Sunday 13 July 2014

UnBrogue postmortem: part 2

It has been over a year since my last UnBrogue post-mortem article, which just goes to show how much time I have spent developing this Brogue variant, which is to say, virtually none.

However, Brogue itself has had two version releases since I last finished the game, and a number of design changes have either vindicated or echoed decisions I made in UnBrogue, so I thought it worth documenting the similarities. I'm not going to be able to draw many conclusions from these, other than 'I told you so' and there's plenty of things in UnBrogue that are not in or intended to be in Brogue, but which I'd like to see preserved and playable at some future date.

The most obvious one in Brogue is that armour now affects your stealth. This was one of the original reasons for creating UnBrogue, however Pender incorporated a significant rewrite of the stealth mechanism for the Brogue release whereas I merely added a stealth modifier for different armour types. I've not played enough Brogue 1.7.4 to know how the stealth mechanics differ, but the effect was significant enough in UnBrogue that in plate armour you'd effectively end up fighting all the monsters on a level near the stairs you entered through. I'd like to think I was somewhat influential here in making a case for linking armour with stealth, although this is harder a revolutionary idea.

The second is that the dagger now has a higher backstab multiplier than other weapons - although a fixed amount in Brogue, whereas it is enchantment dependent in UnBrogue. Again, hardly a revolutionary idea, but it was one that I was criticised for by some long term Brogue players, so I had a small moment of satisfaction in seeing this in the release notes for 1.7.4.

The third is that immunities and slays now include categories of monsters rather than individual monster types. Again, Pender differed significantly here in that the categories of slaying and immunity are far broader than the ones I chose, although there are some of the same choices (jelly immunity in both versions). I'm less willing to claim any influence here, because I believe Pender already had intended to do this.

The fourth is one of the new puzzle rooms is similar in theory to a number of the puzzle rooms in UnBrogue. Again, this is because I built it from the same elements already present in Pender's game - and Pender would have noted direct inspiration in the source code.

The fifth is the addition of further weapon types. The designs are radically different (and much more inspired on Pender's behalf), and the flail is likely influenced by games like Hoplite, but it was clear that there was a need for more variety, and what I've played of the whip is especially fun.

As for UnBrogue, I'd like to fix up the last remaining bugs and release a full 1.1.7 rather than the release candidate it has been stuck at for 2 days less than a full year; however the bugs I've introduced were pretty resistant to being fixed (I can't duplicate the wand/staff crash bug) and it is likely to be some time before this will happen.

And the attraction of working on UnBrogue has weakened far faster and further than UnAngband. I think there's several reasons for this. Firstly, as noted above, a big part of the initial reasons for UnBrogue have been incorporated into the main game, even though UnBrogue has gone a lot further than just these initial changes. Secondly, Brogue has diverged in a number of systemic ways that it would probably require that I start reimplementing UnBrogue from scratch on the new code base. Thirdly, Brogue has fallen into this unusual space for me, game play wise. After playing the likes of 868-Hack and Hoplite, Brogue feels in many ways both too big a game: the early decisions are spaced 'too far apart' rather than being turn by turn, and I now find myself on autopilot through the first few levels which inevitably gets me killed on level 4 or 5. Whereas Brogue doesn't have enough strategic headroom to be a truly sprawling game that I can lose myself in (like Unangband).

That's not to say that I don't think Brogue is an amazing game. And I also would like to preserve some of the crazy ideas that UnBrogue has, to continue as a variant of Brogue moving forward. But between child rearing, starting work on Unangband again, and my other current obsessions (like play testing the 3rd edition of High Frontier), I don't see a time or a place in the short term for a new UnBrogue release.

PS: I've updated the Rogue Basin UnBrogue page and added a page on the Brogue wikia to include the links to the latest UnBrogue downloads if you are still interested in playing the game. I think it is well worth it, and I've spent a huge amount of satisfying time adding stuff. I'm especially proud of all the new puzzle room designs I was able to come up with.

Monday 7 July 2014

The Real Roguelike Renaissance

There has been a mini-spate of 'explainer' articles aimed at easing the neophyte gamer into the world of roguelikes: both on Gamasutra ('Roguelikes': Getting to the heart of the it-genre), and IGN (Roguelikes: The Rebirth of the Counterculture). Both simultaneously praise the old school nature of high difficulty levels, permadeath, procedural generation, and point towards the emergence of the genre into the mainstream as a fresh new trend, with interviews with well known indie game developers such as Edmund McMillen (The Binding of Isaac) and Daniel Cook (Road Not Taken) as well as up and comers like Teddy Lee (Rogue Legacy), Paul Morse and Duncan Drummond (Risk of Rain) and Keith Burgun (100 Rogues, Auro).

