Posts for DrJones


1 2
21 22 23
Experienced Forum User
Joined: 4/3/2005
Posts: 575
Location: Spain
As Fast as possible is the most important as competition goes. That's why submissions come with the time lenght, duh. Even a movie of a fighting game in which the player performs always the same attacks is interesting, because it sets the minimum time required to beat the game. On the other hand, movies processed are meant to be watched. They shouldn't be boring. It's more interesting to process a movie which also entertains. I personally find it okay if stunts don't take too much time. Could a TAS be accepted just for record purposes, with no intention to publish a video movie of it? Not that I'm suggesting anything. I'm just just curious.
No.
Experienced Forum User
Joined: 4/3/2005
Posts: 575
Location: Spain
Doesn't this game abuses programming errors? What happens then with most of the monsters at the end of the game, which die as soon as they appear?
No.
Post subject: *correcting mistake*
Experienced Forum User
Joined: 4/3/2005
Posts: 575
Location: Spain
xebra wrote:
DrJones: None of what you are saying makes any sense. I'm so tired of arguing with people that insist on talking about things they simply don't understand. I won't bother to refute your latest post, I tire of it, and Nitsuja already tried to reason with you, and you ingored him completely. But it's painfully clear from all of your babble that you have absolutely no clue what you are talking about.
I will pass over this paragraph, thank you.
xebra wrote:
Expressing your opinion is wonderful, but do you actually have any programming experience or training, like Nitsuja? I program for a living, but it's just higher level BS web applications, so when Nitsuja talks about programming, I listen carefully and respect his opinions more than my own misguided preconceptions. You would do well to do likewise. Do you actually have any experience or training with algorithms, computational methods, or numerical analysis, like I do? I am not the end all be all in this field, but it was my specialization in college. Even after all my training I still make mistakes, sometimes simple, sometimes complex, so when my peers tell me that what I am saying does not make sense, and ask me to critically examine my ideas, I listen. You would do well to do likewise.
I'm also a programmer myself, and AI is my computer-related favourite topic. I just don't understand why are you writing so strongly negative posts. I'm in no way being aggresive with you, nor with anyone here. It would help more if you wrote here what are you reading from our words. These posts are dense and it's easy to miss something. That could be considered off-topic, but I prefer this to a long discussion about who misunderstood who.
No.
Experienced Forum User
Joined: 4/3/2005
Posts: 575
Location: Spain
nitsuja wrote:
DrJones: It's a little unclear what you're suggesting and how you think it could actually be implemented. It sounds to me like it would take enormously more programming effort to get anything. Sure, better search algorithms than pure randomness can (and possibly should) be used. But this talk about playing more than small segments and actually affecting the strategy... unless you have some damned amazing AI planned, that is quite impossible for a game like this, especially if you want it to be able to handle other games like this without being completely remade. Maybe a pruning method makes sense, but it has to be incredibly good pruning to reduce the computation time enough. So far nobody has come close to suggesting what that pruning method might be, probably because it's very difficult to figure out. Bisqbot (after being modified with some much-easier-to-figure-out heuristics) can already handle improving both tactics and skill within a practical timeframe, and the high-level strategy has already been optimized independently of the low-level playing by some humans with actual intelligence. Doing even better than that might be a nice goal, but maybe we should start small and consider something that could be completed within the next 6-12 months of part of the free time of a few individuals.
