Posts for rchokler

1 2 3
Posted: 1/25/2015 11:57 AM
Quote
rchokler
Experienced Forum User, Published Author, Player (145)
Joined: 11/14/2011
Posts: 68
Location: Brookline, MA
In that case, here is another frame down: http://tasvideos.org/userfiles/info/20351047306603533
Posted: 1/24/2015 9:12 PM
Quote
rchokler
Experienced Forum User, Published Author, Player (145)
Joined: 11/14/2011
Posts: 68
Location: Brookline, MA
I will try to give the fight one more push. If it doesn't produce any result then I think we can safely say that the frame war is over. Edit: Okay, I admit that the previous post may be a bit unfair to Samsara as the one frame improvement was made by trivially moving the last line of the input up one spot where Kuros still would gladly fire his weapon and defeat the last elemental of the fight. Thus it was a trivial adjustment. The following is a more fair response that shaves another 7 frames over the other one by a not-so-trivial modification. Now we are clocking at less than 7:16. http://tasvideos.org/userfiles/info/20338379539004966
Posted: 1/24/2015 11:23 AM
Quote
rchokler
Experienced Forum User, Published Author, Player (145)
Joined: 11/14/2011
Posts: 68
Location: Brookline, MA
And 1 more frame saved over Samsara's attempt. Now we probably will need to discover a different route to get more improvement. http://tasvideos.org/userfiles/info/20328404088202288
Posted: 1/24/2015 3:17 AM
Quote
rchokler
Experienced Forum User, Published Author, Player (145)
Joined: 11/14/2011
Posts: 68
Location: Brookline, MA
I will try to squeeze more out in the fight tomorrow. I 'm not sure if even the final version of the fight will be optimal though, due to so much uncertainty.
Posted: 1/23/2015 6:39 PM
(Edited: 1/23/2015 6:41 PM)
Quote
rchokler
Experienced Forum User, Published Author, Player (145)
Joined: 11/14/2011
Posts: 68
Location: Brookline, MA
Not sure how many more frames could be saved on the final boss, but regardless of the outcome when the submission is accepted, I am glad that we will now have a much more optimized publication for the game. Like the others, I voted Yes.
Posted: 1/23/2015 1:32 PM
Post subject: Boss improvement
Quote
rchokler
Experienced Forum User, Published Author, Player (145)
Joined: 11/14/2011
Posts: 68
Location: Brookline, MA
I was playing around with the boss fight and it turns out to be improvable, if only just slightly. I verified my attempt by playing back my version to make sure it syncs. Frame wars anyone? Rerecord count is now 21997 Movie length now: 07:16.31 (26222 Frames) http://tasvideos.org/userfiles/info/20308457084957722 Great job Alyosha! I almost was about to tear my hair out in search of the improvement. It took me several hours yesterday and today with many close calls along the way. That is how well optimized it was.
Posted: 1/14/2015 6:18 PM
Post subject: Continuation of the TAS
Quote
rchokler
Experienced Forum User, Published Author, Player (145)
Joined: 11/14/2011
Posts: 68
Location: Brookline, MA
It seems that goofydylan8 abandoned TASing this game. I am not sure what his opinion is by now, but I personally think that a TAS for this game would land in the moons. It is quite interesting to watch the main character defeating enemies so fast that they stand absolutely no chance at fighting back, especially if many enemies don't get a chance to enter the screen. However, as goofydylan8 mensioned, this TAS should not be hard to make. So I would say that a published TAS for this game would most likely be in the Moons teir. Here is a complete WIP for level 1-1: It is about 2/3 goofydylan8's work and 1/3 mine: http://tasvideos.org/userfiles/info/20112191051370022 Also, I am not so sure that a complete TAS of this game would be 25 minutes long. I think a well made TAS of this could be sub-20 since 2:03.26 (7408 frames) into the game, the WIP I just posted enters level 2 and there are 3 levels with 3 stages each. However, we can be reasonably sure that the length of the final TAS will be more than 18:31.20 (66672 frames) in length.
