Additionally, another aspect to consider is that having a real timer going instead of /played can help admins detect splicing/clips/changes in a way a /played timer cannot.

I disagree with this, Balaar. I think you're overrating the importance of a timer - IMO, it's just a tool for the runner and doesn't assist with verification whatsoever. A bad actor could easily overlay a timer over a spliced run with editing if they wanted to.

The only effective method of catching splices is by looking at an audio spectrogram of a run, but I think that you'd agree that it would be an unnecessary verification step, even for top runs, since PoE is a smaller, tighter community where people actually trust each other.

I'm not going to speak towards the livesplit extension because I've never used it before and don't know if you could get banned for it like Wafu is saying. However, using /played specifically has two problems:

#1. Most runners thinks that PoE short categories are best run without breaks, in order to emulate the racing environment. This means no changing filters, checking poelab, pob'ing, or taking a breather on the character select or login screen.

#2. We have no way of knowing if /played is accurate (and according to JSTQ, who tested it a long time ago, it's actually inaccurate).

I could easily test #2 by doing my own framecounting in my spare time but #1 is hard to argue with so I've never bothered.

So, in my opinion, PoE could never use IGT, at best it would have to be RTA with load times removed via the livesplit extension. Of course, the livesplit extension doesn't solve the issue of ICT, which has historically always been bs, but that's not an argument to not implement it if it's actually effective at removing hardware-dependent loads and doesn't get you banned.

Thanks for clarifying what the cause of the discrepancy is. Now I see my mistake with submitting the time for MqcCheeze's level 2 run (I should've used the final level 1 time as the start time, not the initial level 2 time, otherwise the IGT ends up being +0.02s compared to the frame count time instead of +0.04s).

Also, I'd like to plug my early game racing discussion discord, The Knights of the Round Table! We're currently in the middle of discussing this topic right now, so feel free to hop in if you'd like to participate! https://discord.gg/4wartpssYV

Proposal: Make the timing for ASPLAB speedruns end on whichever last relevant objective is completed (res penalty if it's Kitava, clicking "ascend" on the plaque if it's lab, consuming the last skill book if it's a skill point)

A while ago, I wrote a wall of text arguing for this, but at the time I felt pretty entitled to having my opinion heard so I came off like an asshole. I'd like to apologize to anyone was offput by my behavior. I admittedly haven't went back and read what I wrote, or I'd probably implode into a black hole of cringe.

These days, I don't really have as strong of an opinion and I'd be fine with the rules staying the same, but I figured I'd still write this so that a record of an actual rational and measured argument exists.

The reason why I think runs should be timed this way is because it makes more sense given the context of how PoE is played. It's a character building game. Getting to your character to x point, regardless of the order, should be the stated objective in a game like PoE, in my opinion.

Others have mentioned that finishing the run by defeating the big bad, Kitava, is thematically appropriate for the speedrun. Indeed, in a category like SM64 120 Star, it's mandatory to collect 119 of the 120 stars before entering the final stage, getting the last red coin star, defeating Bowser, and rolling the credits. If you don't follow this order, your run isn't a valid 120 Star run. However, I don't feel the need to ascribe the same importance to Kitava that SM64 ascribes to Bowser. In PoE, defeating Kitava is merely an unlock (namely, A6 or maps) in an infinitely-scaling game.

It should be noted that I'm not particularly interested in making the run easier, flexible, or more forgiving. There's nothing inherently wrong with disqualifying a run that forgets a skill point, as there's plenty of downtime where you can type /passives. There's also nothing inherently wrong with enforcing an order of merc lab -> A10 Kitava. However, given how character progression is the central part of PoE, particularly in ASPLAB (in a boss rush, defeating the boss is undoubtedly the central goal - it's literally in the name), I feel like these restrictions, while not inherently flawed, are unnecessary and could be removed.

