Digital Joysticks provide better control than Analog Joysticks

[ledzep]

That paddle stuff was what I was finishing up from your previous reply. I mainly want to discuss the analog joystick vs. digital joystick. The arrow keys/ctrl key is like a digital joystick except it's harder to use. But you are free to argue that as well since that's part of the poll. However, paddles are not the same as an analog joystick. Nor is that in the poll list of items to vote for as the games using them are few and nor can you play the games that work with digital joystick with a paddle.

 

I am free to argue paddles because I am certified due to my experience driving trucks. I don't know why Divya16 would allow such a loophole but she did. Your statement that arrow keys/ctrl key is "like a digital joystick except it's harder to use" validates my statement that a paddle controller is like an analog joystick only it's one POT instead of two and allows its inclusion in the discussion whether you included it in your poll or not since your poll, which erroneously includes arrow keys, does not match the unfounded, unproven statement you made that

 

"Digital Joysticks provide better control than Analog Joysticks. It's a scientific fact; let's see who can refute it."

 

You did not include arrow keys in that regrettable theory nor did you support, in any way, the opinion that it's a "scientific fact". This after dozens of opportunities provided to you to link to a scientific study, attach a PDF of a scientific study or provide data of a comprehensive controlled experiment that you have conducted that would support your theory. You will agree that we have demonstrate the patience of saints on this score.

 

It would ruin the controlled experiment if I mixed in other arcade contraptions that work for specific games. I can say you haven't flown an airplane so you can't compare analog joystick vs. digital joystick.

 

I didn't say anything about mixing in other arcade contraptions. And, please be honest with us. You have never ever conducted any controlled experiments using video games.

 

You mixed in other arcade contraptions the second you claimed to know that digital and analog joysticks "don't do the job well" for playing Star Wars when compared to that arcade game's native controller that you have never seen. By admitting that you have never even laid eyes on a Star Wars arcade game, much less played one, you have disqualified yourself from judging or comparing any other controllers to that game's native controller. Go play the arcade version of Star Wars a few times. No useless simulations, no imperfect home conversions, no hallucinations, play the actual game as it was meant to be played. Then and only then will anyone care what you think works well for the home version of Star Wars. Not one second before.

 

Yeah, I forgot to attach the data file in the previous post. So it's attached in this file. There are 3 joystick recordings for the same game. One is using the Atari 2600 digital joystick in the picture in post #1, one is the gemini digital joystick (looks similar to Atari 2600 joystick), and third file is the Gravis joystick (analog joystick). Only one variable changes, the type of joystick used. The target machine is Atari 800 in this case. The game is Popeye for all 3 joysticks.

 

Hypothesis: subject title

Observation: analog joysticks fail more often

Conclusion: Hypothesis.

 

Wait, what? WHAT?! You just finished claiming to have "megabytes" of data and now you grudgingly hand over data for just 3 joystick recordings for one game?!? Do you have a short-term memory disorder? Did you not recently read the definitions for controlled experiments and scientific method? Did you forget them already?

 

Where is the accompanying proof that you actually played those games in the manner in which you claim to have played them? What were the parameters, all the parameters, of the comprehensive experiment? What were the standards you used to record those three joysticks? Where is the video proving that you are actually running the experiment and not simply generating random numbers to feed into a data file? What are the specs of those 3 joysticks?

 

It's there for everyone to repeat. Try to guess the state of the joystick with your eyes closed and either have a witness if you are dishonest or open your eyes after you guess and see if that's the state on the screen. I get 100% right on the digital joystick, but not for analog joystick.

 

You have the poorest command of English of anyone on this thread. I never asked you to tell he how I could repeat your worthless experiment, I asked you to prove that it is by nature harder to get exact states on the analog joystick. Prove it. Prove that you are not mistaking the state of the joystick with the state of the program or the state of the video display. You can keep your eyes open, prove that you know which state you are referencing (or guessing). Prove that you are not throwing the results of your guesses in order to make digital joysticks score better. Prove it.

 

Or... admit that you don't know what you're talking about and you've just been spitting out random opinions. Otherwise if you claim something is a scientific fact then you must satisfy the definition of what a scientific fact is. Since you have demonstrated, again, that you don't understand or forgot what a scientific fact is I will supply it again -

 

"An observation that has been confirmed repeatedly and is accepted as true (although its truth is never final)."

 

[some irrelevant stuff deleted because it's meaningless or already answered]

 

I quoted a lot of your words. Are you saying they're all meaningless? It can't be already answered since you have failed to supply the data from your experiments (handing out one incomplete instance from one part of the experiment doesn't count), describe the parameters of your experiments, list who was in the control and experimental groups in your experiments, describe how you gathered the data or had any corroborating proof that you amassed the data (which we all know doesn't exist) in the manner that you claim to have amassed it.

 

That's rubbish. If I simulate a joystick using arrow keys + ctrl and the game only uses those 5 bits of information, I have perfectly simulated the joystick.

 

You have just proven that you don't know how a joystick works. After I just showed you why arrow keys aren't the same as a digital joystick. Now that's bias.

 

Arrow keys cannot be used as a joystick simulator because -

 

1) There is no accounting for the effects of the length of the joystick handle, the amount of leverage or torque that the handle can apply to the switches nor the speed that those switches can be activated using just arrow keys.

2) It is possible to presses down 3 or 4 arrow keys at once. It is impossible to activate more than 2 directional switches on a properly built, unbroken digital joystick.

3) It is possible to press opposite directions (left & right, up & down) with arrow keys. This, too, is impossible to do with a real joystick.

4) It is not a given that the switches used in computer keyboards function in the same way and at the same speeds as joystick switches.

5) It is not a given that signals from keyboard buttons are interpreted the same way as digital joystick signals.

 

That's not "perfectly simulated", that's "approximated".

