Journey to silius

2013-03-27 02:47:51 am

Scroll down a bit on his nico-community page and there are links to his youtube channel, twitter account etc. No need to learn Japanese for that. :-)
I don't know how likely it is that he will respond to a message in English (or google translated into Japanese) though...

Btw, I watched my own run again. I couldn't help but notice how many places I had to stop or take hits, especially in stage 3 and 4. I know that was part of my routing and that there were good reasons for pretty much everything. However, that's not to say that one couldn't try harder and come up with ways to iron out some of those losses of momentum. It wouldn't surprise me if there were a few more seconds to be gained from just small optimizations that add up in the end.

ZakkyTheHybrid

2013-08-17 08:35:34 pm

Goatragon!

Quietly started working on this a couple of weeks ago.

Here's my best run so far, a 12:13.28, played on console.

I've added annotations mostly to point out quite a few places where I made mistakes (I had way more brain cramps than I actually noted), but also to just mention some various things. I apologize for the quality of the video and for any annotations that are difficult to read (which hopefully isn't all of them).

There are a ton of places where I need to iron out the wrinkles and I had a lot of brain cramps during this run. I also screwed up a weapon select navigation toward the end. I'm hoping that once I get all of those worked out I can get a run really close to 12 minutes.

ktwo

2013-08-21 09:36:14 am

Not bad. What's your goal for this game? Are you just trying for sub 12 or will you try to push the time down as low as you can (it can go quite a bit below 12 and it looks like you have the capacity to take it there, should you be willing to invest in this game)?
As you said, there are a bunch of things that can be improved and it's therefore difficult to give feedback at this early stage (in case you're interested in that in the first place, of course). In any case, I'll keep an eye out for developments in this thread.

ZakkyTheHybrid

2013-08-24 12:13:40 pm

Goatragon!

My first goal is sub-12. That's possible with good execution alone. I think that beyond that it starts getting more and more dependent on drop luck, but after I hit sub-11 I'd like to push it as far as I can.

Being busy this week set me back a little because I found myself a little off when I first picked it up again yesterday, but I did manage after quite a few hours of attempts to improve on my personal best, so I'm getting there. And I'm getting much more consistent at my execution, especially on 3 and 4, which is where it's needed.

I'll put up my improvement in the next day or so. I think it's a little bit more well-executed than my previous run, but I'll have to look over it to see how much more so.

ZakkyTheHybrid

Edit history:

ZakkyTheHybrid: 2013-08-26 11:59:00 am

2013-08-25 05:57:41 pm

Goatragon!

Apologies for the double-post, but I made some really good progress on this today that I thought was worth sharing.

11:52.40

Some notes:
Stage 1 - Nothing to report here. I've been consistent at this for a while.

Stage 2 - This is the smoothest run of this stage I've had. I wasn't as close to the miniboss when I killed it as I would have liked to have been.

Stage 3 - This was a pretty decent run of this stage overall. I was out of position when I jumped onto the first elevator which cost me a pause at one of the flame mines, but I was able to keep pace otherwise. After the trip down the vertical shaft I got hung up on the first ledge which more than likely resulted in that bullet I took from the turret after I made it up on the ledge. The gauntlet after the last elevator went remarkably smooth. I only took two unintended shots and the last one turned out beneficial as it saved me taking a hit from the bot and allowed me to kill the bot during the invincibility period.

Stage 4 - The path through to the second elevator went smoothly, as it should, but I had trouble on the trip down the first vertical shaft. The next real trouble spot was the last vertical shaft, but I managed to recover decently. I started off smoothly after the fall and then got pegged by a bullet from one of the ground turrets when I was slow on the jump button. After that the only real trouble was on the miniboss in taking one more hit than intended, but no wasted shots.

Stage 5 - My daring strategy for the first boss is to use five grenades on the first core and then finish it off with the hand gun to save an extra grenade for the final boss. I've lost a lot of good runs messing that up, so I played it safe this time and just completely destroyed the core with the grenades. However, I mistimed my first dodge from the turret under the second core and then was careless and took a second shot. The final boss wasn't perfect, but still went well. I took one hit, but my jump timing was otherwise very good and I hit with all grenades and missed only one shot with the hand gun.

I feel like this was a decent run overall and it was good to get a sub-12 time. That being said, I'm definitely not done with this, yet.

EDIT: I think this might be an 11:48 by SDA timing.

Apologies for the double-post, but I made some really good progress on this today that I thought was worth sharing.

11:52.40

http://www.youtube.com/watch?v=NktskEw9jks

Some notes:
Stage 1 - Nothing to report here.  I've been consistent at this for a while.

Stage 2 - This is the smoothest run of this stage I've had.  I wasn't as close to the miniboss when I killed it as I would have liked to have been.

Stage 3 - This was a pretty decent run of this stage overall.  I was out of position when I jumped onto the first elevator which cost me a pause at one of the flame mines, but I was able to keep pace otherwise.  After the trip down the vertical shaft I got hung up on the first ledge which more than likely resulted in that bullet I took from the turret after I made it up on the ledge.  The gauntlet after the last elevator went remarkably smooth.  I only took two unintended shots and the last one turned out beneficial as it saved me taking a hit from the bot and allowed me to kill the bot during the invincibility period.

Stage 4 - The path through to the second elevator went smoothly, as it should, but I had trouble on the trip down the first vertical shaft.  The next real trouble spot was the last vertical shaft, but I managed to recover decently.  I started off smoothly after the fall and then got pegged by a bullet from one of the ground turrets when I was slow on the jump button.  After that the only real trouble was on the miniboss in taking one more hit than intended, but no wasted shots.

Stage 5 - My daring strategy for the first boss is to use five grenades on the first core and then finish it off with the hand gun to save an extra grenade for the final boss.  I've lost a lot of good runs messing that up, so I played it safe this time and just completely destroyed the core with the grenades.  However, I mistimed my first dodge from the turret under the second core and then was careless and took a second shot.  The final boss wasn't perfect, but still went well.  I took one hit, but my jump timing was otherwise very good and I hit with all grenades and missed only one shot with the hand gun.

I feel like this was a decent run overall and it was good to get a sub-12 time.  That being said, I'm definitely not done with this, yet.

EDIT: I think this might be an 11:48 by SDA timing.

ktwo

Edit history:

ktwo: 2013-09-01 08:53:16 am

2013-08-27 08:33:22 am

I had a look again at the Japanese speedrun wiki-site that lists Hotarubi's 11:17. I noticed there was a comment attached to his time. Googletranslated, it became "It seems to cut 10 seconds, subtle use walking briskly". I actually remember having read that before, just not being able to make any sense of it. I decided to take a closer look at it and I ended up finding some things I think can be useful.
- The horizontal speed is monitored by RAM-address 0504
- Acceleration on the ground is 4
- When at full speed on the ground, the speed will oscillate between 22 and 26 every other frame (= in average 24)
- Releasing the directional button on the ground will bring the speed down to 0 on the next frame
- Acceleration in the air is 1
- When at full speed in the air, the speed will cycle through 22, 23, 24 and back to 22 etc every other frame (= in average 23)
- Upon landing, it's possible to use the ground acceleration to get the speed up to 27 for one frame, but it then drops back to 22 on the following frame
- Holding up on the d-pad will keep the current speed and prevent the oscillations (even when jumping)

There are a few conclusions that can be made based on the above observations:
- There is a fairly noticeable difference between 22 and 26, so if you switch to pressing up on the d-pad instead of left or right, you should be able to see which one it is. If it's 26, continue holding up and gain a considerable speed boost compared with the average of 24. If it's 22, just try again until you get a speed of 26. It's kind of tricky to do this all the time, so some parts will of course have to be done at 22/26 speed. If this can be done perfectly (like in the tas), I think it would save around 40 seconds or so. It's pretty obvious that this is what Hotarubi meant by his comment and the estimation of cutting 10 seconds over the course of a run doesn't sound completely unreasonable (some time needs to be spent on routing which stretches this can be used on of course).
- Getting speed 27 is a one frame trick (and missing it could mean any one of the speeds in the range 22-26), so tas only in my opinion
- For the parts where 26-speed is not possible, jumping is slightly slower than running. While this will probably be difficult to control and any potential time saving is likely in the order of a few frames, it's still worth to keep in mind.
- It's possible to run in "slow motion", which looks funny. Can be used while waiting for bosses to spawn for entertainment purposes (or in level 5).

edit: spelling

I had a look again at the Japanese speedrun wiki-site that lists Hotarubi's 11:17. I noticed there was a comment attached to his time. Googletranslated, it became "It seems to cut 10 seconds, subtle use walking briskly". I actually remember having read that before, just not being able to make any sense of it. I decided to take a closer look at it and I ended up finding some things I think can be useful.
- The horizontal speed is monitored by RAM-address 0504
- Acceleration on the ground is 4
- When at full speed on the ground, the speed will oscillate between 22 and 26 every other frame (= in average 24)
- Releasing the directional button on the ground will bring the speed down to 0 on the next frame
- Acceleration in the air is 1
- When at full speed in the air, the speed will cycle through 22, 23, 24 and back to 22 etc every other frame (= in average 23)
- Upon landing, it's possible to use the ground acceleration to get the speed up to 27 for one frame, but it then drops back to 22 on the following frame
- Holding up on the d-pad will keep the current speed and prevent the oscillations (even when jumping)

There are a few conclusions that can be made based on the above observations:
- There is a fairly noticeable difference between 22 and 26, so if you switch to pressing up on the d-pad instead of left or right, you should be able to see which one it is. If it's 26, continue holding up and gain a considerable speed boost compared with the average of 24. If it's 22, just try again until you get a speed of 26. It's kind of tricky to do this all the time, so some parts will of course have to be done at 22/26 speed. If this can be done perfectly (like in the tas), I think it would save around 40 seconds or so. It's pretty obvious that this is what Hotarubi meant by his comment and the estimation of cutting 10 seconds over the course of a run doesn't sound completely unreasonable (some time needs to be spent on routing which stretches this can be used on of course).
- Getting speed 27 is a one frame trick (and missing it could mean any one of the speeds in the range 22-26), so tas only in my opinion
- For the parts where 26-speed is not possible, jumping is slightly slower than running. While this will probably be difficult to control and any potential time saving is likely in the order of a few frames, it's still worth to keep in mind.
- It's possible to run in "slow motion", which looks funny. Can be used while waiting for bosses to spawn for entertainment purposes (or in level 5).

edit: spelling

ZakkyTheHybrid

Edit history:

ZakkyTheHybrid: 2013-08-29 11:42:55 pm

2013-08-29 04:03:28 pm

Goatragon!