Posted: 1/14/2015 11:18 AM
Post subject: One more optimization
Quote
rchokler
Experienced Forum User, Published Author, Player (145)
Joined: 11/14/2011
Posts: 68
Location: Brookline, MA
Your level 3 is good. In fact I like your use of the steam eruptions to save frames over Aglar's run. However, I noted that during the lava boss fight, you seemed to hit the boss one frame after the boss resumed being vulnerable since I can see the eyes go white for a frame on a couple of moments. That costs frames of course. However, in level 4, my descent after the Earth Scorch is collected is a couple of frames faster than Aglar by once again modifying the trajectory to land on an unavoidable platform closer to it's left edge. Keep that in mind as you do that level. I don't think my WIP 3 has any more to offer beyond that though. I could try to route level 5 to see if I can beat Aglar's route for it but most likely I will look for some other project. Good luck!
Posted: 1/12/2015 12:03 AM
Quote
rchokler
Experienced Forum User, Published Author, Player (145)
Joined: 11/14/2011
Posts: 68
Location: Brookline, MA
You are welcome. I figured I would contribute on it since it was one of the games I enjoyed beating when I had an NES console.
Posted: 1/11/2015 9:19 PM
Post subject: Link to WIPs
Quote
rchokler
Experienced Forum User, Published Author, Player (145)
Joined: 11/14/2011
Posts: 68
Location: Brookline, MA
Here is a link to my WIPs: http://tasvideos.org/userfiles/user/rchokler
Posted: 1/11/2015 7:52 PM
Post subject: Comparisons if WIPs
Quote
rchokler
Experienced Forum User, Published Author, Player (145)
Joined: 11/14/2011
Posts: 68
Location: Brookline, MA
Alyosha, I compared your WIP to my WIP 3 from a few months ago and there are a few possible optimizations you can make on level 2. Note that overall, your WIP is faster and gains 100% mana on the first boss fight so good job there. 1. After the Golden Fly pickup, I save 4 frames on the descent to the entrance into the frog room. You will need to test out the frame rule on entrance to the frog room since I save another 4 frames from it but your outcome might wind up different. 2. In the descent to the entrance into the water lair, I save 8 frames. 3. You can ignore the part where I pick mana up since you don't need it, however, I have a different, more interesting way to fight the boss. I save 3 frames during the fight and it may be possible to save even more frames. My method is to get in the boss' face and attack while jumping constantly and staying as far right as the game allows. In my earlier, but slower samples, from WIP 1 and 2, I also allow the boss to attack and miss me multiple times for added entertainment. Implementing this strategy did prove to be quite frustrating though, but was worth the effort. As you do the other levels, I will see if I can offer some more advice where possible. Good luck on the game.
Posted: 9/22/2014 12:14 AM
Post subject: Boomerang Kid
Quote
rchokler
Experienced Forum User, Published Author, Player (145)
Joined: 11/14/2011
Posts: 68
Location: Brookline, MA
Here is a rough attempt at Boomerang Kid. It is under 5 minutes. There are some inevitable long waits for now, but mostly it's pretty fast. http://tasvideos.org/userfiles/info/17583277889243121 Note that this movie uses a ROM where selecting boomerang kid will start Super robin Hood and Vice versa.
Posted: 10/22/2012 1:55 AM
Quote
rchokler
Experienced Forum User, Published Author, Player (145)
Joined: 11/14/2011
Posts: 68
Location: Brookline, MA
partyboy1a wrote:
You can't get a run published if you rush those runs like that. Good runs require a lot of time. A few weeks or months is quite normal. Some take years. Monopoly is a really bad game for a beginner. The game itself is quite uninteresting, but heavy luck manipulation counterfeits that. Manipulating luck like in the published runs is a really difficult task. Choose some more interesting game please, there are plenty of it. Fast platformers are a good choice to start with.
