Game Time vs Real Time

User avatar
JoeTheRapper
Youngster
Posts: 67
Joined: Thu Apr 24, 2014 8:57 pm

Re: [TALK] Game Time vs Real Time

Post by JoeTheRapper » Fri Apr 25, 2014 7:48 am

The post that showed that if we removed units of time, such as seconds or minutes, the wr becomes easier to tie and situations where one run is clearly better than the other can be considered a tie. There is nearly a minute gap between xx:00 and xx;59. Considering these two to be the same exact time seems like a poor idea in my eyes, as the minute can be lost on execution, causing the run to be objectively worse than the other. If we used seconds, we become capable of giving the title of wr to one single person. Game time doesn't even show accurate time whatsoever, I fail to understand why it is used. There are games where the game timer can be 10 minutes off of the real time. This is all for now.
WR is easy just don't fuck up.
twitch.tv/joetherapper
Showdown ladder peaks (old rating): XY OU 2000, BW OU 2300, XY LC 1800.

User avatar
jolteontrainer
Bug Catcher
Posts: 17
Joined: Thu Apr 24, 2014 12:23 am

Re: [TALK] Game Time vs Real Time

Post by jolteontrainer » Fri Apr 25, 2014 7:24 pm

werster wrote: tl;dr Records being tied is not a bad thing at all and anyone who disagrees needs to learn to think as an individual.

The game timer is accurate, even if it has a slight disparity to what we call a minute that is 100% irrelevant, the accuracy within itself and it's comparisons to other versions of itself are all that matter

Runs that S+Q have always used real time why is this even in the cons

The majority of other games either don't have an IGT or they have a terrible one this is a terrible point
Stop being a vicious cunt and just say this.

Based on this, there would be more competition to break that tie for more godly runs.
mew

User avatar
Masteri Mori
Bug Catcher
Posts: 14
Joined: Thu Apr 24, 2014 7:15 am

Re: [TALK] Game Time vs Real Time

Post by Masteri Mori » Fri Apr 25, 2014 11:38 pm

Another point to add is that the SGB has a considerable advantage over the regular GameBoy because it runs at 61.17fps instead of 59.73fps (source). Therefore, if we switch to RTA, the 'fastest' runs will be done on the SGB, but these are not necessarily the best, and will likely beat better runs, which is stupid. Also, this pretty much forces everyone to run on the SGB, which means that potential new runners will be alienated and dissuaded from picking up the games, and current (good) runners will have to needlessly spend money just to make their times competitive again.

So we could just ban the SGB, and force everyone to play on the GameBoy Player/Gambatte, right?
Well, this is exactly the same argument as above: What if someone got the God Run on SGB? Do we let a clearly better run not have 'WR' because of some arbitrary (and stupid) reason? I think not.

As you see, both options are stupid, and therefore so is switching to RTA. End of discussion.

User avatar
Shenanagans
Site Admin
Posts: 116
Joined: Fri Apr 11, 2014 3:59 am
Location: Ohio, USA
Contact:

Re: [TALK] Game Time vs Real Time

Post by Shenanagans » Sat Apr 26, 2014 3:20 am

We simply ban SGB outright as a community anyways mori. There are currently Gen 1/2 RTA speedruns (see 151, NSC, gold glitched) and it is against category rules to use a SGB for them. If we switched to RTA we would still ban SGB.
Feel free to follow me for all your Pokemon Resetting Needs!

