trivia212005's Feed

trivia212005
RSS
06-25-2020 at 10:27 PM
8 Comments
Rate this Entry

Future Speedrun Questions

Here's a good question for future speedruns.


Well first, once I get over 4000 credibility points, I am going to disqualify myself on many games so I can submit them to the right time.


Now here's the question. Do you suggest I should submit and re-submit my speedruns by the actual time (ex. 20:18.9) or should I still submit my speedrun to the nearest second?

ThanksJJT_Defender thanked this post
LikesJJT_Defender liked this post
Comments
  1. starcrytas's Avatar

    I would leave off the milliseconds.

    Example: A time of 20:18.9 would be submitted as 20:18.

    Milliseconds should only be used if the game displays it like on Mario Kart.

  2. Barthax's Avatar

    The historical reason for the seconds accuracy of speed runs at Twin Galaxies is because the technology simply isn't available to be perfectly accurate below 1 second's resolution. 30 FPS is close to 0.1 accuracy but still inaccurate to 0.1 sec.

    Old submissions (including retired members that submitted under TGSAP) do not have the option to become more accurate with their evidence. So where seconds-level accuracy is used already on a track, future submissions should also be seconds-level accurate to maintain the fair level playing field that is the backbone of TG.

    The issue of rounding will always be a factor when attempting time-based manual scrutiny - the number of decimals per second is not going to change that.

  3. Barthax's Avatar

    Quote Originally Posted by starcrytas
    Example: A time of 20:18.9 would be submitted as 20:18.

    I really don't get the rounding down culture. A 20:18.9 is almost a second slower than 20:18. Why credit someone with something they didn't achieve? Once the timer reaches 20:18, they failed to surpass 20:18.

    Anyway, that's one of the reasons I don't adjudicate manually timed runs anymore.

    Likesnads, Fly liked this post
    Updated 06-26-2020 at 12:48 AM by Barthax
  4. swaggers's Avatar

    This 100%. It should either be rounded up always or; 4 is down and 5 is up. NEVER always rounded down.

    Quote Originally Posted by Barthax

    I really don't get the rounding down culture. A 20:18.9 is almost a second slower than 20:18. Why credit someone with something they didn't achieve? Once the timer reaches 20:18, they failed to surpass 20:18.

    Anyway, that's one of the reasons I don't adjudicate manually timed runs anymore.

    Thankstrivia212005 thanked this post
    Likestrivia212005 liked this post
  5. swaggers's Avatar

    Also don't dispute your own scores for something that was allowed. Either there will be a global rule change or not. Don't hose people on something that was valid in the past.

    Thankstrivia212005 thanked this post
    Likestrivia212005, Barthax liked this post
  6. Tompa's Avatar

    20:18.9 is still a time under 20:19 and is therefore not a 20:19. The best is always to have an exact time, if such can be measured. Otherwise it is certainly a time under 20:19.

    Thanksstarcrytas thanked this post
    Likesstarcrytas liked this post
  7. Barthax's Avatar

    Quote Originally Posted by Tompa

    20:18.9 is still a time under 20:19 and is therefore not a 20:19. The best is always to have an exact time, if such can be measured. Otherwise it is certainly a time under 20:19.

    The reverse is not true, however. If someone is going for a longevity record the 20:18.9 has not reached 20:19 and should not be credited with 20:19. Similar for scores, if someone gets 20100 they didn't get 21000. For speed runs, 20:18 has been failed by the time 20:18.00000001 has been reached.

    Thankstrivia212005 thanked this post
    Likestrivia212005 liked this post
  8. Tompa's Avatar

    Which is why exact times should be used whenever possible. This is where the timing of TASes is superior to anything else. You have an exact frame count, anyone can download and view the input file to verify both the run and the time. Precise and impossible to fake/cheat.

Join us