Honeycomb Factory Frenzy - January 17, 2023

Edit history:

SDAVerification: 2023-01-17 02:34:09 pm

2023-01-05 07:12:56 am

Game Page: Doesn't exist yet

Honeycomb Factory Frenzy () (pc) [Any %] [Single Segment]

Decision: Accept

Congratulations to 'RockoSonicFan'!

Thread title:

SDAVerification

Edit history:

SDAVerification: 2023-01-16 12:24:50 am

2023-01-05 07:12:56 am

Run Information

Honeycomb Factory Frenzy () (pc) [Any %] [Single Segment]

Verification Files

http://v.speeddemosarchive.com/HoneycombFactoryFrenzy-20230104/

Please refer to the Verification Guidelines before posting.

Please post your opinions about the run and be certain to conclude your post with a verdict (Accept/Reject). If you wish to remain anonymous, you can also send a pm with your reply to 'sdaverification' (please state clearly in that case which run you have verified). This is not a contest where the majority wins - Each verification will be judged on its content.

ktwo

2023-01-07 03:29:02 am

Not to sound disrespectful, but for my understanding, is there anything that can be considered "difficult" or random in this run? I've watched the run once and from that quick look, it looks like a very straight-forward platformer, where you basically only need to hold right and jump at the right time. From watching, the jump timing doesn't seem particularaly tight. Sprinkled in here and there are also turbo shoes that are worth making small detours for. The health management also needs to be taken into account for the route. From that quick glance, both of these seemed mostly thought-through to me.

But to get back to the difficulty level of this run, are there any tricks, jumps, random events or similar that lower the success rate of the attempts? If not, I think an "optimized" run should look pretty much perfect, without any clearly visible mistakes. But there is e.g. a missed jump (at 3:21). There are also minor questions to the health management. For example, why did you opt to take damage at 4:13 instead of at 3:57? The former hit didn't give you any benefit and looks like it could have been avoided. The latter would have given you a bigger speed-boost from the ice (?) patch.

There are more things that can be commented on, but I at least wanted to get the discussion going with these questions.

Not to sound disrespectful, but for my understanding, is there anything that can be considered "difficult" or random in this run? I've watched the run once and from that quick look, it looks like a very straight-forward platformer, where you basically only need to hold right and jump at the right time. From watching, the jump timing doesn't seem particularaly tight. Sprinkled in here and there are also turbo shoes that are worth making small detours for. The health management also needs to be taken into account for the route. From that quick glance, both of these seemed mostly thought-through to me.

But to get back to the difficulty level of this run, are there any tricks, jumps, random events or similar that lower the success rate of the attempts? If not, I think an "optimized" run should look pretty much perfect, without any clearly visible mistakes. But there is e.g. a missed jump (at 3:21). There are also minor questions to the health management. For example, why did you opt to take damage at 4:13 instead of at 3:57? The former hit didn't give you any benefit and looks like it could have been avoided. The latter would have given you a bigger speed-boost from the ice (?) patch.

There are more things that can be commented on, but I at least wanted to get the discussion going with these questions.

RockoSonicFan

2023-01-07 06:17:58 am

4:13 & 3:57: I wanted to make sure I had enough health to take damage and realized I had more than enough health to do that after taking the damage.
3:21: That kind of stuff happens all the time. I jump and sometimes I don't get it.
Overall: I'm pretty happy with the run, but yeah I get what you're saying.

RockoSonicFan

2023-01-07 07:26:01 am

I redid the run not missing the jumps and taking the damage at both points that you talked about and submitted it.

ktwo

2023-01-07 09:58:43 am

The new run is around a second faster than the previous iteration. However, it basically only addresses the specific points I made. It was produced a few hours after those ideas had been pointed out. The run we're all looking for when making attempts can of course happen any time, but it's very unlikely to only take a few hours from starting a new grind. If an improvement comes that quickly, it's in my opinion (and experience) more likely that even more time can come off.

I've watched the new run and also fired up the game myself to try a few things out. The more I look at it, the more room for improvement I see. Let me give a few examples. Why stop at 1:07, 2:41, 2:44, 2:48 and 4:48? The stop at 1:07 didn't serve any purpose that I can see. The other stops might have to be compensated by stops elsewhere, but it looks like a few of the hits in the section after 2:49 could have been avoided for minimal loss of time instead. And the stop at 2:48 might have been possible to avoid by jumping over the shower from top of the pyramid?

I could go on, but I feel at this point it needs to come from you if you're interested in pushing this game as far as you can or if you're essentially happy, but now just want to do the absolutely bare minimum to get a run accepted here.