 

Wow, I pointed out the land mine to you, warned you about it, and you still decided to walk over to it and stand on it. Fantastic. I can almost see Divya16 in slow-motion with her mouth in a round "NNNNNNNNOOOOOOOOOOOOOOOOOOOOOOOO!" shape as she tries to reach out to stop you. Maybe she can track down what's left of your legs and make some kind of meat sculpture that celebrates your ignorance.

[kool kitty89]

You are talking about some theoretical joysticks. Even in that case, the digital one would read more consistently at those positions SNIP

WRONG - the digital one couldn't even reach those positions. On or off, remember, with 100% certainty. Scientific fact. I guess that makes them inferior because they cannot do what another product can.

No, that's only in the context of an 8-way joystick (be it analog or digital), my statement was regarding a fully digital joystick with a HUGE range of positions: ie instead of potentiometers you have a lot of digital switches to toggle through, say a rotary switch with many positions (which are not uncommon) or a sliding mechanism in the case of sliding potentiometers.

 

See the context of my previous posts, but in the immediate case:

http://www.atariage.com/forums/topic/171339-digital-joysticks-provide-better-control-than-analog-joysticks/page__st__300__p__2133261#entry2133261

 

It's not just hypothetical as some devices do just that, albeit perhaps not with 8-bit precision, but far more than 2-bit... and my point was that you'd still have a wide range to sort through and consistency would not be significantly different from analog sticks... from the computer/programmer's POV it would be no different at all as you'd still have straight up digital values either direct I/O or analog to digital. (or either converted to a serial data stream and then decoded) You do still have issues with calibration (potentially) of pot based systems, but for the biggest issues tied to that (dirt and wear) you have similar issues with a fully digital system from dirty contacts or similar -you'd have jitter/skipping)

 

 

256 positions is a bit overboard, but 16 positions would not be unreasonable at all (same precision as a 4-bit DAC) and that would only really require 8 contacts per axis with a slide or rotary mechanism with either 1 or 2 contacts being grounded at any time (that would also avoid gaps between switches as you'd always be grounding at least 1 switch). That would result in a 16x16 grid of possible values for 2 axes, or 256 possible values (8 bits) compared to 256x256 common to most ADC/DAC stuff. (though in many, many cases 4-bit per axis would be totally fine if not overkill: paddles are one major exception as you need pixel exact positioning, but for most general flight sims or modern consoles -ie mid 90s onward- you'd probably notice the difference very rarely and even in most of those cases it wouldn't be critical -almost all games used variable control for speed sensitivity or vectored values rather than fixed position tracking as with paddle based games)

 

And again, mechano-optical mechanisms are somewhat similar to that in the sense of being digital (more or less), as ball mice and the N64 controllers use, though not direct digital switches. (technically it's closer to tapping a digital joystick in some respects as the photo-diode reads on/off values at a variable rate from how fast the perforated disc is rotated with varying rate of flicker and thus a 1-bit data stream not unlike that of a PWM ADC/DAC or the action of tapping a simple 8-way stick in a rough sense)

Of course, it was primarily Nintendo who used that with aftermarket controllers using conventional analog potentiometers --the hardware/software didn't care what was used so long as the correct serial data stream was transmitted.

And do note that that mechanism has no relation to the wear problems on the springs of the N64 controllers, that's a separate issue tied to the centering mechanism.

[atariksi]

With a standard digital stick, you have 100% control-- the best the world has to offer. Now on the dark side, there's the analog joystick. Some modern ones even look like those darth vador spacecrafts. You may have gotten the worst of both worlds if the mapping is sort of like a software method of checking thresholds. Add to that the problem of having to move the stick a longer distance to switch states since they have to allow for long enough distance for range of values. As an analogy, you know the switching theory of transistors is that the smaller they are the faster the switching times.

You have 100% control of ON or OFF - nothing else. What if you want 50%? Where is the control? That is missing, therefore since something is missing, you don't have 100%.

 

We don't want around 50% and introduce uncertainty. I'm sure if you transmit a video in analog format, you can squeeze in more features at same bandwidth than a digitized version of the video. But the digitized video provides better control. Also, for analog requirements, you also have the option of paddle controllers/mice/etc. which provide better control than analog joysticks. Just because you want some inexact feedback reliant 50% feature, doesn't mean you should switch to analog. Koolkitty has a good idea regarding extending the digital states on the axes-- that would still be better than analog joystick.

[atariksi]

That's rubbish. If I simulate a joystick using arrow keys + ctrl and the game only uses those 5 bits of information, I have perfectly simulated the joystick.

 

LOL!

 

Seriously? I mean SERIOUSLY?

 

With all due respect, you wouldn't last 10 seconds in a real research environment.

 

Ya, that deserves a +1, haahaahaha. I read that like 5 times to make sure he really wrote that.

 

I don't think you even understand what he wrote nor read my reply given your other drivel.

 

Let's go back to catagories and see where this comment of yours fits in:

 

(1) logical deduction/mathematical analysis (undeniably true)

(2) experimental data (which leads to the scientific fact idea)

(3) mental speculation-- something out of the blue according to emotional attachment, or mistaken views with no evidence

(4) blind following the blind-- because most people seem to be doing that so that must be right

 

 

Item (4) would be best fit for your remark. I already answered that I'm not trying to make the arrow keys like a physical joystick so handle lever, stand, etc. don't matter. You can generate the same signals as the joystick using the arrow keys and CTRL key for the 5-bits of information that the Atari 2600 style digital joystick provides. Yeah, joystick is easier to generate those signals but only the signals matter here.

[atariksi]

I didn't say anything about mixing in other arcade contraptions. And, please be honest with us. You have never ever conducted any controlled experiments using video games.