Of course, I was talking about teorethical ways of implementing a bot able to play games. Splicing up the task between various levels of abstraction seems to me a better method than brute forcing, and helps us having a better understandment of the problems that may arise. For example, how a sub-optimal path lends to a better overall time later in the game could be explained because another tactic/strategy was being used. Not all layers have to be implemented. Of course, it can have some pathfinding routines and some kind of genetical algorithms for battles (as they usually rely on few combinations that are very effective), but they can also be directed by human players. Also, it should be noted that most of the strategic layer is predefined by each game (that's why they have to explain their rules and the goal on their manual), which makes more difficult creating a generic playing bot. All this explanation was just to make clear that most of the work of bisqbot is directed to optimization of paths, which, as the rest of layers needed to "play a game" are given by an human player. About the pruning: An actual pruning method would be comparing the actual path with other path obtained by an human player. I have said that I think that, if all paths use the same tactic and strategy, we can consider that a path which creates a small delay that it's not corrected within X frames is just bad and can be pruned. The main problem I have found, is that it is very hard for bots to "see" what's happening in the game, and detecting if they have reach the goal, and such. I have an heuristic idea. Collisions are very interesting. I think that having a short memory (2-3 seconds) of positions in which collisions happened, and if they were good (i.e.powerups), or bad (i.e. walls or shoots) would help the bot avoiding the bad ones. Some collisions can be detected because the input didn't affect the outcome (i.e. walking against a wall), or because it changed the state of the player (i.e. life meter drops). As many collisions are produced by not static elements, the memory shouldn't last too much time.
No.
Experienced Forum User
Joined: 4/3/2005
Posts: 575
Location: Spain
Baxter wrote:
I only see 2 different timesavers: - Strategy (pretty much your first and second point, and a part of your third point) - Luck manipulation (I guess it's a part of your third point) I think Bisqbot is able to help a great deal on the luck manipulation part.
Your way is not less correct than mine (it could be even more correct, in fact, as it englobes everything), but I think it is less appropiate for coding a bot (unless, of course, your sole purpose is having a bot that manipulates luck).
No.
Post subject: About Timesavers
Experienced Forum User
Joined: 4/3/2005
Posts: 575
Location: Spain
The efficiency of Bisqbot is measured in frames. That's it, the fewer frames used for achieving a certain goal, the better. If there's a known input that achieves the goal, and Bisqbot finds a faster one, then it's said that Bisqbot has found a timesaver (I love making up terminology). I've identified three levels for Timesavers: 1. Those that come as a result of a Strategic decision. By strategy, I mean choosing the goals to accomplish, and the order in which they will be achieved. 2. Those that come as a result of a Tactic decision. By Tactics, I mean choosing the path/method that will be used to achieve each goal. 3. Those that come as a result of Improved Skill. By Skill, I mean how well each path/method is accomplished. Changes at an higher level have more weight on the result. A change of strategy will usually look unpredictable working from a lower level. A small delay at the start may cause an huge timesaver much much later. (This would rend useless boco's idea) However, changes at a lower level have a lot less impact. I can figure that a change of gameplay to achieve a simple goal (such as jumping into a platform) will only repercute on, maybe, just the next X frames. We could use then ramification & pruning, if we looked ahead X frames before "cutting branches".
No.
Experienced Forum User
Joined: 4/3/2005
Posts: 575
Location: Spain
People taking into account the number of buttons should consider that waiting is an acceptable movement, too. Action buttons can be pushed while in motion, too. Heuristics are useful, but they take up time and could be useless or worse than that. Even something as simple as "don't death" would remove shortcuts from many games, and would stop the bot from beating Prince of Persia 2 (except in some versions, in which you can glitch the game to the end) Boco's suggestion of using a pruning method makes sense. Maybe Xebra hasn't understood that you don't generate all valid/possible inputs that way. Instead, you look at the possible start from all sequence inputs (a "small" number), and decide whether they are good starts, or not. Then, you continue exploring just the good ones. This method has very serious disadvantages, still, but I'll talk more about it in my next post. Finally, I think the best game for bot-testing purposes should be one like Deja Vu.
No.
Experienced Forum User
Joined: 4/3/2005
Posts: 575
Location: Spain
I don't know if this game is like the C64 version, but in that version, you could pick sand to deactivate a bomb, scissors to deactivate a string-activated pistol, umbrellas to deactivate the acid bucket, and pincers to deactivate a spring that may be placed inside the furniture. I had also a demo of Spy vs Spy 2, in which both spies crashed in a desert island. As a demo, it couldn't be finished, but it didn't matter.