If you're truly interested in getting a run accepted on SDA, I'd like to quote the following section from the verification guidelines:

Quote from Verification guidelines, https://kb.speeddemosarchive.com/Verification_Guidelines:

Overall, imagine you were in the shoes of the runner. If you think you would have told yourself after the run:
”Wow, that went really well. There were only a few mistakes, bosses were behaving well and the RNG was mostly favorable – this time will be hard for me to beat without a considerable effort”,
then it's probably a good fit for SDA!

When you completed the last run, did you feel like you had turned every stone you could imagine in terms of routing (health management etc)? Was the completed attempt the pinnacle of a hard grind of hundreds or even thousands of attempts, where everything came together in one run?

Don't get me wrong, I think your run is pretty good. But I'm sure you can improve the time even further. Route every section in detail. Know exactly where you can take hits and not. If some of the tricks are tight to execute or if avoiding damage is unreliable in some spots, have back-up plans. And if things don't work out in an attempt, it's easily fixed with the reset button!

And finally, please move the mouse away from the gameplay area! :-)

The new run is around a second faster than the previous iteration. However, it basically only addresses the specific points I made. It was produced a few hours after those ideas had been pointed out. The run we're all looking for when making attempts can of course happen any time, but it's very unlikely to only take a few hours from starting a new grind. If an improvement comes that quickly, it's in my opinion (and experience) more likely that even more time can come off.

I've watched the new run and also fired up the game myself to try a few things out. The more I look at it, the more room for improvement I see. Let me give a few examples. Why stop at 1:07, 2:41, 2:44, 2:48 and 4:48? The stop at 1:07 didn't serve any purpose that I can see. The other stops might have to be compensated by stops elsewhere, but it looks like a few of the hits in the section after 2:49 could have been avoided for minimal loss of time instead. And the stop at 2:48 might have been possible to avoid by jumping over the shower from top of the pyramid?

I could go on, but I feel at this point it needs to come from you if you're interested in pushing this game as far as you can or if you're essentially happy, but now just want to do the absolutely bare minimum to get a run accepted here.

If you're truly interested in getting a run accepted on SDA, I'd like to quote the following section from the verification guidelines:
[quote="Verification guidelines, https://kb.speeddemosarchive.com/Verification_Guidelines"]Overall, imagine you were in the shoes of the runner. If you think you would have told yourself after the run:
”[i]Wow, that went really well. There were only a few mistakes, bosses were behaving well and the RNG was mostly favorable – this time will be hard for me to beat without a considerable effort[/i]”,
then it's probably a good fit for SDA! [/quote]
When you completed the last run, did you feel like you had turned every stone you could imagine in terms of routing (health management etc)? Was the completed attempt the pinnacle of a hard grind of hundreds or even thousands of attempts, where everything came together in one run?

Don't get me wrong, I think your run is pretty good. But I'm sure you can improve the time even further. Route every section in detail. Know exactly where you can take hits and not. If some of the tricks are tight to execute or if avoiding damage is unreliable in some spots, have back-up plans. And if things don't work out in an attempt, it's easily fixed with the reset button!

And finally, please move the mouse away from the gameplay area! :-)

RockoSonicFan

2023-01-07 10:25:44 am

The stop at 4:48 had to happen because I had to save a little bit of health to go past the showers. I also had to stop at the other points for the same reason to save a little bit of health

RockoSonicFan

2023-01-07 03:29:04 pm

Ok I re-did the run again and cleared the spots you mentioned.

ktwo

2023-01-08 04:51:30 am

So the new run is an 11s improvement over the previous run. Partly by implementing the improvement ideas above, but also including a few new ideas. For a short run like this, every time you're not running in a straight line to the right, you need to ask yourself questions like this:
* Even if it's a detour, could I end up saving time by collecting those shoes? (sometimes you do this, but there are some shoes you don't collect and it's not immediately obvious whether they would save time or not)
* How much time do I save by taking damage here compared to the alternatives?
* What are the alternative death abuses in this situation and how much time does each one cost?
* Can I skip any cycles (e.g. moving platforms)?

As the time improvement indicates, the new run is much leaner, but there are still execution mistakes and the routing can be improved. You haven't provided any run comments indicating how much work went into the routing, but given rapid improvements in such a short time span, I think you're rushing it. I suggest instead that you take your time and work out every section of the run in detail. This will give you a solid routing foundation to stand on when trying to get everything together in one good run without major execution mistakes. You can definitely do it!