Thanks for the tips! I've been working with the speed trick. I think one time I managed to save about 4-5 seconds on stages 1 and 2 alone using that. I haven't gotten any full runs yet. Mostly I've been taking the time to work out adjustments to my route.

What I've found so far is that it's definitely possible to work it in almost everywhere, but the catch, as you said, is that it can be tricky to get. Here are my routing thoughts so far based on my practice sessions:

Stage 1:
- I've found that the spots where I really should always get this are the start of stage, right after the mine after the first missile launcher, right after the final missile launcher, and then right before the miniboss (the speed actually allows me to hop over both mines in one jump). They're spots I don't really have to concentrate on anything other than setting up the speed. Beyond that, any spots I'm lucky enough to get the boost is bonus. I also have to be careful I'm not focusing too much on setting up the speed boost when I need to make jumps or shoot things or both.

Stage 2:
- Right after getting up the first stairway is a really great place for it. As long as the ground laser doesn't decide to fire at the wrong time (which is extremely rare anyway), the boost can be sustained all the way up to the downward passage.
- I haven't experimented with it much on the long run after the first elevator, but if I need to maintain the same screen position in order to avoid getting shot by the turret, I would have to stop at least for the last bot on the floor as I would run into it before I could kill it. Then again, even if I have to stop, it might still be faster.
- The long run off of the final elevator is a definite must for this.

Stage 3:
- This stage seems very tricky for this, given all the shots and enemies flying around. I tried at the start, but I kept catching the back edge of the flying pinwheel that was passing overhead when I tried jumping over the first pit, even when waiting until the last moment to jump.
- The passageway with the two bots and two jumpers right after the first ground turret seems to be a good place. I'd love to be able to get it for the next passage with the flame-shooters and turrets, but if I don't get it right away I can't worry about it
- Right off the first elevator after killing the bot there's enough time to work it in and then after the fall, but before the falling platforms.
- After the two bots and the flying bot, if I can get the boost and time a jump properly, I can fall, unimpeded, all the way down the shaft.
- After that, the last place it might help is on the final approach to the elevator. I haven't been able to get it there yet, but it may allow jumping over those two flame-shooters in one bound, provided their cycles are cooperatives.

A rather unfortunate glitch I've noticed about those flame shooters is that the hitbox of the flames seems to extend all the way to the top of the screen. It's not noticeable everywhere else since there's not really any way to jump that high over them until those two by the elevator.

Stage 4:
- This stage will be where the most time is saved using this. Virtually everywhere there isn't some tricky pit jump to be made or that isn't a downward passageway, this will help.

Any thoughts on the spots I've mentioned or on ones I might have missed that might be good?

EDIT: Started full runs and was able to get an 11:34 by SDA timing. I also am beginning to consistently get sub-12 times, at least.

The bulk of the savings on my 11:34 came from stages 2, 3, and 4, with 4, 3, and 4 seconds respectively. I managed to shave off 1 second in stage 1 and having a better final boss fight helped me save 1.5 seconds.

My next goal is to get sub 11:30. If I can get the run trick to work in a few places I wasn't able to in this current PB, have smoother stages 3 and 4, and a better 2nd to last boss fight, I'll be in good shape.

Thanks for the tips!  I've been working with the speed trick.  I think one time I managed to save about 4-5 seconds on stages 1 and 2 alone using that.  I haven't gotten any full runs yet.  Mostly I've been taking the time to work out adjustments to my route.

What I've found so far is that it's definitely possible to work it in almost everywhere, but the catch, as you said, is that it can be tricky to get.  Here are my routing thoughts so far based on my practice sessions:

Stage 1:
- I've found that the spots where I really should always get this are the start of stage, right after the mine after the first missile launcher, right after the final missile launcher, and then right before the miniboss (the speed actually allows me to hop over both mines in one jump).  They're spots I don't really have to concentrate on anything other than setting up the speed.  Beyond that, any spots I'm lucky enough to get the boost is bonus.  I also have to be careful I'm not focusing too much on setting up the speed boost when I need to make jumps or shoot things or both.

Stage 2:
- Right after getting up the first stairway is a really great place for it.  As long as the ground laser doesn't decide to fire at the wrong time (which is extremely rare anyway), the boost can be sustained all the way up to the downward passage.
- I haven't experimented with it much on the long run after the first elevator, but if I need to maintain the same screen position in order to avoid getting shot by the turret, I would have to stop at least for the last bot on the floor as I would run into it before I could kill it.  Then again, even if I have to stop, it might still be faster.
- The long run off of the final elevator is a definite must for this.

Stage 3:
- This stage seems very tricky for this, given all the shots and enemies flying around.  I tried at the start, but I kept catching the back edge of the flying pinwheel that was passing overhead when I tried jumping over the first pit, even when waiting until the last moment to jump.
- The passageway with the two bots and two jumpers right after the first ground turret seems to be a good place.  I'd love to be able to get it for the next passage with the flame-shooters and turrets, but if I don't get it right away I can't worry about it
- Right off the first elevator after killing the bot there's enough time to work it in and then after the fall, but before the falling platforms.
- After the two bots and the flying bot, if I can get the boost and time a jump properly, I can fall, unimpeded, all the way down the shaft.
- After that, the last place it might help is on the final approach to the elevator.  I haven't been able to get it there yet, but it may allow jumping over those two flame-shooters in one bound, provided their cycles are cooperatives.

A rather unfortunate glitch I've noticed about those flame shooters is that the hitbox of the flames seems to extend all the way to the top of the screen.  It's not noticeable everywhere else since there's not really any way to jump that high over them until those two by the elevator.

Stage 4:
- This stage will be where the most time is saved using this.  Virtually everywhere there isn't some tricky pit jump to be made or that isn't a downward passageway, this will help.

Any thoughts on the spots I've mentioned or on ones I might have missed that might be good?

EDIT: Started full runs and was able to get an 11:34 by SDA timing.  I also am beginning to consistently get sub-12 times, at least.

The bulk of the savings on my 11:34 came from stages 2, 3, and 4, with 4, 3, and 4 seconds respectively.  I managed to shave off 1 second in stage 1 and having a better final boss fight helped me save 1.5 seconds.

My next goal is to get sub 11:30.  If I can get the run trick to work in a few places I wasn't able to in this current PB, have smoother stages 3 and 4, and a better 2nd to last boss fight, I'll be in good shape.

ktwo

2013-08-30 06:59:55 am

Congrats on your new pb! That's a very respectable time.

I'd like to take a brief moment to talk about routing and how low this can go. If we start out by taking my run here on the site as a baseline. 13:48 on the pal-version translates* to 11:28 ntsc (+ maybe a few seconds that come from more lag in ntsc). In my run comments, I tried to list areas that could be improved by taking more risk and how more time could be saved by better luck and less mistakes. All-in-all almost 20 pal seconds (so ~16 s ntsc). While I can't recall from the top of my head what it was, I believe I found one or two more seconds of savings. There are also a few other runs (including your own) available on the net (some of them linked in this thread). Let's say 5 more seconds (just a wild guess) could be cut off by using the strats from everyone combined. Finally, let's assume Hotarubi's comment about 10 s of savings from the fast walking is correct. That would mean a "perfect" console speedrun would be around 10:55-11:00 (sda-timing).

I think we can both agree on that it would be very ambitious with our current knowledge to aim for such a low time though. You'll never get perfect luck with lasers at the end of level 3 or with all the "flame mines" and some time has to be traded by using safer strats etc. However, unless you have redone many of the strats in your new run, I think it would be a good idea to make individual level emulator videos to show your route. It would be easier to comment on them (and also to identify suitable areas for fast walking). I could spot several things in your previous videos that would have saved time without adding risk. I just didn't want to point out things that could already be seen in the existing runs. Anyways, it'd be sad to get a good run without having considered all the known strats first. Take it as you will, it's just my 5 cents...

*

Just to be perfectly clear, I don't advocate comparing my pal-time with your ntsc-time. That would of course be unfair. However, they can be compared in a theoretical sense, meaning the game mechanics are the same so what's possible gameplay on pal is also possible to do on ntsc.

Congrats on your new pb! That's a very respectable time.

I'd like to take a brief moment to talk about routing and how low this can go. If we start out by taking my run here on the site as a baseline. 13:48 on the pal-version translates* to 11:28 ntsc (+ maybe a few seconds that come from more lag in ntsc). In my run comments, I tried to list areas that could be improved by taking more risk and how more time could be saved by better luck and less mistakes. All-in-all almost 20 pal seconds (so ~16 s ntsc). While I can't recall from the top of my head what it was, I believe I found one or two more seconds of savings. There are also a few other runs (including your own) available on the net (some of them linked in this thread). Let's say 5 more seconds (just a wild guess) could be cut off by using the strats from everyone combined. Finally, let's assume Hotarubi's comment about 10 s of savings from the fast walking is correct. That would mean a "perfect" console speedrun would be around 10:55-11:00 (sda-timing).

I think we can both agree on that it would be very ambitious with our current knowledge to aim for such a low time though. You'll never get perfect luck with lasers at the end of level 3 or with all the "flame mines" and some time has to be traded by using safer strats etc. However, unless you have redone many of the strats in your new run, I think it would be a good idea to make individual level emulator videos to show your route. It would be easier to comment on them (and also to identify suitable areas for fast walking). I could spot several things in your previous videos that would have saved time without adding risk. I just didn't want to point out things that could already be seen in the existing runs. Anyways, it'd be sad to get a good run without having considered all the known strats first. Take it as you will, it's just my 5 cents...

*[hidden]Just to be perfectly clear, I don't advocate comparing my pal-time with your ntsc-time. That would of course be unfair. However, they can be compared in a theoretical sense, meaning the game mechanics are the same so what's possible gameplay on pal is also possible to do on ntsc.[/hidden]

ZakkyDraggy

2013-08-30 01:00:08 pm

Quote from ktwo:

Congrats on your new pb! That's a very respectable time.

Thanks!

Quote from ktwo:

However, unless you have redone many of the strats in your new run, I think it would be a good idea to make individual level emulator videos to show your route. It would be easier to comment on them (and also to identify suitable areas for fast walking).

I'll definitely see about recording emulator videos to demonstrate the strategies. They may not perfectly reflect what I intend to do on console, but hopefully they can show my intent well enough.

I did change a strat in stage three, though. After the fall through the shaft with the wall-walkers, I don't backtrack to the flame mine anymore to shift the screen. I keep pressing forward and hopefully take a shot from the turret so I can jump through the claw and then kill the flying bot during the invincibility. Then I can proceed forward and I'm far up enough on the edge of the screen that I can actually get past the laser without stopping. What I do need to work on is getting past the flame-mines smoothly as possible. One drawback is that I have way less time to try and shoot them or set up a precise jump, but it's certainly not impossible.

