Forums  /  Castlevania series  /  Castlevania III: Dracula's Curse  /  Castlevania collection and New Runs
  disdonndisdonn

Just a question about the new collection. Will the new runs be added as new platforms. Will we get collection categories if major changes in gameplay speed/emulation or will they be filed under emu runs? I'm mostly looking forward to practicing on Switch.

 
  PianotoriousPianotorious

Good question! Hopefully it's just going to run on a relatively standard emulator and we can handle it with the "emu" tag... otherwise we'd have no choice but to make it a separate category. Unless the differences are obvious we'll probably have to spot check the frame count on some easy-to-replicate sections -- ex. the falling bridge before Death since it's straightforward and also lags.

disdonndisdonn likes this. 
  disdonndisdonn

That was my guess. I'm hoping it's fairly straightforward emulation so I can get realistic practice on the go.

 
  disdonndisdonn

Good news, looks like we're getting both Famicom and NES (soon) in an update. That means both versions woo

 
  Chambers_NChambers_N

In my opinion, console should be its own category and everything else should be in an emulator category. One of the things that has always bothered me about people using emulator with Alucard route is that on console it is easy to fail the Down+A to transform into a bat, many of my runs have failed by not being precise enough, as well as the accidental up input which is more common with the D-Pad causing Alucard to activate the watch failing heart management. You can map two inputs to a single key on the keyboard and use that as your perfect bat transformation, even worse is not having an input display so that we can discern if this is occurring or not. Having a WR on this leaderboard done on emulator without input display has established a dangerous precedent and if such is retained then I suggest different categories because there are not checks to ensure level playing field actually exists. Even with input display I would still suggest a separation for additional reasons.

disdonndisdonn likes this. 
  Chambers_NChambers_N

Its not that its just easier to cheat, Emulators in general are not as accurate as people often say they are, so much so it offers significant advantages in games that have uninitialized RNG-related memory addresses, and games that are influenced by PPU/CPU alignment, among other things. As I had mentioned in another thread, Emulators start from an artificial, fabricated, Ram State as opposed to a natural state which is indicative of the Ram within the console. Emulators also fix the relationship between the PPU and CPU in a manner that is not representative of the variability that exists between these two chips in console.

How a player interfaces with the game has always been a topic of discussion, and I have maintained that in certain cases the use of a keyboard, for instance, can give a player a means of interfacing with the game that could minimize if not eliminate the type of input errors that are common with thumb on a D-Pad.

I understand that not everyone will concern themselves with these matters but some people do, so I can understand that the debate will continue to exist, but when time permits I like to add my own opinion to the discussion and hope that some may find it helpful.

Edit 1: I would prefer that we have a demonstrable level playing field instead of an assumed level playing field, and I think there is sufficient reason to suggest we employ what is easily available to us to make explicit that which we assume.

Edit 2: Relevant post where I talk about original hardware and emulators... https://www.speedrun.com/Batman_The_Video_Game_NES/thread/etb21/2#cqubm

 
  Chambers_NChambers_N

Initial Ram State - Hard Reset (Power on)
This presets the fabricated, artificial state. i.e. FF which is default in Bizhawk, or alternating 00 and FF every four addresses which is default in FCEUX.

Initial Ram State – Soft Reset (Reset)
The following relates to RNG-related, uninitialized memory addresses with Metroid in mind as a case. As the code runs it will alter the ram, meaning that whatever value was in the memory address initially, FF for instance, will change progressively down a very specific RNG track. If you reset, the last value in the memory address is retained since it is not initialized in our example, and will progress from that basis. The problem of requiring a soft reset for emulator players is that it changes the playing field once again, but this time a little in favor of console players. While the “randomized” reset does alter the fixed state, you can reset on a specific frame and pick right back up from what you had before, a means of manipulating the fabricated state. While console players could try a few consoles, and in the case of getting a perfect Kraid pattern can be found to be 99.9% consistent from power on without frame perfect resets, etc. A soft reset changes the scenario but it does not create a level playing field.

