Going Rambo on D1 level 6 (And other levels)

[Marvin]

Level 2, 9secs improvement. I should be able to go under 2:30 if I try more, but I'm fine with this run. I don't plan playing more this year. Maybe I will come back in future:).

[vision]

I should be able to go under 2:30 if I try more...

Realistically any time better than this has to be all RNG. That was as close to a perfect run as I can imagine.

[Marvin]

Good RNG is a must. But since the level in general is simple, it is not unimaginable to do. Besides this 2:30 run I did twice 2:31. In comparison, improving Level 3 I posted few days ago is harder.

[Marvin]

Realistically any time better than this has to be all RNG. That was as close to a perfect run as I can imagine.

That is because you don't speedrun it. I see many small imperfections in execution in that vid, which resulted in few lost seconds.

And I actually improved it to 2:26 now:). Still not perfect, that will never happen. At least not by me. Long time ago I would say such a time has to be perfect run, but while improving you simultaneously learn that it is always improvable.

[Marvin]

Once I was 5 secs faster than this 2:26 run while entering reactor room. That actually felt really close to perfect. Unfortunatelly last part was total mess and I finished with same time at the end..

[Darth Wang]

How fast do you think you could do it if you didn't have to kill everything?

[Marvin]

To be honest, I have never tried. But twice as fast should be doable on insane. If it is on hotshot it can be done in 1 minute, long time ago I speedrunned all levels this way.

[Marvin]

So, I kept going, did multiple 2:25 and 2:24 runs until I did this 2:22. The start (from beginning to hostages) was close to perfect. Best pace I ever had. Rest wasn't that perfect, but there wasn't significant time lost anywhere neither. Reaching exit was a bit sloppy, I though I missed last green robot . Luckily I didn't. Usually I have nearly no vulcan ammo at the end, so I switch to laser for last 2 robots. This time I had enough, it was a mistake I didn't check it, using laser was unnecessary risk. It happened multiple times I missed here before.

[vision]

Here is a question, since shielding is a resource just like any other, how often do you deliberately take damage to save time?

[Marvin]

It happens rarelly, because usually it means instant death, but for example start of secret level 3. Most of the time it means loosing nearly all shield, but it is a must, because there is no other fast solution. Same with green platforms in level 26 (before blue key). But in most cases it is possible to go nearly as fast without taking extra damage, because you have to destroy robots. Very different story would be if there are no cloacks and invulns in the game. I think I would have to change strategy drasticaly many times.

[Marvin]

Back with new run!

I remember, when I did L1 (1:26) a few years ago, I thought 1:25 is max for me. I tried that and it felt like even tie my 1:26 is not possible at all, haha. Took me 2 days to understand better all the details and 1:25 felt possible again. And not only that, I believed 1:24 aswell. And after doing 1:25 three times, I managed. And actually it was close to 1:23, lol. But to reach that time, so many things and details have to go well, it is close to limits.

[Marvin]

Level 5, 11sec improvement. Had some better runs, but never got as good ending as here. A bit lucky, since my shield was very critical. Btw the only strategical differences from old run were using vulcan for opening one door and (and that turned out to be very important) switching for concussion missiles for brown hulks at the end. Homings there are just a luckfest, because of that wall infront of hulks. About 3:30 is max for me in this level I believe, so 3:35 is not that bad at the end I guess:).

Edit: oh and I forgot, this time I took L4laser, which is slightly out of a way (about 1 sec of loosing time to take it), but overally I believe it pays of. The way how I killed some robots might not work with L3. But I'm just guessing. If somebody is into nerdy descent stuff, you can tell me the science

[Marvin]

Ok, let's go to the hell. Level 19. Improved by 20sec to 5:11. After that I still kept trying, because I was on better pace many and many times (best runs were 10sec faster), but the amount of luck needed to finish it off without loosing time is demoralizing. In 99,9% you die, or RNG makes you loose time. Roaming is problematic basicly in all parts of the level.

Anyway, although I'm not fully satisfied, this is still a solid run hard to achieve.

[vision]

this is still a solid run hard to achieve.

Great job. It really hurts to see those couple spots where things didn't go right, especially by the yellow key.

[Marvin]

Thanks. What is triggering me more is first floor. I was 5sec faster entering blue door quite often (best tries were 10, but that is very rare). After that it is pure lottery. All parts can be done better, but having perfect RNG everywhere is close to impossible. In this case it was overally very good RNG. I guess I can't ask for more:).

[Marvin]

Ok, I moved to L13. More or less same route, but I changed a bit fighting tactics in some places. I managed improve from 5:24 to 5:11, then ridiculous series of events happened (during whole week). I have finished it 5 times under 5:10 (1x 5:07, 3x 5:06, 1x 5:05), but in all cases one baby spider behind fake wall in the end survived!! That was quite demoralising, lol. And I don't even know how many times I have died infront of exit..

Yesterday I finally managed. First 5:10, then 5:08 and then shoking and satisfying jump to 5:02. Most annoying about this level is ammount of spiders. Random baby spiders spawns can (and VERY often do!) easily ruin whole run. Very annoying is also yellow key area. One of the heavy drillers loves to come from behind while you are fighting drillers and red mech guarding the key, which is basicly a game over. Killing that heavy driller in advance is not an option, you have to fight him later during recharge, so you don't waste any time. So you are just hoping for a good RNG here. In this run overal RNG was REALLY good, if I avoided some small mistakes (especially in reactor room) I could (maybe) break the barrier of 5mins.

[vision]

OMG this has to be my favorite run of yours, so much went absolutely right. Great job. Really exciting to watch.

[Marvin]

Thanks! I think now it is time to move on and practice for the ultimate descent challenge!

[Marvin]

I have a route for insane-max full game. Yesterday 3 deaths, ingame time 2h47min. Making it deathless will be really hard, there is always some unexpected deadly roaming happening or you simply make a mistake during such a long run. But I will keep trying!

[Sum0Beatz]

Make a demo next time you try,would love to watch!

[Marvin]

Make a demo next time you try,would love to watch!

When I make it deathless, I will post whole run on youtube.

Meanwhile I got new run on level 27. Above all expectations. I knew I can do theoretically better than current 2:40, but my Boss fight there was so good, that even when I was 5 sec ahead before the fight many times, I was always slower at the end. Until this shocking 2:33 run happened (actually very close to 2:32). I'm far from sure, but I tend to believe, when the Boss teleported basicly on the same spot I was on, it caused him some extra damage from colision, so he died faster. Also I think he got full damage from my smart missiles. Green blobs normally goes for random bots, but I think in this case they instantly damaged the Boss. Also, in the secret area at the end there are 6 cloacked lifters that behave differently every time. So it is quite lottery, sometimes you loose quite a lot of time. In this run it worked well, so overally this run is almost perfect.

[Sum0Beatz]

Damn that level looks fun to run lol