Experienced Forum User, Published Author, Experienced player
(699)
Joined: 2/19/2006
Posts: 742
Location: Quincy, MA
The thing is with popular runs like this, it gets quite a few votes very quickly and they are all "Yes" votes, which is most likely why they get published so fast.
Reguardless of whether it was published or not I like to add my opinion of the movie by voting. With a run like this it is obvious that it will be published because it is a significant improvement of a Mario game. Any new Mario that is a faster run gets about 98% "yes" votes and 2% "no" votes, so in this case I really don't think a vote matters because I think it would be published anyways.
Super Mario Bros. console speedrunner
- Andrew Gardikis
Experienced Forum User, Published Author, Experienced player
(699)
Joined: 2/19/2006
Posts: 742
Location: Quincy, MA
I figured this would come along quickly after mister posted the 1/1 trick. Very suprised that it was improve by almost 1 second though., considering half the improvements weren't even from the 1/1 trick. Time to watch it. ;)
Super Mario Bros. console speedrunner
- Andrew Gardikis
Experienced Forum User, Published Author, Experienced player
(699)
Joined: 2/19/2006
Posts: 742
Location: Quincy, MA
I forget who was telling me, but someone said they can get 361's about 1/3 of the time on that level by running to the end and letting go of the D-pad and doing a spin-jump off the breakable blocks at the end.
Super Mario Bros. console speedrunner
- Andrew Gardikis
Experienced Forum User, Published Author, Experienced player
(699)
Joined: 2/19/2006
Posts: 742
Location: Quincy, MA
Xkeeper wrote:
Tetris is quite popular for AI, at least I thought so. It's a fairly simple game at the core.
But:
AI agents who learned with the most successful policy had an average score of 8186 points over 50 games. The average score of five humans, who each played 10 games, was 8064. In the non-learning experiment when agents used random policies, the average score was just 676. Other policies produced scores that fell within this range.
I can usually score a decent average of 70,000, and I am by no means good...
Wouldn't it have made more sense to have, say, 10-15 humans and take the top 5 to get a better guess?
I seem to remember me scoring around 100,000 points when I first played ms. pacman. I'm sure the people they used for their research do not play videogames.
I thought that an average score of the first level being beaten would be a score of around 8,000? maybe I'm wrong.
http://www.pacmangame.info/ms_pacman.html
Well, if this game scores the same way the original does, I got a score of 11,300 for the first level trying my best with one life.
Super Mario Bros. console speedrunner
- Andrew Gardikis
Experienced Forum User, Published Author, Experienced player
(699)
Joined: 2/19/2006
Posts: 742
Location: Quincy, MA
I'd like to test a few things out, but it's odd that left+right while on the left side of a vine sends you somewhere. Most likely my theorys would prove to be slower, but still could be an entertaining thing to include if it meets the 21 frame rule.
Super Mario Bros. console speedrunner
- Andrew Gardikis
Experienced Forum User, Published Author, Experienced player
(699)
Joined: 2/19/2006
Posts: 742
Location: Quincy, MA
So I was pretty suprised to find out that holding left+right while on the left side of the vine warps you forwards. Could this be used for any levels in the 100% run?
Super Mario Bros. console speedrunner
- Andrew Gardikis
Experienced Forum User, Published Author, Experienced player
(699)
Joined: 2/19/2006
Posts: 742
Location: Quincy, MA
Probably some sort of glitch occured, or you saw it incorrectly.
One time my score for skeet shooting in track and field was 1 point higher than is possible, which I assumed was some glitch. It read "28,201".
Super Mario Bros. console speedrunner
- Andrew Gardikis
Experienced Forum User, Published Author, Experienced player
(699)
Joined: 2/19/2006
Posts: 742
Location: Quincy, MA
I had to vote no.
The movie itself was slow, because the game doesn't allow you to do anything useful. Although it may be hard, it's slow, even if it's the fastest possible thing. It's a good run through, but just a horrible hack to run.
I kind of hate things that involve using the koopa like that because it takes so long to execute, and you had to do it a few times in this hack in order to beat it.
Super Mario Bros. console speedrunner
- Andrew Gardikis
Experienced Forum User, Published Author, Experienced player
(699)
Joined: 2/19/2006
Posts: 742
Location: Quincy, MA
Interesting, although I assume it would be a bit slower. (time taken to grab the coin, time taken to fail the bonus)
EDIT: I suggest a level with Toad because he picks up items the fastest from the ground. I would think level 1-2 would be a good place to check it out.
Super Mario Bros. console speedrunner
- Andrew Gardikis
Experienced Forum User, Published Author, Experienced player
(699)
Joined: 2/19/2006
Posts: 742
Location: Quincy, MA
Bisqwit wrote:
andrewg wrote:
I actually don't think you can do that.