Don't speculate. You are completely lost and getting emotional and throwing up incoherent remarks some of which are self-contradictory. You have opportunity to repeat my experiment, you didn't. You have opportunity to use a joystick simulator, you didn't. You have opportunity to look at some of the data of my experiment, you didn't. I have you a simplified experiment with BASIC program (which doesn't involve joystick simulator), you dismissed it by shoving it under the rug. You dismissed the mathematics as well. Here's some more:

 

C(t) = d(t) where the control you have over the joystick in human reaction times over the directions is 100% for a digital joystick. Now for your analog joystick case even for your simple 8 directions and center, there's instability due to inconsistent readings (inherent flaws), thresholds being used by the software, etc. There's also slower switching times but even leaving that out, you have:

 

C(t) = (1/L)(d(t)) where 1/L is the probability that you directions you want control over you actually achieved. This low probability gets lower as you increase the levels you want to control on the analog joystick. Then the limit as L->infinity, C(t) = 0.

 

Rest of emotional biased rubbish deleted.

[atariksi]

256 positions is a bit overboard, but 16 positions would not be unreasonable at all (same precision as a 4-bit DAC) and that would only really require 8 contacts per axis with a slide or rotary mechanism with either 1 or 2 contacts being grounded at any time (that would also avoid gaps between switches as you'd always be grounding at least 1 switch). That would result in a 16x16 grid of possible values for 2 axes, or 256 possible values (8 bits) compared to 256x256 common to most ADC/DAC stuff. (though in many, many cases 4-bit per axis would be totally fine if not overkill: paddles are one major exception as you need pixel exact positioning, but for most general flight sims or modern consoles -ie mid 90s onward- you'd probably notice the difference very rarely and even in most of those cases it wouldn't be critical -almost all games used variable control for speed sensitivity or vectored values rather than fixed position tracking as with paddle based games)

Although 4-bit digital positioning on each axes may be something a human could get used to if it was consistently done for every digital joystick and the distance was enough, it still would require a greater distance travelled for switching 9-state games. Perhaps, some way of turning off the distance travelled would help. You couldn't really just use the first position, since an atari style joystick, you hit a wall as joystick doesn't move anymore. And sometimes you break joysticks as you quickly go back and forth/up and down. So that would be a theoretical joystick that doesn't exist.

 

Looks like Atari did the best job with control on their digital joysticks as well as their analog controllers namely the paddles. Both outdo the analog joystick in control. And they did try to minimize the analog errors by fixing the ranges by making all their analog controllers use 0..228. However, I have seen these ranges reach early on some A5200 joysticks and not reachable on others. Paddles do seem to use that entire range.

[atarian63]

Digital is better for "Pac-man." Analog is better for "Jane's WWII Fighters." What's the argument about, again?

bingo! right again!

[kool kitty89]

Although 4-bit digital positioning on each axes may be something a human could get used to if it was consistently done for every digital joystick and the distance was enough, it still would require a greater distance travelled for switching 9-state games. Perhaps, some way of turning off the distance travelled would help. You couldn't really just use the first position, since an atari style joystick, you hit a wall as joystick doesn't move anymore. And sometimes you break joysticks as you quickly go back and forth/up and down. So that would be a theoretical joystick that doesn't exist.

 

Looks like Atari did the best job with control on their digital joysticks as well as their analog controllers namely the paddles. Both outdo the analog joystick in control. And they did try to minimize the analog errors by fixing the ranges by making all their analog controllers use 0..228. However, I have seen these ranges reach early on some A5200 joysticks and not reachable on others. Paddles do seem to use that entire range.

That was my point, analog vs digital isn't the issue, it's variable control with a wide range and when that's desirable or when it's not.

 

Getting "used" to anything isn't the issue... an 8-bit (or 16-bit -way overkill) resolution per axis wouldn't matter in most cases save for lower resolution being a bit granular and the player noticing the steps too much (ie noticing that there's not a seemingly infinitely smooth transition), but for most modern 3D games the speed (vector) variable input would be accepatble if limited to a 4-bit range, though I'm sure most modern controllers use 8-bit minimum internally (it's all converted to serial I/O of course).

 

But back to the 70s and 80s, I agree, there are far fewer games (especially on home consoles) that necessarily need such variable control and if you had to choose: 8-way (with a good, reliable, not-sloppy controller) would be much more foolproof, and that held true for the majority of console genres into the mid 90s. (simple 8-way control with certain 3D games gets awkward fast, but racing games were always a problem, even fully 2D ones like out run)

And you can have good and bad control design in a game regardless of being analog or digital and it's up to the programmer to manage things properly. (Checkered Flag on the Jaguar made things a good deal worse than normal hard turning with an 8-way pad/stick due to the funky physics... granted analog/wide range control wouldn't solve that either)

[atariksi]

Digital is better for "Pac-man." Analog is better for "Jane's WWII Fighters." What's the argument about, again?

bingo! right again!

 

Choose the joystick that provides better control overall. I would rather have 100% control over 95+% of the games and lacking a feature in less than 5% than have 70-80% control in all the games. Remember that the joystick controllers are solely meant for controlling. And paddles are not analog joysticks. i.e., you can certainly do better in SuperBreakout with paddles than with an analog joystick.

 

Although 4-bit digital positioning on each axes may be something a human could get used to if it was consistently done for every digital joystick and the distance was enough, it still would require a greater distance travelled for switching 9-state games. Perhaps, some way of turning off the distance travelled would help. You couldn't really just use the first position, since an atari style joystick, you hit a wall as joystick doesn't move anymore. And sometimes you break joysticks as you quickly go back and forth/up and down. So that would be a theoretical joystick that doesn't exist.

 

Looks like Atari did the best job with control on their digital joysticks as well as their analog controllers namely the paddles. Both outdo the analog joystick in control. And they did try to minimize the analog errors by fixing the ranges by making all their analog controllers use 0..228. However, I have seen these ranges reach early on some A5200 joysticks and not reachable on others. Paddles do seem to use that entire range.

That was my point, analog vs digital isn't the issue, it's variable control with a wide range and when that's desirable or when it's not.

 

Yes, analog vs. digital isn't the issue here but analog joystick vs. digital joystick is the issue. Just to be clear, I was talking 4-bit axis like -8..7 with 0 at center and all digital not ADC-based which would once again introduce the uncertainties. That sort of digital joystick would still be superior in control to an analog joystick although it would slow down switching times.

