United StatesAfter 4.5 years as an admin with site staff, I am resigning from site staff effective immediately.
I made site staff aware of my plans to exit about a month ago, and I will make myself available to site staff for consultation for some additional period of time. There are a couple ongoing items I'm advising on in the background. Please do not contact me to help with site issues any longer, I will probably refer you to this post. I no longer have site admin powers to handle these issues. Contact the full mods or Pac, depending on the type of issue.
I need to move on with my life, and I need less stress in my life. The ongoing conflict and disagreement in maintaining the site for the community has taken its toll on me over the years both physically and mentally. I've lost plenty of sleep over the years dealing with site conflicts, crisis management, and whatever else goes along with that. In recent months I've been having substantial issues with stress to a much larger extent than in the past. More recently I've been having moderate health problems for about 6 weeks that I'm still trying to get through. I should be fine in a couple months.
Being an admin on the site was never a position well-suited for someone like me. I don't gain any sort of happiness from resolving conflict or helping people resolve issues. I'm far happier in life as an engineer at my day job than as a customer service representative or a community manager. My primary goal was always to help organize a successful system for maintaining leaderboards for the community.
The early organizational phases of leaderboards, which I was involved with prior to speedrun.com had substantial failures. At that time I was extremely stressed out and disappointed with how things had turned out. I felt a personal obligation to help get the site going due to those previous failures. When I first came on to site staff, I was intending to help with the start-up phase and exit a few months after. I ended up filling numerous roles doing things that needed done to keep the site running, and here we are 4.5 years later. One regret I have is that I failed to help get involved many capable people who have reached out over the years
I don't claim to have been the best person for this job. I think I did alright. There are people out there who would've done a better job. There are a substantial number of people who would've done a worse job.
I've spent multiple thousands of hours of my life handling the day-to-day for the site, and I'm down about 80 dollars for server upgrades for a couple months a few years back. I've reviewed almost 10,000 game requests since April 2016, and many more prior to that. I handled almost all of the moderation requests and disputes up until the last couple years. This is something I did in my free time for over 4 years, and I never wanted to treat it like a job. I'm very cognizant that I've inherently burdened myself too much. I'm very burnt out.
I personally find public relations and community management to be very stressful. Community managers at various gaming related organizations generally don't last for more than a few years. The social media environment can be very toxic, and even aside from that, the hyper-connectivity of the social environment just puts conflict and disagreement on your doorstep daily. I find that most people I've interacted with are genuine people who have reasonable concerns. It turns out that reasonable people still disagree with each other frequently. There's just an infinite amount of negativity out there, and I think people should actively try to not contribute to it. I think we amplify the negatives in life all too often.
I think the site is still lacking various features to make life easier for staff, users, and moderators. Many problems that manifest as people problems are attributable to system shortcomings. These things obviously take time, effort, and prioritization. I feel some responsibility for failing to solve some of the more complex organizational issues with the site that still exist today.
It is my opinion that the site needs to work towards functioning more like an organization. The site is extremely large compared to the size of staff, and it needs more tools for people to handle issues at lower levels. The site needs to be careful and deliberate with the design decisions it makes, because small changes can have very significant impacts on moderators and users. As a primarily community driven website, I think the site should only be in the business of doing things that the majority of users want it to do.
I'm very thankful for everyone that has helped the site get to this point. Pac, it's been a wild ride. There's so much more to say than that, which I can't adequately put into words. Kabuki, Lighnat0r, and more recently Volvagia, thank you for helping solve some serious and complex problems over the years. To the rest of staff, thank you for carrying forward, and I sincerely apologize for any additional burden caused on you by my exit. To the community, thank you for helping build the site to what it is today. Through the site, a substantial amount of communities have greatly expanded to connect people with common interests and to provide a wealth of knowledge on various games.
I hope this post meaningfully articulates my personal experiences, and provides some understanding of why I need to leave staff. It has been an extreme learning experience for me personally, and I wouldn't be the person I am today without having been through this. I'm not going to go so far as to say I'll never rejoin staff, but I will say that I currently have no future intentions of rejoining staff. Out of respect for the rest of staff, I intend to minimize my public discussion of my opinions on site direction going forward.
