Run Information

Run Information
Completion Date			Aug. 3, 2025
State			Processing
Tag			Deaths
Links
Verification Thread

Run Comments
Disclaimer I did this run on a Powerpak. In case it matters for future reference, I used a ROM with checksum 36f9c268d5152085c914eb732392848e. SDA history 29-Jun-2018 - 0:11:29 by ktwo (http://www.archive.org/download/2018-08-12-SDA_Speed_Runs/) 04-May-2023 - 0:10:52 by ktwo (http://www.archive.org/download/2023-12-30-SDA_Speed_Runs/) Speedrun information For information about speedrunning Beetlejuice and to get a better understanding of what's going on in this run, go to the guide in the SDA knowledge base (referred to as sdakb in my comments below): https://kb.speeddemosarchive.com/Beetlejuice_(NES) Background I was relatively happy with my previous effort in this game, but there were three strats from the TAS that seemed worth investigating a bit more for real-time attempts: * The TAS uses a few different ideas when fighting the cavern monsters in the sewer ("storm drain") level. A bit difficult to estimate how much could be saved, because one can pick and choose a bit among the strats. But somewhere around 2 (3?) seconds seemed to potentially be there. * In the second room with Delia's art pieces (a reddish room with three stools in it), the TAS let's the hammer defeat two of the art pieces off-screen. The time save depends on how well the different methods are executed and is therefore difficult to estimate accurately, but it should be around 6s. The downside is that it also relies quite a bit more on RNG since the art pieces move around randomly. * The fastest fight against the graveyard level boss is 4 attacks in 2 jumps. This requires a bit of tight execution, as well as spending time in the face of the boss, with the risk of getting hit in case the boss attacks (RNG). The time save from this one is again a bit difficult to estimate because it cascades into additional, albeit minor, changes elsewhere, such as which scares to purchase and menuing. But let's say 2s? One second from the faster boss fight and one from the other benefits. Given I'm the TAS author :-), I've been well aware of these strats for quite some time, but I dismissed them earlier as being too inconsistent. When I picked up the game again, the art piece strat actually didn't seem that bad after some preliminary testing though. Definitely more attempts being lost to RNG than before, but still seemed acceptable considering the time saved. The fast graveyard boss fight also quickly seemed more feasible than I had previously thought. With two nice time saves ready to be implemented, there were two major troublespots that I wanted to address before doing attempts again. The first was the first boss fight. I previously used "method 2" (see sdakb), but I noticed that pretty much all other speedrunners of this game were using the first method. There was a little bit of muscle memory to re-learn this fight, but I soon started to get much better consistency and method 1 was clearly the better choice. The second issue was the sewer stage (or "Storm drain" as the manual calls it). You have to fight three cavern frogs and each one can one-hit kill you for 2-3s lost (each one very roughly a 50/50). I started by looking into the code to better understand what determines how the cavern monsters attack (see sdakb for more information), which was informative, but didn't really provide any key to making the fights noticeably more consistent. I then started experimenting with various death abuse methods to better control the time losses in these fights by taking advantage of the invincibility frames. This seemed to stabilize the consistency a bit and I was actually saving time in this level over my pb. With a first new draft of the route in place, I fairly quickly lowered my pb to 10:48 (see the forum game thread). There was clearly more time to be saved though, so only a step towards what I was looking for. When preparing to start working on my next pb, I was still not happy with the sewer stage and experimented with alternative strats (both with and without death abuses). I never really found that perfect strat to settle for and changed approach several times. I was also looking at back-up strats in the tower in the graveyard in case I didn't get an invincibility flower (RNG). I had some convoluted solutions already in place, but it seemed like something else should be possible and it didn't take long to come up with a new strat. In rapid succession, I then found a faster overall route through the tower and a faster way through the first part of the graveyard. Both methods were even faster than the routes used in the TAS and overall looked pretty smooth imo. While the new routes are more difficult, don't save that much time (~2.5s in total) and come quite late in the run, I decided to go for them. More attempts (and changing back and forth between strats in the sewers) eventually resulted in a 10:35. I was somewhat happy with this attempt, but it had two really bad-looking sections (the sewers and the mouse room in the overhead section). If it weren't for these sore spots, I might have ended with this run, even though I already at this point had a few attempts going into the graveyard on 10:1x pace. I decided to try for one more pb though. The next stretch turned out to be one of the more painful in my speedrunning experience. I had 12 failed attempts at the last RNG part of the run, where two enemies need to be defeated with one potion scare, before I finally got another shot on pb pace at the last level. Getting the "1-hit kill" should be around 1/3 or 1/4, so having 12 straight failed attempts is just unbelievable bad luck. In-between, I obviously also got to this spot above pb pace and then managed to get several 1-hit kills... While it was very frustrating to not get the last piece of RNG to cooperate, it did allow me to further hone some of the strats. I found a death warp strat for the very first screen of the game, I made some minor tweaks to the menuing and finally decided to go all-in in the sewers. On my 13th attempt at the 1-hit kill, the game finally cooperated and I got a shot at the last level on pb pace. This resulted in the submitted run. As I will describe below, I do have some regrets about the submitted run. I've had a handful attempts going into the graveyard that could have been 10+s faster than the submitted run. Getting into the graveyard on a good pace is however not the same as closing out a run without noticeable