Quote from ktwo:

I could spot several things in your previous videos that would have saved time without adding risk. I just didn't want to point out things that could already be seen in the existing runs. Anyways, it'd be sad to get a good run without having considered all the known strats first. Take it as you will, it's just my 5 cents...

I am in complete agreement. Actually, what my intent has been is to get good at the strategies I planned out and get my time as low as I can with those strategies, then go back and examine the other strategies, see which ones are faster, and then work them in. In fact, I'm glad you brought that up. That thought had slipped my mind recently.

Feel free, though, to point anything out that I'm consistently not noticing. I may have missed some of the videos and there are a lot of times where there are things I just don't spot without them being pointed out to me.

Thanks for all the feedback. Much appreciated and it's been immensely helpful.

[quote="ktwo"]Congrats on your new pb! That's a very respectable time.[/quote]

Thanks!

[quote="ktwo"]However, unless you have redone many of the strats in your new run, I think it would be a good idea to make individual level emulator videos to show your route. It would be easier to comment on them (and also to identify suitable areas for fast walking).[/quote]

I'll definitely see about recording emulator videos to demonstrate the strategies.  They may not perfectly reflect what I intend to do on console, but hopefully they can show my intent well enough.

I did change a strat in stage three, though.  After the fall through the shaft with the wall-walkers, I don't backtrack to the flame mine anymore to shift the screen.  I keep pressing forward and hopefully take a shot from the turret so I can jump through the claw and then kill the flying bot during the invincibility.  Then I can proceed forward and I'm far up enough on the edge of the screen that I can actually get past the laser without stopping.  What I do need to work on is getting past the flame-mines smoothly as possible.  One drawback is that I have way less time to try and shoot them or set up a precise jump, but it's certainly not impossible.

[quote="ktwo"]I could spot several things in your previous videos that would have saved time without adding risk. I just didn't want to point out things that could already be seen in the existing runs. Anyways, it'd be sad to get a good run without having considered all the known strats first. Take it as you will, it's just my 5 cents...[/quote]

I am in complete agreement.  Actually, what my intent has been is to get good at the strategies I planned out and get my time as low as I can with those strategies, then go back and examine the other strategies, see which ones are faster, and then work them in.  In fact, I'm glad you brought that up.  That thought had slipped my mind recently.

Feel free, though, to point anything out that I'm consistently not noticing.  I may have missed some of the videos and there are a lot of times where there are things I just don't spot without them being pointed out to me.

Thanks for all the feedback.  Much appreciated and it's been immensely helpful.

ZakkyTheHybrid

2013-08-31 07:37:36 pm

Goatragon!

Okay, again, apologizing for a double-post, but I put together some movies on emulator to demonstrate my intended strategies. I recorded the movies and then went through each movie while comparing it to several different runs to see what those runs were doing better than I was. Hopefully I managed to pick up on most of the things I needed to improve on. I do have input files for the movies available, as well.

Stage 1 Route Plan:
- I don't think that there are any major issues with this route, as it looks pretty good and seems to match up well with all the other first stage runs that I've seen. I won't rule out the possibility, though.

Stage 2 Route Plan:
- Shortly after hitting the bottom of the downward passageway, there's a jump during which I fire three shots to kill a spinning turret up ahead and then duck, wait for a laser, and then proceed. I noticed that in ktwo and Retro's runs, they were able to jump immediately after landing from that first jump to proceed without waiting. Right there that's 1 - 1.5 seconds saved. Though it's not shown in this demo vid, I am aware of it and will be implementing it.
- Beyond that, I'm not sure, at this point, if there's anything else major to address here

Stage 3 Route Plan:
- I made a video of my strategy. Then I compared it to other runs and ended up totally revising my plan and making a new video. My original demo video was 2:04. My revised video 1:55. The credit goes to ktwo here. The lasers in stage 3 make me really nervous, but I decided to suck it up and go for it. So I studied ktwo's stage 3 route and worked it in. That is a major time saver. I also referred to that same route for the path to the last elevator. NOT killing the flying bot makes me really nervous, like with the beams, but as ktwo's run showed me, it's more optimal, so I took the plunge on that. Then I got really good luck on those lasers at the end. It seems to me that those two parts of the stage are going to be really critical to get right.
- Before I moved on to stage 4, I decided to test out trying to destroy the stage 3 boss at close range. A nearly, if not, perfect scenario is part of this stage 3 demo. I definitely believe this is possible to pull off, mainly since you need to time the missiles similarly for the 2nd boss. And pulling this off perfectly saved about 4 seconds over my previous strategy. I plan on giving this a test on console to confirm it's possible.
- I also recorded a video with 5 different fights on this boss. These were tests I was running to see approximately how much time I would save by getting more up close and personal.

Stage 3 Boss Tests:
- 1st Fight: The strategy I've been using (Control)
- 2nd Fight: Up close test 1 (a not perfect, but not terrible fight) - Despite the mistakes it appears to be about 1s faster than the control
- 3rd Fight: Up close test 2 (a more cautious approach) - Despite being cautious it also appeared to be about 1s faster than the control
- 4th Fight: Up close test 3 (going all out and getting it pretty much perfect) - This appeared to be about 4s faster than the control
- 5th Fight: Up close test 4 (going all out and being very sloppy) - Despite being horribly sloppy and running out of missiles, this was still almost 1 sec faster. Go figure.

Stage 4 Route Plan:
- I'm not sure on this one. It feels like different approaches worked almost equally as well. I feel like there have to be some things I can improve, but it feels like I might have to scrutinize this one to find them.

Anyway, that's the plan so far. Definitely interested to know what I've overlooked and how these routes might further be refined. The planning's also been a nice break from running the game and I'm looking forward to getting back to it with a good plan going forward.

Okay, again, apologizing for a double-post, but I put together some movies on emulator to demonstrate my intended strategies.  I recorded the movies and then went through each movie while comparing it to several different runs to see what those runs were doing better than I was.  Hopefully I managed to pick up on most of the things I needed to improve on.  I do have input files for the movies available, as well.

Stage 1 Route Plan: [url]http://youtu.be/7vgPILPyEcw[/url]
- I don't think that there are any major issues with this route, as it looks pretty good and seems to match up well with all the other first stage runs that I've seen.  I won't rule out the possibility, though.

Stage 2 Route Plan: [url]http://youtu.be/81I6apyQL1Y[/url]
- Shortly after hitting the bottom of the downward passageway, there's a jump during which I fire three shots to kill a spinning turret up ahead and then duck, wait for a laser, and then proceed.  I noticed that in ktwo and Retro's runs, they were able to jump immediately after landing from that first jump to proceed without waiting.  Right there that's 1 - 1.5 seconds saved.  Though it's not shown in this demo vid, I am aware of it and will be implementing it.
- Beyond that, I'm not sure, at this point, if there's anything else major to address here

Stage 3 Route Plan: [url]http://youtu.be/6tP7HBEuEGI[/url]
- I made a video of my strategy.   Then I compared it to other runs and ended up totally revising my plan and making a new video.  My original demo video was 2:04.  My revised video 1:55.  The credit goes to ktwo here.  The lasers in stage 3 make me really nervous, but I decided to suck it up and go for it.  So I studied ktwo's stage 3 route and worked it in.  That is a major time saver.  I also referred to that same route for the path to the last elevator.  NOT killing the flying bot makes me really nervous, like with the beams, but as ktwo's run showed me, it's more optimal, so I took the plunge on that.  Then I got really good luck on those lasers at the end.  It seems to me that those two parts of the stage are going to be really critical to get right.
- Before I moved on to stage 4, I decided to test out trying to destroy the stage 3 boss at close range.  A nearly, if not, perfect scenario is part of this stage 3 demo.  I definitely believe this is possible to pull off, mainly since you need to time the missiles similarly for the 2nd boss.  And pulling this off perfectly saved about 4 seconds over my previous strategy.  I plan on giving this a test on console to confirm it's possible.
- I also recorded a video with 5 different fights on this boss.  These were tests I was running to see approximately how much time I would save by getting more up close and personal.

Stage 3 Boss Tests: [url]http://youtu.be/1Gqr8yQ7Agc[/url]
- 1st Fight: The strategy I've been using (Control)
- 2nd Fight: Up close test 1 (a not perfect, but not terrible fight) - Despite the mistakes it appears to be about 1s faster than the control
- 3rd Fight: Up close test 2 (a more cautious approach) - Despite being cautious it also appeared to be about 1s faster than the control
- 4th Fight: Up close test 3 (going all out and getting it pretty much perfect) - This appeared to be about 4s faster than the control
- 5th Fight: Up close test 4 (going all out and being very sloppy) - Despite being horribly sloppy and running out of missiles, this was still almost 1 sec faster.  Go figure.

Stage 4 Route Plan: [url]http://youtu.be/sw6Gsa1Qh_8[/url]
- I'm not sure on this one.  It feels like different approaches worked almost equally as well.  I feel like there have to be some things I can improve, but it feels like I might have to scrutinize this one to find them.

Anyway, that's the plan so far.  Definitely interested to know what I've overlooked and how these routes might further be refined.  The planning's also been a nice break from running the game and I'm looking forward to getting back to it with a good plan going forward.

ktwo

2013-09-01 08:52:13 am

First of all, it's pretty difficult for me to see if you run at 26 or 22/26 speed based on these youtube videos. My comments assume you run at 22/26, as that's what I have some experience with.

Stage 1
-----------
- 0:25, 0:39: Were the quick stops intentional? They're not necessary.
- I assume you will go for machine gun strat against the boss? It looks easier and less luck-based than what I did in my speedrun, so it's understandable. Just out of curiosity though, have you timed the two methods (including the weapon change in stage 2).
- In your full runs, you started attacking the helicopter too late.

Stage 2
------------
- 0:21-0:28: You can get through this section faster (see my video). The tricky part is to arrive at the first rotating turret so you don't immediately get hit. In order to sync this, you need to make a short jump over the laser turret (0:22) and land right behind it on the same platform (not immediately jump down like in your video). It is difficult though. You just have to practice it from a save state. I think I was something like 1/3 or 1/4. I didn't reset if didn't get it, but I always tried to get through it the fast way.
- 0:31: I see you're already aware of that the stop you made was unnecessary.
- 0:43: It's nothing wrong with shooting the robot, but I remember sometimes not firing fast enough and running into it. I found it more reliable to just jump over it. Up to you...
- 0:51: You can technically run a bit more to the right and then jump back when the elevator approaches. This way you skip some screen scrolling when you arrive at the top. Probably very minor. Try it out and see if it's useful.
- 1:09: Why do you duck?
- Boss: Not shown in your video, but in your full runs, you started attacking it too late.

