Thursday, July 24, 2025

"n% faster/slower" is misleading

Suppose you wanted to promote an upgrade from an RTX 3080 card to the RTX 5080. To do this you could say:

"According to the PassMark G3D score, the RTX 5080 is 44% faster than the RTX 3080."

However, suppose that instead you wanted to disincentivize the upgrade. In that case you could say:

"According to the PassMark G3D score, the RTX 3080 is only 30.6% slower than the RTX 5080."

Well, that can't be right, can it? At least one of those numbers must be incorrect, doesn't it?

Except that both sentences are correct and accurate!

And that's the ambiguity and confusion between "n% faster" and "n% slower". The problem is in the direction we are comparing, in other words, in which direction we are calculating the ratio between the two scores.

The RTX 3080 has a G3D score of 25130.

The RTX 5080 has a G3D score of 36217.

If we are comparing how much faster the latter is to the former, in other words, how much larger the latter score is than the former score, we do it like:

36217 / 25130 = 1.44118  →  44.1 % more (than 1.0)

However, if we are comparing how much slower the former is than the latter, we would do it like:

25130 / 36217 = 0.693873  →  30.6 % less (than 1.0)

So both statements are actually correct, even though they show completely different percentages.

The fundamental problem is that this kind of comparison is mixing ratios with subtractions, which leads to uneven results depending on which direction we are making the comparison. When only one of the ratios is presented (as is most usual), this can skew the perspective of how much the performance improvement actually is.

A more unambiguous and accurate comparison would be to simply give the factor. In other words:

"According to the G3D score, the speed factor between the two cards is 1.44."

However, this is a bit confusing and not very practical (and could also be incorrectly used in comparisons), so an even better comparison between the two would be to just use example frame rates. For example:

"A game that runs at 60 FPS on the RTX 3080 will run at about 86 FPS on the RTX 5080."

This doesn't suffer from the problem of which way you are doing the comparison because the numbers don't change if you do the comparison in the other direction:

"A game that runs at 86 FPS on the RTX 5080 will run at about 60 FPS on the RTX 3080." 

Sunday, July 20, 2025

What the 2008 game Spore should have been like

Like with so many other examples, the 2008 video game "Spore" generated quite some hype prior to its launch, but ended up with a rather lukewarm reception at the end. Unsurprisingly, the end product was nowhere near as exciting and as expansive as the prelaunch marketing hype made you believe.

It was supposed to be one of those "everything" games: It would simulate the development of life from unicellular organisms to galaxy-spanning civilizations, and everything in between, giving players complete freedom on how the species would evolve in between, building entire planetary and galactic civilizations!

The hype was greatly enhanced by the creature editor from the game being published as an independent application prior to launch, giving a picture of what would be possible within the game. And, indeed, by 2008 standards the creature editor was quite ahead of its time, and literally millions of players made their own creations, some of them absolutely astonishing and awesome, and likely better than even the game developers themselves could ever imagine. (For example, people were creating full-sized Tyrannosaurs Rex-like fully animated creatures, which seemed impossible when you first started using the editor, but players found tricks and ways around the limitations to create absolutely stunning results.)

Unsurprisingly, the game proper didn't live up to the hype at all.

Instead of an "everything" game, a "life simulator" encompassing all stages of life from unicellular organisms to galaxy-spanning civilizations, it was essentially just a collection of mini-games that you had to play completely linearly in sequence (with no other options!) until you got to the actual "meat" of the game, the most well-developed part of it, ie. the final "space stage", which is essentially a space civilization simulator.

The game kind of delivered the "simulate life from the very beginning" aspect, but pretty much in name only. Turns out that the game consists of five "stages":

  1. Cell stage, where you control a unicellular creature trying to feed, survive and grow.
  2. Creature stage, where you jump to a multi-cellular creature, where the creature editing possibilities start kicking in.
  3. Tribal stage, at the beginning of which the final form of the creature is finalized and locked, and which is a very small-scale "tribal warfare" simulator of sorts.
  4. Civilization stage, which now has turned into a somewhat simplistic, well, "Civilization" clone, where you manage cities and their interactions with other cities (trading, wars, etc.)
  5. And finally the actual game proper: The space stage, where you'll be spending 90+ % of your playthrough time. This is essentially a galaxy civilization simulator, and by far the most developed and most feature-rich part of the game.

The major problem with all of this is that every single stage is completely independent of every previous stage. Indeed, it literally doesn't matter what you do in any of the previous stages: It has pretty much no effect on the next stage. The only major lasting effect is a purely cosmetic one: The creature design you created at the transition between the creature and tribal stages will be the design shown during the rest of the game. And that's it. That's pretty much the only thing where one stage affects the next ones. And it is indeed 100% cosmetic (ie. it's not like your creature design affects eg. how strong or aggressive the creatures are, for example.)

