I don't like the idea of requiring continuous real-time input and visual updates. This disqualifies a lot of applications which I would definitely consider video games.

For example, this rules out anything turn based which listens to input only when it is the user's turn and updates the screen only when turns are taken. This could be anything from a simple board game to a highly complex turn-based 4X strategy game.

I would also argue this doesn't allow for any type of text based adventure game in the vein of Zork: they may listen to input most of the time, but there's no indication of a continuous logic or render loop running while waiting for user input.

I think, the distinction is necessary for determining the first video games. As the genre matures, we may broaden the definition. (E.g., to anything sold and/or reviewed under this title. The problem arises only as long as there is no recognition or general concept of video games.) I would also suggest to exclude anything that is "just" a simulation of an existing, physical game the same way. (Is a billiards simulation a video game? Well, is then a dice game with realistic animations of dropping dice a video game?)

Definitions of game are an old pastime within video game design, and are roughly as useful as definitions of chair; they engage us in a language game, but mostly don't serve a "therapeutic" purpose in the Wittgensteinian sense. All too often when these definitions get specific, they get stuck on a hypothesis around competition, simulation, technical attributes or other concepts that end up eliminating half of the things that are called video games. And equally as often, the proponents of such definitions will, rather than accepting a critique suggesting that the definition is unrealistic, smugly remark, "well, they are not Real Games."

But if I were to take a crack at some of the major boundaries that come up around video games specifically - the phrase "scene and feedback" comes to mind. That is, there is a scene presented, real or imaginary, and player actions automatically produce rules-mediated feedback on that scene through computation alone.

So that excludes mechanical actions like a pinball game, or human game masters(which do come up in online multiplayer often enough to make this a tenuous definition, but somewhat recognizable - we tend to not treat social networks as video games, even though they do resemble MMOs). It says little else about implementation, but the aspect of scene then comes to the forefront. Is Nim a "fictional scene" or is it just data? And are "real scenes", like controlling a robot from afar, actually still video games?

I think the scene question is primarily reflected by focus and player mastery. A game that's been solved like Tic-Tac-Toe is perceived without mystery or anticipation: to win play here, to draw play there. But something with a lot of essential complexity like Chess can draw players into obsession. And something like a painting app explicitly positions the content as data, something to be manipulated in totality, rather than as a scene that is "interacted" with according to rules. For productivity apps to be as interesting as a game, you have to provide your own ruleset, which is a good deal harder than just following one that's handed to you from a designer. Likewise, if the scene is literally just a real-world scenario, the game has no coherent focus and stops being game-like.

But, video games that are "quintessential video game" - the Marios and Minecrafts - adhere pretty strongly to this abductive hypothesis: they have a strong notion of scene, and a focus on direct, rule-based, computationally automated interaction. They have a specific lens through which the scenes are defined that supports the types of interactions being explored - Mario has levels that let him run and jump, Minecraft worlds are designed to support building and exploring.

What makes people willing to swallow Higinbotham's "Tennis for Two" or "Spacewar!" as video game and not OXO(1949) or Nimatron really is about the notion of scene and its associated fiction. Adding an action element immediately makes it more of a scene, one that is limited by your ability to react and manipulate the controls. Graphics and real-time operation accomplish this most easily, but text adventures and strategic wargames with no substantial graphics and turn-based gameplay likewise also produce enough of a fiction to be lumped in the category, albeit reluctantly so by the more graphics-obsessed.

> What makes people willing to swallow Higinbotham's "Tennis for Two" or "Spacewar!" as video game and not OXO(1949) or Nimatron really is about the notion of scene and its associated fiction.

The keyword here is diegesis. In other words, there's a fictional world exposed by the apparatus and things happen in this world, by means of interaction with objects that are occurring both in this fictional world and the rendition on the screen. Game mechanics, rules and narrative content necessarily overlap, by this creating an investment on the side of the player. (Compare the somewhat peculiar importance put by Spacewar! authors to provide an explanation for why shots are not subject to gravity, i.e. weightless "photon torpedoes", or why hyperspace is prone to fail eventually, i.e. experimental "Mk I hyperdrive".)

The aspect of note here is, how this investment is established. As in all games, this is subject to things like riddle, challenge and mastery. So, how is this presented in a specific, typical way? I'd suggest to emphasise the presence of an iconographic viewport and seamless interaction in order to separate video games from other kinds of computer games. Notably, video games present a challenge, which is both tactical and strategical in nature. There are necessarily multiple levels (a hierarchy, not game levels) to the puzzle, with motor skills – as introduced by the temporary aspect of seamless, frame-based interaction – at the basic level. Also, there's an incessant need on the player's side to update her/his/* orientation, of what is going on in the diegesis and what hints may be exposed on the viewport.

(We may even expand on this to establish sub-genres: E.g., first person shooters introduce a specific, subjective instability to the viewport, which adds to the riddle and the challenge of mastery.)

Arguably, in early video games, there was on emphasis on the motor skill and orientation challenge, which contributed to the popularity of person-to-person two player games: For the first time humans of widely differing size, weight, strength, age, and abilities could match on an equal level in a competitive way. (In a way, video games didn't abstract their in-game world as much as they did abstract human players by the very mode of interaction.)