Stage 3
-----------
Which lasers are you talking about? The ones between the first elevator and the second elevator or the ones near the end?
- 0:22, 0:28: You're always gonna get hit here. Isn't it better to turn around just before the impact, so you get boosted forward instead of backward?
- 0:39: I used make a small jump here and turn around just before getting hit by the fork thing. Should be faster than how you do it.
- 0:50: Are you sure it's faster to take this hit than trying to avoid it?
- 0:52: It's possible to jump over this enemy. Quite difficult though. Try it out and see if you can do it reliably.
- 1:11: Was this bad luck? If not, try and mimic how I did it in my video. From what I remember, that turret did usually not cause any problems (but it's possible that my memory fails me on this detail?).
- 1:32: Why duck in this spot?
- I don't have any suggestions regarding the boss fight. They're all much faster than how I did it in my run.

Stage 4
-----------
- 0:19: Why not get boosted forward instead?
- 0:39 (falling block and blue hovering robot): This looked a bit slow. Have you compared with my run?
- 0:42: Was it really necessary to turn around?
- 0:43: You can take a hit and clip through the falling block. Your method looked slow.
- 0:46: I don't recall exactly how, but somewhere around here, turn around mid-air and shoot the flying enemy that approaches from behind. It can be done without losing much speed and it saves you a hit from that same enemy a little later on. (don't look at my run for this, I didn't do it the right way)
- 0:52: Is it possible to trigger the spike block before you turn around and deal with flying enemy?
- 1:19-1:45: I can see you tried to get the screen position right so you didn't end up like I did in my run (running back and forth at the end to scroll the screen enough to avoiding freezing the screen position when jumping down). Still, it looks like you spent a lot of time to avoid this. I think my method looks faster. Have you compared?
- 1:59: With better timed shots, I don't think you need to stop here.
- 2:04: It looks like you hit the corner of the ceiling. You can jump earlier and still avoid the blue robot.
- Pre-boss: I remember in one or both of your full runs that you managed to get on the right side of the boss. That way you could shoot the lasers faster since they' faster go off the screen. I don't know if it adds risk, so just take it as an idea to try out.
- Boss: I remember that missiles could take the boss out one cycle faster than grenades. You need two missiles to connect in each phase, so it's also slightly riskier. Try it out and see if you can do it reliably.

Stage 5
----------
- Ship: I'd like to see your strat here. There are different approaches depending on where you take damage and which weapons you use. In the speedruns that have been done so far, I think several different strats are used, so you should compare them with how you do it.
- Boss: This fight depends on how much weapon energy you have left, so it should of course be taken into account when looking at the ship strats.

I have tried to point out as many improvement ideas as possible in my comments. Since they're already quite long, I haven't bothered writing down all the good things that you should just keep doing. So don't take my nitpicking as me being negative. Overall, you're definitely on the right track and there are lots of good ideas in your videos.

And finally, I recommend recording the full stages (including the bosses) and start the next video from where the last one ended. This way, you can add them together to get a total "perfect" time. That will be helpful when you start recording again to see what a realistic target time is.

First of all, it's pretty difficult for me to see if you run at 26 or 22/26 speed based on these youtube videos. My comments assume you run at 22/26, as that's what I have some experience with.

Stage 1
-----------
- 0:25, 0:39: Were the quick stops intentional? They're not necessary.
- I assume you will go for machine gun strat against the boss? It looks easier and less luck-based than what I did in my speedrun, so it's understandable. Just out of curiosity though, have you timed the two methods (including the weapon change in stage 2).
- In your full runs, you started attacking the helicopter too late.

Stage 2
------------
- 0:21-0:28: You can get through this section faster (see my video). The tricky part is to arrive at the first rotating turret so you don't immediately get hit. In order to sync this, you need to make a short jump over the laser turret (0:22) and land right behind it on the same platform (not immediately jump down like in your video). It is difficult though. You just have to practice it from a save state. I think I was something like 1/3 or 1/4. I didn't reset if didn't get it, but I always tried to get through it the fast way.
- 0:31: I see you're already aware of that the stop you made was unnecessary.
- 0:43: It's nothing wrong with shooting the robot, but I remember sometimes not firing fast enough and running into it. I found it more reliable to just jump over it. Up to you...
- 0:51: You can technically run a bit more to the right and then jump back when the elevator approaches. This way you skip some screen scrolling when you arrive at the top. Probably very minor. Try it out and see if it's useful.
- 1:09: Why do you duck?
- Boss: Not shown in your video, but in your full runs, you started attacking it too late.

Stage 3
-----------
Which lasers are you talking about? The ones between the first elevator and the second elevator or the ones near the end?
- 0:22, 0:28: You're always gonna get hit here. Isn't it better to turn around just before the impact, so you get boosted forward instead of backward?
- 0:39: I used make a small jump here and turn around just before getting hit by the fork thing. Should be faster than how you do it.
- 0:50: Are you sure it's faster to take this hit than trying to avoid it?
- 0:52: It's possible to jump over this enemy. Quite difficult though. Try it out and see if you can do it reliably.
- 1:11: Was this bad luck? If not, try and mimic how I did it in my video. From what I remember, that turret did usually not cause any problems (but it's possible that my memory fails me on this detail?).
- 1:32: Why duck in this spot?
- I don't have any suggestions regarding the boss fight. They're all much faster than how I did it in my run.

Stage 4
-----------
- 0:19: Why not get boosted forward instead?
- 0:39 (falling block and blue hovering robot): This looked a bit slow. Have you compared with my run?
- 0:42: Was it really necessary to turn around?
- 0:43: You can take a hit and clip through the falling block. Your method looked slow.
- 0:46: I don't recall exactly how, but somewhere around here, turn around mid-air and shoot the flying enemy that approaches from behind. It can be done without losing much speed and it saves you a hit from that same enemy a little later on. (don't look at my run for this, I didn't do it the right way)
- 0:52: Is it possible to trigger the spike block before you turn around and deal with flying enemy?
- 1:19-1:45: I can see you tried to get the screen position right so you didn't end up like I did in my run (running back and forth at the end to scroll the screen enough to avoiding freezing the screen position when jumping down). Still, it looks like you spent a lot of time to avoid this. I think my method looks faster. Have you compared?
- 1:59: With better timed shots, I don't think you need to stop here.
- 2:04: It looks like you hit the corner of the ceiling. You can jump earlier and still avoid the blue robot.
- Pre-boss: I remember in one or both of your full runs that you managed to get on the right side of the boss. That way you could shoot the lasers faster since they' faster go off the screen. I don't know if it adds risk, so just take it as an idea to try out. 
- Boss: I remember that missiles could take the boss out one cycle faster than grenades. You need two missiles to connect in each phase, so it's also slightly riskier. Try it out and see if you can do it reliably.

Stage 5
----------
- Ship: I'd like to see your strat here. There are different approaches depending on where you take damage and which weapons you use. In the speedruns that have been done so far, I think several different strats are used, so you should compare them with how you do it.
- Boss: This fight depends on how much weapon energy you have left, so it should of course be taken into account when looking at the ship strats.

I have tried to point out as many improvement ideas as possible in my comments. Since they're already quite long, I haven't bothered writing down all the good things that you should just keep doing. So don't take my nitpicking as me being negative. Overall, you're definitely on the right track and there are lots of good ideas in your videos.

And finally, I recommend recording the full stages (including the bosses) and start the next video from where the last one ended. This way, you can add them together to get a total "perfect" time. That will be helpful when you start recording again to see what a realistic target time is.

ZakkyTheHybrid

Edit history:

ZakkyTheHybrid: 2013-09-02 03:21:09 am

2013-09-02 01:19:52 am

Goatragon!

Okay, very long and detailed post upcoming...

I decided to do a full test run on emulator addressing the suggestions:

In most cases I addressed the suggestions to the word, but in a few cases I experimented, with mixed results. I've included the times in the demo that are equivalent to the times mentioned in the individual videos for suggested improvements. The few individual notes about the route demo are mixed in with the response notes.

I've also noted times where I used 26 speed as I hadn't anticipated last time how virtually impossible it is to notice that in the video.

Stage 1
----------
26 Speed: 0:07-0:16, 0:22-0:29, 0:38-0:43, 0:45-0:50

Quote from ktwo:

- 0:25, 0:39: Were the quick stops intentional? They're not necessary.
- I assume you will go for machine gun strat against the boss? It looks easier and less luck-based than what I did in my speedrun, so it's understandable. Just out of curiosity though, have you timed the two methods (including the weapon change in stage 2).
- In your full runs, you started attacking the helicopter too late.

- (0:29, 0:44) Fixed the stops. I was going full-tilt at 26 during the last demo. Doing that was what necessitated the stops. So I tried slowing down to 22/26 right before those points, which worked.
- Yeah, I'll be going with the machine gun strat for now. I did time it out and it is slightly slower, but unfortunately it's the only strat I've ever been (consistently) successful with on the boss.
- I started attacking late in this video too. That's just going to require practice, to be honest. I'll go back and work in the emulator again to try and figure out how to time it and then try to put it into practice on the console. I just never time it right. I either fire too early or start a little late. What I did notice is that the further back I stood, though, the early I could fire. I figured that by standing closer I could get more shots in faster, but I guess that's offset by having to wait a little longer to fire.

Stage 2
----------
26 Speed: 1:52-2:05, 2:17-2:20, 2:28-2:31, 2:42-2:51

Quote from ktwo:

- 0:21-0:28: You can get through this section faster (see my video). The tricky part is to arrive at the first rotating turret so you don't immediately get hit. In order to sync this, you need to make a short jump over the laser turret (0:22) and land right behind it on the same platform (not immediately jump down like in your video). It is difficult though. You just have to practice it from a save state. I think I was something like 1/3 or 1/4. I didn't reset if didn't get it, but I always tried to get through it the fast way.
- 0:31: I see you're already aware of that the stop you made was unnecessary.
- 0:43: It's nothing wrong with shooting the robot, but I remember sometimes not firing fast enough and running into it. I found it more reliable to just jump over it. Up to you...
- 0:51: You can technically run a bit more to the right and then jump back when the elevator approaches. This way you skip some screen scrolling when you arrive at the top. Probably very minor. Try it out and see if it's useful.
- 1:09: Why do you duck?
- Boss: Not shown in your video, but in your full runs, you started attacking it too late.

