Discussions about videogames often hinge on the relative value of the game – what it’s offering to the player and for how much. But when are games good value? What are the criteria for determining the value of a game?

One way of considering this is in comparison to other media. A movie, for instance, retails on DVD for about $15 (say) providing two hours of entertainment. This sets our baseline for judging game value (given that DVDs run in a similar environment to typical videogames) – about $4 per hour if you assume the DVD is watched twice. On this metric, console videogames may seem like great value – some might offer 100 hours of entertainment for $60, which is just $0.60 an hour. But this conclusion hides a hidden cost that needs to be taken into consideration: the purchase price of the console itself.

For a gamer hobbyist who buys and plays many different games, the cost of the hardware is actually not a very significant factor. If we call the cost of a power console like the Xbox 360 or PS3 $250 (while bearing in mind that early adopters paid more), this cost needs to be shared out with all the games played on it. But for a hobbyist, they are likely to play 20, 50 or even 100 games over the lifespan of that console – so the hidden platform cost is just $2.5 to $10 per game. A title offering a hundred hours might be pushed up to $0.70 per hour, but the value is still good.

Now look at what happens when we switch our attention to the mass market. A typical midcore gamer (who plays, say, FPS and driving games on a power console) will only purchase three or four games for their console over its entire lifestyle. This means that each game costs an extra $60-80. What’s more, such a gamer finishes very few of their games – a title that hobbyists might play for 100 hours might only see 10 to 20 hours of play for this typical mass market player. Now the boxed game is looking very expensive indeed at $120 total expenditure (including the hidden console cost) and a consequent $6-12 per hour of entertainment. This is a lot more expensive than DVD movies, hour for hour.

The same kind of value comparisons can be made in other contexts. Early arcade games, for instance, charged $0.25 for play that would last an unskilled player about 5-10 minutes, while an expert could last for 30 minutes or more. (In my arcade while I was growing up I had at least two arcade games – Nemesis and Rastan Saga – I could play for half an hour on one credit). So the unskilled player is netting $1.50 an hour, while the expert is getting $0.50 an hour. These actually look pretty good next to the $4 per hour estimate for a DVD, although you can see why those people invested in arcade revenue moved to the pump and play model (typified by the 1985 title Gauntlet) in order to get more coin drops. Once again, however, the gamer hobbyist gets better value for their games than anyone else – paying less than a dollar an hour.

What about social games, like those offered on Facebook by companies like Zynga? Well in principle these can be played for free so there is a sense in which these represent truly excellent value (you can’t beat free!). But of course, you can’t play at your own pace without paying for the privilege. Let’s consider a hypothetical title which charges $0.20 for an energy refill that is provides 15 minutes of play (I welcome any actual figures for social games that someone might have, but this will do as a rough estimate). The paid version of this game is $0.80 an hour. For the hobbyist, the probably looks like a con since they can get a much higher quality experience on their console for a very similar price. But for the mass market player, this is a bargain – the console game was offering them $6-12 per hour, after all. Social games, as would be expected, offer excellent value to mass market players – which is precisely why Nintendo is scared of them.

All of this discussion serves to demonstrate that the value proposition for any videogame is not something that can be determined independent of the player in question. Gamer hobbyists get their digital fun at excellent rates because they are bulk purchasers of this kind of entertainment, and enjoy the corresponding discount. Mass market players, however, get much better value for money out of social games than they do from the alternatives – even if they are paying to play – which is one of the many reasons that these kinds of games are becoming an ever-bigger player in the market for digital games.

…so I was able to hold the first edition of Imaginary Games in my hands this afternoon. Don’t know if this means it’s available to buy now or not, but there are definitely copies in existence now.

How could we best recount the history of digital game design? Although genres seem to provide a framework for discussing the progression of various designs, are the categories being applied reflective of actual historical forces, or merely convenient collections of otherwise spuriously connected games?

A recurring but often unnoticed problem in game studies is the inability to discuss the historical lineage of a title, or only being able to do so by assuming the validity of often dubious genre categories. For the most part, the obsession with research that “smells like science” has reduced historical investigations to a lesser status. This is unfortunate: a culturally-situated critical discourse about games could enrich game studies immeasurably.

There are numerous complications. The direct influences on any game designer are often unknown, plus many other team members on most projects have input towards design considerations, making it impossible to unravel a chain of antecedents. Similarly, any commercially successful title is affected by and has effects on the expectations and assumptions of players (including those players who make future games) that affect which titles succeed, and this applies differently in each region of the world. The ZX Spectrum, for instance, directly influenced British games development (including the Grand Theft Auto franchise) but was not a part of the local game histories in the US or Japan.