The other major problem is that the first four stages are relatively short, they have to be played in linear order, and are pretty much completely inconsequential. While each stage is longer than the previous ones, the first four are still quite short (you could probably play through all four of them in an hour or two at most, even if you aren't outright speedrunning the game.)

In other words, Spore is essentially a galactic civilization simulator with some mandatory mini-games slapped at the start. In no way does it live up to the "everything" hype it was marketed as.

Ironically, the developers originally planned for the game to have even more stages than those five (including an "aquatic stage", which I assume would have been between the cell and creature stages, as well as a "city stage" which would have been between the tribal and civilization stages.)

How I think it should have been done instead:

Rather than have a mandatory and strictly linear progression between the stages (which is a horrible idea), start directly at the galactic simulator stage.

In this stage there could be hundreds or even thousands of planets with life forms at different stages (including those that were planned but not implemented). The player could "zoom in" into any of these planets and observe what's happening there and even start playing the current stage on that planet, in order to affect and speed up the advancement of that particular civilization, and create all kinds of different-looking creatures on different planets.

In fact, the player could "seed" a suitable planet by adding single-celled organisms there, which would start the "cell stage" on that planet (which the player could play or just allow to be automatically simulated on its own). If the planet isn't suitable, then he could have one of his existing civilizations terraform it.

The stages themselves should have been developed more, made longer and more engaging and fun, which would entice the player to play them again and again, on different planets.

Moreover, and more importantly: Every stage should have strong effects on the next stages on that particular planet: The choices that the player makes in one stage should have an effect on the next stages. For example certain choices could make the final civilization very intelligent and peaceful, and very good at trading. Other choices made during the different earlier stages could make the civilization very aggressive and prone to conquer other civilizations and go to war with them. And myriads of other choices. (These traits shouldn't be too random and unpredictable: The player should be allowed to make conscious choices about which direction the species goes towards. There could be, for example, trait percentages or progress bars, and every player choice will display how much it affects each trait.)

That would actually make the "mini-games" meaningful and impactful: You shape a particular civilization by how you play those mini-games! The choices you make in the earlier stages have an actual impact and strongly shape what the final civilization will be like, what it's good at, and how it behaves. 

Monday, July 14, 2025

Would The Truman Show have been better as a mystery thriller?

The 1998 film The Truman Show is considered one of the best movies ever made (or if not perhaps in the top 100, somewhere in the top quarter of all movies at least, depending on who you ask).

The film takes the approach where its setting is made clear to the viewers from the very beginning, from the very first shots. In fact, it's outright explained to the viewer in great detail, so there is absolutely no ambiguity or lack of clarity. The viewer knows exactly what the situation is and what's happening, and we are just witnessing Truman himself slowly realizing that something is not as it seems. There isn't much suspense or thriller about the movie because of this, and it's more of a comedy-drama.

The environment in the movie is also quite deliberately exaggerated (because there's no need to hide anything from the viewers, or try to mislead them in any way). The town doesn't even look like a real town and more like an artificial amusement park recreation "town". And that's, of course, completely on purpose: This isn't even supposed to look like an absolutely realistic place. (Of course Truman himself doesn't know that, which is the core point of the movie.) In other words, the movie veers more towards the amusing comedy side than on the realistic side, deliberately. The viewer's interest is drawn to how Truman himself reacts when he slowly starts suspecting and realizing that everything is not right or normal, and how the actors and show producers try to deal with it.

But one could ask: Would it have made the movie worse, the same, or perhaps even better, if it had been a mystery thriller instead? Or would that have been too cliché of a movie plot?

In other words:

  • The viewer is kept in the dark about the true reality of things, and knows exactly as much as Truman himself does. The true nature of what's happening is only revealed as Truman himself discovers it, with the big reveal only happening at the very end of the movie. Before that, it's very mysterious both to Truman and to the viewer.
  • The town is much more realistic and all the actors behave much more realistically, so as to not raise suspicion in the viewer. Nothing seems "off" or strange at first, and it just looks like your regular small American town with regular people somewhere in some island. It could still be a very bright and sunny happy town, but much more realistically depicted.
  • The hints that something may not be as it seems start much more slowly and are much subtler. (For example, no spotlight falling from the sky at the beginning of the movie. The first signs that something might be off come later and are significantly subtler than that.)
  • For the first two thirds of the movie the situation is kept very ambiguous: Is this a movie depicting a man, ie. Truman, losing his mind and becoming paranoid and crazy, or is something else going on? The movie could have been made so that it's very ambiguous if the things Truman is finding out are just the result of his paranoia, or something else. The other people in the movie are constantly "gaslighting" both Truman and the viewer in a very plausible and believable way that he's just imagining things.
  • The reveal in the end could be a complete plot twist. The movie could have been written and made in such a way that even if the viewer started suspecting that Truman isn't actually going crazy and the things he's noticing are actually a sign of something not being as it seems, it's still hard to deduce what the actual situation is, until the reveal at the very end.