time losses (which the submitted run is a good example of)... However, I definitely feel I could realistically improve by a few seconds, even though it would risk becoming another grind to get all of the RNG to line up one more time. While thinking about whether to settle or try and go for one more pb, an external factor seems to have taken the decision for me... The DVD-function of the HDD-recorder I'm using broke down very shortly after I achieved this run (at least I managed to transfer this run before that happened!). This makes my current recording setup essentially useless since any recording on the HDD-recorder no longer can be transferred without an unacceptable quality loss. I'm not going to say this was my last (NES) speedrun ever, but I don't think I will look into a new recording setup at the moment. So as far as Beetlejuice goes, I will stop here. Run comments Town * New strat outside the house, death warping back to the door after collecting the key. ~1s saved. See sdakb for an explanation. * Essentially perfect beetle farming outside the shop. Around 2-3s faster than what I usually settled for. Storm drains * Because of a new shop strat, I could avoid jumping on the second fly in the raft section and take another hit instead. Saves around 1s. * Got lucky with the first two cavern monsters and didn't take any hits. When running back to the left after the first cavern monster, I used a more aggressive route that turned out to be consistent once I understood the timing of the different enemies. In my previous pb, I got hit by both the first two cavern monsters. Overall, I saved almost 8s until the shop, compared with my 10:52 run. * I changed the purchase strategy to cut out one of the shop visits (see sdakb for a discussion about scare purchases). * Just like in my 10:52 run, I managed to jump through the last octopus without taking a death, which would have cost ~3s. This is a 50/50 event (see sdakb). * I got bad luck against the boss and had to realign after the jump (ideally, the boss attacks earlier, so Beetlejuice can jump over the football and land next to the boss, instead of retreating). The first big execution mistake of the run then happened. One of the scare attacks missed and I had to activate another birdman scare. Other than the time loss of another menu trip (~1s), this also proved to be a very costly mistake towards the end... The first overhead section * Minor change in the sweeper fight. Instead of using 3 potions, I'm now using 2 potions. See the sdakb for more details on this fight and some advantages of the 2-potion fight. * The first room with Delia's art pieces went as well as I could have hoped for. Could have been faster, but the most important is to not have to chase down any of the art pieces on the left side with the hammer. * Big time save in the second room with Delia's art pieces thanks to a new strat. I collected a health beetle on the way back, which was slightly out of the way. It's debatable how far out of the way one should go to collect a beetle or avoid damage, but other than potentially reducing the number of deaths over the course of the run, it's on average also slightly better to enter the next room with one hit than two. * Overall, three rooms in this section that heavily rely on RNG and all of them went well. Not perfect, but at least according plan. The second overhead section * A minor mistake in the first auto-scroller section, where the d-pad was acting up during a diagonal press and I shortly got stuck. Ugly, but not a huge time loss (~0.5s). * The (dreaded) fly room went very well. This is possibly the worst RNG-part of the run and there is almost always more time that can be saved. I'd say this was a top 5-10% completion of this room though. I even managed to re-collect the swatter, which I could use in the last room of this section. * The mouse room is another very random room, almost on the same level as the fly room. Not as great result as the fly room, but I'd still say better than average for me and better than my 10:52 run. * Overall, very good stage of the most RNG-heavy section in the run. Graveyard * New route through the first part of the graveyard and the tower for a time save of around 2.5s * I missed a tight jump before the tower with a 2-frame window. I'm not completely consistent with this jump and this negated the time saved from improved routing. * I barely missed the 4-attack/2-jump boss fight and had to activate another birdman scare and jump a third time to finish the boss off. Together with the mistake against the second boss, this left me with only birdman scare left, which unfortunately meant an additional time loss in the last level. Third overhead section The 1-potion fight against the sweepers comes down to RNG and this is where I had 12 (!) consecutive attempts on pb pace that didn't take out both of them. Other than that, this section is very straight-forward. Afterlife * After the execution mistakes against the 2nd and 3rd bosses, I entered this level with only one birdman scare left. This prevented my from death abusing to collect ticket #1 (see route in the sdakb), which would have saved ~2.5-3s. * A mistake around ticket #2. I was meant to jump over the "magnet crane" and land on the platform to the right, but didn't get enough momentum and had to make another jump. Maybe 1s lost. * With only one life left, another death would have triggered a continue screen. For that reason and because I already had taken two hits, I collected ticket #5 before #6. In my 10:52 run, I first collected #6 and then used a death abuse to collect #5 right before the end, which is faster. * Very slow level. One unnecessary execution mistake, but most of the time loss came from execution mistakes earlier in the run. Final thoughts The time of this run is thanks to great RNG, which is then offset by execution mistakes. I would probably have been very happy with this time if it instead had relied on better execution, but worse RNG. I therefore certainly have an itch to get a better-looking run. However, with my recording setup being out of commission, I take it as a sign to stop here. There is also another aspect when there is this much influence on the time and success rate by RNG. If only looking at this particular attempt, better execution would have resulted in completely different RNG and most likely a slower time. So one can argue that even bad execution contributes to a pb in a game like this! (at least that's what I'm trying to tell myself right now...)