But by pointing at successful examples of commercial roguelikes and focusing on pull quotes from current developers, both articles miss the real roguelike renaissance which predates the examples given by a number of years. Listeners of Roguelike Radio and members of the close knit roguelike community will be aware of the much more titanic shifts in roguelike development in the noncommercial space from a long spanning tradition of the big four roguelikes (NetHack, Angband, ADOM and Dungeon Crawl), through to the relative stasis of the early to mid 2000s followed by the explosive reinvention of the genre inspired by the 7DRL competition, and coffee break roguelikes in general.

I would like to trace a through line from this early roguelike Renaissance to the antecedent of almost all recent commercial roguelikes, Spelunky, which in its freeware form was a huge influence on the early successes of the Binding of Isaac and FTL, but I'm not sure whether such a path can be followed. If it is, Derek Yu would be the person to talk to, but neither explainer article includes an interview with Derek, and I don't recall him talking about whether any of the more recent games influenced Spelunky. There is a direct connection between Derek Yu and DoomRL, but I suspect Spelunky was much more fashioned by the general rise of the indie game and Derek's fondness for NetHack than his participation or awareness of early 7DRL efforts.

But even if there is no connection, it is an important part of the story that the community should be telling and a narrative that is far more interesting than the 'difficulty is in again' message that the mainstream press is perpetuating. It is also a way of including the unsung heroes of the rise of the roguelike - the authors of countless freeware roguelikes who influenced and built the platforms upon which we stand, as well as the original authors of Rogue who founded the genre.

Sunday 16 February 2014

Review: High Frontier Colonization

Are you willing to gamble an hour of time, sweat and frozen tears on the roll of a dice? That's the question High Frontier will ask of you, and if you're not willing - and the randomness this game tempts you with can always be overcome with more turns and more water (its unit of currency) - the steely eyed missile man next to you will be, and he (or she) will win the game as a result.

Phil Eklund designs board games by building a simulation then letting his play testers tell him what is fun; and High Frontier is the simulation closest to his heart and some twenty years in the making. The basic game (sadly out of print) outlines the template of colonising the solar system by flying a rocket to a planet or planetoid which comes in five spectral classes, prospecting and (if successful) industrialising it using a robonaut and refinery, and then ET producing advanced technologies in the resulting factory which allow you to fly further, faster with less mass and get more water when you get there. The Colonization expansion takes this template and turns it into a compelling, deep, rewarding game.

Each turn you get two types of actions: operations, which lets you progress towards your goal by doing things like earning income, researching patents (cards) in auction, boosting cards into orbit; and moves, which lets you form and move a spacecraft. You begin with one crew operation, and one rocket move, and by acquiring additional technologies, many of which are presented in the Colonization rules as self-contained modules, you can get additional moves as you build and promote freighters, and promote your Bernal - a self-contained space habitat that I strongly recommend you include the rules for if you are playing with the political and event tracks, and additional operations by recruiting and boosting colonists.

The political track comes into its own with the colonist module which lets you recruit would be space farers from the mundane (Islamic Refugees) to the theoretical (Blue Goo Symbonts) - each of which has political allegiance to and votes for one of the five factions in the game, and not necessarily the faction who is its employer. The current politics determines how much you get when you free market patents back to the bottom of the deck; but also who is in power and able to antitrust duplicate card types out of any other player's hand. Extreme politics also restrict some or all players in the game and otherwise mix up the rules; intervening politics takes the form of anarchy or war, which allows felonies and even combat between factions.

High Frontier has been criticised for being inaccessible: and this inaccessibility takes two main forms. The first of which is mechanical. In the perhaps 25 turn basic game, you'll spend the first 15 to 20 turns doing nothing but auctioning for patent cards and boosting rocket parts in order to build and fuel your first rocket. The basic rockets in the game use water for fuel, and 40 tonnes of water in low earth orbit (1 WT) is the basic currency of the game, so you literally are burning dollars as you move around the solar system. The last few turns involve a rapid and risky land grab and industrialisation and then the basic game is over just as it feels like it is beginning - a fault Colonization rectifies but unfortunately by extending what is otherwise a briskly paced 2 hour game.

The second point of inaccessibility is conceptual: and it involves understanding the map. The High Frontier map is an elegant and beautiful artefact - the London Underground map of the solar system out to Sedna in the Oort cloud. Moving about this map, once you grok it, is both graceful and frustrating, as planned routes fall short by one tank of fuel, or the mass of the fuel required to get your rocket to its destination instead brings it to a dead stop. For beginners that frustration manifests in failed missions, where the price of failure often involves many turns of effort to reassemble the rocket and recover the stranded crew.

