Okay, it's the time for those Awards already. Let's add a new award to the list of the current awards: Crappiest Game of 2014
The award for this category is awarded to the crappiest game TASed of this year!
To be elligible to this category, the TASed game must be a crappy game and the TAS must be published. The crappiest the game is, the better its chances of winning!!
I'll begin with some nominations:
[2756] NES The Adventures of Rocky and Bullwinkle and Friends by TASeditor in 03:54.16[2595] A2600 Spider-Man by Noxxa in 03:04.04[2693] Genesis Sonic Jam 6 by True in 07:53.69Please note something: This is not a official award and it's not intended to be a serious award. It's intended to make some laugh.
Watched Haikyuu.... Noticed that is pretty new since there is just 25 episode out.
Already waiting for the next episodes(that's going to take some time, sadly. :( )
oui c'est le temp de deplacement j'avais demandé dans le forum anglophone est spikestuff m'a dit que la valeur etait le temps de deplacement je remercie spikestuff de m'avoir dit a quoi servait la valeur
Je suspecte que Spiky a utilisé google translation. Alors je dirais plus que la valeur que tu as est probablement le déplacement de Rayman dans le temps. Soit la vitesse...?
Mais dans quelle direction?
Concernant la recherche d'addresse, je vais tenter une approche un peu plus universelle:
pour trouver une valeur, tu dois utiliser une approche scientifique. Considère tous les jeux comme étant un paquet de variables(Toutes les addresses sont des variables). Pour trouver une variable, il faut savoir cibler une recherche.
Pour ce faire, tu peux:
-T'organiser pour que la variable recherchée ne varie pas du tout.
Avec ceci, tu pourras éliminer toutes les variables qui ont variées!
-T'organiser pour que la variable recherchée varie
Avec ceci, tu pourras éliminer toutes les variables qui n'ont pas variées.
-Suivre la valeur de la variable dans le temps
Avec ceci, tu pourras éliminer toutes variables qui ont variées de façon très bizarre! Par exemple, si tu cherches une position, et que si, en bougeant, une addresse passe d'une valeur négative à une valeur positive très éloignée pour ensuite redevenir négative(Par exemple: -24 à 53 à -46 en 3 frames), tu pourras éliminer cette addresse!
-Suivre le nombre de fois que la variable a varié selon le nombre d'images(frames) écoulée.
Avec ceci, si tu cherches une variables qui devrait varier à tous les images, tu pourras éliminer ceux qui ont varié moins souvent!
-Supposer une valeur que la variable recherchée ne soit pas égale à
Avec ceci, on peut éliminer les variables qui sont égale à la valeur recherchée. Mais attention, ce n'est pas très pratique de faire cette étape!
Donc, rechercher une addresse consiste à réaliser plusieurs étapes composé des 5 méthodes de recherche mentionnés plus haut. Il est libre à toi de réfléchir laquelle utiliser en premier, laquelle utiliser ensuite, etc.
Afin de m'assurer que gamerretro2 comprenne mon approche universelle, je vais expliquer ce que Dica voulait dire:
tu sautes(Ici, on fait varier la vitesse de sorte que l'on ait une valeur différente de 0)
tu mets le jeu en pause en étant en phase ascendante/descendante*(Ici, on s'assure que la variable ne soit pas égale à 0)
tu commences à chercher une valeur différente de 0 (vu que tu montes/descends la vitesse est forcément différente de 0)(On élimine les variables égale à 0)
tu dé-pause le jeu et attends de retomber au sol(On s'assure que la variable soit égale à 0)
arrivé au sol, la vitesse est nulle, donc suffit de chercher une valeur nulle(On élimine les variables non égale à 0)
maintenant faut recommencer à l'étape 1 jusqu'à trouver
Évidement, on a supposé que la valeur de la vitesse au repos est de 0!
Si la valeur n'était pas égale à 0, il faut utiliser une autre approche.
Ainsi, faire une recherche, il faut réfléchir sur ce que ferait normalement l'addresse!