As a final note, I also think races (if they ever happen again lol) should be changed in the same way, so that in my ideal world, speedruns and races are consistent with how they are timed. You could argue that it makes the spectator experience worse if it's possible for two players to both be on Kitava, but progressed to different degrees. However, the optimal order is to save Kitava for last anyways, so this won't be a problem in 99% of situations, especially with top players, who usually won't forget anything.

Response to the counterargument: "Oh, so you mentioned how the category is about creating a map-ready character, like during leaguestart? Then why aren't capped resistances required then? What about HP, bestiary progress, bench crafts, etc? The way you are redefining the category opens it up for more arbitrary requirements like those to be added."

Pre-emptive response to the response to the counterargument: I don't LITERALLY mean the category should be a "map-ready" category, I just think completing the skill points, labs, and final boss kill in any order just captures the spirit of the run so much better, and it's hard to perfectly describe why.

Obviously replying very late, apologies for reviving a dead conversation, but I completely agree with Timmy here and I thought I'd give my two cents, because it seems like the initial rules (which I disagree with) were kept in place.

I see two arguments for why you'd want to make people do all skill points / labs before Kitava:

The verification argument: The current ruleset ensures that verification occurs at a single point in the run (i.e. you see Kitava dying, /passives being typed, and /deaths being typed all around the same time, with the /passives always having the correct amount of skill points)

Counterargument: Even if you were allowed to do skill points / labs after Kitava, you would still only have 2-3 points of verification in the run at maximum. It would still be a very streamlined verification process.

For example, let's say you killed A10 Kitava first, then did Merc Lab, then realized you forgot to pick up the Dweller skill point so you do that last. All you would have to do is type /passives and /deaths after the Dweller skill point, and then the verifier can just go through the video and look for the Kitava kill and the acquisition of the merc lab ascendancy points. This doesn't suddenly become a situation where the verifier has to watch the entire run, it's still very fast for them to go through.

The philosophical argument: The run should end with killing the final boss, and there is a great deal of precedent for this in other games. For example, in Super Mario 64 120 star, the run always ends after killing Bowser in the Sky, even though it could be viewed as an "objective collection" category.

Counterargument: This is a way stronger argument and harder to counter, but I have always distinctly felt that 100% categories are purely about getting your character to X level of progression, with the order taken to get there not mattering. I guess this kind of speaks to Timmy's point about how All Skill Points + Lab essentially mimics league start. Whatever route you take, you are producing the same end product: a map-ready character, and it just makes so much more sense in my head for THAT final product to be the end goal of the run.

I actually think getting the two skill points from Lani in the A10 run SHOULD be required, because in my mind that's absolutely a requirement to get your character to the "maximal" level of progression that you can obtain from all ten acts of the story.

I feel like I'm not arguing this point very well and it frustrates me that I can't do it in a more logically bulletproof way, but I guess that makes sense because it's a purely philosophical disagreement lol. Maybe I'll come back and add to this once I can better express my thoughts.

Sorry for the wait. I've updated the categories and rules and changed the times on the level 1 leaderboard, and over the next few days I will update the other three leaderboards.

What do you guys think about changing IL category timings to start when the level fully loads in (instead of on first input or touching the ground) to be more consistent with full game timing? This would involve:

1. Retiming all IL runs (which I'll volunteer to do since I haven't logged in to verify anything in a while)
2. Unverifying all IL runs that do not contain any frames of loading into the level
3. Adding a fifth IL category, level 3 with the floor clip allowed, so that the WRs in level 1, level 2, level 4, and this new category can be added up to create an accurate community SOB for full game runs

This timing change would not affect/penalize full game runs on the leaderboard that do not show loading into level 1, because timing rules were less clear back in the day, and the +0.33s penalty given to those runs is completely fair and accurate as far as I know.

I feel like whether you consider PoE speedruns to have to be deathless or not is based on your personal perspective. When I was first learning PoE the official race seasons were all HC races, and once race seasons ended the people who were still around like Alkaizer, Helman, Fitegarr, and DeadandDoom were still doing purely HC speedruns. So when I started doing my own runs, it was natural to consider every run where I died (which was a lot) to be null and void.