As for me, I'm working towards improving my lifestyle. I intend to substantially reduce my social media presence going forward. I'm finding it unhealthy for a person like me. Spending half my life on the internet hasn't been the best for me. I'll still be around playing games with friends and showing up to events and things like that.
I'm not really looking for thanks or anything, I'm just seeking to inform and offer insight. I'm leaving town for a while later today due to additional life issues.
New pages now exist with substantial text describing how the site handles things.
https://www.speedrun.com/rules https://www.speedrun.com/moderationrules https://www.speedrun.com/requestgame https://www.speedrun.com/gamerequestrules
This is a work in progress to document all of the confusing things that go on around here. There should not be anything particularly surprising in there, but there is some clarification on a few gray cases we've run into over the years.
If you have feedback, questions, or additions for the documents, or if you disagree with the wording of something, please let me know.
This current set of documents does not attempt to cover the topics of FAQs, Leaderboard Setup, Getting Started, Recording Guides, Speedrunning Glossarys, Recommended Emulators, Recommended Console Guidelines. Those may be separate pages in the future.
The focus of this is on the rules and the game requests right now.
A few game moderator accounts were compromised on April 1, in a very similar manner to what happened back in November. About 3 to 5 accounts moderating a few prominent boards were compromised. The cause is still understood to be a few users using or re-using passwords that were compromised from other sites years ago. We implemented password re-use stipulations and reset everyone's password again. We also rolled back the database to shortly after 2019-03-30 23:37 US Eastern. The description of November's issues is here: https://www.speedrun.com/The_Site/thread/l1r3e
If your password was a password that was not previously used on a compromised website, your account was never at risk. We're acutely aware that resetting passwords and rolling back the database is a disappointing approach. In order to get the site up today, that's what we went with. We have every intention to implement further safeguards against this method of account compromise.
The site's leaderboards as a whole are only as secure as the least secure user, and we're very aware this is a problem. We intend to implement further protection against known compromised passwords from other websites. We currently intend to utilize the pwnedpasswords API, which offers built-in anonymity in password match requests: https://www.troyhunt.com/ive-just-launched-pwned-passwords-version-2/
It is our current understanding that every compromised password was gained from that list. We're looking into the logistics for implementing 2-factor authentication as well. 2FA could potentially be mandatory for game moderators, if that is seen by the community as acceptable. If it's optional it only protects people who want to be protected.
Hours prior to all of this, a couple users also took the opportunity for April Fools to trash a couple boards. Those were also cleaned up by the database rollback. We'll address that facet for April 1 2020. There is some additional work to make damaged board cleanup much easier, such as site staff being able to restore from backups easily. Restoring a single board is currently a very manual process that only a couple people know how to do currently.
I brute forced the encounter counter behavior. It's essentially exactly what I had speculated in the past, but I had not verified it previously.
I think whether or not you get into the battle has to do with an internal formula checking if the game can lay down the battle grid or not. I've never understood exactly how this works, but it's well-known that sometimes you won't get into an encounter up against certain walls.
50 step units is about 0.83 seconds or about 5 Brian steps. Brian's speed is 2 step units per 1/30th of a second (one frame).
Exactly every 50 step units it rolls for EncounterChance = (Step Units / 2000)
50 step units = 0.025 = 2.5% 100 step units = 0.05 150 step units = 0.075 200 step units = 0.1 250 step units = 0.125 300 step units = 0.15 350 step units = 0.175 400 step units = 0.2 450 step units = 0.225 500 step units = 0.25 550 step units = 0.275 600 step units = 0.3 650 step units = 0.325 700 step units = 0.35 750 step units = 0.375 800 step units = 0.4 850 step units = 0.425 900 step units = 0.45 950 step units = 0.475 1000 step units = 0.5 1050 step units = 0.525 1100 step units = 0.55 1150 step units = 0.575 1200 step units = 0.6 1250 step units = 0.625 1300 step units = 0.65 1350 step units = 0.675 1400 step units = 0.7 1450 step units = 0.725 1500 step units = 0.75 1550 step units = 0.775 1600 step units = 0.8 1650 step units = 0.825 1700 step units = 0.85 1750 step units = 0.875 1800 step units = 0.9 1850 step units = 0.925 1900 step units = 0.95 1950 step units = 0.975 2000 step units = 1.0
Cumulative math: P(A or B ) = P(A) + P(B ) – P(A)P(B )
0.025 + 0.05 - 0.025 x 0.05 = 0.07375 0.07375 + 0.075 - 0.07375 x 0.075 = 0.1432
Since the counter resets, it's harder to exactly calculate the chance of two encounters over a set distance. The following covers chance of at least 1 encounter.