Image Image Image

Vulajin
Schoolkid
Posts: 46
Joined: Thu Apr 24, 2014 4:47 am

Re: [TALK] Game Time vs Real Time

Post by Vulajin » Sat Apr 26, 2014 8:27 am

JoeTheRapper wrote:Game time doesn't even show accurate time whatsoever, I fail to understand why it is used. There are games where the game timer can be 10 minutes off of the real time. This is all for now.
This literally does not matter. Is game time consistent with itself? If you run the game twice in a row the *exact* same way, and complete the game in the exact same amount of real time/frames/whatever on the same console twice, do you get the same in-game time? If so, then the fact that it doesn't accurately reflect real time is completely irrelevant. No one in any other game community cares about this as long as the game time permits accurate comparison of runs (i.e. if game time is lower, the run was faster in a way that matters).
Chuckolator wrote:Hypothetical situation. Say humanity's accepted way of measuring time was different, and it was different in a way that the programmers for Pokemon games decided to display the final time in intervals of what we consider five minutes. Your Yellow 1:55 is now a 0:23, and so is gunner's 1:58. Would you call them equally good runs that should be listed the same? What if it were fifteen minutes? What if it displayed only the hour? The HG WR is now "4", and now someone with an equivalent of 4:59 also gets "4". Now they're tied for WR. Is there a problem with that? If so, why? They're both the same displayed time, is it such an atrocity for another person to get the title of WR holder?
This argument is fundamentally flawed because it assumes that the minute granularity is not a factor in our consideration of game time at all. Obviously if Pokemon games only timed to hour granularity, that would be not be sufficient to differentiate runs of differing quality. The size of the units matters a lot, and the fact that we are talking about seconds, which are in many cases largely inconsequential in comparing two Pokemon runs, completely changes the discussion from if we were talking about minutes.

There seems to be this common argument tactic where people seem to assume that, if someone advocates position X in situation Y, they must advocate position X in all possible situations. That doesn't make sense and assumes that people are incapable of understanding context. Every single game is different, and every category is different. Super Mario Bros. runs are counted down to the exact number of frames, and in fact, they even rely on a special measurement technique using RNG replication on an emulator to get an exact count instead of using a video count (which could include lag frames or frame rate conversion issues). Are you suggesting we do that for Pokemon? Would that not be ludicrous? If you agree that it would be, then I hope you would agree that context matters.

For Pokemon speedruns, we have an in-game timer that we have used for years. It does not provide second granularity, but that is okay because seconds are not an especially compelling level of granularity for a speedrun on the order of 2+ hours. We could switch to real-time to obtain that granularity, but that introduces a host of issues with regard to platform differences that we simply don't have to deal with right now. Switching to real-time would make some of our games more friendly to newcomers, but if you seriously think we have a problem getting new runners, you must be willfully ignoring the huge number of those over the past year. Some categories use glitches that prevent the use of the in-game timer, but these categories already use real time and there's no reason they can't continue to do so.

Unless someone can suggest a very strong reason that the status quo should be overtuned - a stronger reason than "some other games use real time", "WRs shouldn't be tied," or "it would help newbies" - I don't see why we should suddenly switch from game time to real time.

Sanqui
Jr. Trainer
Posts: 93
Joined: Thu Apr 24, 2014 10:05 am
Contact:

Re: [TALK] Game Time vs Real Time

Post by Sanqui » Sat Apr 26, 2014 9:33 am

Can't we just switch to real time but count all times within a 60 second range as ties
It's less stupid than game time because in game time you can have two times that are off by a second not be a tie so you should like this if you only like game time for ties

Like I'd seriously be in favour of this, no Kappa
I am the person who has ruined Gold Glitched forever and seconhandedly, Red NSC too
Ask me to make a dumb romhack for you!
Image

Jul, a board.

User avatar
heero_fred
Youngster
Posts: 70
Joined: Thu Apr 24, 2014 2:37 am
Location: Mexico
Contact:

Re: [TALK] Game Time vs Real Time

Post by heero_fred » Sat Apr 26, 2014 1:35 pm

For Pokemon speedruns, we have an in-game timer that we have used for years. It does not provide second granularity, but that is okay because seconds are not an especially compelling level of granularity for a speedrun on the order of 2+ hours.
I have to disagree with that statement, seconds are important no matter how long or short the run is, losing over 30 seconds because of excecution is as big as a time lose in a half an hour game as well as a 5 hours game.

now in the case of ties:

I really don't mind them when both runs are equal in excecution,

example: run A =1:55(:10) IGT and run B= 1:55(:50) IGT

run B is 40 seconds slower because of luck in getting unecessary crits texboxes, in that case, I don't see anything wrong with a tie, but there is a scenario I haven't seen being mentioned:

what if run A lost over 20 seconds to the same reason and missed the 1:54 IGT?

