Effects of Display Settings on User Performance in First-Person Shooters Frame Rate and Resolution on Movement Related Tasks

1
Effects of Display Settings onUser Performance
inFirst-Person ShootersFrame Rate and
Resolution on Movement Related Tasks

• Tim Connor
• Adam Fiske
• Ryan Kennedy
• with
• Professor Mark Claypool

2
Overview

• We tested the effects of frame-rate and
  resolution on user performance in first-person
  shooters.
• Specifically, we tested the effects they have a
  on players movability.
• Three basic movements
• Walking
• Running
• Jumping
• Recognition

3
Related Works

• Several studies were done the effects of
  frame-rate and resolution on streaming video. In
  this case, framerate can be very low and users
  can still percieve what is going on just fine,
  but higher resolutions make things much better.
  This is passive media, whereas our study deals
  with active media.
• McCarthy, John D, Sasse, Angela, and Miras,
  Dimitrios. Sharp or Smooth? Comparing the
  Effects of Video Quantization vs. Frame Rate for
  Streamed Video. University College London,
  London, United Kingdon, 2004.
• Tripathi, Avanish, and Claypool, Mark. Improving
  Multimedia Streaming with Content- Aware Video
  Scaling. IMMCN, Durham, North Carolina, 2002.
• Other studies, specifically Reddys The Effects
  of Low Frame Rate on a Measure for User
  Performance in Virtual Environments, show that
  users performance at different movement tasks
  begin to greatly increase at 15 fps.
• Reddy, Martin. The Effects of Low Frame Rate on
  a Measure for User Performance in Virtual
  Environments. University of Edinburg, Scotland,
  UK, 1997.
• Swartz, Merryanna, and Daniel Wallace. "Effects
  of Frame Rate and Resolution Reduction on Human
  Performance." Silver Spring, Maryland, 1993
• A previous study was done very similar to ours.
  Quake 3 Arena was used to study the effects
  frame-rate and resolution had on users ability
  to aim and shoot targets.
• Claypool, Mark, Claypool, Kajal, and Damaa,
  Feissal. The Effects of Frame Rate and
  Resolution on Users Playing First Person Shooter
  Games. MMCN, San Jose, California, 2006.

4
Test Parameter and Map Development

• Frame-rates
• 15 fps
• 7 fps
• 3 fps
• Resolutions
• 640 x 480
• 512 x 384
• 320 x 240

• Four Maps
• Walking
• Running
• Jumping
• Recognition

A test harness was developed in java to randomize
the order of maps and run them.
5
Walking / Running map
6
Jumping Map
Resolution Map
7
Test Setup

• Specs
• AMD Athlon64 3700
• 2 GB RAM
• 300 GB HHD
• nVidia GeForce 6800
• 17 Monitor

• User space Normal computer setup
• Second keyboard and monitor behind a divider away
  from user
• This was necessary to monitor the user and to
  control the test harness

8
User Performance Rating

• After the testing we developed a user performance
  rating
• For frame-rate maps it was based on two things
• Time taken to complete a map
• Amount of health lost during this map
• (Time)(Health Lost) Performance Rating
• For resolution maps
• Time taken

9
Analysis of Performance on Walking Maps
As frame rate improves, worst performances
improve and the spreads of data points become
smaller, possibly indicating that frame-lag
effects are having less of less impact on
performance.
None of the confidence intervals overlap,
indicating that, with 95 confidence, frame rate
makes a significant difference in performance
rating for the frame rates tested.
10
Analysis of Performance on Jumping Maps
As was the case in the walking maps, as frame
rate increases, the worst performances improve
and the spread of data lessens.
The overlapping of the 7fps and 15fps confidence
intervals indicates that performance at 3fps is
significantly different from performance at 7fps
and 15fps, but performance at 7fps is not
significantly different from performance at 15fps.
11
Analysis of Performance on Running Maps
Similar to the results for the walking and
jumping maps, as frame rate improves, worst
performances do as well. Additionally, the spread
of the data lessens.
Like the analysis of the jumping map, the
confidence intervals for 7fps and 15fps overlap.
This tells us that there is a significant
difference between performance at 3fps and
performance at 7fps and 15fps, but not between
7fps and 15fps.
12
Analysis of Performance on Resolution Maps
On the lowest two resolutions, 320x240 and
512x384, respectively, participants completed the
maps with similar distributions of time. At the
highest resolution, however, 640x480, worst
performance improved and the spread of data
lessened.
95 confidence intervals for 320x240 and 512x384
overlap with each other, but 640x480 does not
overlap with either. This indicates a significant
difference in time to completion for the
resolution of 640x480.
13
Conclusions

• Frame rate impacts performance in walking,
  jumping, and running.
• Performance in walking maps improves
  significantly with each jump in frame rate
  tested.
• Performance in jumping and running maps improved
  significantly from 3fps to 7fps, but not from
  7fps to 15fps.
• Resolution impacts performance in recognition
  tasks.
• Time taken did not significantly improve from
  320x240 to 512x384. But, 640x480 is significantly
  better than both of the lower resolution.

14
Future Work

• Things that could be expanded upon in this
  particular study
• Better demographic range to draw conclusions
  between gender or age
• More recent first-person shooter to make it more
  relevant to today.
• Effects of frame-rate and resolution on user
  performance in other genres.

15
The End
Questions?