So the new run is an 11s improvement over the previous run. Partly by implementing the improvement ideas above, but also including a few new ideas. For a short run like this, every time you're not running in a straight line to the right, you need to ask yourself questions like this:
* Even if it's a detour, could I end up saving time by collecting those shoes? (sometimes you do this, but there are some shoes you don't collect and it's not immediately obvious whether they would save time or not)
* How much time do I save by taking damage here compared to the alternatives?
* What are the alternative death abuses in this situation and how much time does each one cost?
* Can I skip any cycles (e.g. moving platforms)?

As the time improvement indicates, the new run is much leaner, but there are still execution mistakes and the routing can be improved. You haven't provided any run comments indicating how much work went into the routing, but given rapid improvements in such a short time span, I think you're rushing it. I suggest instead that you take your time and work out every section of the run in detail. This will give you a solid routing foundation to stand on when trying to get everything together in one good run without major execution mistakes. You can definitely do it!

RockoSonicFan

2023-01-08 11:06:28 am

Ok so I re-did the run with some new notes for it.
Level 1 was pretty straight forward. The speed shoes that were there in the level where you have to jump off of the honeycomb platform wouldn't save you time as I tried to do it before recording.
Level 2 was pretty straightforward as Level 1. The death is definitely worth it after grabbing the checkpoint since you'll need all the health you'll need for the rest of the level. The speed shoes near the checkpoint wouldn't be worth to go for because you'd have to go back up platforming which would be a waste of time.
Level 3 is completely optimized as I was able to not miss a single movement in the run. The deaths are pretty much the same explanation in Level 2. And I'm very happy how well it went.
Level 4 is also pretty straightforward as Level 1.
This run is definitely the best I've done since my first submission. I think this is the perfect run for Speed Demos Archive.

ktwo

2023-01-15 03:35:24 am

A/V: Thanks for getting rid of the cursor in the latest iteration!

The new run, so iteration #4, is around 4s faster than the previous one. I must say that this run starts to look pretty good. I will therefore invest a bit more time in my reply this time and go more in detail in my comments:

Level 1
-------
Looks well played and I can't come up with any routing questions that could be worth exploring (given that you have already investigated the shoes at 1:42).

Level 2
-------
* Was the quick turnaround at 2:42 a mistake? You took a hit anyways right after. A few tenths lost in case the hit was unintentional.
* The death abuse makes sense and looks good
* I'm looking at the section around 3:40 and thinking if there is any way to slide on one more ice patch? You slide on patches 1, 3 and 5. Sliding on 1, 3, 5 and 7 would not be faster since you would then miss out on one of the patches on the floor (or maybe even end up in the milk (?) patches). And something like 1, 2, 4 and 6 would probably not work because you can't jump to the directly next patch?

This level doesn't look easy to fully optimize in the sense taking full advantage of ice patches, timing jump chains to minimize bumping into the environment and similar. It looks like the execution could be even tighter in some spots, but overall I'd say it's well played (unfortunate if the first point above was unintentional though, but it's a small time loss).

Level 3
-------
* Have you really timed out that turning around at 4:51 is faster than running into the spikes in front of you? Judging by the looks of it, the pit on the left and the spikes on the right are the same distance from the checkpoint. And avoiding turning around removes the risk of overshooting the checkpoint (or undershoot and lose the attempt).
* At 5:28, have you tried skipping one platform and make a big jump to the next (the one you land on at 5:31)? If that jump is too long, maybe wait slightly for the platforms to get closer? Your current method looks a bit slow, both because you use the intermediate platform, but also because you lose momentum when you land at 5:31.
* If the above point doesn't work out, can you run down from the platform at 5:28 to the next instead of jumping?
* Same questions as the two points above for the platforms at 5:32-5:34.

No visible execution mistakes that I could spot, but several questions about the routing that I think would be worth clarifying.

Level 4
-------
* Instead of taking the hit at 6:55, could you not have taken a hit by jumping on the spikes and then use invincibility frames to get past the hook? The advantage would be one less jump from backwards-moving conveyor belt.
* What happens if you make a max jump at 6:58? Even if you hit the spikes above, it looks like it would save a bit of time compared to running against the conveyor belt.
* Unintentional hit at 7:12. The health lost in this spot should have been possible to use elsewhere to avoid jumping/running against the conveyor belts. The idea being that you jump on spikes elsewhere instead of staying on conveyor belts that slow you down. Avoiding running against the conveyor belt at 6:58 is probably the biggest potential time save, but it could also work for example at 6:48.

Execution seemed according to the route, but the route contains questions marks as far as I'm concerned.

A/V: Thanks for getting rid of the cursor in the latest iteration!