Would the movie have been better this way, or would it just have been way too "cliché" of a plot? Who knows. 

Thursday, July 3, 2025

Why I don't watch many speedruns nowadays

About twenty years ago I was a huge fan of watching speedruns. At the time, speedruns of Quake, Doom, Half-Life and Half-Life 2 were some of my all-time favorites. And, in fact, still are (particularly the Quake ones.) Back then I used to watch as many speedruns of as many games I could, as most of them were really interesting. Before YouTube this was a bit more inconvenient, but especially after YouTube was created and speedrunners and speedrunning sites started uploading there, it was a treat.

Glitch abuse was significantly rarer back then (for the simple reason that speedrunners were yet to discover most of the ones that are known today), but even then I found it fascinating. I oftentimes read with great interest the technical descriptions of how particular glitches worked and how they were executed.

The one thing I loved most about speedruns, particularly those of certain games, was the sheer awesome skill involved. Quake and Half-Life 2 (back then) were particularly stellar examples, with the runner zooming through levels at impossible speeds, doing maneuvers that felt almost impossible. It was like watching an extremely skilled craftsman perform some complicated task at an incredible speed and precision. It was absolutely fascinating.

But then, over the years, slowly but surely things started changing.

The main thing that started changing was that speedrunners became enamored with finding and using glitches to make their runs even faster. In fact, many speedrunners became outright "glitch hunters": They would meticulously research, study and test their favorite speedrunning games in order to see if they could figure out and find glitches that would help in completing the games faster. It became a source of great pride and accomplishment when they could announce yet another glitch that saved time, or a setup to make an existing glitch much easier to perform.

Thus, over the years glitch abuse in speedruns started becoming more and more common.

And the thing is, the domain for glitch hunting started being expanded more and more. Not only were they trying to find glitches that could be exploited from within the gameplay itself, using the in-game mechanics themselves, but they started looking more towards the outside for ways to glitch the game: Rather than keep the glitch abuse restricted purely within the confines of the gameplay proper, they started hunting glitches outside of it: In the game's main menu, the game's save and load mechanics, in the options menu, sometimes even completely outside the game itself (which became particularly common in console game speedrunning, ie. trying to find ways to manipulate the console hardware itself in order to affect the game.)

It was precisely Half-Life 2 speedrunning where I started to grow a dislike for these glitches for the first time. You see, back then speedrunners had found that a particular skip could be performed by quick-saving and quick-loading repeatedly. And not just a few times, but literally hundreds of times! That's right: At one point it became so bad that a Half-Life 2 speedrun could literally spend something like 10 minutes doing nothing but quick-saving and quick-loading hundreds of times in quick succession. (I believe that other techniques have since been found that have obsoleted this particular mechanic. I haven't really checked. Doesn't make much of a difference to my point, though.)

I grew a great distaste for this particular glitch execution because it just stepped so far outside the realm of playing the game itself. It was no longer showing great skill at playing the game, and playing within the confines of the gameplay proper. Instead, it was stepping outside of the gameplay proper and affecting it effectively from the outside (after all, quick-saving and quick-loading are not part of the gameplay proper, part of playing the game, advancing towards the end goal. They are a meta-feature that are not part of the gameplay itself.)

I endured this for some years, but at some point I just outright stopped watching Half-Life 2 speedruns. They had become nothing but boring glitch-fests with very little of the original charm and awe left in them anymore. Sure, the runner would still fly through stages at impossible speeds, but this would be marred by boring out-of-game glitch abuse. I just lost interest.

While Half-Life 2 was one of the first games where extremely heavy glitch-hunting and glitch-abuse, particularly of the "abusing non-gameplay meta-features" kind happened, rather obviously it was not the only one. The habit started spreading among all speedrunners of most games like wildfire.

One particularly bad example made me outright want to puke: In a speedrun of the game The Talos Principle, the runner at one point would go to the main menu, go to the game options, set the framerate limit to 30 frames per second, return to the game, perform a glitch (that could only be done with a low framerate), and afterwards go back to the options menu and set the framerate back to unlimited. This was so utterly far-removed from gameplay proper, and was just so utterly disgusting, that I just stopped watching the speedrun right then and there.