Attempts to provide a taxonomy of game genres founder on the lack of consistent criteria, and usually have to be arbitrarily assigned. Connecting ‘shooters’ into a lineage suggests scrolling shooters were direct influences on first person shooters, for instance. But there's no evidence suggesting Zaxxon has any connection with the design of DOOM, or that Space Invaders inspired Zaxxon. As a historical tool, genre categories can provide some useful connections – DOOM certainly did influence GoldenEye 007, for example – but genre cannot be used as a unifying framework for game history because the genre lineages are narrowly valid and do not constitute a complete description of game history.

An alternative approach can begin by considering the historical constraints that acted on any given title, grouping games by common constraints into related clusters. The considerations that follow from hardware are along similar lines to what Montford and Bogost discuss in Racing the Beam, but this is not the only (nor even the most important) constraint to be identified. For many games, market, cultural or contextual factors had a greater effect on the design. We can call the former hard and the latter soft constraints on the design.

For example, 8 bit arcade games from 1980-1984 share hard constraints in terms of computing power, display technology and control devices, as well as the soft constraints of an arcade coin-op: for commercial reasons, games were required to be of short average duration to drive many coin drops per hour. Galaxians and Pac-man belong in the same cluster on this approach, despite the obvious differences in genre. Gauntlet, in 1985, marks a change in conditions, away from one coin, one play and into a many coined ‘pump and play’ cluster, fostering multiple coin drops in each game (what could be called the original microtransaction model). Gradius/Nemesis straddles the two clusters, originally one-coin but later modified with a continue to encourage multiple coin drops per game. The change in soft constraints here is more significant than the change in hard constraints, and this is frequently the case.

Throughout the nineties, the most significant shift in hard constraints resulted from the increasing focus on polygonal 3D representations. Here, the hardware development was lead by a demand for 3D graphics so there is a sense in which the hard constraints were driven by soft constraints. The 2000s were marked by a radical increase in touchscreens, first with the DS (about to become the best selling console of all time) and then with the iPhone, iPad and all the copycats since. Once again, soft constraints led hardware: a correctly perceived need to simplify the intensely complex controllers that resulted from the constraint history of the preceding decade produced the DS.

Other hard constraints have grown up more organically. Despite all the efforts at creating online entertainments in the early 2000s, the technology was not up to speed until the middle of the decade. The easing up of constraints on bandwidth moved clusters of games that had grown up in LAN play, such as the multiplayer FPS format codified in Quake, into the open internet. This brought significant new soft constraints into play, including the ongoing shift to games-as-services that has emerged from the mass market demand for easy-yet-fun games coupled with the hard constraints of browser-based, persistent content.

The rise of browser-based games also highlights the significant transformation in hard constraints in recent years: the specifics of physical hardware now matter less than ever before, since content is defined for software platforms (such as Flash) rather than specific hardware. Although some hardware-exclusive games are still made, the number of these is vanishingly small next to multi-platform or platform agnostic titles. Software is the new hardware, or at least the origin of the new hard constraints affecting game design.

This historical approach to game design offers a broader understanding of how various constraints, both technical (hard) and circumstantial (soft) have shaped the development of the contemporary digital games industry. Grouping games into clusters based on common constraints provides an alternative way to present the history of games without resorting to ad hoc genre categories. I have not found any better approach for game design history, and I am convinced that without something of this kind, game studies is lacking a vital piece of the story of games.

During one of the talks at DiGRA, the person sat in front of me seemed far more intent upon the puzzle game he was playing on his iPad than he was in the talk – although he asked perfectly cogent questions afterwards. Later that day, another delegate showed me the same game – Drop7, now owned by Zynga – and discussing what had happened earlier we discovered it was Eric Zimmerman who’d been playing it in the session. He said it helped him to concentrate, and asked if I thought it rude to play in the talk. I pointed out half the delegates were typing notes or tweeting from their phones, so it scarcely mattered. But I became interested in this game that he'd used as an aid to concentration. What made Drop7 tick?

The essential play of Drop7 draws from the well established tile-stacking genre that flowers from Tetris. Whereas Pajitov’s classic relies on the simplicity of it's mechanics for its appeal (anyone can work out how to stack tetronimos), later tile-stackers such as Dr. Mario, Super Puzzle Fighter, Baku Baku Animal and Puyo Pop focus on combos to drive interest. The player learns to stack tiles in such a way that, when correctly prepared and then triggered, a giant chain reaction occurs. This reaction is the core of Drop7’s appeal as well, but unlike earlier stacking games Drop7 has a secret weapon in its rules.

Failure in tile stacking games occurs when the board becomes stable, which results from a configuration where the reactive combinations of tiles cannot be made to come into contact e.g. in Super Puzzle Fighter, tiles of the same colour don't touch. Conversely, it is possible to stack tiles chaotically in such a way that a small landslide will trigger large (accidental) chain reactions. This quality, volatility plays almost no role in Tetris but has been core to all other tile-stackers since.