I agree. A good TAS requires a lot of time, particularly because it is heavily based on trial and error analysis. My TAS of NES Puzznic took about three months to finish with the first draft, and even then I had to look through and do some additional optimization before it was finally published. Also a good TAS showcases either constant performance of tricks that are only achievable by a superhuman or human doable tricks in ways/circumstances that require a superhuman. In my Puzznic TAS, none of the solutions to levels can possibly be rivaled by a human player at an NES console as the fastest complete console playthroughs are over an hour long, while my TAS is just under half an hour long. So if you want to make a publishable NES Monopoly TAS in new categories, you will want to analyze thoroughly how the game works and read in the forums about the game features the expert players have discovered so far.
Posted: 5/22/2012 3:05 PM
Quote
rchokler
Experienced Forum User, Published Author, Player (145)
Joined: 11/14/2011
Posts: 68
Location: Brookline, MA
It would be interesting to see a console verification for my NES Puzznic TAS.
Posted: 12/5/2011 3:56 PM
Quote
rchokler
Experienced Forum User, Published Author, Player (145)
Joined: 11/14/2011
Posts: 68
Location: Brookline, MA
Voted Yes. I liked this TAS a lot more than the currently published one due to several places where you did interesting tricks that the other TAS didn't have. The railroad stage got more life as a result of pointing the Bō staff in different directions while jumping over a series of bottomless pits.
Posted: 12/5/2011 11:03 AM
Quote
rchokler
Experienced Forum User, Published Author, Player (145)
Joined: 11/14/2011
Posts: 68
Location: Brookline, MA
Thanks. It would be interesting to disassemble or decompile the ROM and see what exactly causes the apparent overflow that rolls the score over and whether 9,999,999 is the highest value admissible, which it presumably is.
Posted: 11/27/2011 3:37 PM
Quote
rchokler
Experienced Forum User, Published Author, Player (145)
Joined: 11/14/2011
Posts: 68
Location: Brookline, MA
Thank You. Yeah, I remember playing Brix too when I was young. I only beat Brix 1 a couple of times and the last level in brix requires good timing. You'll see what I mean when you enter the password for that level and try it out. Because Brix 1 and 2 are Puzznic derivatives, a number of the levels in both games are taken directly from Puzznic, however, Brix can't be TASed nearly as impressively as the blocks there don't have to all land for clearings to take place which means in Brix you can't arrange for big combos.
Posted: 11/20/2011 10:28 PM
Quote
rchokler
Experienced Forum User, Published Author, Player (145)
Joined: 11/14/2011
Posts: 68
Location: Brookline, MA
Yeah. The patterned blocks are cleared by placing 2 or more blocks with the same pattern together in a group and the goal is of course to clear all of them from the field. EssentiaFour has a youtube walkthrough showcasing the human-doable solutions for every level. If his solutions were recorded in a TAS manner, they would still result in a TAS that is considerably longer than 29:02.75 and so I always wondered how he would react when observing the solutions in my run.
Posted: 11/20/2011 10:14 PM
Quote
rchokler
Experienced Forum User, Published Author, Player (145)
Joined: 11/14/2011
Posts: 68
Location: Brookline, MA
I like the run a lot. Interesting shooting tricks including near miss shots. I wonder what the run would look like if these tricks were performed in a real life missions. They would possibly scare many enemies into surrendering. Yes vote!
