For my last hour of Diablo 3, before I vanquished the titular antagonist from the world, I decided to install some freeware data mining applications.
The first was IOGraph, designed to draw simple lines to plot where the mouse pointer explores and stops. This is what the history of my mouse movements looks like, for the final hour in Diablo 3 as a level 30 Demon Hunter.
It would be impossible to deduce I was playing a game by looking at this image, but it does plot the experience of Diablo 3
The next step for this concept of plotting my time with Blizzard's latest was to track the same areas with two different character classes. Classes, in all games, are designed to change the experience. Of course, this means characters have different stories and abilities, but does the core pattern of Diablo 3
change based on a selected class?
For this experiment I decided to return to the Cathedral in the very beginning of Diablo 3
.Tristram CathedralDemon Hunter - Level 33 / Nightmare Difficulty
Section Completion: 11:16
With a focus on ranged attacks, my mouse tracks for the Cathedral [above, red outline indicates character position] show a wide radius of attack. Usually, I would plant my Sentry into the ground, hole up in a section of the map and fire into the distance. Every section of the Cathedral was explored in this run and I was partnered with the AI-controlled Templar to ensure success. [Non-overlay path image
Using a program that maps mouse movements as a heat map, it's easy to see how far out of the center position my cursor would focus on to attack enemies.Key and Mouse Click Statistics [WinOMeter]
- Keyboard Presses: 148
- Mouse Trajectory: 58.154m
- Mouse Clicks [LMR]: 305, 0, 36
My next focus was the Barbarian class, a warrior that prefers to get in the middle of the action -- very much unlike the Demon Hunter.
Barbarian - Level 5 / Normal Difficulty
Section Completion: 12:08
According to the data compiled, playing as a Barbarian is far more focused [above]. The range for attack is small, so my mouse was centralized around my character for the majority of the playthrough (the exception being when I would load my inventory or skills screens). [Non-overlay path image
The heat map of my Barbarian session highlights the focus of playing as a melee-based class. The mouse trajectory for both characters is very similar, which show similar ways of movement were completed in both runs. Also of note was my reliance on mouse-mapped attacks for the low-level character; however, that's directly linked with the lack of skills available to me on my keyboard.Key and Mouse Click Statistics [WinOMeter]
- Keyboard Presses: 72
- Mouse Trajectory: 55.756m
- Mouse Clicks [LMR]: 775, 0, 48
The patterns for these two classes are quite different between these two classes. Of course, the difference in character level could confound the data, so I intend to map the first chapter as two new characters (and new classes). The aim in the long term is to explore how different, at a completely stripped down and core level, Diablo 3
playthroughs can be.
Returning to the final hour of Diablo 3
[overlay below, heat map under]: though the action is much more intense and far more time was tracked, the pattern for the Demon Hunter remains fairly similar in the final hour to its appearance in the first act's Cathedral section.
According to key and mouse click tracking software, my left mouse button was clicked an average of 38 times per minute (roughly 2,282 cph or one click every two seconds) in the final hour of Diablo 3
. My mouse trajectory within the hour was 334m (5.57m per minute). With Diablo 3
abilities mapped to keyboard presses – and the final hour involving a slew of enemy encounters – it's no surprise I hit 1,632 keys in that time (27.2 per minute).
Though I didn't track the game from start to finish – I'm working on that now – you could estimate [based on the above data] that my left mouse button was clicked over 68,000 times; keys hit almost 50,000 times; and my mouse traveled over 6 miles throughout my first complete 30 hour run.
There is more that can be done with this software. After exploring more of Diablo 3
from start to finish, my next goal is to take these tracking applications into the MMO space where we will ask: Do most free-to-play MMOs all share a similar pattern? Are they designed, at their core, to be the same type of playing experience?
The plan is to examine five leading F2P MMO titles to examine what similarities the games share when stripped down to nothing but simple lines and heat signatures. And maybe, eventually, we'll be able toguess what genre a game is simply by examining how we interact with it.