Now, I'm not saying we should change to RTA (screw version differences) but a standard timing would be ideal, if everyone starts and end the timer at the same points, a better comparison can be achieved, because a 59 seconds range is a huge range

User avatar
Masteri Mori
Bug Catcher
Posts: 14
Joined: Thu Apr 24, 2014 7:15 am

Re: [TALK] Game Time vs Real Time

Post by Masteri Mori » Sat Apr 26, 2014 3:17 pm

Shenanagans wrote:We simply ban SGB outright as a community anyways mori. There are currently Gen 1/2 RTA speedruns (see 151, NSC, gold glitched) and it is against category rules to use a SGB for them. If we switched to RTA we would still ban SGB.
Oh, that's true, but I don't like the idea of a good run being discounted simply because it's on the wrong console. I'm the kind of person who's stubborn enough to grind out runs on the SGB until I get an amazing run just to make a point (faster resets Kappa )

We're also assuming that our timers measure time perfectly, when in fact they do not, as there is some variation due to clock speed (?) and whatnot. I don't know what the difference can be over the course of a 2 hour run, but it may be significant. The best timer, for any speedrun, should be one that literally counts the frames, which is what the in-game timer already does.
Sanqui wrote:Can't we just switch to real time but count all times within a 60 second range as ties
It's less stupid than game time because in game time you can have two times that are off by a second not be a tie so you should like this if you only like game time for ties

Like I'd seriously be in favour of this, no Kappa
Either you mean all 1:59:xx runs are counted as ties, for example, but we do this already with the in-game timer, which literally counts frames, as discussed above, or you mean that a 1:59:40 is tied with a 2:00:10, and a 2:00:10 is tied with a 2:00:40 but the 1:59:40 is not tied with a 2:00:40, and ties not being transitive is a really bizarre concept, and will likely confuse people.

Sanqui
Jr. Trainer
Posts: 93
Joined: Thu Apr 24, 2014 10:05 am
Contact:

Re: [TALK] Game Time vs Real Time

Post by Sanqui » Sat Apr 26, 2014 3:37 pm

Masteri Mori wrote:
Sanqui wrote:Can't we just switch to real time but count all times within a 60 second range as ties
It's less stupid than game time because in game time you can have two times that are off by a second not be a tie so you should like this if you only like game time for ties
Either you mean all 1:59:xx runs are counted as ties, for example, but we do this already with the in-game timer, which literally counts frames, as discussed above, or you mean that a 1:59:40 is tied with a 2:00:10, and a 2:00:10 is tied with a 2:00:40 but the 1:59:40 is not tied with a 2:00:40, and ties not being transitive is a really bizarre concept, and will likely confuse people.
I do mean the latter. And a tie only makes sense if it's the WR, so it's just the WR that would be a tie between everybody holding a time between the best time and a time a minute worse.
I mean it's the best of both worlds. And if the ties are the only reason you like game time, it's more accurate and fair this way.
I am the person who has ruined Gold Glitched forever and seconhandedly, Red NSC too
Ask me to make a dumb romhack for you!
Image

Jul, a board.

User avatar
Shenanagans
Site Admin
Posts: 116
Joined: Fri Apr 11, 2014 3:59 am
Location: Ohio, USA
Contact:

Re: [TALK] Game Time vs Real Time

Post by Shenanagans » Sat Apr 26, 2014 4:03 pm

Masteri Mori wrote: We're also assuming that our timers measure time perfectly, when in fact they do not, as there is some variation due to clock speed (?) and whatnot. I don't know what the difference can be over the course of a 2 hour run, but it may be significant. The best timer, for any speedrun, should be one that literally counts the frames, which is what the in-game timer already does.
If we switched to RTA we would have measures to to determine time much more accurately then we currently do. We would re-time runs at the top level as well to get a more accurate reading similar to what other communities do (like OOT Any%)

My fear is that some emulators might run faster but Gambette and Bizhawk seem to be very accurate. But at the top level if a run comes down to frames They will likely be tied to the second anyways.
Feel free to follow me for all your Pokemon Resetting Needs!

Image Image Image

User avatar
JoeTheRapper
Youngster
Posts: 67
Joined: Thu Apr 24, 2014 8:57 pm

Re: [TALK] Game Time vs Real Time

Post by JoeTheRapper » Mon Apr 28, 2014 5:18 pm

"This literally does not matter. Is game time consistent with itself? If you run the game twice in a row the *exact* same way, and complete the game in the exact same amount of real time/frames/whatever on the same console twice, do you get the same in-game time?"