 

Getting "used" to anything isn't the issue... an 8-bit (or 16-bit -way overkill) resolution per axis wouldn't matter in most cases save for lower resolution being a bit granular and the player noticing the steps too much (ie noticing that there's not a seemingly infinitely smooth transition), but for most modern 3D games the speed (vector) variable input would be accepatble if limited to a 4-bit range, though I'm sure most modern controllers use 8-bit minimum internally (it's all converted to serial I/O of course).

In order for joysticks events to be known a priori, you would have to get used to the joystick so you know you are at -8 or -5 or 0 or +3 or +7. I guess extremes and center could be made consistent so they would always be known in a digital setup.

 

But back to the 70s and 80s, I agree, there are far fewer games (especially on home consoles) that necessarily need such variable control and if you had to choose: 8-way (with a good, reliable, not-sloppy controller) would be much more foolproof, and that held true for the majority of console genres into the mid 90s. (simple 8-way control with certain 3D games gets awkward fast, but racing games were always a problem, even fully 2D ones like out run)

And you can have good and bad control design in a game regardless of being analog or digital and it's up to the programmer to manage things properly. (Checkered Flag on the Jaguar made things a good deal worse than normal hard turning with an 8-way pad/stick due to the funky physics... granted analog/wide range control wouldn't solve that either)

 

You also have the problem that many games which were made analog because joysticks happened to be analog on the original system but could easily have been made with digital joysticks.

[carmel_andrews]

Question is, would we be talking about this if all Atari standard joysticks were like the bbc/pc analog(ue) type

[atariksi]

Question is, would we be talking about this if all Atari standard joysticks were like the bbc/pc analog(ue) type

 

That's more true the other way around. However, taking your hypothetical case, you would still end up seeing the lesser control experimentally due to the existence of the keyboard. And of course, logically you would also end up to the same conclusion even if neither keyboard nor digital joysticks existed.

[ledzep]

That's rubbish. If I simulate a joystick using arrow keys + ctrl and the game only uses those 5 bits of information, I have perfectly simulated the joystick.

 

LOL!

 

Seriously? I mean SERIOUSLY?

 

With all due respect, you wouldn't last 10 seconds in a real research environment.

 

Ya, that deserves a +1, haahaahaha. I read that like 5 times to make sure he really wrote that.

 

I don't think... [some irrelevant stuff deleted because it's meaningless or already answered]

 

That's all you need to say.

 

Let's go back to catagories and see where this comment of yours fits in:

 

(1) logical deduction/mathematical analysis (undeniably true)

(2) experimental data (which leads to the scientific fact idea)

(3) mental speculation-- something out of the blue according to emotional attachment, or mistaken views with no evidence

(4) blind following the blind-- because most people seem to be doing that so that must be right

 

 

Item (4) would be best fit for your remark.

 

Nope, it's (1). When you've driven a truck as much as I have you'll see. Ask your AWOL XO.

 

I already answered that I'm not trying to make the arrow keys like a physical joystick so handle lever, stand, etc. don't matter. You can generate the same signals as the joystick using the arrow keys and CTRL key for the 5-bits of information that the Atari 2600 style digital joystick provides. Yeah, joystick is easier to generate those signals but only the signals matter here.

 

It doesn't matter what you are or are not trying to make. And, be honest, you would never make one anyway, that would take more effort than conducting the experiments you've never conducted. But, back to your latest error, while it's true that you can generate the same signals as the joystick using the arrow keys and CTRL key for the 5-bits of information that the Atari 2600 style digital joystick provides, it is also true that you can generate signals using the arrow keys and CTRL key that the joystick cannot generate. Which means you wouldn't "perfectly simulate" a digital joystick as you ignorantly claimed, you would merely approximate it.

 

Go ahead, explain how the differences between a digital joystick and keyboard arrow keys don't contradict your latest unfounded claim. What differences? Glad you asked. I'll quote Post #326 which you just read (come on, dude, it's like 3 posts above your flailing response) but are pretending didn't refute your claim -

 

Arrow keys cannot be used as a joystick simulator because -

 

1) There is no accounting for the effects of the length of the joystick handle, the amount of leverage or torque that the handle can apply to the switches nor the speed that those switches can be activated using just arrow keys.

2) It is possible to presses down 3 or 4 arrow keys at once. It is impossible to activate more than 2 directional switches on a properly built, unbroken digital joystick.

3) It is possible to press opposite directions (left & right, up & down) with arrow keys. This, too, is impossible to do with a real joystick.

4) It is not a given that the switches used in computer keyboards function in the same way and at the same speeds as joystick switches.

5) It is not a given that signals from keyboard buttons are interpreted the same way as digital joystick signals.

 

That's not "perfectly simulated", that's "approximated". Simulators are worthless for generating test data. But since you never planned to generate test data nor conduct any comprehensive experiment that obvious issue doesn't really concern you, does it.

 

By the way, what happened to Divya16? Is she still sifting through the smoking dirt for the last bits of your legs for that meat sculpture? If you look at her avatar it looks like she can actually see the pink cloud of mist from the land mine taking your legs out. Explains the smile, too, haahaha.

[ledzep]

I didn't say anything about mixing in other arcade contraptions. And, please be honest with us. You have never ever conducted any controlled experiments using video games.

Don't speculate. You are completely lost and getting emotional and throwing up incoherent remarks some of which are self-contradictory. You have opportunity to repeat my experiment, you didn't. You have opportunity to use a joystick simulator, you didn't. You have opportunity to look at some of the data of my experiment, you didn't. I have you a simplified experiment with BASIC program (which doesn't involve joystick simulator), you dismissed it by shoving it under the rug.