Cumulative: 50 step units = 0.025 100 step units = 0.07375 150 step units = 0.1432 200 step units = 0.2289 250 step units = 0.3253 300 step units = 0.4265 350 step units = 0.5269 ~50% 400 step units = 0.6215 450 step units = 0.7067 500 step units = 0.7800 550 step units = 0.8405 600 step units = 0.8883 650 step units = 0.9246 700 step units = 0.9510 ~95% 750 step units = 0.9694 800 step units = 0.9816 850 step units = 0.9894 900 step units = 0.9941 950 step units = 0.9969 1000 step units = 0.9985
Just documenting this. In Jungle 1 after the second monkey, you can zip up the middle rising platform by holding jump. It's probably possible in a lot of other places. (Maybe the technique is already used in the run, I admittedly haven't watched it.)
New route: (Foot power updated Jul 29, 2018)
1: Go up to floor 33 2: Walk on glass for 8 seconds. (Lose 4 Foot boxes plus ~240/256 sub-units) 3: Take stairs 11 times. 4: Arrive on floor 32 at 39 seconds on lock two, hear the elevator 5: Go to floor 31, shoot a bullet to lure out Heinrich faster. (hope he's in the second room on the bottom), grab the first aid, get the detonators. 6: Take stairs to floor 35 7: Kill roof key crook (hope he's in the room on the right without checking the door.) 8: Grab the C4 near the roof door. 9: Go to roof, place C4, run to the top center.
Important: Do not hold the A or B button during the following cutscenes, you will lose about a quarter second while it loads the cutscene: -- John checks his pistol to find 15 bullets in the clip. -- John uses the First Aid Kit -- Ending victory.
Notes:
- It is probably a few frames faster to keep the pistol instead of grabbing the machine gun if you can manage low ammo, but it is very minimal.
- Grabbing the first aid appears to be just slightly faster than not grabbing it.
A mediocre TAS of this was able to get 1:52 from power-on.
I think I have better luck with RNG when I alternate 33 and 32 instead of 32 and 31, but this could be nonsense. I still don't understand how the enemy spawns in this game work. This is something worth looking into with the debugger.
More thorough discussion:
This route change does 2 things from the current route: 1: Don't go to floor 35 early. 2: Walk on glass on floor 33.
The most important time point in the run is being on the 32nd floor at around 39 seconds remaining on lock two. All current top 5 runs arrive at this time point currently. If you are a little late it's okay. If you leave to the 31st floor early, Heinrich will be in a bad spot, coming up from the elevator.
If you have the radio, this time point is where Hans says "Heinrich check the 31st floor.", which triggers an elevator with Heinrich whether or not you have the radio. By being on floor 32 when the elevator starts, when you go to floor 31 Heinrich will be in a semi-random spot. It seems as though if the elevator starts while you're on the 31st floor, Heinrich will be by the elevator which is very slow. (Using the elevator is slower than the stairs.)
Note that there is also a way to accidentally cancel the Heinrich event by triggering two events on one staircase. Going from 40 seconds down past lock two will trigger two events, and only the second event occurs. Heinrich never shows up with the detonators, or at least not for quite a while, I can't recall.
Approximate Game Time consumed per staircase at given foot power increment lost: 0: 15 seconds 1: 19 seconds 2: 22 seconds 3: 27 seconds 4: 31 seconds 5: 35 seconds
Walking a staircase cycle takes around 188 frames (3 seconds) for 1 cycle (slightly more or less depending on foot power.)
