Some data on DC Halls density

This data is a sample from zones (halls123) that i full cleared.

I use the term full clear very strictly. I discount the zones where i left a random room in a corner unexplored because that room could lead to other 3/4 rooms full of mobs. I only count zones where I actually full cleared so I can either see myself entering every room and/or i use /remaining.
I still count zones where occasionally a random mob that didnt get caught in the flameblast stays alive or i skipped an additional life/es/purity rare.

Yes, i triple checked that 370k halls2 clear.

Most of the data is from earlier sigs when i was trying to get a better understanding of DC.

The vast majority of the numbers are my personal races although there might be a few janis' aswell can't remember exactly.

There are more Halls3 numbers than any others because thats the zone i tend to full clear the most.
Similarly, the Halls2 has the fewest sample size because its the zone i tend to skip the most.

Don't pay too much attention on eXP, specially halls3 where i spend some time crafting gear for grav and making sure i full clear everything if i have extra time.

Keep in mind most of this is only concerned for classes that can consistently full clear (witch/templar)...

I usually arrived at halls1 within the lvl17-19 range.
Note that I never have xp penalty in halls.

The difference between 17 and 19 is 245903 xp. That is enough to push a 3M run into a record run or to compensate for an average hall.
You had two options: you could try to full clear everything past zombie zone and hoped you had enough time to full clear all 3halls (sometimes the layouts can cause you to lose a lot of time), or you could lets say clear 80% of zombie/misty cave and then full clear after stormy. Then, you enter halls1 with a little less xp but you have a higher threshold where if you get very good density but bad layouts (with a lot of deadends) you can still break the record. I tend to prefer the second option but i understand people that prefer the first approach.

It's about playing with averages. Knowing how much time you need on average for halls123 and clearing as much as you can in earlier zones before reaching that time.

On my best run this season, i was 18,75% entering halls1. I only full cleared halls3, halls1 and 2 had a few rooms left that i didnt have time to clear but the density was so good that it was still my best run this season.

On a run that i could potentially hit 3.4m if i didnt fucked up, i was barely 18 going into halls1. I also didnt full clear all 3 halls but amazing density = good run.

The point is, there were multiple times where i arrived in halls1 with higher xp and more time left to clear halls123. I full cleared all 3 halls and it ended up on a 2.7-2.9m run.
I'm sure others had similar races.

I guess thats why i personally prefer to rush more early and just hope for good halls.

For reference, based on my (small) sample of data: if you arrived at halls at lvl 20 and had average halls123 your final experience would be: 2857105.43 which is nowhere near the record - consequence of so many people playing witch/templar and so many people being capable of achieving those numbers compared to other classes.

Again, its a very small ammount of data, so keep that in mind.

Regarding possible cheating, i think DC is one the few races where cheating actually doesn't have as big an impact as it does on others. There's obviously still an impact, you can probably full clear or clear very efficiently everything after rukuku because you will backtrack a lot less so your threshold for average time in halls123 is higher.
Which means you can compensate for a less than great halls123, but you still need that density rng though.
For example, i think cheating in vanilla races (even fixed seed ones) has much higher impact.

I guess one of the most frustrating things for me in DC is: i knew it doesnt rly matter what lvl or what time you enter halls1. If you are faster/higher lvl you have a little more leeway but in the end the race will be almost entirely decided in just halls123.
In every other race type, this aspect is not nearly as prevalent.

Despite having other glaring problems, the density is one the reasons i think these types of "no-reset" races are always terrible choices for a signature race.

Saying the exact same thing happens in vanilla races is so naive in my view.
Sure, you getting an amazing warehouses might make you reach docks 2m faster... and i'm sure if you run the numbers for a crypt or docks instance you might get similar results.
It's the fact that you farm a zone multiple times that helps normalize this variance.

A balance should be achieved where you have enough farming time in a zone to average your splits while not having too much time that you fall asleep farming that zone.

I wasn't rly planning on writing this "my personal thoughts segment" because i dont think it really matters but whatever...