Considering the mistakes made in the start of the movie and everythings that can improve the start of the movie, I'll cancel it finally.
If the legendarymudkip ever finishes it, so be it, that's what mistakes bring! If not, I'll edo it later this year, once I finish university. I've learned a lot of things. If I do, I'll post WIPs in the forum before submitting.
Niamek
3600 rerecords in 36000 frames is 6 rerecords per second, still far above this list. Re-record counts are too much of a snap judgement to be trusted: They're easy to inflate, especially in BizHawk where you can just change it in a text file. Game complexity changes how often re-records are needed, sometimes there are a lot of cutscenes or text advancing which usually doesn't require more than 1 or 2 re-records per scene/box, sometimes people just get lucky and are able to get complicated tricks done really quickly, some people work off of just one file and do multiple attempts of a game within that same file, leading to a much higher count than was actually used for the final run, there could be problems with the emulator in determining re-record counts... Anyway.
What I'm wondering is why the re-record count dropped to 3400 in the updated run: User movie #18627730843176147
I haven't watched either run yet so maybe there's an obvious reason I'm missing.
I used the wrong movie file when I corrected the first level. Instead of redoing it on the correct movie file, I copied the input that were missing fro the good movie file to the movie file that I used.
Do you know what causes this glitch to occur?
Honestly, no. The walls that you can climb o it has weird physics.
Sorry, I change my vote to Yes for vault. And I will recheck it.
That's okay, mtvf1. ^^ I myself did a mistake with the movie file because I stopped it later than I had to.(Between, when the movie file will be changed?) But I still have to end the movie 13 seconds later than the time claimed in both movie: yours and Tompa. Since the timing were different, without checking carefully, it was easy to make mistake. :)
Mothrayas wrote:
ALAKTORN wrote:
Well SMB warp has 9739 for 5 minutes and that game is just pressing right (I think).
I like how your metric of comparison for rerecord count is literally one of the most optimized movies on the site.
And no, that run is definitely not just pressing right. If it were, it wouldn't have been improved and published over 15 times. 2D Mario games generally have quite a bit more movement complexity and tricks than similar 2D platformers, and SMB1 is no exception.
I'd like to point out that this game is not that hard to advance TAS quickly: 4 pictures of 19 seconds + the story scenes that comes with them is about 1 minute and 20 seconds... Plus the two autoscrollers... That's a lot of time that's easy to TAS without adding a lot of rerecords.
I'd say TASing Darkman is pretty similar to TASing SMW if we remove ALL the tricks except the 49 speed of SMW: You manage to get the highest speed possible, then go through the majority of the level. TASing SMW like this is very fast and doesn't require a lot of rerecording.
Also, this game has a lot of run to the right. But, there is things to consider: Avoiding enemies, avoiding balloons, etc. So I'd say this game is not mostly run right for justice as it seems.
Look at the first level of the pink t shirted guy, at the start I managed to climb quickly on the left side of the "tree". I could climb on the tree or pass the enemy without climbing.
Look at the first level of the last character, I ran left to climb a wall instead of going only left. Look at the first level of the bazooka guy, I could land on the balloon instead of waiting of the clown attack.
Look at the small glitch of the 3rd level.
So, this game is not only running right for justice.
Good luck improving the 3rd level.
The main improvements are in the 3rd level mainly. And not by much: The first section surely, then the balloon rope.
The main addresses I used are speeds during the TASing.
I've found other, such as position, timer counter(useless), potion counter, picture counter(useless).
The explanation behind this sentence:I went "too fast" during the first moving platform and that all change the rest of the first level.
It's because when you jump at this platform, you are at a X position. Because of your speed, you can jump at that X position. But If I do it in my movie, because of my higher speed, the X position where I can jump is Y which is more on the left than you. If I wait ONE frame, I simply can't jump since I'd move to X+1 position, which makes me fall of the platform.
You intend to improve it? Or to continue testing?^^
I intended to put them in the submission, but I didn't since I don't know how to makes tables...