No.
Experienced Forum User
Joined: 4/3/2005
Posts: 575
Location: Spain
This was ony of my favourite games back in the 80s! The gameboy version still costs 60€ in a shop I know ·_· Have to watch it, I'll vote later.
No.
Experienced Forum User
Joined: 4/3/2005
Posts: 575
Location: Spain
In the Amiga version (not sure about the PC version), you see a sword in the place where your mirror image appears later. One would think you will get another sword, but when you reach the room again, the sword has dissapeared (it has been taken by your mirror image). It looks like the NES version allows you to skip the trap where the mouse comes to rescue you. On SNES that can't be done.
No.
Post subject: Re: The Technical Aspects of Game Causal Connections
Experienced Forum User
Joined: 4/3/2005
Posts: 575
Location: Spain
Yes, this exactly is what I was thinking when looking for a way to solve mazes by divide-and-conquer, and my own studies about AI. This is your point. You assume that there is an optimal path between points A and B of a given game. Then, if there is a point in-between, C, in the optimal path, then the optimal path from A to C, and the optimal path from C to B appear in the optimal path from A to B. This property work only if these segments are independent (or just slighty dependent, with some luck), but could be enough for some games. But you must know that some optimal paths do sequence-breaking, so instead of a stream-lined game, you would have a graph (or a labyrinth), in which you could reach the same point from different places. You can't assume beforehand that your optimal path to B will pass from a given point C (just like mazes). You just know that it has to pass from one of the paths that reach B.
 A. To solve this maze:
############
...##....#.#
#.#...##.#.#
#.#.#.#....#
#.#.#.#.####
#...###..#..
#.##...#...#
#....#..####
#.#####....#
############

These two maze must be solvable
A
############
...##....#.#
#.#...##.#.#
#.#.#.#....#
#.#.#.#.####

B
#...###..#..
#.##...#...#
#....#..####
#.#####....#
############
And their solutions must be connected.
Of course, you can keep partitioning the labyrinths until they are trivial problems, and then interconnect them. But this wastes a lot of time solving parts that won't be part of the main solution. On the other side, many computers may work at the same time on different parts of the main problem. Those sub-labyrinths show that there can be multiple solutions to a single segment, but there is maybe only one that can be connected to the next sub-labyrinth. There can be many starts to a single segment, too. In games, this will be starting the segment with different values of: lives left, ammo, starting position, items equipped, etc... So, the same segment must be solven min-maxing those variables until they are no longer solvable: that way, you get the min time needed to pass the segment (wasting any number of resources), and to know the min values required to pass the segment, (though maybe losing time). Some segments are highly dependent of others. For example, there is a trigger in Prince of Persia 2 (PC Game) that creates a skeleton on the bridge that triggers another event which makes you lose your sword. There's a way to pass without activating the trigger, and thus the skeleton is never created. The path to solve that segment changes drastically. If you don't know beforehand about that trigger/glitch, you could end with a sub-optimal path. Note: If you pass the bridge without losing your sword, the next level is a lot more easy. Though, you need to move very quickly in order to pass.
No.
Experienced Forum User
Joined: 4/3/2005
Posts: 575
Location: Spain
I have found this run a bit more entertaining than the previous one, and it is faster, too. Voting yes.
No.
Experienced Forum User
Joined: 4/3/2005
Posts: 575
Location: Spain
How these bugs work? The more I think about them, the more odd they are. The only way I believe a sprite collision could affect the destination of a warp would be if the collision changed somehow the register used to store the target location. Could someone explain me why those odd things happen? They look interesting.
No.
Experienced Forum User
Joined: 4/3/2005
Posts: 575
Location: Spain
This may look weird, but some days ago I had a dream in which I was trying to solve this very problem. In the dream, I discovered that the problem of beating some games can be seen as a problem of solving a maze. I also realized then that mazes can be solved by using Divide & Conquer, a detail that I wasn't aware of until then. This could lead to an "efficient" algorythm to beat games, but I still have to experiment more with the idea.