Run Comments

Disclaimer
I did this run on a Powerpak. In case it matters for future reference, I used a ROM with checksum 36f9c268d5152085c914eb732392848e.

SDA history
29-Jun-2018 - 0:11:29 by ktwo (http://www.archive.org/download/2018-08-12-SDA_Speed_Runs/)
04-May-2023 - 0:10:52 by ktwo (http://www.archive.org/download/2023-12-30-SDA_Speed_Runs/)

Speedrun information
For information about speedrunning Beetlejuice and to get a better understanding of what's going on in this run, go to the guide in the SDA knowledge base (referred to as sdakb in my comments below):
https://kb.speeddemosarchive.com/Beetlejuice_(NES)

Background
I was relatively happy with my previous effort in this game, but there were three strats from the TAS that seemed worth investigating a bit more for real-time attempts:
* The TAS uses a few different ideas when fighting the cavern monsters in the sewer ("storm drain") level. A bit difficult to estimate how much could be saved, because one can pick and choose a bit among the strats. But somewhere around 2 (3?) seconds seemed to potentially be there.
* In the second room with Delia's art pieces (a reddish room with three stools in it), the TAS let's the hammer defeat two of the art pieces off-screen. The time save depends on how well the different methods are executed and is therefore difficult to estimate accurately, but it should be around 6s. The downside is that it also relies quite a bit more on RNG since the art pieces move around randomly.
* The fastest fight against the graveyard level boss is 4 attacks in 2 jumps. This requires a bit of tight execution, as well as spending time in the face of the boss, with the risk of getting hit in case the boss attacks (RNG). The time save from this one is again a bit difficult to estimate because it cascades into additional, albeit minor, changes elsewhere, such as which scares to purchase and menuing. But let's say 2s? One second from the faster boss fight and one from the other benefits.

Given I'm the TAS author :-), I've been well aware of these strats for quite some time, but I dismissed them earlier as being too inconsistent. When I picked up the game again, the art piece strat actually didn't seem that bad after some preliminary testing though. Definitely more attempts being lost to RNG than before, but still seemed acceptable considering the time saved. The fast graveyard boss fight also quickly seemed more feasible than I had previously thought.