But If I look at our input difference, you've saved some frames in the very beginning of the movie. You did D+A in the same frame where I didn't... Again a stupid mistake, I'll fix it.
edit: I won't edit it today, since it messes up the glitch of the third level.
I'll leave it as is it right now, no further modification will be made by me.
The legendary Mudkip still saved some frame in comparison of me, and that is because of speed management. I went "too fast" during the first moving platform and that all change the rest of the first level.
As always, I hope the current movie in the WIP section(Someone should update it, it's the last one) is still publishable.
I intend to correct my mistakes later, but not right now.
Thank you the legendarymudkip for your work, it's appreciated.
Now that I retry it, It seems that this jump wasn't properly tried when I did the movie in the start. It's a stupid mistake to make indeed. I've tried the lastest frame possible to jump, not all.
Well, I'll improve at least this oversight, give me a second.
EDIT: Currently verifying it syncs.
EDIT: Here it is: User movie #18627730843176147
But hey, it's the same time than before.
As you can see, The legendary mudkip, I now jump over the 2nd platform.
Concerning the balloon of the rope section, if you can manage to pass it without waiting, awesome, that would save me some future headache and if you explain what you did, I can try and edit that part now... While I'm bored of my school duties. :P
I found the run entertaining in general, although I thought that more entertainment could be added in thelevels where you do nothing but shoot nothing. However, I also feel that the run is very suboptimal. I did a quick TAS of the first level and managed to beat it by 56 frames. Voted no for the levels where you do nothing but shoot nothing and lack of optimization.
I'm skeptical about the improvement you claim concerning the first level.
The first jump should be not improvable.
Then comes the rope part. If you jump down, not only you're slowed down considerably(passing the rope reset your speed), but you're going to be waiting for the platform. Trying to jump from the first platform directly to the next floor is impossible unless you wait a whole cycle of the platform. The jumps after the platforms are frame perfect and should not be improvable.
Unless you specifically describe how you made the improvements, I'll don't believe you.
I'm confused, why I can't skip the second platform in my run and you could without that much waiting if any?
Well... I believe you know.
Between, for the next times with other submission, I'd suggest you explain the diffrences between your WIPs and the actual submission in your futures comment instead of just saying :"it's suboptimal".
* Frame count might be iffy because the author just uploaded a file which is 9:39.71 but shows frames of 34840... which means, niamek fix your final input for instage, I'm faster than you by 2 frames ♥.
I stopped the input in the lastframe that reaches the end of the level in the fastest way possible. That's why I stopped later than you. ^^
Oh okay, but there is still one small problem:
To access the credits, I need to press A somewhere after giving the last hit to the last boss. So, where do I stop the movie? Last input of the last level? or last input before credits?
I'll upload both movies, since I won't have time later today to come back.
User movie #18625863636125814User movie #18625839008334444
@Spikestuff: Its camera part, not gun part. :P
But yeah I agree it's pretty boring there...
I found the run entertaining in general, although I thought that more entertainment could be added in thelevels where you do nothing but shoot nothing. However, I also feel that the run is very suboptimal. I did a quick TAS of the first level and managed to beat it by 56 frames. Voted no for the levels where you do nothing but shoot nothing and lack of optimization.
I'm skeptical about the improvement you claim concerning the first level.
The first jump should be not improvable.
Then comes the rope part. If you jump down, not only you're slowed down considerably(passing the rope reset your speed), but you're going to be waiting for the platform. Trying to jump from the first platform directly to the next floor is impossible unless you wait a whole cycle of the platform. The jumps after the platforms are frame perfect and should not be improvable.
Unless you specifically describe how you made the improvements, I'll don't believe you.
Concerning the picture minigames, I'm personally not entertained by a moving camera taking random picture that misses, I'd prefer a fix camera.
mtvf1 wrote:
I vote NO.
max12187566 has made a run in 10:01 before 6 years.
(video)
Again, my last input is after the credits and not from the last hit on the last boss. Exactly like the last video that Tompa pointed to me. So, I beat his run by a whole 30 seconds about. I pressed "A" after the credit so the game can return to the title screen. So that's add about 30 seconds to the movie.