You do raise a good point here. However the game I mentioned has a game timer that's not consistent with itself. It's not pokemon, though. I merely dislike the Idea of having two seperate runs at far ends of a minute count as the exact same time. I do think that we would encounter large inconveniences by switching to real time, so I actually would agree that as long as we can't recalc the current records, we should stick to ingame time.
WR is easy just don't fuck up.
twitch.tv/joetherapper
Showdown ladder peaks (old rating): XY OU 2000, BW OU 2300, XY LC 1800.

danschemen
Preschooler
Posts: 1
Joined: Mon Apr 28, 2014 5:50 pm

Re: [TALK] Game Time vs Real Time

Post by danschemen » Mon Apr 28, 2014 5:53 pm

I say just switch to real time. The same discussion was going on just a little while ago in the RE community. Real time is just better to go by even if it makes it harder for new people that just play casually to join.

User avatar
BobChao87
Bug Catcher
Posts: 11
Joined: Thu Apr 24, 2014 4:23 pm

Re: [TALK] Game Time vs Real Time

Post by BobChao87 » Mon Apr 28, 2014 9:24 pm

While I do not intend to weigh in on the matter at hand, a major point of confusion is the inaccuracy of the timers. It is already common practice to refer to runs in the community as "1 5 7" as opposed to calling it "one hour fifty-seven minutes". There is a notable lack of units in the common usage. This is good as, in order to avoid confusion, it would be best to not think of the IGTs are being in hours or minutes, but rather separate units entirely.

1 hour prime = 60 minutes prime
1 minute prime = 60 seconds prime
1 second prime = 60 frames
59.7275 frames = 1 traditional second

The times are not inaccurate, but rather are in a slightly different unit base.

User avatar
Sup3rH3nry
Preschooler
Posts: 9
Joined: Thu Apr 24, 2014 2:59 pm

Re: [TALK] Game Time vs Real Time

Post by Sup3rH3nry » Tue Apr 29, 2014 1:25 am

In my opinion it shouldn't be this big of a deal. In game time is much more convenient. While it is more accurate to use real time to break ties, having a tied world record isn't necessarily a bad thing. 2 people got the same in-game time, both should be happy. The person who did it slightly faster shouldn't bitch about that they were faster and so it isn't a tied WR (Making up a scenario, nobody's done this Kappa ) Both people should be fine. Basically by putting in in-game time you are saying that we want less people to be satisfied with their runs.

One vote for In game time from me.

User avatar
JoeTheRapper
Youngster
Posts: 67
Joined: Thu Apr 24, 2014 8:57 pm

Re: [TALK] Game Time vs Real Time

Post by JoeTheRapper » Tue Apr 29, 2014 7:10 am

After reviewing this a bit more, I think that we should stick to IGT. We can't really change it, this is a popular game to run. Knowing this, nearly everyone that has had a run would need to either find their RTA time, or have to recalc their time, making their run seem less impressive than before. Using IGT lets people be happier with their runs, and a minute lost to RNG can be completely ignored if IGT is kind.
WR is easy just don't fuck up.
twitch.tv/joetherapper
Showdown ladder peaks (old rating): XY OU 2000, BW OU 2300, XY LC 1800.

tetonator
Preschooler
Posts: 9
Joined: Thu Apr 24, 2014 10:34 am

Re: [TALK] Game Time vs Real Time

Post by tetonator » Fri May 02, 2014 11:06 am

The sad part is that there actually is a second timer (@0xDA44 in ram) in the game, but it is not shown at the end :(

Sanqui
Jr. Trainer
Posts: 93
Joined: Thu Apr 24, 2014 10:05 am
Contact:

Re: [TALK] Game Time vs Real Time

Post by Sanqui » Fri May 02, 2014 1:49 pm

I've actually made a hack at some point which shows the exact time, including seconds and frames, at the hall of fame. It was just some fifteen bytes.
I am the person who has ruined Gold Glitched forever and seconhandedly, Red NSC too
Ask me to make a dumb romhack for you!
Image

Jul, a board.

User avatar
JoeTheRapper
Youngster
Posts: 67
Joined: Thu Apr 24, 2014 8:57 pm

Re: [TALK] Game Time vs Real Time

Post by JoeTheRapper » Fri May 02, 2014 4:15 pm

Sanqui wrote:I've actually made a hack at some point which shows the exact time, including seconds and frames, at the hall of fame. It was just some fifteen bytes.
That sounds pretty cool, but how could we get an entire community to go by that timing?
WR is easy just don't fuck up.
twitch.tv/joetherapper
Showdown ladder peaks (old rating): XY OU 2000, BW OU 2300, XY LC 1800.