Need to drain about 375 seconds of game time, which is 25 cycles at foot 0 20 cycles at foot 1 17 cycles at foot 2 (3 cycles is 9 seconds) 13 cycles at foot 3 (4 cycles is 12 seconds) 12 cycles at foot 4 (1 cycle is 3 seconds) 10 cycles at foot 5
Walking on glass loses 8-11 foot power per 3 frames, or about 90 frames roughly to lose 1 foot power. FeetUnit = 0x0797 Feet = 0x0798 --8 feet bad, 0 feet good
1: As discussed on Discord due to the recent run, Low% is kind of a bad label, and the rules are a bit ambiguous. Since people seem to like both misc categories, I think we could add "minimum collection" and "minimum level" It's a bit of category bloat, but there aren't exactly a ton of ways to run this game, so I tend to think it's fine.
2: I'd like to see "Uses OoB" and "Uses Spirit Dupe" combined into either "Uses Glitches" or "Glitchless". I don't think having both indepdently adds much value to the board. I'm probably biased by having a glitchless run, but I feel like that's something viewers would still want to see.
3: There will probably be a "Uses RNG Manip" Flag eventually. Theoretically most of the early game for a spirit dupe and possibly through Solvaring could be maintained with RNG manip. Additional, crazy things could probably be done from a save in Brannoch, though I haven't investigated that much yet.
The question of emulators was brought up in Discord.
Referencing some other SNES games on the site (particularly Super Mario World https://www.speedrun.com/smw/thread/8w7yc ), I think the following is a good start.
Suggested:
- bSNES Compatibility Core (Bizhawk, Higan, lsnes)
- SNES9x v1.5x and higher
Less preferred but probably acceptable:
- bSNES Performance Core (Bizhawk, Higan)
For this game, the big things to pay attention to for time accuracy are loading time between rooms, lag after midboss 1, and lag in the last part of stage 5.
I think emulating on another console such as an Xbox or a Wii introduces too many unknowns into the equation. In order to understand the legitimacy of a run or investigate nuances of the effects of the emulation, one would potentially need to fully recreate the hardware setup. Most people playing this game either have a PC or a SNES.
That's just my personal opinion. If other runners feel differently, please comment.
I'm putting the moderators from Bloons Tower Defense 5 on this game since we've had issues before: https://www.speedrun.com/btd5/thread/s8gn4
I do not think recreating a known previously conflicting situation by creating the game with the moderators as currently listed was a good idea, and given that one of the users involved doesn't moderate any of the other games here (and in fact intentionally removed themselves at one point), I'm adjusting moderation accordingly.
If the other series moderators have a different opinion on this, feel free to get a hold of me or one of the site full mods.
We're going to be issuing bans for users who submit fake runs to "test" moderators. I see this behavior as a detriment to the quality of the site's leaderboards and to the people moderating the site. Letting a few people get by doing it would just encourage more people to engage in this behavior. If people do it on an alt account and we find out about it, we're going to issue a ban on the main account.
I don't wholly disagree with sentiments of encouraging people to be more thorough with moderating, but I find "testing moderators" to be a highly detrimental way to go about it.
I'm aware a rules page would help clarify the expectations. I also think that the lack of a formal rules page doesn't mean people can go around engaging in detrimental behavior.
I reviewed Squid's 43:15 run for time loss primarily to gauge how much further this game could be pushed. Squid is definitely playing better than I've ever played, and this run definitely sets a pretty high bar for anyone looking to take the record. Even though this run is quite good, there is still about a minute to be squeezed out of this game with the current knowledge.