With two nice time saves ready to be implemented, there were two major troublespots that I wanted to address before doing attempts again. The first was the first boss fight. I previously used "method 2" (see sdakb), but I noticed that pretty much all other speedrunners of this game were using the first method. There was a little bit of muscle memory to re-learn this fight, but I soon started to get much better consistency and method 1 was clearly the better choice.

The second issue was the sewer stage (or "Storm drain" as the manual calls it). You have to fight three cavern frogs and each one can one-hit kill you for 2-3s lost (each one very roughly a 50/50). I started by looking into the code to better understand what determines how the cavern monsters attack (see sdakb for more information), which was informative, but didn't really provide any key to making the fights noticeably more consistent. I then started experimenting with various death abuse methods to better control the time losses in these fights by taking advantage of the invincibility frames. This seemed to stabilize the consistency a bit and I was actually saving time in this level over my pb.

With a first new draft of the route in place, I fairly quickly lowered my pb to 10:48 (see the forum game thread). There was clearly more time to be saved though, so only a step towards what I was looking for.

When preparing to start working on my next pb, I was still not happy with the sewer stage and experimented with alternative strats (both with and without death abuses). I never really found that perfect strat to settle for and changed approach several times. I was also looking at back-up strats in the tower in the graveyard in case I didn't get an invincibility flower (RNG). I had some convoluted solutions already in place, but it seemed like something else should be possible and it didn't take long to come up with a new strat. In rapid succession, I then found a faster overall route through the tower and a faster way through the first part of the graveyard. Both methods were even faster than the routes used in the TAS and overall looked pretty smooth imo. While the new routes are more difficult, don't save that much time (~2.5s in total) and come quite late in the run, I decided to go for them.

More attempts (and changing back and forth between strats in the sewers) eventually resulted in a 10:35. I was somewhat happy with this attempt, but it had two really bad-looking sections (the sewers and the mouse room in the overhead section). If it weren't for these sore spots, I might have ended with this run, even though I already at this point had a few attempts going into the graveyard on 10:1x pace. I decided to try for one more pb though.

The next stretch turned out to be one of the more painful in my speedrunning experience. I had 12 failed attempts at the last RNG part of the run, where two enemies need to be defeated with one potion scare, before I finally got another shot on pb pace at the last level. Getting the "1-hit kill" should be around 1/3 or 1/4, so having 12 straight failed attempts is just unbelievable bad luck. In-between, I obviously also got to this spot above pb pace and then managed to get several 1-hit kills...

While it was very frustrating to not get the last piece of RNG to cooperate, it did allow me to further hone some of the strats. I found a death warp strat for the very first screen of the game, I made some minor tweaks to the menuing and finally decided to go all-in in the sewers. On my 13th attempt at the 1-hit kill, the game finally cooperated and I got a shot at the last level on pb pace. This resulted in the submitted run.

As I will describe below, I do have some regrets about the submitted run. I've had a handful attempts going into the graveyard that could have been 10+s faster than the submitted run. Getting into the graveyard on a good pace is however not the same as closing out a run without noticeable time losses (which the submitted run is a good example of)... However, I definitely feel I could realistically improve by a few seconds, even though it would risk becoming another grind to get all of the RNG to line up one more time. While thinking about whether to settle or try and go for one more pb, an external factor seems to have taken the decision for me... The DVD-function of the HDD-recorder I'm using broke down very shortly after I achieved this run (at least I managed to transfer this run before that happened!). This makes my current recording setup essentially useless since any recording on the HDD-recorder no longer can be transferred without an unacceptable quality loss. I'm not going to say this was my last (NES) speedrun ever, but I don't think I will look into a new recording setup at the moment. So as far as Beetlejuice goes, I will stop here.

Run comments

Town
* New strat outside the house, death warping back to the door after collecting the key. ~1s saved. See sdakb for an explanation.
* Essentially perfect beetle farming outside the shop. Around 2-3s faster than what I usually settled for.