Izraill
Preschooler
Posts: 6
Joined: Fri May 02, 2014 6:56 pm

Re: [TALK] Game Time vs Real Time

Post by Izraill » Fri May 02, 2014 7:03 pm

JoeTheRapper wrote: That sounds pretty cool, but how could we get an entire community to go by that timing?
That hack could be made into a separate timer program that looks into the game memory, and that timing could be used. However, it would probably be for emu only, or require a lot of stuff to be able to use that program on a console.

Sanqui
Jr. Trainer
Posts: 93
Joined: Thu Apr 24, 2014 10:05 am
Contact:

Re: [TALK] Game Time vs Real Time

Post by Sanqui » Fri May 02, 2014 7:47 pm

JoeTheRapper wrote:
Sanqui wrote:I've actually made a hack at some point which shows the exact time, including seconds and frames, at the hall of fame. It was just some fifteen bytes.
That sounds pretty cool, but how could we get an entire community to go by that timing?
Given how many people play on console, there's no way (and it's probably not a good idea to make the community switch to a hack, either).
But it was pretty cool.
Image
I am the person who has ruined Gold Glitched forever and seconhandedly, Red NSC too
Ask me to make a dumb romhack for you!
Image

Jul, a board.

luckytyphlosion
Pokémon Trainer
Posts: 182
Joined: Thu Apr 24, 2014 2:26 am

Re: [TALK] Game Time vs Real Time

Post by luckytyphlosion » Fri May 02, 2014 11:42 pm

plz, just pre-8f the save file to show seconds (and frames Kappa)
Jack-Flys are OpieOP.

"GO ASK SANQUI" - luckytyphlosion whenever someone asks about how something in Pokémon Red/Blue works. (this means you exarion Kappa)

tetonator
Preschooler
Posts: 9
Joined: Thu Apr 24, 2014 10:34 am

Re: [TALK] Game Time vs Real Time

Post by tetonator » Sat May 03, 2014 8:29 am

Well you coul release it for people using emulator, doenst hurt :P

Sanqui
Jr. Trainer
Posts: 93
Joined: Thu Apr 24, 2014 10:05 am
Contact:

Re: [TALK] Game Time vs Real Time

Post by Sanqui » Sat May 03, 2014 8:40 am

luckytyphlosion wrote:plz, just pre-8f the save file to show seconds (and frames Kappa)
That's actually just not possible :P Only way I can think of is hooking up the OAM DMA routine, but that would require you to run 8F ingame, and might lag the game.
tetonator wrote:Well you coul release it for people using emulator, doenst hurt :P
http://sanqui.rustedlogic.net/etc/poker ... e-timer.gb
I am the person who has ruined Gold Glitched forever and seconhandedly, Red NSC too
Ask me to make a dumb romhack for you!
Image

Jul, a board.

tetonator
Preschooler
Posts: 9
Joined: Thu Apr 24, 2014 10:34 am

Re: [TALK] Game Time vs Real Time

Post by tetonator » Sat May 03, 2014 8:50 am

Sanqui wrote:
luckytyphlosion wrote:plz, just pre-8f the save file to show seconds (and frames Kappa)
That's actually just not possible :P Only way I can think of is hooking up the OAM DMA routine, but that would require you to run 8F ingame, and might lag the game.
tetonator wrote:Well you coul release it for people using emulator, doenst hurt :P
http://sanqui.rustedlogic.net/etc/poker ... e-timer.gb
Thanks buddy :D

Thomaz
Preschooler
Posts: 5
Joined: Fri May 09, 2014 9:15 am

Re: [TALK] Game Time vs Real Time

Post by Thomaz » Fri May 09, 2014 9:23 am

I have one question that's important for me.

I do segmented runs of Blue and Red. Right now, as far as I know, I have the segmented WR for glitched Red any% (using Item Underflow). How would you guys handle timing such a run? Especially interesting because there are no real segmented runs on the site at all (a few S+Q's here and there doesn't count as a segmented run imo).

Locked

Return to “Generation 1”

Who is online

Users browsing this forum: No registered users and 2 guests