The new run, so iteration #4, is around 4s faster than the previous one. I must say that this run starts to look pretty good. I will therefore invest a bit more time in my reply this time and go more in detail in my comments:

Level 1
-------
Looks well played and I can't come up with any routing questions that could be worth exploring (given that you have already investigated the shoes at 1:42).

Level 2
-------
* Was the quick turnaround at 2:42 a mistake? You took a hit anyways right after. A few tenths lost in case the hit was unintentional.
* The death abuse makes sense and looks good
* I'm looking at the section around 3:40 and thinking if there is any way to slide on one more ice patch? You slide on patches 1, 3 and 5. Sliding on 1, 3, 5 and 7 would not be faster since you would then miss out on one of the patches on the floor (or maybe even end up in the milk (?) patches). And something like 1, 2, 4 and 6 would probably not work because you can't jump to the directly next patch?

This level doesn't look easy to fully optimize in the sense taking full advantage of ice patches, timing jump chains to minimize bumping into the environment and similar. It looks like the execution could be even tighter in some spots, but overall I'd say it's well played (unfortunate if the first point above was unintentional though, but it's a small time loss).

Level 3
-------
* Have you really timed out that turning around at 4:51 is faster than running into the spikes in front of you? Judging by the looks of it, the pit on the left and the spikes on the right are the same distance from the checkpoint. And avoiding turning around removes the risk of overshooting the checkpoint (or undershoot and lose the attempt).
* At 5:28, have you tried skipping one platform and make a big jump to the next (the one you land on at 5:31)? If that jump is too long, maybe wait slightly for the platforms to get closer? Your current method looks a bit slow, both because you use the intermediate platform, but also because you lose momentum when you land at 5:31.
* If the above point doesn't work out, can you run down from the platform at 5:28 to the next instead of jumping?
* Same questions as the two points above for the platforms at 5:32-5:34.

No visible execution mistakes that I could spot, but several questions about the routing that I think would be worth clarifying.

Level 4
-------
* Instead of taking the hit at 6:55, could you not have taken a hit by jumping on the spikes and then use invincibility frames to get past the hook? The advantage would be one less jump from backwards-moving conveyor belt.
* What happens if you make a max jump at 6:58? Even if you hit the spikes above, it looks like it would save a bit of time compared to running against the conveyor belt.
* Unintentional hit at 7:12. The health lost in this spot should have been possible to use elsewhere to avoid jumping/running against the conveyor belts. The idea being that you jump on spikes elsewhere instead of staying on conveyor belts that slow you down. Avoiding running against the conveyor belt at 6:58 is probably the biggest potential time save, but it could also work for example at 6:48.

Execution seemed according to the route, but the route contains questions marks as far as I'm concerned.

RockoSonicFan

Edit history:

RockoSonicFan: 2023-01-15 12:21:04 pm

2023-01-15 12:07:10 pm

2:42 was a tiny mistake.
3:40 is always random because you can fall off the platform and lose a life.
4:51 is faster because if you hit the spikes you'll be bouncing up which loses time before the death.
I tried skipping one platform at 5:28 and making a big jump to the next at 5:31 but it definitely failed. Plus I tried running down on them to the next but is impossible to make same goes for 5:32 to 5:34.
6:55 only takes away 1 piece of your health while the spikes take away 2 pieces of your health.
6:58 is definitely worth a try
7:12 I can understand.

RockoSonicFan

2023-01-15 02:06:58 pm

Ok I did the max jump mentioned at 6:58 and I fixed 7:12 by jumping on conveyor belts at that point.

ktwo

2023-01-15 02:10:35 pm

The hit at 6:55 took away 2 health, so the same as hitting spikes.

About 4:51, I believe you if you say you've tried this. However, I would just like to point out that bouncing up like that doesn't necessarily mean losing time. In some games, the time to starting at the checkpoint is timed from the moment you lose the last health and not from the moment you leave the screen. Again, not saying that's the case in this game, but it's something I've encountered myself when speedrunning, so wanted to at least mention it.

RockoSonicFan

2023-01-15 02:20:32 pm

I know that doesn't necessarily mean losing time and yes, I've seen some games have the checkpoint timed from the moment you lose the health and not leave the screen. Thanks for mentioning it, nonetheless.

ktwo

Edit history:

ktwo: 2023-01-16 07:25:11 am

2023-01-16 01:31:19 am

Ok, I've watched iteration #5. Around 1 second faster than #4.

I like the new strat around 1:10, well done!

Levels 1-3 look great now in my opinion (and thanks for addressing my questions in one of your previous posts). But a few of the questions in level 4 still remain. You took 2 damage at 6:59 (and not 1, like you said). I would like to see you try jumping on the spikes instead of running against the conveyor belt. The jump pattern will be different for the last part, but I think it can even be beneficial with this change (I'm thinking at the platforming around 7:01).