Author's note: I'm estimating tenths of seconds purely based on experience. I might be off by a few, but on average I should be close. I didn't look at the "gold splits doc"
Bomber: Battle Room: Near Optimal Hyper Room: Lost 0.5-0.8 due to mediocre bomb throw. Probably lost 0.1-0.2 due to no "Yashi slide" on the stairs. Heavy Room: Lost 0.3 or so at the shield barrier. Sky Room: We still don't fully understand the final pillar timing. There's probably a way to do this about 1 second faster, but it's not known. Blue Cave: Near Optimal Hole Lake: It's possible but tricky to get one more boost here and save about 0.2 Red Cave: Uses "slow jump" setup, which is about 0.5 slower, but much more consistent. Dark Wood: It's possible but highly inconsistent to get no ledge grab here and save about 0.8 seconds. Nitros 1: Near Optimal Clown Valley: Near Optimal Great Rock: Second jump was slow, losing about 0.3. The bombing of the platform was slow, losing about 0.4. There are probably faster ways to handle the first fire pillar for 0.4 or so. Fog Route: No skip, loses 6 seconds I think? The angles at the end were kinda bad, losing about 0.3 Endol 1: Near optimal
Planet Bomber: Squid: 7:42 Clean Run: 7:34 (+8) Reasons: 2 seconds could've been saved through cleaner play, and 6 seconds from Fog Route Skip. There may be another 1 second to save from Sky Room, but it's not clear. Strong Bomber, but no Fog Route skip.
Primus: Groog Hills: Near Optimal Bubble Hole: Uses left side at the end instead of the right, losing probably 0.5-0.8 seconds. Erars Lake: Missed corner boost before Manta. Got the two corner boosts after the warp, nice! If you want to be risky, you can corner boost in the holes the orbs spawn out of. 0.2 for the first corner boost. WaterWay: I think the last "river" before the warp you want to drop the bomb going off the ledge to save 0.1-0.2. It's possible but rare to skip the ledge grab before the warp for about 0.8. Rock N Road: Missed an attempted corner boost and got bad dice position, losing about 0.6. Killing the door and specifically entering the exit was quite slow, losing about 0.4 Water Pool: Got caught on the wall at the beginning jump losing about 0.3. Pretty good otherwise. Warp Room: First try warp room skip. Nice! You could setup a few tenths faster by landing better off the wall (0.3). I think it's slightly faster to take the hit from the barrel at the end (jump into it), but health management is also important. You would've died in Dark Prison if you did. Dark Prison: Bad bomb drop, lost about 1 second. Hit self on enemy, lost about 1 second. Used aggressive ending, good. Nitros 2: Let him get away, lost about 2 seconds. Killer Gate: Having to grab the health here probably loses about 0.5, but it's really risky not to. I think you manipulated the cannon, but angled in a bit too much. Getting hit lost about 0.5 seconds. Spiral Tower: Near optimal. Snake Route: Near optimal. Baruda 1: Going to say you could saved 1-2 seconds, but this was a really good fight.
Planet Primus: Squid: 16:40 Clean Run: 16:25 (+7) (+8 Bomber) Reasons: Dark Prison and Nitros lost about 3 seconds to mistakes, about 4 seconds were lost due minor things or being too careful. Strong Primus with a couple key mistakes.
Kanatia: Hades Crater: Near Optimal Magma Lake: Near optimal Magma Dam: Lost about 3 seconds on skip. It is possible to get a "super skip" by being near pixel perfect, which saves a couple seconds by letting you drift into the exit while falling. I've gotten it like 5 times ever. Just pointing out this small detail: After the yellow heart, you can gauge the enemy position to see if you can jump in front of him or not to save a few tenths. In this position, you could not. Crysta Hole: Lost about 0.2 at the beginning, the platform at the end isn't spawned until you walk right for 1-2 seconds. Death Temple: You can bomb the switch a little faster by dropping the bomb while running to save about 0.3. Death Road: First Section, I'm curious if the method used is faster the going to the left and corner boosting. I think you lost about 0.3 here. 0.2 seconds in the second section from being slow off the portal and using not great angles through the middle. Another 0.4 lost on the door. Float Zone: Lost about 5 seconds. 3 seconds due to missing the switch, 1 second due to missing the enemy, and about 1 second in the last section. Aqua Tank: Ending could be a little cleaner, 0.3. Aqua Way: Near optimal Nitros 3: Pillars lost about 0.5 seconds. No double hit, but it's really hard (4 seconds) Lost about 3 seconds on the last hit. Hard Coaster: Pretty good, missed the first bomb throw at the end losing about 0.3 seconds. Dark Maze: A little sketchy at the end, losing about 0.5. Mad Coaster: You can jump a little sooner at the end, but it's scary for 0.5. Move Stone: I'm going to pretend the 2 push isn't a thing. It's dumb. For the third block, I think you should bomb the pillar when switching sides of the block for 0.3. The last pillar I don't think you meant to do that, but you might be onto something. Regardless, whatever happened lost about 0.4 or so. Bolban 1: Quite good but lost about 1 second.
