going. The idea is simple; After figuring out how runes are rewarded in PvP, you can use that to find out the level of other players you defeat. So if you just do a lot of PvP and log the levels of players you encounter, you can start figuring out some of the matchmaking level ranges that way.
Here's what my initial results look like, from my perspective as a level 62 invader:
Every rune amount (+ level) that's listed is something I've encountered once, and every individual "x" afterwards means a 2nd, 3rd, etc, encounter.
Now you might be thinking that this is a horribly inefficient way of testing level ranges, and you'd be right! Controlled tests where one player levels up around another would get much faster results, but I know people are working on that already and so I wanted to do my weird "brute force" method for independent verification.
There's a few pros to this method:
- If two players trying to meet each other don't meet some extra, unknown criteria that affects the level range, this is something that would likely be found in random encounters.
- If two players trying to meet each other run into issues where the invader keeps invading too many random/wrong hosts by accident, this can ensure that wasn't the reason a set of failed matches happened.
But it is horribly inefficient in that you need a really big sample size to sort out what's possible and what isn't. I feel pretty confident in hypothesizing that the downwards range for invaders is 10% beneath their level, as I invaded down to level 56 at the lowest from 62. The math of course would be 62 - 6 = 56.
I feel fairly confident in that with a really non-scientific look at the data, because I see what I like to call "the wall". See all those x's by 56? And then a good number for 57, and again 58, 59, etc? If I invaded all of those levels a bunch of times, but it WAS somehow possible to invade down to 55, then it starts to look more and more like it'd have to be some really bad luck for it to not have happened. We start to see a "wall" because it likely is a hard cut-off for the level range.
As you can see, the upper end of the range is a lot thinner and doesn't have a wall, so I am NOT confident in declaring that yet. It does match some preliminary expectations though, so I think we're on the right track.
There are other issues with this testing to bear in mind; I'm also just not very good at PvP! I'm also doing this all in my first playthrough, which has largely been blind, so I haven't made anything that's even remotely optimized or OP. So getting this data in the first place is something of an uphill battle, but I also enjoy it a lot. The video linked here is a demonstration of how I've been winning some encounters in Raya Lucaria when I'm underpowered and overwhelmed.
I also finally hit a point where I wanted to progress further into the game (I also have the game unpatched on PC and I can jump around/ test stuff with cheat engine, but I'm not playing the game "for real" there), which meant I had to say goodbye to staying at level 62. It was getting a little difficult and I ran out of unique stuff to buy from vendors. I was dumping runes into everything besides leveling up but figured I might as well just... you know, play the game normally again for a bit? 😄
So now I'm at level 75 and I think this is where I'll start a second batch of invasion testing. The "5" in 75 will also provide a good test on what happens with rounding, as different games in the past have handled what the 10% part means differently. My first guess is that 10% of 75 will round up to 8, so a level 75 invader should be able to invade down to a level 67 host at the lowest. But you never know!
This also doesn't even scratch the surface of weapon level matchmaking, for that random invasions are out the window and we'd have to use the same testing methods that others will be using for that...
Comments
illusory wall
Ah, so this has all been Bloody Finger invasions. I'd love to see the results! I have a much bigger sample size from level 75 now that I'll be sharing soon.
Hollow Invasion
Which Finger are you using? Recusant or Festering Bloody. I've started doing the same testing, but at the upper levels (225)