 

No, I didn't dismiss your BASIC program, I pointed out where you had badly coded the analog joystick program since it wasn't testing for the same type of outputs that the digital joystick was being tested for. I didn't shove it under the rug, I put it in the spotlight. And then you dismissed the fixes. Tsk tsk. You need at least two sets of BASIC program tests. One set tests the speed and accuracy of both digital and analog joysticks to output signals for the 4 cardinal directions and the other set tests the speed and accuracy of both digital and analog joysticks to output signals for the in-between states. You would probably also need a BASIC program that tests the speed and accuracy of both digital and analog joysticks to output signals for speed of movement of the joystick handles. So go ahead, you have the opportunity to correct your coding errors. I'm betting you won't.

 

Don't speculate. I'm not getting emotional at all, unless doubling over with laughter as I watch you try harder and harder to change the subject from your lack of supporting data and lack of understanding of basic scientific terms is emotional. It's not my responsibility to repeat your experiment (that you never ran), it's your responsibility to post the parameters of your imaginary experiment and to post the supporting data. Why? Because you're the one who claimed to have run an experiment using faulty methods (simulator) on only one inadequate group (yourself) that generated insufficient data (one pass for only 3 different joysticks used to play just one game on one system).

 

But, again, don't speculate. I decided to follow your trailblazing formula to scientific oblivion and prove that analog joysticks in actuality provide better control than digital joysticks. But, unlike you, I will spell out my steps -

 

1.) I played "hundreds" of video games using both simulated digital and analog joysticks. I also included paddles since I can "perfectly simulate" one axis of an analog joystick with the paddle's single POT.

2.) I restricted the control and experimental groups to one person - me.

3.) I noticed that I consistently scored higher when playing those games with the simulated analog joysticks and decided that met the requirement for "data".

4.) I didn't bother to record or store any useful data during the times I was playing all those games.

5.) I also didn't bother to outline any parameters for those game-playing sessions so that others could possibly reproduce my methods or results since it would then be possible for someone to run the same experiment that I had yet come up with a result contradictory to my own and that simply would not do.

6.) I talked to some people who also score higher when playing video games with analog joysticks, confirming the truthiness of my own results to a fanatical degree.

7.) I went on a forum and claimed that analog joysticks provided better control than digital joysticks and, to the shock of all those on that forum, stated that was a scientific fact.

8.) When those other forum members requested that I either provide references to this claim of "fact" or supply data to prove my position I answered by stating that I had run these experiments and had the data to back it up and then repeatedly provided neither even as I claimed that my joystick simulator could effortlessly gather data that would support my theory.

9.) As demands for proof increased I side-stepped the issue by telling them that they could repeat my experiments while willfully ignoring the obvious fact that I had not provided any of the parameters and methodology for my own experiments, making it impossible for anyone on Earth to repeat my experiment.

10.) I would respond to any logical refutations using a series of stock answers involving claims that those who questioned my methods or results were illogical or emotional or illiterate or confused or blind or spouting drivel or spouting rubbish or a combination of those.

11.) I would also mix in sciency terms I'd heard on shows like "Chuck" such as "scientific fact", "a priori", "perfectly simulate", "controlled experiment", "data", etc., with complete disregard for how badly I had misunderstood the terms along with occasionally including jpegs of random, badly photographed screen grabs of games I may or may not have played in the manner I claimed to have played them.

 

Therefore I have proven that Analog Joysticks provide better control than Digital Joysticks. It's a scientific fact, let's see who can refute it. Thank you for the tips on effectively forming this proof.

 

You dismissed the mathematics as well. Here's some more:

 

C(t) = d(t) where the control you have over the joystick in human reaction times over the directions is 100% for a digital joystick. Now for your analog joystick case even for your simple 8 directions and center, there's instability due to inconsistent readings (inherent flaws), thresholds being used by the software, etc. There's also slower switching times but even leaving that out, you have:

 

C(t) = (1/L)(d(t)) where 1/L is the probability that you directions you want control over you actually achieved. This low probability gets lower as you increase the levels you want to control on the analog joystick. Then the limit as L->infinity, C(t) = 0.

 

Rest of emotional biased rubbish deleted.

 

No way! Equations? That's like, so sciency. Do you have an equation that will solve for the total number of times you will avoid providing all the data you claim to have that would have proven your position days ago? I would expect that it would have a lot of variables in it, too, like L(t) representing when the moderator will lock this thread so you can claim that you were just about to post the data that doesn't actually exist but now you can't, aawwww man! Maybe Sc can represent the total number of sciency terms you can misunderstand and F(n) is the amount of friends who still think you know what you're talking about the longer you make excuses (represented by E(t), right?) so as Sc * E(t) -> infinity, F(n) = 0.

[potatohead]

clusterfuck = 1

[atariksi]

Arrow keys cannot be used as a joystick simulator because -

 

1) There is no accounting for the effects of the length of the joystick handle, the amount of leverage or torque that the handle can apply to the switches nor the speed that those switches can be activated using just arrow keys...

Read post #315 where I clearly stated I'm not trying to simulate a physical joystick. In fact, if you read post #1, I clearly put arrow keys/ctrl key in a separate item to be voted and also if you read the thread you would know that I argued the superiority of digital joystick even over the digital arrow keys/ctrl key.

 

2) It is possible to presses down 3 or 4 arrow keys at once. It is impossible to activate more than 2 directional switches on a properly built, unbroken digital joystick.

So what. It's a superset of the signal functionality of a digital joystick. As long as it can generate the pulses a joystick can generate, it can simulate it.

 

4) It is not a given that the switches used in computer keyboards function in the same way and at the same speeds as joystick switches.

5) It is not a given that signals from keyboard buttons are interpreted the same way as digital joystick signals.

Interpretation is what the software makes sure are exact 1:1. It can even avoid more than three keypresses. Also, did you notice that on some joysticks you can push the whole thing down and generate all bits = 0.

[atariksi]

No, I didn't dismiss your BASIC program, I pointed out where you had badly coded the analog joystick program since it wasn't testing for the same type of outputs that the digital joystick was being tested for. I didn't shove it under the rug, I put it in the spotlight.

Where are the errors/badly coded area here:

 

10 ? PADDLE(0),PADDLE(1),PTRIG(0):GOTO 10

 

...group (yourself) that generated insufficient data (one pass for only 3 different joysticks used to play just one game on one system).