Storm drains
* Because of a new shop strat, I could avoid jumping on the second fly in the raft section and take another hit instead. Saves around 1s.
* Got lucky with the first two cavern monsters and didn't take any hits. When running back to the left after the first cavern monster, I used a more aggressive route that turned out to be consistent once I understood the timing of the different enemies. In my previous pb, I got hit by both the first two cavern monsters. Overall, I saved almost 8s until the shop, compared with my 10:52 run.
* I changed the purchase strategy to cut out one of the shop visits (see sdakb for a discussion about scare purchases).
* Just like in my 10:52 run, I managed to jump through the last octopus without taking a death, which would have cost ~3s. This is a 50/50 event (see sdakb).
* I got bad luck against the boss and had to realign after the jump (ideally, the boss attacks earlier, so Beetlejuice can jump over the football and land next to the boss, instead of retreating). The first big execution mistake of the run then happened. One of the scare attacks missed and I had to activate another birdman scare. Other than the time loss of another menu trip (~1s), this also proved to be a very costly mistake towards the end...

The first overhead section
* Minor change in the sweeper fight. Instead of using 3 potions, I'm now using 2 potions. See the sdakb for more details on this fight and some advantages of the 2-potion fight.
* The first room with Delia's art pieces went as well as I could have hoped for. Could have been faster, but the most important is to not have to chase down any of the art pieces on the left side with the hammer.
* Big time save in the second room with Delia's art pieces thanks to a new strat. I collected a health beetle on the way back, which was slightly out of the way. It's debatable how far out of the way one should go to collect a beetle or avoid damage, but other than potentially reducing the number of deaths over the course of the run, it's on average also slightly better to enter the next room with one hit than two.
* Overall, three rooms in this section that heavily rely on RNG and all of them went well. Not perfect, but at least according plan.

The second overhead section
* A minor mistake in the first auto-scroller section, where the d-pad was acting up during a diagonal press and I shortly got stuck. Ugly, but not a huge time loss (~0.5s).
* The (dreaded) fly room went very well. This is possibly the worst RNG-part of the run and there is almost always more time that can be saved. I'd say this was a top 5-10% completion of this room though. I even managed to re-collect the swatter, which I could use in the last room of this section.
* The mouse room is another very random room, almost on the same level as the fly room. Not as great result as the fly room, but I'd still say better than average for me and better than my 10:52 run.
* Overall, very good stage of the most RNG-heavy section in the run.

Graveyard
* New route through the first part of the graveyard and the tower for a time save of around 2.5s
* I missed a tight jump before the tower with a 2-frame window. I'm not completely consistent with this jump and this negated the time saved from improved routing.
* I barely missed the 4-attack/2-jump boss fight and had to activate another birdman scare and jump a third time to finish the boss off. Together with the mistake against the second boss, this left me with only birdman scare left, which unfortunately meant an additional time loss in the last level.

Third overhead section
The 1-potion fight against the sweepers comes down to RNG and this is where I had 12 (!) consecutive attempts on pb pace that didn't take out both of them. Other than that, this section is very straight-forward.

Afterlife
* After the execution mistakes against the 2nd and 3rd bosses, I entered this level with only one birdman scare left. This prevented my from death abusing to collect ticket #1 (see route in the sdakb), which would have saved ~2.5-3s.
* A mistake around ticket #2. I was meant to jump over the "magnet crane" and land on the platform to the right, but didn't get enough momentum and had to make another jump. Maybe 1s lost.
* With only one life left, another death would have triggered a continue screen. For that reason and because I already had taken two hits, I collected ticket #5 before #6. In my 10:52 run, I first collected #6 and then used a death abuse to collect #5 right before the end, which is faster.
* Very slow level. One unnecessary execution mistake, but most of the time loss came from execution mistakes earlier in the run.

Final thoughts
The time of this run is thanks to great RNG, which is then offset by execution mistakes. I would probably have been very happy with this time if it instead had relied on better execution, but worse RNG. I therefore certainly have an itch to get a better-looking run. However, with my recording setup being out of commission, I take it as a sign to stop here. There is also another aspect when there is this much influence on the time and success rate by RNG. If only looking at this particular attempt, better execution would have resulted in completely different RNG and most likely a slower time. So one can argue that even bad execution contributes to a pb in a game like this! (at least that's what I'm trying to tell myself right now...)