Planet Kanatia: Squid: 29:00 Clean Run: 28:25 (+20) (+15 Primus/Bomber) Reasons: 11 seconds were lost between Magma Dam, Nitros 3, and Float Zone. For 4 more seconds, Nitros 3 double hit is important but difficult. The other 5 seconds are little things but notable things. Kanatia had some mishaps, but it's a difficult portion of the run.
Mazone: Hopper Land: Hesitated for a second before the cliff jumps, and then if you angle properly in at the end (slight angle up of straight right, you can hold up. This probably all only lost 0.3 or so. Junfalls: Near Optimal Cool Cave: Small hesitation jumping on the first iced guy for 0.2. Snowland: Near optimal Storm Valley: Near optimal Snow Circuit: Beginning is not quite right. Hold right, let go of right to straighten out slightly, jump hit B, hold left. Should save 0.5-1.0 seconds. The angles throughout the level weren't great. Time is lost when you turn too much, I'd say about 0.3. Heaven Sky: Missed a corner boost at the beginning for 0.1. If you aim for the bottom of the yellow pentagon on the boss, the bombs will hit sooner instead of waiting for the blast radius. I forget if you can accidentally hit his heads like this or not. I'd say 1 second can be saved there. Eye Snake: Near optimal. Nitros 4: Near optimal. Air Room: Slow out of the fan. Count 6 pillars and hold the direction earlier for 0.5. The method of landing on the center may or may not be faster, I never timed it, not going to say time was lost there because I don't know. Zero G Room: Letting go of directions when falling onto the CD will cause you not to skid forward as much saving about 0.4 or so. Mirror Room: I think it could be about 0.4 seconds faster on the right side of the stage. Particularly just before bombing the switch and then the jumps right after the portal. Bumping the wall after the portal means you could've lined jumps up better. Natia 1: No Cronus double hits and did not knock Natia off the level. I think you can only do one of those two things, and I'm not sure what's fastest, but there's about 3 seconds to be saved on first phase probably. Lost 4-5 seconds on Natia phase. We need to find a better way to do this.
Planet Mazone: Squid: 38:11 Clean Run 37:25 (+11) (+35 First 3 planets) Reasons: The Natia 1 fight can probably be about 7 seconds faster. 4 seconds of minor stuff. Good Mazone, but the Natia 1 fight has a lot of room for improvement.
Garaden: Endol 2: Near optimal. Got the eating animation, which is important for style. Baruda 2: Maybe 0.2-0.3 to save, but pretty good. Cronus: Could be 5 seconds faster with 3 double hits, but it's difficult. Nitros 5: Got the double hit through some absurd series of events. Basically the best fight you could get. 9090 lol. Bolban 2: Missed a hit, losing about 4 seconds. Natia 2: Lost about 3 seconds to not respecting the whip. Pretty good though. Bagular Phase 1: Pretty good, but a little slow on throwing bombs after he teleports. I think about 1 second can be saved. Bagular Phase 2: Pretty good, not sure about the TV meta these days, maybe 2-3 seconds to be saved. Bagular Phase 3: Mediocre hits on the left-side arm, losing about 2 seconds, but overall pretty good dealing with the arms. Missed his face once losing a second. That last hit was something else. You basically almost died.
Planet Garaden: Squid: 43:15 Clean Run: 42:11 (+18) (+46 First 4 planets) Reasons: The missed hit on Bolban, missing Natia, and not getting the extremely fast Cronus kill lost about 12 seconds. The Bagular fight could be about 6 seconds faster maybe. Solid Garaden overall, but a couple key issues lost a chunk of time.