And even if there would still somehow be a difference in damage taken, you ended level 4 with one health to spare, so that would not have been a problem.

RockoSonicFan

Edit history:

RockoSonicFan: 2023-01-16 06:51:21 am

2023-01-16 06:51:14 am

Ok I jumped on the spikes and it did go faster.

ktwo

2023-01-16 07:41:46 am

It actually looks like iteration #6 is a tenth of a second or so slower than #5. You're faster by around half a second in level 4, probably thanks to jumping on the spikes. I think most of that gain was lost by getting stuck a bit at 2:21 though?

If you're willing to try one more time to see if you can beat iteration #5, I have a few minor ideas:
* Maybe take a hit at 6:52 instead of at 7:11?
* It looks to me like the conveyor belt is moving you forward at 6:56. I'm thinking that after taking the hit at the spikes, you could therefore just run as far as you can before the conveyor belt changes direction again. Maybe, maybe you could then jump down directly from the middle platform (6:57) instead of jumping up to the platform below the spkes. Not sure how that works out with the next jump sequence though.
* If you decide to take the hit at 7:11, why not run a bit more on the spikes before jumping to the conveyor belt? I think you could then cut out one of the two jumps on the next conveyor belt.

It actually looks like iteration #6 is a tenth of a second or so slower than #5. You're faster by around half a second in level 4, probably thanks to jumping on the spikes. I think most of that gain was lost by getting stuck a bit at 2:21 though?

If you're willing to try one more time to see if you can beat iteration #5, I have a few minor ideas:
* Maybe take a hit at 6:52 instead of at 7:11?
* It looks to me like the conveyor belt is moving you forward at 6:56. I'm thinking that after taking the hit at the spikes, you could therefore just run as far as you can before the conveyor belt changes direction again. Maybe, maybe you could then jump down directly from the middle platform (6:57) instead of jumping up to the platform below the spkes. Not sure how that works out with the next jump sequence though.
* If you decide to take the hit at 7:11, why not run a bit more on the spikes before jumping to the conveyor belt? I think you could then cut out one of the two jumps on the next conveyor belt.

RockoSonicFan

Edit history:

RockoSonicFan: 2023-01-16 08:40:01 am

2023-01-16 08:36:58 am

From my failed attempt

Attachment:

2023-01-16 09-59-00.mp4

ktwo

2023-01-16 09:57:23 am

Interesting. Do you really need to jump already at 7:00 though? Isn't the conveyor belt moving forward at that time or is it just about to change direction? If the lower belt is moving forward, you could delay the jump to the middle platform and then jump closer to the right edge and avoid the spikes in the ceiling? (note the question marks here - I'm just commenting on how it looks to me, but you know the game better)

I'm still wondering if you really need to jump at 6:59 since it looks like the conveyor belt is anyways moving forward?

ktwo

2023-01-17 11:36:47 am

Dude! I didn't see that you uploaded a new run (iteration #7). You nailed it! It's a little over a second faster than #5 and #6. From what I can tell, it's the faster strats in level 4 that paid off, combined with essentially no mistakes in the run.

Seriously, congrats! Looking back at the first uploaded video, you've come a long way. I'm glad that you kept at it all the way to the finish line. I hope you also feel that you now have a really solid run.

ACCEPT

RockoSonicFan

2023-01-17 11:43:12 am

Thank you so much.

SDAVerification

2023-01-17 02:34:10 pm

Decision posted.

LotBlind

2023-01-19 04:24:29 pm

Well, I also took a look at this today. Not bad! I think there's gonna be no shame in putting this up on the front page when we get that far.

You might have been told already but you can either try to crop it precisely yourself (incl. the window top bar; you can do this in Yua I think) or leave it to us. Either is fine.

A/V fine, no cheating detected. Just at a glance might have looked like there could be an earlier cycle in one part but this was already discussed in detail here so I'm not gonna stir it any further!

accept

RockoSonicFan

2023-12-10 04:03:49 pm

Quote from LotBlind:

Well, I also took a look at this today. Not bad! I think there's gonna be no shame in putting this up on the front page when we get that far.

You might have been told already but you can either try to crop it precisely yourself (incl. the window top bar; you can do this in Yua I think) or leave it to us. Either is fine.

A/V fine, no cheating detected. Just at a glance might have looked like there could be an earlier cycle in one part but this was already discussed in detail here so I'm not gonna stir it any further!

accept

Thanks, btw I did a failed attempt thinking that I may have found new strategies for it but turned out that it's completely slow. Here's the vid.