Of course there are myriads and myriads of other examples which, while not as disgusting as that, just make the speedruns boring. For example I was watching a speedrun of Dark Souls 3, and every few minutes, even several times a minute, the speedrunner would quit to the main menu and immediately load back in. Why? Because loading times did not count towards the total time of the speedrun, and doing that quit&resume would move the player to a particular location within the level.

The thing is, those particular locations were usually just a few seconds of running from where the speedrunner would quit&resume, and while quit&resuming took something like 10-15 seconds, that loading time wasn't counted towards the speedrun's time, and thus while they made the speedrun overall longer in duration, it saved a couple of seconds each time by the speedrun's official clock. And the speedrunner would do this literally hundreds of times during the run. This was so utterly boring and outright annoying to watch that, once again, I just stopped watching mid-way through.

There would be literally dozens and dozens such examples I could write about, but let me add one more, a very recent one: Very recently Legend Of Zelda: The Wind Waker speedruns have been completely overhauled by a new glitch that has become possible.

Not "discovered". Not "found a new setup that allows doing it more easily." But outright became possible. And what made it possible? The Switch 2, that's what. You see, the Switch 2 runs the game under a new internal emulator which has this curious feature that if the emulated game crashes, the user is allowed to just keep the emulated game running rather than resetting the emulator. And it so happens that one out-of-bounds glitch in Wind Waker causes the game to crash in the original GameCube, but not in the Switch 2, if you opt to allow the game to keep running after the crash. Turns out that after a minute or two the game somehow "recovers" and starts running again... with the playable character being in a completely different room, allowing for big skips.

That's right: This glitch now abuses emulation to make it possible, and it's only possible on the Switch 2, not in the original console. And this is allowed only because the emulator is an official one by Nintendo (such emulator-only glitch abuse is never allowed if using third-party emulators. But apparently it somehow makes it different if the emulator is an official one.)

I can't decide if this is less or more disgusting than the Talos Principle glitch abuse. They are closely matched.

Overall, this is just the top of the iceberg: Glitch abuse has become so utterly prevalent in speedrunning, and such a huge portion of it abuses glitches that effectively affect the gameplay "from the outside", via non-gameplay means, that it has pretty much ruined speedrunning for me. With the exception of just a handful of games (such as Quake, thankfully), long gone are the days when speedruns would just run through the game via sheer skill, without resorting to disgusting outside-of-the-game glitch manipulation.

But what about speedruns that only use "within-the-game" glitches and at no point venture out of gameplay proper? In other words, they never use saving and loading, never to go to the game's menu, never affect the gameplay proper in any way from "outside" of it? Are those A-ok in my books?

Well, for the longest time I didn't mind those glitches and those speedruns. After all, they were effectively "legit" in my book, by my own standards. What's there to complain?

Yet, in later years I have grown tired of those too. The more the speedrun glitches the game, even if it happens fully from within gameplay proper, the more boring I tend to find it. Out-of-bounds glitches in particular I find boring, particularly those that skip huge chunks of levels. Many of them just bypass what made speedruns originally so great entertainment in the first place: Seeing an extremely skillful player beat the game with astonishing precision and speed.

I like to use an analogy for this: Suppose you are going to watch a top-tier professional sports event, like a basketball match: You go there in order to witness the absolute best players in the world show their utter skill at playing the sport. You are expecting 2 hours of sheer excitement and wonder. However, suppose that one of the teams finds an obscure loophole in the rules of the game that allows them to effectively cheat a victory for themselves without even playing the game, the other team having no recourse: The first team just declared victory at the start of the game abusing the loophole, the match ends, and that's it. It's over, everybody go home.

Well, that would be an utter disappointment, and utterly boring. You didn't go there to watch a team abuse a rulebook loophole in order to snatch a quick technical victory without even playing the game. You went there to watch a game! The spectators would be outraged! You would certainly demand your money back!

Well, for me most speedruns that skip major parts of the game are the same: I don't find them interesting in any way. They skip what I was wanting to watch the speedrun for in the first place! They skip the most entertaining part! I didn't "sign up" to watch a player skip the entire game: I did it so that I could see an extremely skilled player play the game, not skip it.

Unfortunately, skips, out-of-bounds glitches and other ways to bypass gameplay proper have become ubiquitous in speedrunning (especially when it comes to 3D games), and only few games have been spared.

That is why I don't really watch much speedrunning anymore. It's just boring. I'm not interested in glitchfests anymore. I would want to see someone play the game skillfully, I'm not interested in seeing someone break the game and skip the majority of it.