PPU/CPU Alignment – Hard and Soft Resets
In this case, neither the power on nor the soft reset addresses the fixed alignment since said alignment is always the starting basis in emulator whether you power on or reset. In this case, nothing is solved, the situation is unchanged.

 
  disdonndisdonn

I'm not entirely for or against the separation of emu runs, but the fact that you can filter them out is very helpful to see where you compare vs other people using the hardware. Sadly, it is difficult to get real hardware in some places (read as outside the US and Japan). It's fairly crummy how some games were never released in some regions, or were released in limited quantities making it harder to find. I still am hoping I can one day send a cart/console for PAL regions to my friend Kutsu (also on the boards) since he has no real options to get them himself due to availability/cost barriers. This problem will also slowly become more and more prevalent as hardware (both carts and consoles) break down over time. But i understand the arguments for/against using a separation between NES and EMU runs

 
  Chambers_NChambers_N

Another problem with the Anniversary Collection is that it has been reported to me from a few individuals that there is lag reduction in the emulation. If this is true then this is another reason to separate categories.

 
  Chambers_NChambers_N

It isn't about how well the leaderboard looks, it is about a level playing field. Nor is this about legitimate categories, if you want to compete within a certain set of parameters then you compete in that category, change the conditions then it goes into a different category. Nor is this about accessibility, its about competing in the category you are able. Therefore, the language of better looking leaderboards, imposing fake "legitimate" language, or speaking of accessibility all have nothing to do with establishing a level playing field for people to compete. If people want to compete against each other on Anniversary Collection then that's great they can compete with one another, but to match runs completed on Anniversary Collection with those who compete on the initial NES release is just not equivalent, you are not comparing apples with apples but comparing apples with oranges. Yes they are both fruit, and they both come from trees but they are not exactly the same. Anniversary Collection and NES releases may both be Castlevania, and they may both have been produced by Konami, but they are not exactly the same, they are not equivalent. Some people may not care about equivalence in this regard but many other people do, and it is hardly an odd notion. I am not a mod here, and the mods here will need to determine the type of leaderboard they want to have, and not everyone will agree with those decisions. All we can do is state our thoughts and then move on, otherwise we will simply be rehashing the same things over and over. In my approach to competition a Level Playing Field must be maintained at all costs, whatever consequences result from this are inconsequential to me, because I maintain that one of the most fundamental principles of competition is a Level Playing Field. This exists in many different shapes and forms such as in weight lifting. There is a category for men who take steroids and extra testosterone Etc and there is a category for people who don't, there's just a difference and therefore it is a different category as it should be. In Castlevania 3 on the Alucard route there is a ton of lag in the skeleton fight just before Frankenstein level, if the Anniversary Collection version greatly reduces the lag then it belongs in its own category because it's different, the playing field is no longer level, and one of the most fundamental pillars a competition is removed. If you remove one of the foundational elements of what makes competition a competition then it is no longer a competition it's only a friendly leaderboard, a fun collection of runs.

 
  disdonndisdonn

Unfortunately there's not enough outright changes to necessitate what happened with the Mario/Luigi U games being revamped for Switch (adding characters, etc) to warrant Anniversary to have it's own leaderboard. The actual changes will, considering most of these games have very minimal players on speedrunning (and likely won't gain a ton of new runners) will probably only be 10-20 seconds at most in the end, considering that the top tier runners are currently very near perfect on many of the games included, so that extra amount of time won't likely be saved by anyone anytime soon (unless everyone who has ever run the games decides to run on the new hardware). I see both sides of the argument and honestly don't know if there's going to be much of any actual impact until we start seeing a ton of new runs in old games with the anniversary collection.

 
  Chambers_NChambers_N

That is the question isn't it, how many changes is a person willing to allow before they would want it to be in its own category. There are tons of lines that can be drawn for sure depending on the principles followed, and mine is any changes at all that makes even one frame of difference. As a speedrunner, its not so much how far the world record holder is ahead of others, or whether it would make a practical difference at a certain point in time, or whether there is any interest in playing the game a certain way, it has to do with every player facing the same elements all the time. For example, for me the level of participation on a leaderboard is not a primary factor, or to say it another way will always be secondary to that fundamental principle of a level playing field. For other people though, they are more concerned about popularity, accessibility, relevance, aesthetics, etc, and while those are interesting aspects of any particular game or leaderboard, I see them as secondary to that which makes competition what it is. But once again that has to do with what particular type of leaderboard you want to have. There are some leaderboards that don't require video proof, some don't require verification, some allow save states, etc, etc, and those mods crafted those leaderboard to function in a very specific way. In this regard there really isn't a "correct" way to do it if it can be anything people want to make it into. In my mind though, I have a certain idea of what a leaderboard should be, how it should function, the type of rule set it should have, etc, and I apply that to everything I moderate on SR, and everyone else does the same.