Posted: 11/20/2011 3:09 PM
Quote
rchokler
Experienced Forum User, Published Author, Player (145)
Joined: 11/14/2011
Posts: 68
Location: Brookline, MA
Yes. When I was developing the solutions, I observed some small miscalculations on a number of levels before those in world 13, where the time bonus was calculated correctly and added to my score but with it failing to add the clear bonus to my score. The calculations of the time bonuses in world 13 were big underestimates by a factor of 2.5 to 3 or so. Finally, after my score got as close as possible to going 8 digits, it dropped after the end of the very next level. The last part may be either a miscalculation or possible overflow... or perhaps it truncates all but the last seven digits and then converts the score to a string which would imply removing leading zeros. In any case, the game's scoring system is glitched, which makes me wonder what type of Smart-Alec programmed it. Another interesting behavior is one with the clock at the end of a solve. This had a 1 in 60 chance and was not worth tripping intentionally, but in level 1-8, for example, immediately after the last 8 blocks on the board are cleared in a combo, the clock flipped from 1'00" to 0'59". This can happen only if the last clear is a combo or cascade because when it makes you wait 47 frames before intermission. The first of those 47 frames counts as still part of game play. In fact you can see this when a combo is used to finish a level that has an elevator, and seeing the elevator move for one more frame after the last clear.
Posted: 11/20/2011 1:05 AM
Quote
rchokler
Experienced Forum User, Published Author, Player (145)
Joined: 11/14/2011
Posts: 68
Location: Brookline, MA
Thank you. In fact none of these solutions can be executed by a normal person playing. Even in the few levels where no combos are formed, the the solutions are designed to include at least one part where a triple is cleared in a prohibitively difficult manner.
Posted: 11/19/2011 10:57 PM
Quote
rchokler
Experienced Forum User, Published Author, Player (145)
Joined: 11/14/2011
Posts: 68
Location: Brookline, MA
Thank's for the vote. I guess a better reason for using the old TAS was that by not making a copy in another directory and using the older file, it saves memory on the hard drive, even though the copy could be recycled when the analysis was done. However, the few levels used in the analysis were not altered for the new implementation of the TAS so it didn't really matter which version of the TAS I used for it.
Posted: 11/19/2011 4:13 PM
Quote
rchokler
Experienced Forum User, Published Author, Player (145)
Joined: 11/14/2011
Posts: 68
Location: Brookline, MA
OK, the flash patterns are not altered if you delay parts of the setup so that the combo starts clearing later. It is probably either the arrangement of the blocks that determines the pattern or else which blocks are being cleared in the combo. I used the older slower TAS for this analysis so that I don't ruin the faster TAS, but it applies to the faster TAS too. Here is an example map of a flash pattern with level 1-3 combo. A__ABAB BA_BB__ _ACC___ The period of a full flash is 6 frames for every block. In for the 11 blocks in 4-10 it is ABBBBB ACCCC_ All have a 4 frame period. In 13-10: B__ ABA _BA ___ ___ BA_ _A_ _B_ BB_ All flash with a 4 frame period.
Posted: 11/19/2011 2:10 PM
Quote
rchokler
Experienced Forum User, Published Author, Player (145)
Joined: 11/14/2011
Posts: 68
Location: Brookline, MA
The improvements take advantage of the two rules in the ROM behavior that Nitrodon said I originally overlooked. This resulted in several levels being cleared faster by about 22 frames and level 3.5 improved by 72 frames. I also wound up busting the 47 frame rule of combo/cascade bonuses in about 3 to 4 levels where noticeable amount of frames get saved from doing so. The 4-10/13-10 glitch (as I call it after the levels I discovered it to occur in) will be analyzed for the future, but for now, the latest run will be the one judged for potential publication. If I crack the root cause for this new glitch and it results in significant improvement in the total time for this TAS, I will make another submission to obsolete this one. The current hint is in the way the blocks flash during clearances, which determines how the 22 frame rule that Nitrodon mentioned works.
Posted: 11/18/2011 11:25 PM
Quote
rchokler
Experienced Forum User, Published Author, Player (145)
Joined: 11/14/2011
Posts: 68
Location: Brookline, MA
Never mind about that hunch on the 4-10/13-10 glitch. I just tried the trick I suspected by applying it on level 1-3 where the finishing combo is 12 blocks but it failed to trigger this glitch.
1 2 3