You don't even understand those 3 files that you dismissed under the rug like other things. You can generate the files for yourself if you don't trust the joystick simulator program.

 

1.) I played "hundreds" of video games using both simulated digital and analog joysticks. I also included paddles since I can "perfectly simulate" one axis of an analog joystick with the paddle's single POT.

Did you put a big circular knob on the POT to help you attain better accuracy?

 

2.) I restricted the control and experimental groups to one person - me.

Why? Don't you want people to repeat it or you want people to blindly accept you-- catagory (4):

 

(1) logical deduction/mathematical analysis (undeniably true)

(2) experimental data (which leads to the scientific fact idea)

(3) mental speculation-- something out of the blue according to emotional attachment, or mistaken views with no evidence

(4) blind following the blind-- because most people seem to be doing that so that must be right

 

3.) I noticed that I consistently scored higher when playing those games with the simulated analog joysticks and decided that met the requirement for "data".

How many bits was your ADC converter?

 

4.) I didn't bother to record or store any useful data during the times I was playing all those games.

That's a shame. Because you already admitted in this thread-- you prefer playing games with analog joysticks and you haven't shown any valid point against the flaws pointed for analog joysticks. Show me in any of my screenshots how a person who is used to both types of joysticks can score higher with a uncertain inexact flawed analog joystick over a digital joystick and I can take you more seriously.

 

You dismissed the mathematics as well. Here's some more:

 

C(t) = d(t) where the control you have over the joystick in human reaction times over the directions is 100% for a digital joystick. Now for your analog joystick case even for your simple 8 directions and center, there's instability due to inconsistent readings (inherent flaws), thresholds being used by the software, etc. There's also slower switching times but even leaving that out, you have:

 

C(t) = (1/L)(d(t)) where 1/L is the probability that you directions you want control over you actually achieved. This low probability gets lower as you increase the levels you want to control on the analog joystick. Then the limit as L->infinity, C(t) = 0.

 

Rest of emotional biased rubbish deleted.

 

No way! Equations? That's like, so sciency. Do you have an equation that will solve for the total number of times you will avoid providing all the data you claim to have that would have proven your position days ago?

The way it works is that logic/mathematics doesn't require experimental data. All you need to understand is that that probability for getting exact directions is <1.0 vs. 1.0 for digital joysticks. Then as you increase the number of levels you want to use on your inferior analog joystick, you will see how the limit goes to ZERO control.

[+poobah]

No way! Equations? That's like, so sciency. Do you have an equation that will solve for the total number of times you will avoid providing all the data you claim to have that would have proven your position days ago?

The way it works is that logic/mathematics doesn't require experimental data. All you need to understand is that that probability for getting exact directions is <1.0 vs. 1.0 for digital joysticks. Then as you increase the number of levels you want to use on your inferior analog joystick, you will see how the limit goes to ZERO control.

Ahhh the trainwreck continues...

 

L'hopital begs to differ =P

[atariksi]

No way! Equations? That's like, so sciency. Do you have an equation that will solve for the total number of times you will avoid providing all the data you claim to have that would have proven your position days ago?

The way it works is that logic/mathematics doesn't require experimental data. All you need to understand is that that probability for getting exact directions is <1.0 vs. 1.0 for digital joysticks. Then as you increase the number of levels you want to use on your inferior analog joystick, you will see how the limit goes to ZERO control.

Ahhh the trainwreck continues...

 

L'hopital begs to differ =P

 

Sorry, L'hopital's rule also takes this limit to 0. The denominator grows faster than the numerator. Just for two levels of center to left for an analog joystick you have a region of uncertainty and probability is less than 1.0 (mainly 1/2). That applies for each of the directions. Then you increase the levels (L), the regions of uncertainty increase so the probability of which region your are in goes toward 1/L. At the end, it's like probabiliyt of picking a number from rational numbers [0..1] which is zero.

 

No trainwreck, you haven't even shown that the joystick signals aren't being simulated by pressing arrow/ctrl keys.

[ApolloBoy]

Sorry, L'hopital's rule also takes this limit to 0. The denominator grows faster than the numerator. Just for two levels of center to left for an analog joystick you have a region of uncertainty and probability is less than 1.0 (mainly 1/2). That applies for each of the directions. Then you increase the levels (L), the regions of uncertainty increase so the probability of which region your are in goes toward 1/L. At the end, it's like probabiliyt of picking a number from rational numbers [0..1] which is zero.

http://www.youtube.com/watch?v=PtuqjFf7-N4

[kool kitty89]

One more comment on the all-digital switch based variable control mechanism I mentioned:

there's a few simple reasons it's probably never been used extensively: there are disadvantages of production cost, size, durability, and flexibility compared to using other mechanisms. (megnetic, resistance potentiometers, mechano-optical, etc)

In the context of a system from the 70s or 80s using bulky pots in any case or similarly bulky joysticks using pots comparable in size to VCS paddles/5200 joysticks you probably could manage a competitively priced and sized unit based on dual rotary switches with 4-bit per axis (8-bit total) output.

But I doubt that such a mechanism would minatureize well and stay durable/reliable compared to other mechanisms and you'd deal with many of the same issues as rotary pots with wearing brushes/contacts with the exception that instead of gradually getting jitter issues you'd lose control all at once or get much more dramatic dead spots than with other mechanisms. (the machano-optical and magnetic systems don't deal with moving parts for the actual measuring mechanism, only the assembly itself, so would be preferable in that sense, but again I think most modern systems have stuck with resistance/pot based 2-axis spring-loaded modules, either that or they use the magnetic mechanism but look very much like conventional pot modules)

 

That and anything much more than 4-bits would get increasingly less practical for the switch mechanism while other methods are pretty variable. (with pots you're only limited by the consistency/accuracy of the pots themselves and the precision of the ADCs used)

 

 

Also, most analog joysticks of the 70s and 80s with home computers that caused the most trouble were long-throw and often lacking self centering... in many cases much more so than the 5200 (and even self centering good quality PC joysticks tend to be problematic for games that want simple restricted 4/8-way control). However, in all cases that could have been very easily addressed by the manufacturers (or 3rd parties) offering simple resistor DAC based "digital" (8-way) joysticks or adapters converting standard atari-compatible controllers into simple analog sticks.

