Username:
B
I
U
S
"
url
img
#
code
sup
sub
font
size
color
smiley
embarassed
thumbsup
happy
Huh?
Angry
Roll Eyes
Undecided
Lips Sealed
Kiss
Cry
Grin
Wink
Tongue
Shocked
Cheesy
Smiley
Sad
1 page
--
--
List results:
Search options:
Use \ before commas in usernames
Game Page: http://speeddemosarchive.com/MarbleMadness.html

Marble Madness (Any %) (Single Segment)

Verifier Responses

Quote from bassdeluxe27:
Cheating: None detected.  The player also has a Twitch channel that he regularly uses, so he's legit.

Play quality: I'll admit, I'm going to be brutal here, but that's because I'm extremely familiar with this game

Stage 1: Well played.  A perfect run could save up to half a second on this, but anything with a 55 in-game timer is very good.

Stage 2: This is pretty much perfect.  Very well done.

Stage 3: First half is perfect.  The pause before going down the final ramps at the finish is the largest time loss in the run, and it cost nearly a second.

Stage 4: Extremely good.  Could probably save half a second at the end with perfect execution, but that would be extremely precise.  The runner's comments make this out to be a bigger mistake than it actually is, in my opinion.

Stage 5: Also extremely good.  The run comments point out the mistake at the end (two wallbounces instead of one), but this is a pixel perfect trick, late in the run, with no chance to retry if it's missed.  This is the signature trick in this run, and the runner makes it look much easier than it actually is.

Stage 6: I'm very happy to finally see a high quality run of this stage in the WR.  This has been a problem stage for past WR runs, and we've finally got a near perfect run of it in the WR.  Any time lost here is minimal, and not worth considering.

Overall: 2-2.5 seconds is the sum of time lost, based on my analysis above, and in all but one case, that's really nitpicking.  Also, actually saving that time earlier in the run could through off patterns later in the run (particularly the stage 4 hammers), so it's difficult to say how much time was truly lost.  This is a fantastic run, and it definitely is the run that we've been waiting for.  Anyone who sets aside time to attempt to beat this run might be insane, because it would be extremely difficult to beat.

Video Quality: I have no complaints.

Final verdict: Accept

Congrats Elipsys!  Awesome run!


Quote from UraniumAnchor:
A/V good, no cheating, blah blah.

This run is faster than the 2:50, that's all you really need to know.

It does look like the 2:50 is a bit faster at the beginning of the intermediate (3rd) race, but this run makes up for it with a cleaner run through the wave section. It seems that most of the rest of the improvement is tighter movement in the final race.

Whether or not it's a 2:47 I'll leave up to IsraeliRD, but this is undoubtedly faster than the 2:50 by the same runner published not too long ago.

Accept.


Quote from CarpetCrawler:
Easy Accept, the time that can be saved is marginal and this run is just ridiculously good. Shoutouts to the magic wand. Wink


Decision: Accept

Reason: This one was for ALL the marbles.

Congratulations to Steve 'Elipsis' Barrios!
Thread title:  
Thanks guys!  I'm really really happy with this run - while it's not flawless, it's as close as to perfect as I think I want to push for without completely losing my marbles.  I had been sitting on a 2:48 for months that I didn't want to submit because I knew that I could grind out a 2:47 eventually.  It has been definitely been a very frustrating process at times, and it took a ton of effort and attempts to cut that last second.

I do disagree about the amount of time loss and the expectations of what  a "perfect" human run would look like.  If you discount splitting errors (done with my feet) and impossible combinations of frame rules, my actual sum of best is 2:46.  Maybe I'm just rubbish at the Intermediate race, but had I really saved a full second there it would have blown away my gold by over half a second.  My gold on the intermediate race is 27.0 (out of thousands of attempts, mind you), and this run has my subsplit at 27.4 here.

Reexamining the splits it looks like you guys are right about the Aerial race, as I looked more closely at the subsplits and this run is only 0.2s short of my gold split.  In my opinion the biggest time loss is the double bounce on silly race.  I think I could have had a 2:47.2X, maybe even 2:47.1X if I did it in a single bounce.  I have never been able to develop technology to make this consistent... so what I do is just play for 5 hours and maybe get one or two successful silly shortcuts... and then try not to choke on the Ultimate race.

As far as timing the run... I feel pretty confident about the figures here down to at least tenths, but please let me know if you do an independent timing what you guys time it out to be.  I started on the first frame of movement, not when the Marble spawns (I am always holding a direction at start).

Thanks again for taking the time to verify and give me feedback.  You guys rock!  After I run this tomorrow for the Best of NES Marathon (my first marathon, really excited!!) - I'm planning to pretty much retire from running the game and finally move on to other projects.

BassDeluxe are you coming to AGDQ?  We could do a 2P game and bash the red and blue marbles together faster than anyone has ever seen before.

Here's the run

Edit history:
IsraeliRD: 2014-10-31 12:28:35 pm
Dragon Power Supreme
Quote from Elipsis:
As far as timing the run... I feel pretty confident about the figures here down to at least tenths, but please let me know if you do an independent timing what you guys time it out to be.  I started on the first frame of movement, not when the Marble spawns (I am always holding a direction at start).


From memory I had 2:47.6X, and your loss on the silly race 2nd bounce was roughly 0.5 seconds, so 2:47.1X would've been your run had you not had the bounce. Either way, 2:47 for SDA.
Quote from IsraeliRD:
Quote from Elipsis:
As far as timing the run... I feel pretty confident about the figures here down to at least tenths, but please let me know if you do an independent timing what you guys time it out to be.  I started on the first frame of movement, not when the Marble spawns (I am always holding a direction at start).


From memory I had 2:47.6X, and your loss on the silly race 2nd bounce was roughly 0.5 seconds, so 2:47.1X would've been your run had you not had the bounce. Either way, 2:47 for SDA.


Sounds like we were close.

I guess you guys don't fuss over the tenths and hundredths?  What's SDA's policy on really minor gains?  For example would a 2:47.4X be a viable candidate to obsolete this run or would it have to be a 2:46.99 or better?
Edit history:
IsraeliRD: 2014-10-31 02:03:21 pm
Dragon Power Supreme
Full seconds, so 2:46.XX. For longer games (probably 3+ hrs) we would prefer at least a minute's improvement but verifiers can reject the run for only being 1 minute better.
Games with ILs where decimal places do count, we do accepted the tenths/hundrerdths, e.g. Max Payne NYM or Marble Blast Gold.
Quote from IsraeliRD:
Full seconds, so 2:46.XX. For longer games (probably 3+ hrs) we would prefer at least a minute's improvement but verifiers can reject the run for only being 1 minute better.
Games with ILs where decimal places do count, we do accepted the tenths/hundrerdths, e.g. Max Payne NYM or Marble Blast Gold.

I thought that changed with the 2013 rules rewrite.
Quote:
We removed the paragraph about the minimum required improvement for runs timed with SDA timing.  In the future, verifiers will decide how small an improvement is too small.  This opens the door for, say, smaller improvements for Super Mario Bros. or some IL tables.
Dragon Power Supreme
Well TIL our own rules. A 2:47.4X would be acceptable then Tongue