No.
Experienced Forum User
Joined: 4/3/2005
Posts: 575
Location: Spain
This is... amazing... just amazing! This game is awesome! I never thought this game would be so different from the Amiga version. It was full of surprises at every level, and you made a whole lot of incredible tricks that seemed imposible. Not only you managed to get a lot of "WOW!"s from my friends, you even managed to get applauses!!! Thank you for making this TAS, SprintGod. I wouldn't have enjoyed the movie so much otherwise. Great work! What are the rest of you waiting? Go and watch this run immediatly!!
No.
Experienced Forum User
Joined: 4/3/2005
Posts: 575
Location: Spain
I raise my vote for frogger and Lunar Lander.
No.
Experienced Forum User
Joined: 4/3/2005
Posts: 575
Location: Spain
I don't know if this would be useful to others, but I would enjoy if each movie had a line with an estimation about how much time does it take to beat the game for a normal player. That way, I would know beforehand if Friday XIII is a short game that takes about 10 minutes to beat (with some practice), or if it really is as long as a full run of Zelda: A Link to the Past. It doesn't have to be very precise, just "half an hour", "a raining day", or "many days" would be enough for me. A more precise estimation would help appreciating the quality of a run, but I think it would be a source of troubles. That's all. EDIT: It can be included a difficulty meter, too.
No.
Experienced Forum User
Joined: 4/3/2005
Posts: 575
Location: Spain
I think that rather than "abuses programming errors", this movie should get "invents programming errors". Very entertaining to watch!
No.
Experienced Forum User
Joined: 4/3/2005
Posts: 575
Location: Spain
I think it would be a good idea to write notes when you finish each level, so it will be easier to hex-edit your run if you find a shortcut later.
No.
Experienced Forum User
Joined: 4/3/2005
Posts: 575
Location: Spain
Phil wrote:
No one should worry that Sleepz won't be credited. Sure he will. I never said that all tricks in my run was dicovered by me. When I started creating this run, I didn't thought this discussion to happen.
I think this settle all the discussion. My suggestion is that if sleepz worked hard to find cool tricks and want to be credited for this - which everyone here agrees it's fair - he can write them on the Mario tricks page (if he hasn't done it yet). We should differenciate techniques and routes from perfomances. Most of these videos have been done reading faqs that other people did, and other were done by many people who helped finding improvements. This is part of the game. I think most of this has been originated because people didn't know about Phil's project until it was done. If Phil opened a thread instead, saying that he was going to try to improve this video, would you have helped him, sleepz? I think there's no need to answer. Cheers to all
No.
Experienced Forum User
Joined: 4/3/2005
Posts: 575
Location: Spain
I would want to see both movies, as I plan to record a "shortcut special" bonus CD of speed runs, with both warp and warpless versions of various games.
No.
Experienced Forum User
Joined: 4/3/2005
Posts: 575
Location: Spain
This movie is weird. I didn't understand anything, except that there is a boy, there is a blob, and there is a cauldron.
No.
Experienced Forum User
Joined: 4/3/2005
Posts: 575
Location: Spain
The game is based on the movie Moonwalker, starring Michael Jackson. There is no replacement for him. And I'm more worried by the fact that all his fortune is managed by a sect than by the accusations of pedophilia.
No.
Experienced Forum User
Joined: 4/3/2005
Posts: 575
Location: Spain
What about Rampage? It's fun to crush down buildings ad infinitum, while eating people and avoiding helicopters.
No.
Experienced Forum User
Joined: 4/3/2005
Posts: 575
Location: Spain
This movie is even more broken if you accidentally have PAL emulation active. At first, I didn't know which one was the correct. Then, I guessed it was the one that finished the game. Haha
No.
1 2
21 22 23