Oh, certainly you can.
You can do a three-frame jump over the goomba at the last possible frame, and you will hit the mushroom questionmark block and not die.
An example of a TAS that did that was supermariobros-timeattack-mana.avi .
supermariobros-tasv2-pom.avi hit the upper question mark.
supermariobros-tas-klmz.avi stomped the goomba.
Almost every other SMB TAS saved the goomba and hit the second question mark instead.
You don't have to slow down at all? That's odd.
Super Mario Bros. console speedrunner
- Andrew Gardikis
Experienced Forum User, Published Author, Experienced player
(699)
Joined: 2/19/2006
Posts: 742
Location: Quincy, MA
Chamale wrote:
Glitcher wrote:
Even after twenty years, people are still discovering timesavers in this game. Mark my words, next year the bar will be lowered to 4:00. ^_~
Words marked.
I think that's unlikely (but by no means impossible) because the old run had been standing for 2 years when this came in. I imagine there might be a couple minor improvements left, but major bugs could take a while before they're discovered.
Now, here's a very nitpicky thing, but one that bugs me big-time. Almost all SMB1 runs start 1-1 by jumping to hit the first coin block, then jumping on the Goomba to kill it. This opening has become way too bland for me. I would prefer it if future runs did something a bit different, even if it earned less points. Here's my suggestion: Jump on the first coin block, then jump from there to the high coin block above all the bricks. Then do a flying leap from there to stay above the screen for a while.
Heh. As I read this, I was thinking "I bet he's going to say to jump off the coin block to the higher one" and I was right.
Well hitting the first coin block and then getting that goomba is the toughest thing so that's probably why they do it.
Super Mario Bros. console speedrunner
- Andrew Gardikis
Experienced Forum User, Published Author, Experienced player
(699)
Joined: 2/19/2006
Posts: 742
Location: Quincy, MA
Well Mario's speed is doubled and he can jump higher in this thing.
I'd like to see that duplictaed at normal speed, but Mario cannot jump that high, so I doubt it's possible.
Super Mario Bros. console speedrunner
- Andrew Gardikis
Experienced Forum User, Published Author, Experienced player
(699)
Joined: 2/19/2006
Posts: 742
Location: Quincy, MA
-In the end of 1-2, couldn't the player use the jump-into-corners trick?
Yes, but it probably does not reach the 21 frame barrier in order to improve the time at all. I assume it is a few frames slower, but it doesn't affect the time.
-And in 4-2 isn't it faster to go the upper path instead of going through walls?
He goes through the wall in order to scroll Mario forward on the screen enough so that the pipe he goes down will take him to the warp area. So yes, it is a necessity.
-Why does the jump-into-corners trick not work in 4-2 at the stairs?
Because I believe you can get into the top first block and not get any farther, if you jump, you will jump out of the block. I don't think Mario can fully go through the blocks there, only partially which is how the flagpole glitch works.
Super Mario Bros. console speedrunner
- Andrew Gardikis
Experienced Forum User, Published Author, Experienced player
(699)
Joined: 2/19/2006
Posts: 742
Location: Quincy, MA
1.4 seconds shaved off of a run that was thought to be perfect is amazing.
If they managed to skip the cutscenes the run would be under 4:00, other than that, it's not happening.
Super Mario Bros. console speedrunner
- Andrew Gardikis
Experienced Forum User, Published Author, Experienced player
(699)
Joined: 2/19/2006
Posts: 742
Location: Quincy, MA
It really is interesting that the flag wouldn't drop when it was touched.
Since the game is so short, I was not expecting there to ever be and improvement of this run.
Does anyone think it could be improved even more? I wouldn't imagine there wouldn't be anything else to improve except if you could skip the fireworks animation in a similar way or something in level 4-2.
Super Mario Bros. console speedrunner
- Andrew Gardikis
Experienced Forum User, Published Author, Experienced player
(699)
Joined: 2/19/2006
Posts: 742
Location: Quincy, MA
Kles wrote:
I'm sorry, but the extent of your posts that I've seen have been nonstop negativity, even if they have been jokes. I rarely get so blunt as to tell someone to outright stuff it, but I kind of got irritated with it.
Kindly stop it please. :)
ANYWAYS ALSO, does anyone else have any other ideas of possible improvements? It would be nice to see something gamebreaking. :P
I doubt there are any other improvements that can be made, heh, but most likely I'm wrong. I think something in 4-2 has potential for beaing faster, but there's the 21 frame rule, so I still doubt another improvement.
EDIT: I can't view the video, could someone possibly upload it elsewhere? ;)
Also... I still am thinking that warping from one side of the screen to the other with the vine in 4-2 still has potential.
Super Mario Bros. console speedrunner
- Andrew Gardikis