Ordinarily, the volatility of any given board is a function of spatial dynamics – reactive sequences must be close to one another, yet incomplete. In Drop7, however, reactions are not triggered spatially but on a tile-by-tile basis. The rules of Drop7 (unusually) state that tiles react based on their individual horizontal and vertical line counts. This creates unique play options: a tile at the base of a stack can be triggered by action elsewhere on the board, whereas the traditional tile-stacker requires an open channel to trigger reactions by direct proximity. A side effect of the per tile reaction paradigm is increased volatility in practically every configuration – Drop7 boards become stable only as low-numbered tiles form a crust above locked tiles.

The locked tiles are key to making the game work. Expert players can position the locked, blank tiles to sustain future chains, but perhaps more significantly each locked tile has a high chance of reacting when it is revealed. There is, crudely speaking, two 1 in 7 chances for each newly revealed tile to react (randomly matching the horizontal or vertical tile counts), meaning each has more than a 1 in 4 chance of such a tile reacting wherever it is revealed on the board. This ensures highly volatile boards. Although previous tile stacking games have occasionally featured locked tiles, the spatial configurations required for reacting made these tiles contributed stability not volatility to the board. In Drop7, these blank tiles are more volatile than any previous game of this style and this adds significantly to the appeal.

Although there is some strategy in Drop7, the game’s compulsive play is rooted in fooling the player into believing they are critical to the outcome of each board (when dispassionate analysis would suggest chance was playing a major role). Key to this is the near miss effect, which causes a release of dopamine (as if winning) when players nearly win, motivating persistence. As with all good tile stackers, most players finish every game with the compelling sense that they could do much better next time, despite success being highly contingent upon luck. The near miss effect is associated with gambling, but is generated strongly by any game with a high role for chance, including tile-stackers, and tile-matchers, like Bejeweled. In all these cases, a high element of luck means even an incompetent player can do well sometimes, while high volatility makes for a more exciting and rewarding game.

Drop7’s design has a few issues, particularly in terms of game length. A competent player can play one game for more than an hour, which is rather long for any tile-based game – especially when the near miss effect is encouraging players to have ‘just one more try’. Having a marathon mode (called ‘Normal’ in Drop7) usually means having a sprint or blitz mode as well, but none is currently offered. This leads to more than a few people having to give up the game because there’s no ‘safe’ way to play it. (I myself deleted it two weeks after downloading it, and not from boredom).

From the accessible simplicity of Tetris, the puzzle game genre became stuck in a rut of tile-stackers based largely on colour matching and time pressure. Bejewelled kept the coloured tile matching but sidelined the arcade-style time pressure as optional and ditched stacking in favour of high volatility boards that automatically restock. Drop7 attains similarly volatile boards with more classic stacking play made interesting by a new approach – per tile conditions instead of spatial conditions. Note also that time pressure, now wholly unnecessary, has gone from optional to entirely absent.

Success for puzzle games lies in the combination of simplicity and volatility, and Drop7 enjoys the benefit of both, delivered with clean lines into a compulsive screen-tapping repetitiveness that easily becomes hypnotic. That focused so-called ‘flow’ state is easily reached by tile games (and solitaire card games), a stream of dopamine-releasing decisions that absorbs attention coupled with near misses to drive repeat play. But to keep the player’s attention, the game must offer, in Sid Meier’s memorable phrase, interesting decisions – or at least, interesting outcomes – and in tile games, it is often volatility that provides that interest.

I met Slaton White at Casual Connect this year – he’s a game design student with a lot of drive and energy who will either do great things or have his spirits crushed by the true horror of the games industry (possibly both!).

Anyway, Slaton has started his own blog, and kicked off with a piece entitled The Future of Games: Can Games Impact your Life? Check it out!

On Tuesday 18th October, I will be giving a presentation on the new book, Imaginary Games, at Oxford University, Room 2 of the Taylorian Institute (St Giles', Oxford, OX1 3NA) starting at 5:30 pm. The details can be found at the Zero Books Lecture Series blog.

If you’re in or near Oxford, you’d be more than welcome to come along!

Cross-posted from Only a Game.

Here’s a link to the paper I presented at DiGRA this year entitled Player Typology in Theory and Practice, and the abstract:

Player satisfaction modeling depends in part upon quantitative or qualitative typologies of playing preferences,  although such approaches require scrutiny. Examination of psychometric typologies reveal that type theories have—except in rare cases—proven inadequate and have made way for alternative trait theories. This suggests any future player typology that will be sufficiently robust will need foundations in the form of a trait theory of playing preferences. This paper tracks the development of a sequence of player typologies developing from psychometric  type theory roots towards an independently validated trait theory of play, albeit one yet to be fully developed. Statistical analysis of the results of  one survey in this lineage is presented, along with a discussion of theoretical and practical ways in which the surveys and their implied typological instruments have evolved.

Thanks to everyone who attended my presentation and the boardgames panel! It was a great shame to only be able to attend one day of the conference.