However, I can understand that new players without the same experiences as me might not share the same perspective that I had. In a vacuum, I don't see any reason for SC runs to be unofficial, and in fact I don't see any reason for SC not to be PoE's main category since it's technically faster and this is speedrunning. The only reason why I personally advocate for deathless-only is because of the strong, prevailing "deathless culture" in the PoE community. The truth is that right now, SC runs are incredibly hard to share or brag about. If you post an SC speedrun on reddit without explicitly specifying that you died, people will accuse of deception; if you do specify that you died, people will trashtalk you and tell you how avoidable your death was. Because I've been shaped by this culture I've ended up regarding SC runs as valueless and I'm sure some of the other people on the mod team will tell you the same.

Anyways, this turned into more of a ramble about the Path of Exile culture than an actual strong argument for deathless-only but I hope you now know where some of us are coming from. There's a case to be made that we should just reject the culture and display SC runs anyways but that is something that I'll have to do more thinking on.

I'd also like to announce that PoE speedrun.com now has its official theme song! You can find the link here: https://soundcloud.com/zorig-urangoo/sptsmvqpvqj4

https://docs.google.com/spreadsheets/d/1-STHtr0WET-paFWklDRRsfasz2-3cg1wHHXM9zCh1KY/edit?usp=sharing

Negative start times either means the runner's timer was started late or it was a situation where 0.33s needed to be added to the time (per the special case in the rules).

If the start time is high it probably means the run didn't have a timer so I just used a livesplit alongside recording the video to get accurate times.

ok this is taking longer than expected and I don't feel like highlighting a 2hr+ vod so I'll just post the spreadsheet I'm using when I'm done instead

I'm confused as to what you mean, there are categories for each of the four levels

Now that b9gus has beaten my WR my motivation to stop being lazy about this game has increased so I will be retiming every run (as well as Mamey_Edal's run in the queue of course) onstream soon to ensure the accuracy of the leaderboard. Will highlight it and post it here once I'm done.

There's no trick to it that makes it consistent. Just try to focus on hitting the ground completely vertical with as little left/right hammer movement as possible (thus maximizing its vertical velocity). Level 3 is basically entirely luck-based once you start using this glitch unfortunately.

Apparently there was an outdated guideline here so I updated the post for correctness. The change made is that runs ending in A1 will not show milliseconds except in the case of a tie.

The handcam video is appreciated but it doesn't change the fact that his supposed "wr" video has a weird anomaly in it that makes it look like it's spliced.

For the Chinese players- I think the next best step in verifying ccfst would be to get him to screen record and record a handcam simultaneously and then play them back to back so we can see both perspectives of a run. Turn the subtitles on as well. Also make sure there's no cuts in the handcam video between him finishing the run and showing what windows he has open, otherwise it doesn't prove that he isn't just watching and replicating a recording (in this video there's a cut at 1:31).

It does appear like ccfst is a pretty skilled player, it's just a matter of whether he's cheating to exaggerate that skill. If by some offchance he is actually legit I wouldn't worry because he'd probably get verified EVENTUALLY as long as you continue sending proof like this.

Hi, here are some tips for submitting that will help make the verification process smoother. Note that they are completely optional to follow.

• Runs will not have their milliseconds displayed on the leaderboard except in the case of a tie.

• Right now, Softcore is being treated as a combined category where the 'scuffed' Softcore runs are included. As a result, deathless runs should be submitted under both HC and SC categories.

• This will generally only apply to A1 runs. If you want to find the exact time of your run you can change frames in a Youtube video by using the < and > buttons. However, you don't have to do this because the moderators will find the exact time for you anyways.

I think it's about the same speed as the optimal tree strat, my best wall climb lvl 3 is 5.80s and my best lvl 3 with the tree jump is 5.58s