Editor, Experienced Forum User, Judge, Published Author, Expert player
(2081)
Joined: 8/21/2016
Posts: 1018
Location: US
Tichnickle,
In less than 10 minutes of testing, I was also able to beat you on the first 3 enemies of the game. I'm not sure if it's as fast as Spike's, but it's close. (My scream input does occur on frame 1737 which appears the same a Spike's image above, but I'm not sure if the video he shared is from that particular input).
Both Spike and I attack the left two enemies first. I believe this yields a faster time due to death animation timing with the game only having to process 1 enemy on the 2nd scream...but that's speculation and I can't (currently) prove it.
Doing it this way requires walking right further before doing the first scream, hence the attack not happening until frame 1737.
I can't comment on the glider as I haven't tested that far (nor am I particularly interested in doing so).
Another question: I know you've at least tested one other character; but have you tested ALL the other characters to confirm that this is indeed the best character to use for the game?
TASing is not about just doing something fast and quickly submitting.
It requires doing something, then doing something else, then something else, then something else, and so on...until you've tested every possibility you can imagine to see which is actually fastest. It's a time-consuming process and not something that typically occurs in a short span of time. I'm not trying to discourage you...simply trying to get you to slow down and pay more attention to the quality of what you're submitting instead of rushing to do submissions as quickly as possible.
I'll reiterate what EZGames69 warned earlier: If you continue to submit runs that are noticeably sub-optimal, you'll put yourself at risk of being banned from submitting.
I'd encourage you to cancel this submission and once again re-evaluate it. Spend more time with it. Try different methods...even actions that may not initially seem faster can ultimately yield a better overall time as the above situation proves -- delaying the first scream and changing it's direction actually yields a faster time to moving onward.
Editor, Experienced Forum User, Judge, Published Author, Expert player
(2081)
Joined: 8/21/2016
Posts: 1018
Location: US
I have to disagree on Star tier.
While this run is rather well made, it's just not entertaining enough to warrant Stars, in my opinion. I wouldn't argue with Moon tier (as it seems to have good general entertainment response so far), but I definitely don't find it on-par with other star tier runs.
Frankly, I find the game itself rather awful and unentertaining to watch. I'd truly be surprised if post-publication ratings (after giving adequate time for them to accumulate) would warrant Star tier.
Editor, Experienced Forum User, Judge, Published Author, Expert player
(2081)
Joined: 8/21/2016
Posts: 1018
Location: US
SonicFan53 wrote:
OK, so, I wanna collaborate on a TAS of this game.
You'll be much more likely to find someone willing to collaborate with you (on any game - let alone this one) if you first prove you are capable of making a publishable (or at minimum, a relatively optimized) TAS on your own.
Otherwise--and this is somewhat based on your posts in other threads of the forum--it comes across that you're just wanting to split the glory from someone else's hard work.
Essentially, show us you're worthy of collaborating with.
Many of us have numerous projects we're working on (within and without the TAS community); and we either don't have the time or simply don't want to hold someone else's hand through the TASing process if they aren't going to be a beneficial collaborator.
I do not intend for this post to come across as rude, more just to-the-point.
Editor, Experienced Forum User, Judge, Published Author, Expert player
(2081)
Joined: 8/21/2016
Posts: 1018
Location: US
Here's an update utilizing the shortcut shown above.
Due to changes in RNG, doors and some weapon firing/movement patterns were not as cooperative and thus I did not quite make the 600 frame savings estimated.
This file is still 448 frames faster than the original submission.
If I didn't perfectly re-sync the inputs after the shortcut, it's possible I may have missed an occasional frame or two somewhere.
Hmm. I'd be amazed if it didn't. From a quick estimate just looking at location, I'd guess probably around a 600-800 frame savings over this submission.
When I originally tested jumping down from there, I died. And given that I don't lose any health before that point, I assumed that falling from there was an instant death situation. I didn't consider that the death resulted from starting the game with less than maximum health.
So to utilize that shortcut, I'd have to pick up some health prior to that point. But the extra time for the health pickup (approximately 40ish frames) would be much shorter than the detour around this shortcut that is used in this submission.
As this is rather close to the beginning of the game, much of the later stuff would likely desync requiring the remainder of the game to be essentially redone/resynced.
I'm willing to redo this run adding in the shortcut. Most of the run would otherwise be the same visually. So...should I cancel this submission....or have you put it on a delayed judgement until I can update the run?
Editor, Experienced Forum User, Judge, Published Author, Expert player
(2081)
Joined: 8/21/2016
Posts: 1018
Location: US
Radiant wrote:
What's with the screen after the boss that you killed the wrong monster?
I believe this is simply an intended joke by the devs.
mklip2001 wrote:
The rerecord tally only says 3, though? Has Bizhawk had rerecord bugs lately?
