Sonntag, 12. Juni 2022

Progress report

There has been no progress as far as a new personal best is concerned. But that doesn't mean that no progress at all has been made since the last post almost 2 years ago.

In early 2020 I made the bold exclamation that I won't be able to PB before fall 2021 and while that was not incorrect, it was still far too short a time frame. I don't dare make an assumption as to how long I will need to get 3.2b, but it won't be short.

What this post is about is that while I have no finished a clean run, I have improved significantly both in terms of optimization as well as in consistency. As it turns out, Keine's first spell is quite easy now, and has no RNG anymore. Keine's third spell is not as had as initially thought and I have a capture rate of about 60% on it. My cancels have also gotten marginally better. My averages for the grazing have overall improved - on some spells quite significantly. Most notable is probably Returning Bridge, but also Xufu and Honest Man's Death. I also die less to things like the second non spell spins. 

I now make it to Mokou a few times per session. Unfortunately the improved play doesn't mean the runs get very far most of the time. Overall I still have work to do on my consistency for the early fight. Optimally I should make it to Xufu on 80% of the runs that get to Mokou. Right now it's closer to 30-40% so a good deal of practice - or an insane amount of runs that get this far - is required in order to minimize pointless failures.

In light of my consistency improvements (even if it doesn't sound like it), I opted out of altering the route any further as to not add more resets. Several improvements have been found, and some of those are monumental. Xufu has a route with currently unknown potential, but it is known to be able to get at least 16,000 time, which is 3,000 more than my current PB, as well as 1,500-2,000 more than my current averages. That means it's at least an improvement of 15m score at 16,000 time. Whether 16,000 is possible to be reached with some level of consistency or whether the average with the new route is gonna be significantly lower remains to be seen. Another such improvement is Returning Bridge. Despite what I said in 2020, 2,800 graze in a run is not unrealistic and is in fact about average for me now. About 1 in 3 runs that capture it are over 2,900 - which alone is about 7m score over my PB. There is a routing possibility that is incredibly risky but would allow for about 3,200 graze. Whether I want to learn that is not something I've decided yet, though. This largely depends on where my next PB stands, and how much extra score I will need to squeeze out of the stage. 

As the route currently stands, a new PB should at least score 3.25b, which is already including the bomb I am going to use on the final non spell. I more than make up for that score loss on Keine alone and most of my actual score gain is on Mokou in theory. A lower score is of course always possible and would not be terrible, albeit grating. The reason it is not terribly tragic is because the end goal is, and has been, for a long time, a score of 3.3b. 

In my last posts I talked about how the extra stage had potential of over 3.3b but thought that it wasn't enough leeway to actually achieve that score. This is different now. After a lot of testing, calculating and strat finding, the potential of the stage is nearing 3.4b and that's definitely enough room to spare to achieve 3.3b. It's been a while since I did actual calculations but the last time I landed in the ballpark of 3.38b, which didn't include everything, partially because other minor improvements were found in the meantime, but also because calculating score gains for grazes I haven't practiced a lot is difficult - such as the Xufu route mentioned above, or the double lane graze on Mokou's first non spell (I mean this). There are also thing I have thought about, but haven't looked into. For example, I have some ideas, for example concerning revamping parts of the Imperishable Shooting route to make 50m spell bonus more common. There are some things I don't think I can improve further compared to my practice results, such as Honest Man's Death or a variety of Mokou's more simple spells (like the first three, or Woo and Phoenix Tail). Basically, I already have a route envisioned for 3.3b, but do not want to make the jump immediately. Whether this is the correct choice or not is not something I know but is probably something most people would agree with. 

For the current route, the main improvements are on Xufu, Honst Man's Death and Imperishable Shooting, as well as Keine's third spell. More minor improvements will be Keine's first spell, Returning Bridge and just about every spell Mokou has that spawns familiars. Yes, I'm serious. For instance, I can gain almost 500 time on the first spell, as well as higher spell bonus (up to 1.5m). The only spell where such a gain is not possible is Woo. If I add all the possible scores together for these spells, it adds up to about 20m. Free score to be gained everywhere. Incidentally, this is not consistent, but it's also possible to get some extra on Possessed by Phoenix. The spell bonus can go to almost 54m if your shooting and grazing are both flawless. My PB gets 52.7m, so another minor improvement possible, but not forced by any means. 

To sum it up, all this means that my current route is significantly harder than what I did in 2016, but is not as hard as possible. This will be a 2 step journey, with each step raising the second digit of the score by 1. With the help of some other IN players, I was able to raise the potential of this stage to a point where I think 3.3b is humanly realistic. I am grateful to all the Nameless IN players of this world. I don't know when I will PB again, much less achieve a score of 3.3b, but I will try for as long as it takes (probably).

Also there's plans of IN lunatic between 3.2b and 3.3b and who knows what else I might want to play but that's not relevant just yet. Thanks for reading

Keine Kommentare:

Kommentar veröffentlichen