- (2:05-2:13) Got the timing and got through the section faster.
- (2:15) I did the jump this time.
- Ah okay. 95% of the time the robot it doesn't cause me any problems, so I never really thought about it. But I wasn't sure, so I thought it best to ask.
- I think running past the elevator and jumping back might be a possible time-saver. I didn't apply it to this demo, but I tested it a little bit. I didn't save much time by doing it, but maybe I can refine it a bit.
- I ducked because I was too lazy to go back and redo the section so that I didn't have to think about trying to avoid a bullet. Shame on me because I know better.
- (3:16) I was slow starting the attack on the boss again. I always either attack too soon or too late. Practice, practice, practice.

Other notes on this stage:
- (2:31) The reason I dropped back to 22/26 here was to not crash into bot and to avoid having to wait for a bullet at the jump over the barrier spike.
- (2:41) This was a completely unintended hesitation here, so it's not something that should typically happen.

Stage 3
----------
26 Speed: 3:56-3:59, 4:01-4:09, 4:17-4:20, 4:26-4:29, 4:36-4:39

Quote from ktwo:

Which lasers are you talking about? The ones between the first elevator and the second elevator or the ones near the end?
- 0:22, 0:28: You're always gonna get hit here. Isn't it better to turn around just before the impact, so you get boosted forward instead of backward?
- 0:39: I used make a small jump here and turn around just before getting hit by the fork thing. Should be faster than how you do it.
- 0:50: Are you sure it's faster to take this hit than trying to avoid it?
- 0:52: It's possible to jump over this enemy. Quite difficult though. Try it out and see if you can do it reliably.
- 1:11: Was this bad luck? If not, try and mimic how I did it in my video. From what I remember, that turret did usually not cause any problems (but it's possible that my memory fails me on this detail?).
- 1:32: Why duck in this spot?
- I don't have any suggestions regarding the boss fight. They're all much faster than how I did it in my run.

- So...I actually forgot which lasers I was referring to.
- (3:53, 3:59) Turned my back for the forward boost this time.
- (4:09) Turned my back for this forward boost this time.
- (4:22) I hadn't thought about it, so I tested. If I timed right it's about 1/10th of a second slower than avoiding the shot. I hadn't tried turning my back to the hit to see if it made any difference, so I tried it here. It's still 1/10th of a second slower, so in the future I'm going to be avoiding that hit.
- You know what? I knew this, but it's not something I can get reliably. I didn't try it here (probably should have for the timing). Thinking about it, though, it might be something I ought to attempt regardless for the times I do hit it and perhaps to maybe be able to get it reliably.
- (4:40) That hit last time wasn't intentional, but it was a bad idea. I watched your run and attempted to duplicate what you did. On my first attempt I jumped too far and ended up under the claw right away. I decided to just keep that. It looks like it matches up well, but I wanted to put it out there for feedback.
- (5:03) Last time I ducked out of habit, this time I think it actually was necessary.

Other notes:
- (5:30) I shot this off on purpose to avoid a weapon switch.

Stage 4
----------
26 Speed: 5:46-5:49, 5:53-6:03, 6:06-6:08, 6:13-6:16, 6:18-6:23, 6:28-6:33, 6:40-6:43, 6:58-6:59, 7:10-7:18, 7:22-7:24, 7:29-7:36

Quote from ktwo:

- 0:19: Why not get boosted forward instead?
- 0:39 (falling block and blue hovering robot): This looked a bit slow. Have you compared with my run?
- 0:42: Was it really necessary to turn around?
- 0:43: You can take a hit and clip through the falling block. Your method looked slow.
- 0:46: I don't recall exactly how, but somewhere around here, turn around mid-air and shoot the flying enemy that approaches from behind. It can be done without losing much speed and it saves you a hit from that same enemy a little later on. (don't look at my run for this, I didn't do it the right way)
- 0:52: Is it possible to trigger the spike block before you turn around and deal with flying enemy?
- 1:19-1:45: I can see you tried to get the screen position right so you didn't end up like I did in my run (running back and forth at the end to scroll the screen enough to avoiding freezing the screen position when jumping down). Still, it looks like you spent a lot of time to avoid this. I think my method looks faster. Have you compared?
- 1:59: With better timed shots, I don't think you need to stop here.
- 2:04: It looks like you hit the corner of the ceiling. You can jump earlier and still avoid the blue robot.
- Pre-boss: I remember in one or both of your full runs that you managed to get on the right side of the boss. That way you could shoot the lasers faster since they' faster go off the screen. I don't know if it adds risk, so just take it as an idea to try out.
- Boss: I remember that missiles could take the boss out one cycle faster than grenades. You need two missiles to connect in each phase, so it's also slightly riskier. Try it out and see if you can do it reliably.

- (6:04) Boosted forward. Made the timing a bit tight here, but that was likely a result of my screen positioning.
- (6:24) I timed it out and I think it just looked slow. But I didn't really like the way it looked either so I tried again. It looks better and quicker this time, but apparently still seems to take the same amount of time.
- It was not necessary to turn away. That was part of the route, but it was a result of me being too cautious.
- (6:27) I don't doubt that it also was slow. I adapted the hit/clip through the block.
- (6:31-6:33) Killing the bug is going to be tricky. I tried a bunch of times in this test and failed, eventually ending up with this here. My reflexes are better with a controller than a keyboard, so I may have better luck working that out on console.
- (6:36-6:38) I tried to trigger that block, but it never seemed to trigger until it was too late to turn and kill the bug. I stumbled upon a different approach that might be good. I managed to kill the upper bug on the left, which allowed me to avoid the lower bug by jumping up on the spike block. The timing seemed tight, but I think it's definitely worth going for every time.
- (7:03-7:25) The last attempt wasn't very good and it definitely was slow compared to yours. I think this attempt improves on it and looks a lot quicker, at the least. I hope it does, anyway.
- (7:38) Sure enough, there's the proof that better shot timing works wonders.
- (7:42) I hit the ceiling again here, but I didn't have a choice. It looks like I set myself up for it by landing where I did from the fall way back. For future attempts I'll have to make sure to not land so far forward.
- (7:46) My typical strat here is to fire four shots from the left to destroy his body, take a hit from the shots, and use the invincibility to slip under his head to the right side. Then from there I finish him off. It's more been to just not get pinned by him, honestly. If that's not what you're referring to, I'm not sure what you mean. I actually attempted to avoid taking hits from the miniboss this time, but I think that wasted a second.
- (7:54) I tested this in this demo. It doesn't feel like it'll be any more difficult than what I have to do on either the second or third bosses.

Quote from ktwo:

- Ship: I'd like to see your strat here. There are different approaches depending on where you take damage and which weapons you use. In the speedruns that have been done so far, I think several different strats are used, so you should compare them with how you do it.
- Boss: This fight depends on how much weapon energy you have left, so it should of course be taken into account when looking at the ship strats.

Stage 5
----------
For this route demo I used the strategy of using all my grenades on the ship and then finishing the last core and the final boss with the handgun. For some reason I second guessed a strategy I came up with a while ago. I went and tested a whole bunch of strategies after finishing this demo and it turns out that the strat I didn't use, was the fastest, and the one I did use was the slowest.

I tested 5 different strategies including the ones mentioned above. Strat 1 was the one I used in the demo route and Strat 3 is the one I second-guessed myself on. Here's the results of the testing, assuming perfect/near perfect execution:

Strategy 1: Ship Core 1 (6 grenades), Ship Core 2 (5 grenades, handgun), Final Boss (handgun)
Result: Ship (39s), Final (21s), Total (60s)

Strategy 2: Ship Core 1 (6 grenades), Ship Core 2 (hand gun), Final Boss (5 grenades, handgun)
Result: Ship (44s), Final (15s), Total (59s)

Strategy 3: Ship Core 1 (5 grenades, hand gun), Ship Core 2 (hand gun), Final Boss (6 grenades, handgun)
Result: Ship (43s), Final (14), Total (57s)

Strategy 4: Ship Core 1 (4 grenades, hand gun), Ship Core 2 (hand gun), Final Boss (7 grenades)
Result: Ship (44s), Final (13), Total (57s)
(Assuming this is doable on console, the timing is very tight on the first core. But since #3 takes approximately the same time, I don't think this is worth the risk.)

Strategy 5: Ship Core 1 (10 missiles), Ship Core 2 (10 missiles), Final Boss (1 missile, handgun)
Result: Ship (39s), Final(20s), Total (59s)

Overall the route tested out to about 11:03 (SDA timing) and correcting the few things that went wrong that I didn't correct, it should place a "perfect" route a little bit under 11.

Putting that all aside for just a moment, though, I stumbled upon something...interesting...while testing Strategy #4 above:

I didn't know what had happened at first and I was so stunned that I didn't have the presence of mind just to finish out the final boss fight.

Anyway, I'm not sure if it's due to the emulator or if it's actually doable on a console. I'm going to do further testing to see first if I can duplicate it and if I can, if I can duplicate it on console. I suspect that it had something to do with the timing of that kill shot there. But that's a considerable gain if it is possible on console.

EDIT: I made a post on the TASVideos forum about the glitch and linked to my input file. Hopefully someone can shed some light on what happened.

Okay, very long and detailed post upcoming...

I decided to do a full test run on emulator addressing the suggestions:

http://www.youtube.com/watch?v=8SfrxZYPzJU

In most cases I addressed the suggestions to the word, but in a few cases I experimented, with mixed results.  I've included the times in the demo that are equivalent to the times mentioned in the individual videos for suggested improvements.  The few individual notes about the route demo are mixed in with the response notes.

I've also noted times where I used 26 speed as I hadn't anticipated last time how virtually impossible it is to notice that in the video.

Stage 1
----------
26 Speed: 0:07-0:16, 0:22-0:29, 0:38-0:43, 0:45-0:50

[quote="ktwo"]- 0:25, 0:39: Were the quick stops intentional? They're not necessary.
- I assume you will go for machine gun strat against the boss? It looks easier and less luck-based than what I did in my speedrun, so it's understandable. Just out of curiosity though, have you timed the two methods (including the weapon change in stage 2).
- In your full runs, you started attacking the helicopter too late.[/quote]

- (0:29, 0:44) Fixed the stops.  I was going full-tilt at 26 during the last demo.  Doing that was what necessitated the stops.  So I tried slowing down to 22/26 right before those points, which worked.
- Yeah, I'll be going with the machine gun strat for now.  I did time it out and it is slightly slower, but unfortunately it's the only strat I've ever been (consistently) successful with on the boss.
- I started attacking late in this video too.  That's just going to require practice, to be honest.  I'll go back and work in the emulator again to try and figure out how to time it and then try to put it into practice on the console.  I just never time it right.  I either fire too early or start a little late.  What I did notice is that the further back I stood, though, the early I could fire.  I figured that by standing closer I could get more shots in faster, but I guess that's offset by having to wait a little longer to fire.