As this run was created over the course of multiple BizHawk releases including dev builds, the final .bk2 was created with a copy/paste of those final inputs. The actual rerecord count is unknown and I really have no way to estimate.
Editor, Experienced Forum User, Judge, Published Author, Expert player
(2081)
Joined: 8/21/2016
Posts: 1018
Location: US
nymx wrote:
Wow! To start off...Yes vote.
I was surprised by this. DrD2k9 with another fine submission. I wondered about some of the slow downs, but I finally figured them out to be "hold ups" from other actions on the screen...correct?
Most visual slowdowns are due to lag from explosions, delaying motion to not take death dealing damage, or motion alterations to make sure the target location on screen is reached. Even still, there are few slowdowns and constant motion is maintained as much as possible.
Editor, Experienced Forum User, Judge, Published Author, Expert player
(2081)
Joined: 8/21/2016
Posts: 1018
Location: US
In the newest release (2.3.3), the option to Clone inputs in TAStudio is broken.
In the below image, the left is what I was trying to clone. The right is what resulted.
This occurs with both keyboard shortcut (Control + Insert) and using the Edit/right-click menu.
Editor, Experienced Forum User, Judge, Published Author, Expert player
(2081)
Joined: 8/21/2016
Posts: 1018
Location: US
I hate to have to say this, but in a way I'm glad I get to.
This movie isn't max score.
ViGadeomes had some ideas that panned out and we are working on some improvements to this run that will increase the score. The resulting video will be longer in time, but will be a higher score.
Editor, Experienced Forum User, Judge, Published Author, Expert player
(2081)
Joined: 8/21/2016
Posts: 1018
Location: US
My $.02 regarding timing convention between end-of-input vs reaching some 'end-game' point in the game's code:
Having the metric for timing to be variable on a game-by-game basis makes much more sense than holding all submissions to either one or the other perspective. For some games, there could be equal value from both perspectives; just different methods of execution. My understanding is that this is how things currently are done on the site, if there's a consensus on the timing method used; otherwise shortest input is typically the default.
As Personman said, the accomplishment of a 5 minute TAS for a 2 hour run is pretty superhuman, even if it may not be as satisfying to watch as a run that got to the end-game in 30 minutes but required input the whole time.
I personally don't have a problem with 2 different Monopoly runs published. I can enjoy and be impressed by either variation for this game.
Editor, Experienced Forum User, Judge, Published Author, Expert player
(2081)
Joined: 8/21/2016
Posts: 1018
Location: US
Asnivor wrote:
DrD2k9 wrote:
For others who may be curious. I've already completed a TAS of this, but there's something wrong with ZX emulation in BizHawk that results in the game's music not playing...I'm delaying submission until this problem is solved (a bug report has already been submitted).
This should now be resolved in the master branch (dev build). With any luck the changes wont have caused any desyncs, but that all depends on whether the game is trying to read from the AY register on the same (some might say non-standard) port that it is trying to write to. This seems unlikely, but check for desyncs anyway.
Yep works great. No desyncs. I'll looking at submitting this once the next official release happens.
For others who may be curious. I've already completed a TAS of this, but there's something wrong with ZX emulation in BizHawk that results in the game's music not playing...I'm delaying submission until this problem is solved (a bug report has already been submitted).
Editor, Experienced Forum User, Judge, Published Author, Expert player
(2081)
Joined: 8/21/2016
Posts: 1018
Location: US
EZGames69 wrote:
The only reason why you see temp encodes with not great quality is because people dont know how to make higher quality ones.
...or because some of us who make less than stellar quality temp encodes aren't as particular about the quality of a temporary encode knowing it's going to be re-encoded for publication anyway.
Editor, Experienced Forum User, Judge, Published Author, Expert player
(2081)
Joined: 8/21/2016
Posts: 1018
Location: US
InfamousKnight wrote:
I dont think so. Super mario bros 64 Zeropaige was recently fixed and it also works on linux. I think it might be the renderer.
SMB 64 doesn't work on the most recent windows release of bizhawk (2.3.2). There are graphical issues. This has been fixed in the developer builds. Is linux BizHawk based on these?
Editor, Experienced Forum User, Judge, Published Author, Expert player
(2081)
Joined: 8/21/2016
Posts: 1018
Location: US
feos wrote:
...Not having strictly capped max score makes it worse.
It may not have a strictly capped max score in the sense of 'the score is not allowed to go higher than xxxx.' But the coded movement mechanics (even as simplistic as they are) will ultimately limit the maximum achievable score to a particular value.
Therefore, a run that utilizes the game's coded mechancis to achieve this score will have achieved the 'max' score.
The only exception to this would be if someone discovered a glitch allowing for surpassing/bypassing the intended mechanics to yield a higher score. This would be analogous to the metroidvania examples mentioned earlier than can achieve greater than 100% due to new discoveries/glitches.