Note on the last hit: The bomb barely leaves Squid's hand, hitting the missile without hitting him. The missile is barely in front of him. Then a few tenths of a second later the radius hits Bagular. If the missile was slightly closer or he threw a few frames later, I think he would've died.
I'm just posting as a person who has played this game a bit, not really here in admin capacity, so the following is just my personal opinion.
I think the categories would make more sense as follows:
Categories: 2-island, 3 island, 4-island Subcategories: Normal, Hard Variable: Squad Type - (Custom, Steel Judoku, etc)
You could separate custom squads and assigned squads into subcategories as well I suppose.
The intention of subcategories is to better layout the redundant combinations, so that the category tab doesn't become bloated.
The IGT situation is kind of unfortunate at the moment. I abused it to get the lightning mech achievement, lol. Not sure what to do about that.
I don't really have a strong opinion on tracking carried over heroes or not. Keeping them disallowed might be preferable due to the time involved it setting it up.
Feedback from some other NES runners is that the categories for this game are really arbitrary. I played the game for a bit and I agree with them.
Most games like this contain categories for the default hi-score listing, a loop, and maybe some more difficult to achieve goals. A loop isn't really reasonable in this game as it's a total autoscroller and the endpoint is not precisely defined.
When a run is submitted as a 10k, a 20k, and a 30k run, and the 5k runs are tied to the frame, it starts to look a little ridiculous.
Also the scores are being duplicated into the "Level Leaderboard" which seems relatively pointless.
The categories for this game should be more reasonable goals like 20k, 100k, 500k.
The goal with these games should not be to create as many short categories as possible. 5k is a ridiculous goal. Why not 1k? Why not 40 points?
Can we clean this up?
There's at least one run that's been pending about 4 weeks for this game. It would seem like this game could use another active moderator or two.
Do any of the users that currently have runs want to help moderate?
They should be fixed now. The posts that were attempted to be made within the period when they were broken are mostly gone. If you really need one of the posts for some reason, it's in the site's logs and relatively easy for us to retrieve.
A bunch of NES runners have voiced opinions to me that the recently added misc categories suck. Using a debug cheat as fast as possible to do an arbitrary thing seems like pretty awful category bloat, and people get really annoyed at the quality of the boards resulting from "adding everything you could possibly do with the game". It doesn't seem like there's value in these categories other than this week's 60 second meme.
It also seems to be confusing that beating the game is listed to the right of beat level 20. Viewing the runs I don't really understand why level 20 and 50 are a thing, since they don't really appear to beat the game or be a notable stopping point.
The first thing my eyes go to is a category that doesn't beat the game; if I look further I find hilariously arbitrary 1 minute categories. Can we add "beat level 1"?
There's some balance here between the board being what the runners of the game want and the board being something more presentable to the people using the site. I don't think partial game categories are the worst thing in the world if people are racing them and things like that, but they shouldn't replace beating the game.
I wrote a Dragster lua script for visual feedback while practicing or for recording a stepmania chart to play against.
Please use responsibly. 5.57 needs to be legit. Using the script in a direct manner would invalidate the run.
EDIT: Newer script here: https://pastebin.com/FrZM5FgD
Also I reviewed darbian's 5.61 video:
- His first shift was on the correct frame, but he held the gas one frame too early at the start, so he already missed 5.57 off the start.
- As a result, all of his gear 2 through 4 shifts were about 0.1 seconds too early if he was doing them visually or by audio.
Emphasis: The first three shifts are everything, and even if you get a 5.61 you might be doing them relatively wrong.
The script is pretty good at telling you if you got the first few shifts right (green) or wrong (red) Wrong makes a 5.57 impossible. You will find it is quite difficult to get them all correct.
To my understanding, the game only reads input every other frame at 60 fps, correct me if I'm wrong. This means every button must be held two frames. This is why it seems like inputs drop sometimes.
I'm changing moderation for this game over to a well-known community member with knowledge of the game.
I'm working with some of the site's NES runners to turn over certain games to active users well-involved in the NES speedrunning community and meticulous with timing runs.
This change is mostly intended to improve reliability of the data on boards.