Stage 2
----------
26 Speed: 1:52-2:05, 2:17-2:20, 2:28-2:31, 2:42-2:51

[quote="ktwo"]- 0:21-0:28: You can get through this section faster (see my video). The tricky part is to arrive at the first rotating turret so you don't immediately get hit. In order to sync this, you need to make a short jump over the laser turret (0:22) and land right behind it on the same platform (not immediately jump down like in your video). It is difficult though. You just have to practice it from a save state. I think I was something like 1/3 or 1/4. I didn't reset if didn't get it, but I always tried to get through it the fast way.
- 0:31: I see you're already aware of that the stop you made was unnecessary.
- 0:43: It's nothing wrong with shooting the robot, but I remember sometimes not firing fast enough and running into it. I found it more reliable to just jump over it. Up to you...
- 0:51: You can technically run a bit more to the right and then jump back when the elevator approaches. This way you skip some screen scrolling when you arrive at the top. Probably very minor. Try it out and see if it's useful.
- 1:09: Why do you duck?
- Boss: Not shown in your video, but in your full runs, you started attacking it too late.[/quote]

- (2:05-2:13) Got the timing and got through the section faster.
- (2:15) I did the jump this time.
- Ah okay.  95% of the time the robot it doesn't cause me any problems, so I never really thought about it.  But I wasn't sure, so I thought it best to ask.
- I think running past the elevator and jumping back might be a possible time-saver.  I didn't apply it to this demo, but I tested it a little bit.  I didn't save much time by doing it, but maybe I can refine it a bit.
- I ducked because I was too lazy to go back and redo the section so that I didn't have to think about trying to avoid a bullet.  Shame on me because I know better.
- (3:16) I was slow starting the attack on the boss again.  I always either attack too soon or too late.  Practice, practice, practice.

Other notes on this stage:
- (2:31) The reason I dropped back to 22/26 here was to not crash into bot and to avoid having to wait for a bullet at the jump over the barrier spike.
- (2:41) This was a completely unintended hesitation here, so it's not something that should typically happen.

Stage 3
----------
26 Speed: 3:56-3:59, 4:01-4:09, 4:17-4:20, 4:26-4:29, 4:36-4:39

[quote="ktwo"]Which lasers are you talking about? The ones between the first elevator and the second elevator or the ones near the end?
- 0:22, 0:28: You're always gonna get hit here. Isn't it better to turn around just before the impact, so you get boosted forward instead of backward?
- 0:39: I used make a small jump here and turn around just before getting hit by the fork thing. Should be faster than how you do it.
- 0:50: Are you sure it's faster to take this hit than trying to avoid it?
- 0:52: It's possible to jump over this enemy. Quite difficult though. Try it out and see if you can do it reliably.
- 1:11: Was this bad luck? If not, try and mimic how I did it in my video. From what I remember, that turret did usually not cause any problems (but it's possible that my memory fails me on this detail?).
- 1:32: Why duck in this spot?
- I don't have any suggestions regarding the boss fight. They're all much faster than how I did it in my run.[/quote]

- So...I actually forgot which lasers I was referring to.
- (3:53, 3:59) Turned my back for the forward boost this time.
- (4:09) Turned my back for this forward boost this time.
- (4:22) I hadn't thought about it, so I tested.  If I timed right it's about 1/10th of a second slower than avoiding the shot.  I hadn't tried turning my back to the hit to see if it made any difference, so I tried it here.  It's still 1/10th of  a second slower, so in the future I'm going to be avoiding that hit.
- You know what?  I knew this, but it's not something I can get reliably.  I didn't try it here (probably should have for the timing).  Thinking about it, though, it might be something I ought to attempt regardless for the times I do hit it and perhaps to maybe be able to get it reliably.
- (4:40) That hit last time wasn't intentional, but it was a bad idea.  I watched your run and attempted to duplicate what you did.  On my first attempt I jumped too far and ended up under the claw right away.  I decided to just keep that.  It looks like it matches up well, but I wanted to put it out there for feedback.
- (5:03) Last time I ducked out of habit, this time I think it actually was necessary.

Other notes:
- (5:30) I shot this off on purpose to avoid a weapon switch.

Stage 4
----------
26 Speed: 5:46-5:49, 5:53-6:03, 6:06-6:08, 6:13-6:16, 6:18-6:23, 6:28-6:33, 6:40-6:43, 6:58-6:59, 7:10-7:18, 7:22-7:24, 7:29-7:36

[quote="ktwo"]- 0:19: Why not get boosted forward instead?
- 0:39 (falling block and blue hovering robot): This looked a bit slow. Have you compared with my run?
- 0:42: Was it really necessary to turn around?
- 0:43: You can take a hit and clip through the falling block. Your method looked slow.
- 0:46: I don't recall exactly how, but somewhere around here, turn around mid-air and shoot the flying enemy that approaches from behind. It can be done without losing much speed and it saves you a hit from that same enemy a little later on. (don't look at my run for this, I didn't do it the right way)
- 0:52: Is it possible to trigger the spike block before you turn around and deal with flying enemy?
- 1:19-1:45: I can see you tried to get the screen position right so you didn't end up like I did in my run (running back and forth at the end to scroll the screen enough to avoiding freezing the screen position when jumping down). Still, it looks like you spent a lot of time to avoid this. I think my method looks faster. Have you compared?
- 1:59: With better timed shots, I don't think you need to stop here.
- 2:04: It looks like you hit the corner of the ceiling. You can jump earlier and still avoid the blue robot.
- Pre-boss: I remember in one or both of your full runs that you managed to get on the right side of the boss. That way you could shoot the lasers faster since they' faster go off the screen. I don't know if it adds risk, so just take it as an idea to try out. 
- Boss: I remember that missiles could take the boss out one cycle faster than grenades. You need two missiles to connect in each phase, so it's also slightly riskier. Try it out and see if you can do it reliably.[/quote]

- (6:04) Boosted forward.  Made the timing a bit tight here, but that was likely a result of my screen positioning.
- (6:24) I timed it out and I think it just looked slow.  But I didn't really like the way it looked either so I tried again.  It looks better and quicker this time, but apparently still seems to take the same amount of time.
- It was not necessary to turn away.  That was part of the route, but it was a result of me being too cautious.
- (6:27) I don't doubt that it also was slow.  I adapted the hit/clip through the block.
- (6:31-6:33) Killing the bug is going to be tricky.  I tried a bunch of times in this test and failed, eventually ending up with this here.  My reflexes are better with a controller than a keyboard, so I may have better luck working that out on console.
- (6:36-6:38) I tried to trigger that block, but it never seemed to trigger until it was too late to turn and kill the bug.  I stumbled upon a different approach that might be good.  I managed to kill the upper bug on the left, which allowed me to avoid the lower bug by jumping up on the spike block.  The timing seemed tight, but I think it's definitely worth going for every time.
- (7:03-7:25) The last attempt wasn't very good and it definitely was slow compared to yours.  I think this attempt improves on it and looks a lot quicker, at the least.  I hope it does, anyway.
- (7:38) Sure enough, there's the proof that better shot timing works wonders.
- (7:42) I hit the ceiling again here, but I didn't have a choice.  It looks like I set myself up for it by landing where I did from the fall way back.  For future attempts I'll have to make sure to not land so far forward.
- (7:46) My typical strat here is to fire four shots from the left to destroy his body, take a hit from the shots, and use the invincibility to slip under his head to the right side.  Then from there I finish him off.  It's more been to just not get pinned by him, honestly.  If that's not what you're referring to, I'm not sure what you mean.  I actually attempted to avoid taking hits from the miniboss this time, but I think that wasted a second.
- (7:54) I tested this in this demo.  It doesn't feel like it'll be any more difficult than what I have to do on either the second or third bosses.

[quote="ktwo"]- Ship: I'd like to see your strat here. There are different approaches depending on where you take damage and which weapons you use. In the speedruns that have been done so far, I think several different strats are used, so you should compare them with how you do it.
- Boss: This fight depends on how much weapon energy you have left, so it should of course be taken into account when looking at the ship strats.[/quote]

Stage 5
----------
For this route demo I used the strategy of using all my grenades on the ship and then finishing the last core and the final boss with the handgun.  For some reason I second guessed a strategy I came up with a while ago.  I went and tested a whole bunch of strategies after finishing this demo and it turns out that the strat I didn't use, was the fastest, and the one I did use was the slowest.

I tested 5 different strategies including the ones mentioned above.  Strat 1 was the one I used in the demo route and Strat 3 is the one I second-guessed myself on.  Here's the results of the testing, assuming perfect/near perfect execution:

Strategy 1: Ship Core 1 (6 grenades), Ship Core 2 (5 grenades, handgun), Final Boss (handgun)
Result: Ship (39s), Final (21s), Total (60s)

Strategy 2: Ship Core 1 (6 grenades), Ship Core 2 (hand gun), Final Boss (5 grenades, handgun)
Result: Ship (44s), Final (15s), Total (59s)

Strategy 3: Ship Core 1 (5 grenades, hand gun), Ship Core 2 (hand gun), Final Boss (6 grenades, handgun)
Result: Ship (43s), Final (14), Total (57s)

Strategy 4: Ship Core 1 (4 grenades, hand gun), Ship Core 2 (hand gun), Final Boss (7 grenades)
Result: Ship (44s), Final (13), Total (57s)
(Assuming this is doable on console, the timing is very tight on the first core.  But since #3 takes approximately the same time, I don't think this is worth the risk.)

Strategy 5: Ship Core 1 (10 missiles), Ship Core 2 (10 missiles), Final Boss (1 missile, handgun)
Result: Ship (39s), Final(20s), Total (59s)

Overall the route tested out to about 11:03 (SDA timing) and correcting the few things that went wrong that I didn't correct, it should place a "perfect" route a little bit under 11.

Putting that all aside for just a moment, though, I stumbled upon something...interesting...while testing Strategy #4 above:

http://www.youtube.com/watch?v=_ly-ZmbXN5U

I didn't know what had happened at first and I was so stunned that I didn't have the presence of mind just to finish out the final boss fight.

Anyway, I'm not sure if it's due to the emulator or if it's actually doable on a console.  I'm going to do further testing to see first if I can duplicate it and if I can, if I can duplicate it on console.  I suspect that it had something to do with the timing of that kill shot there.  But that's a considerable gain if it is possible on console.