Editor, Experienced Forum User, Judge, Published Author, Expert player
(2081)
Joined: 8/21/2016
Posts: 1018
Location: US
To everyone who cares, and for what it's worth: This submission is definitively sub-optimal. I have indeed found input improvements that beat this submission's score. So submitting just those would show this run not to be max score.
Further, this submission uses PAL sync settings making the timing slower than it should be; the 1500m should take 3:33 according to the in-game timer, but it actually takes about 4:15 based on the encode. /the game should have been run in NTSC sync settings to begin with (I'm not even sure if I even knew much about sync settings at the time of this runs creation. If I did, this setting slipped my attention somehow).
Finally, I am currently re-working this game in the proper NTSC mode while also attempting to break down the events at a code level to determine a true maximum score per event based on the game mechanics. Once/If I can accomplish this, a new submission or userfile will be uploaded. That update should be definitively max score and therefore vault eligible.
Editor, Experienced Forum User, Judge, Published Author, Expert player
(2081)
Joined: 8/21/2016
Posts: 1018
Location: US
feos wrote:
The problem is quite simple and precise: the actual completion requirements are missing from the game.
While I do wish there were ways to make these kinds of games acceptable, I see this perspective and understand where it comes from.
I understood the original rejection reason, and I understand why the lack of clearly defined completion criteria could result in another rejection. Ultimately, you're the senior judge and the decision is up to you.
I, for one, feel it would be nice to have such runs on the site, although I don't have any further suggestions at this point on how to add/modify the rules to allow such games with undefined completion criteria. Side Note: I am open to contribute to discussion on helping find a way to make these type of games acceptable.
If nothing else, the re-judging of this run helps to further clarify the current rules we have by offering an example of what may not qualify.
Editor, Experienced Forum User, Judge, Published Author, Expert player
(2081)
Joined: 8/21/2016
Posts: 1018
Location: US
I'll admit that the following may appear biased because I'm the author of this submission, but I do believe the arguments would be valid for any game such as this.
I'd argue that a high score in itself is a meaningful completion criteria. For many older games, the whole design of the game was about achieving (and competing) for a high score table. While this game doesn't yield a high-score table, the completion criteria of achieving the highest possible score is still the end goal.
For this reason, games such as this (other decathlon games and likely some other genres) are achieving the intended completion goal...even if poor performance doesn't yield a different visual outcome.
We could consider it this way for TASing purposes: A run that fails to achieve a maximum score, fails to achieve the completion criteria of achieving the highest score for the various events. A run that achieves the highest possible score has achieved the completion criteria.
The biggest issue is "what is the highest possible score?"
Without the knowledge to break down the game code to mathematically determine this from the game's code, the only way to know the highest possible score is to achieve it through performance.
So for judging such games, I'd suggest that a run which seemingly can't be beaten from a score standpoint stands as an adequate value for this theoretical 'maximum' score and could be published. If someone else were to find a way to surpass that score (after the original has been published) then we learn that the original 'maximum' score wasn't actually the best and the new run would obsolete the old. This is essentially how we treat speed based runs already. We publish what we perceive as the fastest completion even if it hasn't been a guaranteed fastest by code breakdown; then we allow for obsoleting by a faster submission that we didn't previously realize was possible.
To put the comparison very simply:
Highest known achievable score = Fastest known achievable completion time
Newer and higher high score obsoletes older lower score = newer fastest completion time obsoletes older (no longer) fastest completion time
To uphold the site's general desire for speed, the fastest method of achieving the highest know score is preferred over a slower method of achieving the same score.
Both Spike and I attack the left two enemies first. I believe this yields a faster time due to death animation timing with the game only having to process 1 enemy on the 2nd scream...but that's speculation and I can't (currently) prove it.
Doing it this way requires walking right further before doing the first scream, hence the attack not happening until frame 1737.
I can't comment on the glider as I haven't tested that far (nor am I particularly interested in doing so).
Another question: I know you've at least tested one other character; but have you tested ALL the other characters to confirm that this is indeed the best character to use for the game?
TASing is not about just doing something fast and quickly submitting.
It requires doing something, then doing something else, then something else, then something else, and so on...until you've tested every possibility you can imagine to see which is actually fastest. It's a time-consuming process and not something that typically occurs in a short span of time. I'm not trying to discourage you...simply trying to get you to slow down and pay more attention to the quality of what you're submitting instead of rushing to do submissions as quickly as possible.
I'll reiterate what EZGames69 warned earlier: If you continue to submit runs that are noticeably sub-optimal, you'll put yourself at risk of being banned from submitting.
I'd encourage you to cancel this submission and once again re-evaluate it. Spend more time with it. Try different methods...even actions that may not initially seem faster can ultimately yield a better overall time as the above situation proves -- delaying the first scream and changing it's direction actually yields a faster time to moving onward.
then
This occurs with both keyboard shortcut (Control + Insert) and using the Edit/right-click menu.