You shouldn't just beat the overall time, but all the internal times, or be able to explain that you were slower in one spot because it gained more time somewhere else.
It's not like that 2.5 year old easily Googled run was a secret, or the basic mechanic of picking up projectiles to use.
Both failing to check for existing runs and not being aware of core game mechanics don't reflect that much effort went into this, even if you did inadvertently end up slightly faster.
You're right, I didn't search enough to find this video before, which I should have since it's easy to find it.
It's true that it's not amcomplicated game compared to the best TASes here, but, hey I gave my best effort with what I knew about this game. I were hoping that my time amd my tricks would show that I put effort on it, but it seems that it didn't show... Too bad. :(
By the way,the video is saying 10:13, but I noticed that he didn't press one input after the credit in which I did. So the time difference is greater.
I don't know exactly where since I've not studied the video in great details, but I can say easily tell some places I lost times and where I gained time.
I already said indirectly where I lost time, so I'll list how I gained time(no precise places, but gemeral places)
He goes for slower paths than me sometimes(ex: the 20th level)
I have better speed management( I said that running gain speed faster than the air for example and that touching the floor reset speeds. Since he jumps a lot with his glitched jumps, his speed isn't increasing as high as mine)
Some place I were more aggressive comcerning tricky jumps(20th, 5th levels are both examples of levels comtaining tricky jumps.
"A lot of improvements can be made" = "Bad TAS".
---
ars4326 wrote:
Correct me if I'm mistaken, but is the real-time run you linked to (9:30) faster than the run you submitted (10:11)?
YT description said that he deleted stages with making photographs.
I'm sad to see that this movie doesn't please you as much as I hoped. I tried my best.
I forgot to say that Tompa pointed this video to me after I submited the movie. P
There is not a lot of improvement that can be made. Therr is some there and there, but it's very minor.
Ars4326: Indeed, the human speedrun removed about 2 or 3 minutes from the video. It not only deleted photographs, but the story cutscenes that came with.
Tompa have pointed to me a video of this very same game. In this video, I sw some improvements that can be made to this run which I'd like to implement in a near future if no one does, but not right now as I restart school soon. (That means maybe in the next holidays)
Here is the movie he llinked to me:
Link to videoNOTE: The author timed the run to the last hit, which is about 32 seconds earlier than where I ended my run. So I bested this run by about 34 seconds.
As you can see, all the levels, he jumps, then land, then jumps very quickly. Idneed to figure out how he does it. It would save frames mainly on the third portion of the game with the ninjas since it will allow you jump less high it seems, thus avoiding the slow floor. I'd like to point out that speed in the air increase only by 5 per frames and the speed is reset to the floor max speed whenever you touch the ground.
Secondly, the boss with a bazooka... well the glitch is interesting and may save some frames, but not much I think.
And finally, in the second level, a oversight of me is that I didn't know you can collect rocks and throw them at enemies. I don't know how it works but it seems very effective with the enemy he threw on.
EDIT: And finally, some balloons could be manipulated finally, which I wasn't able to do so, need more research... The main balloon to manipulate would be the balloon of the rope section. of the 3rd level.
But in the end, I still beat his movie surprisingly.
Edit: finally, It's notnsurprising finally, see next posts.
I hope this runs is still acceptable to the site even with these improvements.
EDIT: Just for comparison, here is a human speed run:
Link to video
and a long play,
Link to video
pour le ram watch tu nous as pas dit comment avoir la vitesse verticale et les position
Je souhaite ajouter qu'il a ,en réalité, dit comment tu dois faire pour trouver la vitesse verticale et les position. Mais il ne l'a pas dis explicitement. Il a donné(je suppose, je n'ai pas regardé la vidéo) un exemple(Donc, il a trouvé l'addresse d'une autre variable). La seule différence est que la variable que tu souhaites trouver n'est pas la même que la variable trouvée dans l'exemple. La logique que tu dois utiliser pour trouver ces addresses est identique par contre!
Ça revient un peu à ce que feeuzz22 veut dire: Tu dois te débrouiller.