EDIT: I made a post on the TASVideos forum about the glitch and linked to my input file.  Hopefully someone can shed some light on what happened.

ktwo

2013-09-02 02:24:34 pm

1:31: Better timing now. I used the helicopter nose becoming visible as a cue for when to start attacking. It looks like you do the same.
1:50-1:52: When trying out this strat, I remember not being consistent at shooting the robot/ascending the steps without stopping and I often ended up with me not killing the robot fast enough and running into it. An idea if you have the same problem, why not keep the machine gun equipped until you are at the top?
3:06-3:16: A crazy strat would be to fire off a few missiles before the boss becomes vulnerable. If you do it correctly, you can time so you run out of weapon energy at the end of the fight and therefore skipping a menu change. I wouldn't do it, but I'm throwing it out there, so it has been said at least...
3:16: If I remember correctly, you can start attacking almost as soon as the "back-pack" is fully visible. It was better this time, but still a little late from what I remember.
5:16, 5:20: Have you had problems with these flame mines in the past? I think you should just jump over them.
6:13-6:36: I made a video demonstrating what I meant, Your new route looks pretty fast and is probably roughly equivalent. It's probably more a matter of personal preference on what feels the most consistent.
6:18: Since I was already doing some testing in this section, I took the opportunity to time jumping down vs running down. Running down is maybe a tenth of a second faster. Obviously no big deal, but since there is no risk involved, there is no reason not to do it.
6:27: Was it intentional not to get a forward boost here?
7:05: I assume this was for the screen positioning and that you concluded this was the fastest method? It seems like it helped you a bit later on in that corridor, so it looks like a good call from your side.
10:41: You should take damage (first from a turret pellet and then once or twice by jumping into the ship) and use the invulnerability to damage the core "from the inside". That's of course moot in case there is a way to glitch kill the ship on the left side consistently (that would surprise me, but nice find either way). There is several seconds to save from this.
Boss strats (stage 5): There is no reason the ship part should be faster with strat 3 than #2 or #4. I don't know what you did, but I'm pretty sure it should be 58s (44+14) for strat 3. That's still maybe the preferred solution (unless #5 is faster when taking boss 4 into account).

About the opportunities for 26 speed. I'm not sure I can contribute much after all. It sounds like some of the sections are very short though (3 seconds?). Is it really practical to try for this kind of 50/50, where you can lose just as much as you can win? Or can you immediately spot if it's 22 or 26 and then just go back to 22/26 if you got a 22?

1:31: Better timing now. I used the helicopter nose becoming visible as a cue for when to start attacking. It looks like you do the same.
1:50-1:52: When trying out this strat, I remember not being consistent at shooting the robot/ascending the steps without stopping and I often ended up with me not killing the robot fast enough and running into it. An idea if you have the same problem, why not keep the machine gun equipped until you are at the top?
3:06-3:16: A crazy strat would be to fire off a few missiles before the boss becomes vulnerable. If you do it correctly, you can time so you run out of weapon energy at the end of the fight and therefore skipping a menu change. I wouldn't do it, but I'm throwing it out there, so it has been said at least...
3:16: If I remember correctly, you can start attacking almost as soon as the "back-pack" is fully visible. It was better this time, but still a little late from what I remember.
5:16, 5:20: Have you had problems with these flame mines in the past? I think you should just jump over them.
6:13-6:36: I made a video demonstrating what I meant, http://youtu.be/LSTEefg9yNw. Your new route looks pretty fast and is probably roughly equivalent. It's probably more a matter of personal preference on what feels the most consistent.
6:18: Since I was already doing some testing in this section, I took the opportunity to time jumping down vs running down. Running down is maybe a tenth of a second faster. Obviously no big deal, but since there is no risk involved, there is no reason not to do it.
6:27: Was it intentional not to get a forward boost here?
7:05: I assume this was for the screen positioning and that you concluded this was the fastest method? It seems like it helped you a bit later on in that corridor, so it looks like a good call from your side.
10:41: You should take damage (first from a turret pellet and then once or twice by jumping into the ship) and use the invulnerability to damage the core "from the inside". That's of course moot in case there is a way to glitch kill the ship on the left side consistently (that would surprise me, but nice find either way). There is several seconds to save from this.
Boss strats (stage 5): There is no reason the ship part should be faster with strat 3 than #2 or #4. I don't know what you did, but I'm pretty sure it should be 58s (44+14) for strat 3. That's still maybe the preferred solution (unless #5 is faster when taking boss 4 into account).

About the opportunities for 26 speed. I'm not sure I can contribute much after all. It sounds like some of the sections are very short though (3 seconds?). Is it really practical to try for this kind of 50/50, where you can lose just as much as you can win? Or can you immediately spot if it's 22 or 26 and then just go back to 22/26 if you got a 22?

ZakkyTheHybrid

2013-09-02 04:46:24 pm

Goatragon!

Quote from ktwo:

1:31: Better timing now. I used the helicopter nose becoming visible as a cue for when to start attacking. It looks like you do the same.

I think picking a good place to stand helps, too. I'm glad I did better on it. I thought I'd done just as bad. I was practicing on console today and I find that it's more just reflexes. I'll spot the visual cue for starting to attack, but my reaction time is slow sometimes. Practice should help.

Quote from ktwo:

1:50-1:52: When trying out this strat, I remember not being consistent at shooting the robot/ascending the steps without stopping and I often ended up with me not killing the robot fast enough and running into it. An idea if you have the same problem, why not keep the machine gun equipped until you are at the top?

I'm fairly consistent, but as that wouldn't cost any time anyway, maybe it's worth doing anyway. The only wrinkle is that I set up a 26 at the top, but I don't see that getting in the way, really.

Quote from ktwo:

3:06-3:16: A crazy strat would be to fire off a few missiles before the boss becomes vulnerable. If you do it correctly, you can time so you run out of weapon energy at the end of the fight and therefore skipping a menu change. I wouldn't do it, but I'm throwing it out there, so it has been said at least...

You know what? I was actually thinking about asking you about this. I don't think I'd do it either. The risk trade off isn't worth it. Also, I think the time it takes for the weapon energy to restore has to be taken into consideration too.

Quote from ktwo:

3:16: If I remember correctly, you can start attacking almost as soon as the "back-pack" is fully visible. It was better this time, but still a little late from what I remember.

Ah. I was going by tread positioning since that's what I've been used to. I'll just take a look at where the treads are when the back-pack is completely visible and note that. More a matter of preference on my part to do that, but it should work just as well. And yeah, it definitely felt slow and then it looked slow when I watched it again. A good thing is that on all my practice attempts today I had perfect fights on this guy every time, any slow starts notwithstanding.

Quote from ktwo:

5:16, 5:20: Have you had problems with these flame mines in the past? I think you should just jump over them.

No, actually I've just always destroyed them. This is a very good point, though.

Quote from ktwo:

6:13-6:36: I made a video demonstrating what I meant, Your new route looks pretty fast and is probably roughly equivalent. It's probably more a matter of personal preference on what feels the most consistent.

I feel more comfortable getting rid of the jetpack guy as things tend to go wrong when I don't. Beyond that I probably will work in your strat, as it looks like it'll work more consistently. Not to mention the fact of getting rid of that flying bug. I've been able to do it on console, but not consistently, so maybe this will help.

Quote from ktwo:

6:18: Since I was already doing some testing in this section, I took the opportunity to time jumping down vs running down. Running down is maybe a tenth of a second faster. Obviously no big deal, but since there is no risk involved, there is no reason not to do it.

Really? Okay, guess I'd better run off from now on.

Quote from ktwo:

6:27: Was it intentional not to get a forward boost here?

I'm really not sure why I didn't at all think about that here.

Quote from ktwo:

7:05: I assume this was for the screen positioning and that you concluded this was the fastest method? It seems like it helped you a bit later on in that corridor, so it looks like a good call from your side.

I found it several seconds faster, but I probably don't need to kill that turret since it's out of range. Unless it's going to cause time-wasting lag if I don't. I won't call this the fastest way, but I really don't want to take the chances of getting to the downward shaft and finding out I'm stuck because I didn't advance the screen far enough.

Quote from ktwo:

10:41: You should take damage (first from a turret pellet and then once or twice by jumping into the ship) and use the invulnerability to damage the core "from the inside". That's of course moot in case there is a way to glitch kill the ship on the left side consistently (that would surprise me, but nice find either way). There is several seconds to save from this.

I probably should take a break for a few days. It seems like I'm not seeing a lot of obvious and common-sense stuff, wow. Anyway, in regards to the glitch, should it indeed turn out to be a glitch in the game, I'd love to be able to duplicate it on console, but I probably won't do that in the first run, as it feels like it's one of those things that's going to be near frame perfect to do and not easy to get consistently. So taking bullets from the turret and jumping inside to get at that core early is a really good idea.

Quote from ktwo:

Boss strats (stage 5): There is no reason the ship part should be faster with strat 3 than #2 or #4. I don't know what you did, but I'm pretty sure it should be 58s (44+14) for strat 3. That's still maybe the preferred solution (unless #5 is faster when taking boss 4 into account).

It's probably just inconsistencies in my timing and execution in the tests. 58 is definitely fair, but it can be faster if I jump inside the ship to get an early shot at the last core. I feel 3 is definitely going to get the final boss down faster.

Quote from ktwo:

About the opportunities for 26 speed. I'm not sure I can contribute much after all. It sounds like some of the sections are very short though (3 seconds?). Is it really practical to try for this kind of 50/50, where you can lose just as much as you can win? Or can you immediately spot if it's 22 or 26 and then just go back to 22/26 if you got a 22?

I'm pretty good at spotting it actually. In the short areas I try once and if I don't get it I forget about it. But in the long stretches, there's really no excuse to not keep trying to get it when the coast is clear.

Thanks again for the feedback. I'll be practicing some more to get the route down and hopefully soon be ready for full runs. Looks like it'll be fun. And hopefully I'll find out more about that glitch, too.

[quote="ktwo"]1:31: Better timing now. I used the helicopter nose becoming visible as a cue for when to start attacking. It looks like you do the same.[/quote]

I think picking a good place to stand helps, too.  I'm glad I did better on it.  I thought I'd done just as bad.  I was practicing on console today and I find that it's more just reflexes.  I'll spot the visual cue for starting to attack, but my reaction time is slow sometimes.  Practice should help.

[quote="ktwo"]1:50-1:52: When trying out this strat, I remember not being consistent at shooting the robot/ascending the steps without stopping and I often ended up with me not killing the robot fast enough and running into it. An idea if you have the same problem, why not keep the machine gun equipped until you are at the top?[/quote]

I'm fairly consistent, but as that wouldn't cost any time anyway, maybe it's worth doing anyway.  The only wrinkle is that I set up a 26 at the top, but I don't see that getting in the way, really.

[quote="ktwo"]3:06-3:16: A crazy strat would be to fire off a few missiles before the boss becomes vulnerable. If you do it correctly, you can time so you run out of weapon energy at the end of the fight and therefore skipping a menu change. I wouldn't do it, but I'm throwing it out there, so it has been said at least...[/quote]

You know what?  I was actually thinking about asking you about this.  I don't think I'd do it either.  The risk trade off isn't worth it.  Also, I think the time it takes for the weapon energy to restore has to be taken into consideration too.