That would also mean full cross-compatibility with analog sticks and allowing the user to select the control based on the game.

It's really odd that that didn't happen though... cheap and simple to design, no reason they shouldn't have.

If you were stuck with only one joystick, something like modern thumbsticks or the vectrex controller is the best middle ground. (springy, responsive short throw that's quick enough to very reasonably approximate a digital thumbstick)

[ledzep]

Read post #315 where I clearly stated I'm not trying to simulate a physical joystick. In fact, if you read post #1, I clearly put arrow keys/ctrl key in a separate item to be voted and also if you read the thread you would know that I argued the superiority of digital joystick even over the digital arrow keys/ctrl key.

 

No thanks, I'll read Post #310 where you clearly contradict yourself. Don't remember your own words? Here they are again -

 

"That's rubbish. If I simulate a joystick using arrow keys + ctrl and the game only uses those 5 bits of information, I have perfectly simulated the joystick."

 

You are stating that you are simulating a physical joystick using arrow keys + ctrl key. You do not state that your are using arrow keys + ctrl key to simulate a simulated joystick nor do you say you are using simulated arrow keys + simulated ctrl key to simulate a simulated joystick. No, by failing to specify that the joystick to be simulated was a simulated joystick instead of a real joystick (as your unfounded, unproven initial claim specifies) you are stating that you would attempt to "perfectly simulate" a real joystick with arrow keys + ctrl key. And you can't. Double damn you, English!

 

I've read the thread. You clearly muddled the debate by including arrow/ctrl keys in your poll but not mentioning them in your error-filled theory.

 

"Digital Joysticks provide better control than Analog Joysticks. It's a scientific fact; let's see who can refute it."

 

Nope, no mention of arrow/ctrl keys there! So why would you screw that up so badly and so quickly? And why lie about it being a scientific fact?

 

So what. It's a superset of the signal functionality of a digital joystick. As long as it can generate the pulses a joystick can generate, it can simulate it.

 

And so long as it can generate alternative pulses a joystick cannot generate, it cannot be called a perfect simulation as you erroneously claimed.

 

Notice that you have immediately contradicted yourself again in your poorly-worded excuse above by specifying that arrow keys + ctrl key can generate the signal functionality of a digital joystick instead of a simulated joystick. Doh! If you really believed that that's what you meant when you tried to avoid admitting your mistake you would have instead worded your excuse like so -

 

"So what. It's a superset of the signal functionality of a simulated digital joystick. As long as it can generate the pulses a simulated joystick can generate, it can simulate it."

 

Just because you can't correctly form English sentences doesn't mean that the rest of us have the same difficulty.

 

Interpretation is what the software makes sure are exact 1:1. It can even avoid more than three keypresses. Also, did you notice that on some joysticks you can push the whole thing down and generate all bits = 0.

 

The software doesn't make sure about anything, the programmer does. And a competent programmer (read: not you) would know that the signals from arrow keys do not end up in the same place as signals from a real (not simulated) joystick and a competent programmer (read: still not you) would also know that the signals from arrow keys do not travel down the same paths as signals from a real (not simulated) joystick.

 

Which is why someone capable of conducting an accurate, comprehensive, controlled experiment (read: definitely not you) would not waste his time by introducing the extra levels of error inherent in a joystick simulator, he would make the minimal effort to read the signals from the actual digital and analog joysticks he claimed to have tested.

 

Who taught you to be so lazy when testing theories? I've never heard of that method, to state a sweeping, error-filled opinion, say you have data to prove it and then tell others if they want to see your proof they can run the experiment themselves.

[ledzep]

No, I didn't dismiss your BASIC program, I pointed out where you had badly coded the analog joystick program since it wasn't testing for the same type of outputs that the digital joystick was being tested for. I didn't shove it under the rug, I put it in the spotlight.

Where are the errors/badly coded area here:

 

10 ? PADDLE(0),PADDLE(1),PTRIG(0):GOTO 10

 

The error was already described to you.

 

...group (yourself) that generated insufficient data (one pass for only 3 different joysticks used to play just one game on one system).

You don't even understand those 3 files that you dismissed under the rug like other things. You can generate the files for yourself if you don't trust the joystick simulator program.

 

I understand them perfectly. They are a small, inconclusive portion of the megabytes of data you claim to have generated yet refuse to present to those who have asked for the data.

 

I cannot generate the files for myself until I know exactly how you generated them. And I cannot know that until you present the parameters and methodology used in your imaginary experiment which you have consistently failed to do.

 

1.) I played "hundreds" of video games using both simulated digital and analog joysticks. I also included paddles since I can "perfectly simulate" one axis of an analog joystick with the paddle's single POT.

Did you put a big circular knob on the POT to help you attain better accuracy?

 

Interesting. Someone on that other forum asked the same question. I told him to stick to the subject which was analog joysticks vs. digital joysticks.

 

2.) I restricted the control and experimental groups to one person - me.

Why? Don't you want people to repeat it or you want people to blindly accept you-- catagory (4):

 

(1) logical deduction/mathematical analysis (undeniably true)

(2) experimental data (which leads to the scientific fact idea)

(3) mental speculation-- something out of the blue according to emotional attachment, or mistaken views with no evidence

(4) blind following the blind-- because most people seem to be doing that so that must be right

 

Wow, someone asked that same question on that other forum, too! I think I said something like I know you can't claim which is better because you don't even address my arguments but want people to blindly accept your views. Oh, and I threw in a that's a Jar-Jar Defense, I think.

 

3.) I noticed that I consistently scored higher when playing those games with the simulated analog joysticks and decided that met the requirement for "data".

