Touchscreens vs. Traditional Controllers in Handheld Gaming Discussion

Synopsis

The article aims to compare input via virtual buttons on a touchscreen (represented here by an iPhone) to input from a physical gaming controller (represented here by a Nintendo DS). Participants were asked to complete a level on “Assassin’s Creed: Altair’s Chronicles” four times on the devices while the researchers measured the difference in the number of deaths, completion time, etc.

Strengths

Traditional controllers outperform virtual touchscreen controllers because of the haptic feedback that they provide.

Tactile feedback emerged as an important aspect for enhancing touchscreen performance. For example, “Brewster et al. demonstrated that vibrotactile feedback in touchscreen keyboards significantly improved text entry speed, reduced errors and improved error correction. Participant feedback strongly favored the tactile display.”

Pitts et al. demonstrated that multi-modal feedback is preferred over visual feedback alone in their study. Combining visual, audible and haptic feedback has consistently been rated as effective. These results indicate the importance of haptic feedback in systems where user attention is divided and visual contact with the interface is limited.

The article mentions several design techniques that have been used with touchscreen devices including, the Shift technique, SideSight, SemFeel, and the Blackberry SurePress.

The study included a very detailed statistical analysis, with hypothesis and suitable test and control groupings (within-subjects design).

The paper references articles on "back-of-device" interaction techniques to avoid finger occlusion, a common touchscreen problem.

Another possible remedy is to use some kind of tactile feedback on the touchscreen, even something like the crude 4iThumbs screen overlay.

Vibration patternsusers can recognize up to 10are another way to provide additional feedback when using a touchscreen (also, "sonic enhancement").

Their conclusion is that game designers should redesign the controls for touchscreen displays, rather than just using onscreen versions of the traditional controller buttons.

I thought that it was a very good idea to grab a bunch of different levels of users to get an overall feel of which one was better to use, but I think that they should have used more beginner gamers so that they would have obtained better results for the learning curve.

It would make sense that the learning curve for the touchscreen device would be much steeper because you would have to keep looking at the screen to make sure that you are hitting the correct button. While with the nintendo ds, you could use muscle memory in remembering what buttons you had to press and when. So the regular buttons cut out one processing step.

The problem with virtual buttons as they said takes up to much room on the screen and is limited to the size of the screen. But i believe that even if the screen was bigger and were able to use more virtual buttons, then the same results would happen.

Shows developers should take into account the hardware on which they developing for may need a different interface than previously designed software on different hardware.

Looked into eliminating the check point time relapse, something that would have a major effect on the results.

They found a game available on the IPhone and the DS with similar controls which is what they wanted to test.

The fact that people improved more on the IPhone is interesting.

I agree with the findings of the paper, which demonstrate that a physical controller is more effective for gaming than a touch screen. The physical controller was both faster and more accurate.

The paper mentions using back-of-device interaction as an additional medium of interaction for touchscreen devices. I think that this could make them much more effective and user friendly if utilized properly.

Stresses the importance of designing alternative control mechanisms for gaming rather than trying to emulate gameplay on other platforms.

Researchers were thorough in their study by considering third-party touch aids like tactile button overlays and vibro-haptic feedback.

Details the strengths and drawbacks of touch screens as well as possible methods to improve upon them like developing apps with certain size touch buttons since most people use their thumbs and that requires a certain surface area for accurate input.

Included a very detailed explanation of the testing procedure and how they examined the results.

Their conclusion that developers should design games for touch screens instead of using traditional controls that have been adapted for use on a touchscreen seems obvious but should be implemented betting in the future.

Weaknesses

The results of the experiment were predictable and expected even from a simple glance of the paper. The researchers could have conducted a more extensive study.

Back-of-device interaction involves adding a touchscreen (or touch surface) on the back of the device to avoid finger occlusion. This concept was mentioned as a possible alternative for directional pad controls. But how exactly would something like this work? How would a player know what buttons he is pressing when he can’t even see the back of the device? Would this really be an effective solution?

“SideSight mitigates the need for user input on the screen, or in fact on any part of the device itself. It instead uses proximity sensors to divert the user input region to the areas on either side of the device. While this solves the problem of occlusion on smaller screens it may be impractical for hand-held gaming as it requires the user to lay the device on a surface while using it.”

The researchers could have used a larger sample size.

Since they largely only compared a direct mapping between traditional controller buttons and onscreen touchscreen buttons, the results were not too surprising: using the traditional controllers gave better results.

Each user only had a limited number of trial and test runs, so there was little time to "get the hand of" the touchscreen controller.

The test used had the subjects starting the game in the middle, so learning to use the system was a problem (although all users had the same problem).

The dominant factor in determining that the traditional controller outperformed the touchscreen was by measuring the number of deaths. They also used time to complete the level, but that time was strongly affected by the number of deaths (due to the need to repeat part of the level after dying).

By the time participants reached device 2, they would have quite a bit of experience with the level meaning they would know when not to die, etc.
This is a direct comparison with similar control schemes but many games use alternative methods of control. This is briefly mentioned but not explored.
There are no user opinion questions mentioned, How did the users feel?

They were saying that the touchscreen marketplace for games is much higher then the traditional handheld gaming systems, and they were comparing that the iPhone had somewhere around 21,000 games while the psp had about 600. What I think they should have compared is the game complexity. Because most of the games that are developed for the iPhone are quite trivial compared to the psp games.

It did not seem like they had a good method in determining the completion time with the deaths involved. A better solution, in which someone would have to sit over there shoulder and time, would be to stop the time if someone dies, and do not resume the clock until they get back to the spot where they died.

They said that they wanted to avoid variables in the game by not letting the user use a feature on the DS that would allow them to roll because the iPhone did not have that capability because it did not have enough buttons. But I think they should have left that in there because then it would show the superiority of a traditional controller because they do not have to fill up the screen with buttons.

Since there was only one game tested, it is hard to show that the fault is not with the developer and that the device causes the results shown. In the beginning of the article it was mentioned that game play with other games that were both offered on the mobile platform and the Nintendo DS were different. This would show that other developers have anticipated the difference between the game play environments and made necessary improvements to ensure better game play.

The results show that there is a learning curve with the Iphone, but not so much with the Nintendo DS. This could have been due to the fact the controls for the Nintendo DS are based from older systems that participants could have used. This effect was not fully disclosed in the paper.

The designers used only one level of one game for the whole study. I think they should have compared performance on more titles from different genres.

Users where given only one hour to practice the game before the trial. This is not enough time to learn the different nuances of a complex game like Assassin’s Creed. The results might have been different if they compared experienced gamers from both platforms in the study.

They didn't talk much about game design and different types of gaming. For example, a racing game is very different than a first-person shooter game, and one may or may not out-perform the other on a given platform. They limited their study by focusing too much on FPS games, offering somewhat of an [obvious] assumption that physical buttons would be the preferred method of control. I don't think their study considered gaming history. Games have long been played using controllers with D-pads, joysticks, and buttons, so the Nintendo DS had an inherent advantage at the gate.

Results might have been skewed since users are more apt at using traditional controllers.

Like some of our other discussions most new HCI tech has to go through a period of a type of learning curve in order for the user to become as acquainted with it as traditional methods so it becomes as useful.

The generalization of the results with how complex the tests are is very opinionated.