[quote="ktwo"]3:16: If I remember correctly, you can start attacking almost as soon as the "back-pack" is fully visible. It was better this time, but still a little late from what I remember.[/quote]

Ah.  I was going by tread positioning since that's what I've been used to.  I'll just take a look at where the treads are when the back-pack is completely visible and note that.  More a matter of preference on my part to do that, but it should work just as well.  And yeah, it definitely felt slow and then it looked slow when I watched it again.  A good thing is that on all my practice attempts today I had perfect fights on this guy every time, any slow starts notwithstanding.

[quote="ktwo"]5:16, 5:20: Have you had problems with these flame mines in the past? I think you should just jump over them.[/quote]

No, actually I've just always destroyed them.  This is a very good point, though.

[quote="ktwo"]6:13-6:36: I made a video demonstrating what I meant, http://youtu.be/LSTEefg9yNw. Your new route looks pretty fast and is probably roughly equivalent. It's probably more a matter of personal preference on what feels the most consistent.[/quote]

I feel more comfortable getting rid of the jetpack guy as things tend to go wrong when I don't.  Beyond that I probably will work in your strat, as it looks like it'll work more consistently.  Not to mention the fact of getting rid of that flying bug.  I've been able to do it on console, but not consistently, so maybe this will help.

[quote="ktwo"]6:18: Since I was already doing some testing in this section, I took the opportunity to time jumping down vs running down. Running down is maybe a tenth of a second faster. Obviously no big deal, but since there is no risk involved, there is no reason not to do it.[/quote]

Really?  Okay, guess I'd better run off from now on.

[quote="ktwo"]6:27: Was it intentional not to get a forward boost here?[/quote]

I'm really not sure why I didn't at all think about that here.

[quote="ktwo"]7:05: I assume this was for the screen positioning and that you concluded this was the fastest method? It seems like it helped you a bit later on in that corridor, so it looks like a good call from your side.[/quote]

I found it several seconds faster, but I probably don't need to kill that turret since it's out of range.  Unless it's going to cause time-wasting lag if I don't.  I won't call this the fastest way, but I really don't want to take the chances of getting to the downward shaft and finding out I'm stuck because I didn't advance the screen far enough.

[quote="ktwo"]10:41: You should take damage (first from a turret pellet and then once or twice by jumping into the ship) and use the invulnerability to damage the core "from the inside". That's of course moot in case there is a way to glitch kill the ship on the left side consistently (that would surprise me, but nice find either way). There is several seconds to save from this.[/quote]

I probably should take a break for a few days.  It seems like I'm not seeing a lot of obvious and common-sense stuff, wow.  Anyway, in regards to the glitch, should it indeed turn out to be a glitch in the game, I'd love to be able to duplicate it on console, but I probably won't do that in the first run, as it feels like it's one of those things that's going to be near frame perfect to do and not easy to get consistently.  So taking bullets from the turret and jumping inside to get at that core early is a really good idea.

[quote="ktwo"]Boss strats (stage 5): There is no reason the ship part should be faster with strat 3 than #2 or #4. I don't know what you did, but I'm pretty sure it should be 58s (44+14) for strat 3. That's still maybe the preferred solution (unless #5 is faster when taking boss 4 into account).[/quote]

It's probably just inconsistencies in my timing and execution in the tests.  58 is definitely fair, but it can be faster if I jump inside the ship to get an early shot at the last core.  I feel 3 is definitely going to get the final boss down faster.

[quote="ktwo"]About the opportunities for 26 speed. I'm not sure I can contribute much after all. It sounds like some of the sections are very short though (3 seconds?). Is it really practical to try for this kind of 50/50, where you can lose just as much as you can win? Or can you immediately spot if it's 22 or 26 and then just go back to 22/26 if you got a 22?[/quote]

I'm pretty good at spotting it actually.  In the short areas I try once and if I don't get it I forget about it.  But in the long stretches, there's really no excuse to not keep trying to get it when the coast is clear.

Thanks again for the feedback.  I'll be practicing some more to get the route down and hopefully soon be ready for full runs.  Looks like it'll be fun.  And hopefully I'll find out more about that glitch, too.

ZakkyTheHybrid

Edit history:

ZakkyTheHybrid: 2013-09-06 05:08:57 pm

2013-09-06 05:08:15 pm

Goatragon!

Set a new pb of 11:24. It surprised me because I had a dreadful stage 3 and 4 and got bailed out on stage 3 only because of a lucky health drop at the end.

I've also verified that the boss skip glitch does, in fact, work on console. It is indeed, however, frame precise. The kill shot needs to hit exactly 2 frames before the core moves off screen and can no longer be damaged. I haven't yet found a consistent way to pull this off or managed to get it to work in a run, yet, but I would like to get this in a good run as it will save roughly 40s over normally fighting the two bosses.

ZakkyTheHybrid

Edit history:

ZakkyTheHybrid: 2013-09-20 01:06:05 am

2013-09-19 03:04:01 am

Goatragon!

I apologize for now triple-posting, but I was a bit curious about something. Anyone know what's currently recognized as the record for the NTSC-U version? I've looked into this a little bit myself, but I just want to find out if what I think it happens to be is actually what it is.

EDIT: I'll be streaming some run attempts late Friday afternoon and maybe Saturday at some point (http://twitch.tv/zakkydraggy). Expect lots of restarts, but hopefully I can get a good run or two in. I'm aiming for sub 11:20.

ktwo

2013-09-21 03:36:13 am

I didn't see you improved your pb until now. Good job!

I see your question about the record, but unfortunately I'm not able to answer it. There is not a universally accepted rule set for speedruns, which means that any attempt to reply to the question "who has the record?" will be subjectively tainted by the one who answers. So either you believe what someone tells you to believe or you look at the reported times and form your own opinion. As for the reported times, those I'm aware of are either mentioned in this thread or on the sunday sequence break page. If you know of any other places, let me know.

ZakkyTheHybrid

2013-09-21 11:23:33 am

Goatragon!

I improved it by another 2 seconds this week, down to 11:22. I think I've kinda set myself up now to where I just can't have anything go wrong anymore to get a better time. My stage 1 was almost perfect (I think I'm still having issues starting my attack on the second boss phase slow) and my stage 2 was near perfect, as well, except for not getting the right timing during the descent on Stage 2 to get hit by the rotating turret's bullet on the fall and run past the second turret. Even my miniboss fight was decent for a change. Stage 3 and Stage 4 are pretty much where all the time is to be saved. There were places I ended up playing things safer than I usually would, due to the run going well and wanting to get a fairly decent showing in a pb. My ideal final run will have much less playing it safe and the glitch (which will undoubtedly take a long time to get in a good run, much less in a run at all!). So while I continue to work on this, I'm going ahead and submitting my 11:22 and seeing what happens.

I improved it by another 2 seconds this week, down to 11:22.  I think I've kinda set myself up now to where I just can't have anything go wrong anymore to get a better time.  My stage 1 was almost perfect (I think I'm still having issues starting my attack on the second boss phase slow) and my stage 2 was near perfect, as well, except for not getting the right timing during the descent on Stage 2 to get hit by the rotating turret's bullet on the fall and run past the second turret.  Even my miniboss fight was decent for a change.  Stage 3 and Stage 4 are pretty much where all the time is to be saved.  There were places I ended up playing things safer than I usually would, due to the run going well and wanting to get a fairly decent showing in a pb.  My ideal final run will have much less playing it safe and the glitch (which will undoubtedly take a long time to get in a good run, much less in a run at all!).  So while I continue to work on this, I'm going ahead and submitting my 11:22 and seeing what happens.

ConHuevos

2013-11-10 09:10:23 pm

Ciento Dos Huevos

http://tasvideos.org/forum/viewtopic.php?p=359026#359026

ZakkyTheHybrid

2013-12-24 02:14:36 pm

Goatragon!

11:13 - http://www.twitch.tv/zakkydraggy/c/3443724

ktwo

2013-12-25 03:03:00 pm

I can't watch twitch videos at the moment but congratulations none the less. Keep it up!

ZakkyTheHybrid

2013-12-25 09:16:48 pm

Goatragon!

Thanks!

Despite some of the mistakes I made during the run and the places I couldn't hit the right frame for the speed trick (which actually resulted in a couple of minor hiccups in Stage 4), I'm overall pretty happy with the run.

It's definitely improvable, though, and likely possible to get it really close to 11 flat if not just barely sub-11 even without the ship skip glitch and just with good luck and perfect execution alone (although luck is now a deciding factor every step of Stage 3 almost). I remember when I started working on this a little over four months ago my original goal was just to get sub-12!

At the moment I'm stepping it down a bit, though. I may attempt another run or so before AGDQ, but my primary focus now is just going to be on keeping sharp. As much as I've started to enjoy run attempts again, I need a break and will most likely be putting any further attempts on hold for a while after AGDQ so I can work on other runs.

ZakkyTheHybrid

2014-03-10 08:42:05 pm

Goatragon!

Okay, so I'm getting back to this and I'm going for the ship glitch, probably starting back up in about the next week or two. But there's a new discovery to report, as well.

BADbrakes discovered a death warp: http://www.twitch.tv/badbrakes/c/3856451

So far my attempts to duplicate this through emulator testing have failed, but I have not yet tested this on console, so it's entirely possible that will make the difference. If I can get this to work, I estimated it will save about 10 seconds.

RetroChristian

2014-03-11 11:11:57 am

Quote from ZakkyDraggy:

Okay, so I'm getting back to this and I'm going for the ship glitch, probably starting back up in about the next week or two.

I just wanted to add that there is a quick kill glitch on the first boss too. Happened to me last year but had no clue how it worked. After seeing the last boss glitch I would guess it's a similar thing, kill it on the correct frame or something but that's just a guess. Has only happened once and I haven't seen it anywhere else.

ZakkyTheHybrid

2014-03-12 10:08:50 am

Goatragon!

I'm guessing you mean when it skips on to the next stage in the middle of the energy restoration?

If that's what you mean, I was thinking the same thing; that it might be a frame perfect kill shot on the first boss that triggers that. I've had that happen quite a few times over the course of run attempts, but never been able to have that line up with a successful run. It might be something else I look into before starting up attempts again.

Also, I have had no luck recreating that death warp that BADbrakes had happen to him and I've tried on console now, too. Unfortunately all I have to go on is the highlight video, so I'm not sure what else to do at this point. If anyone has any thoughts or ideas or the know how to look into this, by all means please let me know.