How many bits was your ADC converter?

 

Ya, that was a tough question. I replied to that one by posting a few jpegs of screen grabs from some video games I had lying around. Like shooting fish in a barrel, man.

 

4.) I didn't bother to record or store any useful data during the times I was playing all those games.

That's a shame. Because you already admitted in this thread-- you prefer playing games with analog joysticks and you haven't shown any valid point against the flaws pointed for analog joysticks. Show me in any of my screenshots how a person who is used to both types of joysticks can score higher with a uncertain inexact flawed analog joystick over a digital joystick and I can take you more seriously.

 

No, not a shame at all because I was never ever going to provide any data anyway. Saved me a lot of time.

 

But, to answer your question, most of the software in the world can do without digitaltivity. Without any experimentation, you can draw this conclusion. If only analog joysticks existed, the few games that need digitaltivity would get adapted for analog. Before you blurt out any more crap and insults (which I will ignore), think about it and calm down.

 

No way! Equations? That's like, so sciency. Do you have an equation that will solve for the total number of times you will avoid providing all the data you claim to have that would have proven your position days ago?

The way it works is that logic/mathematics doesn't require experimental data. All you need to understand is that that probability for getting exact directions is <1.0 vs. 1.0 for digital joysticks. Then as you increase the number of levels you want to use on your inferior analog joystick, you will see how the limit goes to ZERO control.

 

You didn't address the argument.

[carmel_andrews]

lets just be glad we are not playing games on a pc or a zx spectrum and not having just a keyboard or mouse option as standard (remembering that you could only use an atari comp. j/s with a speccy if you purchased an additional piece of hardware)

[kool kitty89]

lets just be glad we are not playing games on a pc or a zx spectrum and not having just a keyboard or mouse option as standard (remembering that you could only use an atari comp. j/s with a speccy if you purchased an additional piece of hardware)

Speccy 128k added the joysticks onboard. And most PC games from the mid 80s onward had joystick support, so pretty much for as long as PC gaming was half decent. (with Tandy machines you had the same 5-pin DIN as the CoCo... hopefully not using the same joysticks though )

 

It would have been neat if 3rd parties had taken it upon themselves to release atari-type joyport adapter cards for the PC and indeed that did happen with Covox's sound master, but that was in 1989 by the time IBM's game+midi port adapter was already in relatively common use and the same year the Sound Blaster came out with integrated IBM compatible gameport and adlib compatibility. (both the covox and SB had an 8-bit DAC with DMA and I think the covox actually had a higher sample rate than the SB1.x, but the Covox used an AY8930 sound chip while the SB used the Adlib's YM3812 OPL2 FM synthesis chip instead... had Covox or another company implemented such a sound card much earlier with similar use of the I/O ports for digital joystick ports it probably would have caught on as a defacto standard -obviously with an AY8910/YM2149 rather than the later 8930, which would of course have sparked interest in maintaining compatibility with that sound chip and you did have the compatible successors with the AY8930 or the more powerful 4-op FM enhanced chips like the YM2203 let alone the fairly high-end YM2608... and it's odd that there weren't simple AY8910 based cards for the PC or any sound cards up to the 1987 Adlib -AY or the simpler SN76489 would have been the obvious choices for lower-end mass market sound chips and the AY8910 is what the Mockingboard used on the Apple II -and of course the ST, MSX, FM-7, Spectrum 128k, CPC, Intellivision, etc)

[atariksi]

Read post #315 where I clearly stated I'm not trying to simulate a physical joystick. In fact, if you read post #1, I clearly put arrow keys/ctrl key in a separate item to be voted and also if you read the thread you would know that I argued the superiority of digital joystick even over the digital arrow keys/ctrl key.

 

No thanks, I'll read Post #310 where you clearly contradict yourself. Don't remember your own words? Here they are again -

 

"That's rubbish. If I simulate a joystick using arrow keys + ctrl and the game only uses those 5 bits of information, I have perfectly simulated the joystick."

And you misinterpreted it. As far as the A8 is concerned, it only sees the signals at the DB9 connector. So that also makes sense along with post #315. Next time instead of looking for fault, try to understand the context. And by the way, post #1 comes before post #310. If I replace the joystick with a PC that is sending those signals via CTRL/ARROW keys, A8 game/software wouldn't be able to tell the difference.

 

You are stating that you are simulating a physical joystick using arrow keys + ctrl key.

Nope. All it says is I am simulating a joystick. Nothing about physical. That's what I wrote to prevent misinterpretation by you since I already know that you are only looking to find fault and have no interest in the truth. You know it's like those emotional biased parties in court that only try to find fault with the other party and have no interest in what is ACTUALLY the case.

 

"Digital Joysticks provide better control than Analog Joysticks. It's a scientific fact; let's see who can refute it."

 

Nope, no mention of arrow/ctrl keys there! So why would you screw that up so badly and so quickly? And why lie about it being a scientific fact?

You are being given a chance to refute and you haven't refuted anything-- neither the experiment nor the logic/mathematics. You are just slinging mud and mocking something that happens to be FACT because it opposes what you believed to be true for most of your life.

 

Just because you can't correctly form English sentences doesn't mean that the rest of us have the same difficulty.

Let me spell it out for you as to whose scraping the bottom of the barrel here trying to pick on English and rely on word jugglery. The CTRL/ARROW keys HAS NO EFFECT on your argument. You are arguing analog joysticks vs. digital joysticks with me and the CTRL/ARROW issue will not effect your argument although you have lost that as well. The joystick simulator also simulates the joystick using various joystick inputs as well which apparently you forgot to read. You are the one living in a life of contradiction-- blindly believing that analog joysticks have better control when the opposite is true.

 

Interpretation is what the software makes sure are exact 1:1. It can even avoid more than three keypresses. Also, did you notice that on some joysticks you can push the whole thing down and generate all bits = 0.

 

The software doesn't make sure about anything, the programmer does.

Again word jugglery. Whatever the programmer did is in the software.