The map is organised in concentric zones with the sun as their centre, named for each of the planets (plus Ceres) and inscribed with helical paths which allow you to move between zones, planets and asteroids. The primary obstacle in your path takes the form of burns. Each burn uses a number of steps of fuel depending on the efficiency of your spacecraft's thruster: the highly inefficient built-in crew module thrusters cost six steps of fuel per burn, the typical workhorse thrusters you build use two to four fuel steps per burn and highly efficient rockets less, to the point where you reach star drives which effectively use no fuel to move (but so do solar sails which you can research immediately, but whose thrust level and dependency on solar power restricts them to the inner solar system unless you know what you're doing). The lighter the rocket, the more steps of fuel you get for each tank of water you invest (but the more fuel carried, the lower the payoff for adding another tank).

Burns appear at the edge of each zone and around planets - the larger the planet, the more burns intervening; and you typically plan starting routes so that you cross no more than three burns before you reach your destination or a refueling point along the way. But you have to balance burns against thrust requirements, because to land on an usable site requires that you have a higher thrust than the size of the site, or you pay a penalty in fuel steps equal to the size of the planet when you land, and again when you take off. The Moon is size 9 in this scheme, and is one of the more inaccessible places in the whole solar system when you take this into account. High thrust thrusters are hugely inefficient (crew modules are thrust 8 or 9, most starting rockets are thrust 3 to 5), so you'll want to carry a second thruster for the landing, but that adds mass which reduces rocket effectiveness...

When you do reach somewhere; and provided you've managed the more complex task of bringing a robonaut or crew with you, you may be able to prospect the site and claim it. You can only prospect if the ISRU rating of the card you're using to prospect is less than or equal to the water rating of the site, both on a 0-4 scale. Crew starts with 4, most beginning robonauts are 2 or 3: the most easily accessible sites are 0, 1 or 2. This is a hard limit, so you have to envisage the board as if sites with lower ISRUs than you can access simply don't exist, for both prospecting and refuelling purposes. But the roll for prospecting depends on the site size, where you roll one dice (or two, with a buggy) to try to get less than or equal to the site size. Comets are a great example of this trade off: they are wet (4 hydration) but small (size 1) and have adjacent hazards which you must spend 4 WT or avoid a roll of 1 to pass through safely, and the majority are synodic comets are only accessible for part of the solar cycle.

A successful claim is only the beginning, because among the heavier cards in the game are refineries, which you combine with a robonaut to industrialise and build a factory on a successful claim (returning both cards to your hand). It is at this point the spectral class of a site matters: because all the technologies you can build also have a spectral class, which allows you to produce the black side of the card but only at your factories with the matching spectral type. Black cards are typically improvements, but the best white cards typically have the most underwhelming black sides, and the most powerful black sides (including the legendary Salt Water Zubrin) are on the black side of the worst white cards. The most easily accessible spectral class is C but few powerful cards are this spectral type (and no gigawatt rockets), the majority and most powerful of cards are spectral class S, but S sites are amongst the smallest and driest in the game (Lunar excepted), and V, D and M spectral classes are distributed in smaller quantities in harder to reach places.

And when you've built the first of your factories and can begin producing black cards that the game of High Frontier Colonization gets going. I can only describe a 2 player High Frontier Colonization game I played recently over at the Board Game Geek forums where I was UN playing NASA in order to capture the flavour of the game.

NASA was able to establish an early claim on Deimos and started shipping water and ET produced D technologies back to his Bernal. I had tried slowing the game down by going anti-nuke, and using my diplomatic immunity to allow me to continue to research and boost reactors but when it became clear that the game was going to run away from me, I forced two elections using Activism, with a threat of going Egalitarianism forcing him to move to Anarchy after he spent an unprecedented 7 WT on the first election to do so to try to counter the political dominance that my Vatican Observers gave me. With my second election I declared war, and hit his Bernal with boosted Kuck Mosquito powered by a Lyman Alpha Trap, killing his Siren Cybernetics Inc colonists who had threatened me consistently with antitrust, along with 7 other cards including two blacks.

I followed this up with a successful water theft on Deimos, avoiding two of his missile counter salvos, and tried to move my Bernal (with the Vatican Observers) to Vesta to avoid further attack which I had already industrialized but it was destroyed moving through the Karin group which I had bust previously.

I was forced to rely on the Ablative Laser and Biophytolytic Algae Farm to moving and prospecting from comet to comet (unsuccessfully), until just before the red synodic comet, which (due to turn order) he was able to get to before me and prospect successfully - I managed to industrialize a non-synodic comet one burn away. I had been able to research and produce both the V freigher and GW thruster which held out some hope, but with a comet dirtside (along with M class Hertha he used to build his terawatt thruster) he was able to boost more colonists, turn his Bernal into a lab and reach Europa. I conceded when he researched the D freighter after a long inspiration burn: it would have been a short step from there to emancipate the robots and build the Beehive future victory with 4 spectral classes Europa would have given him he needed for the Epic Hazard op.

I want to stress that I conceded at what would normally be the beginning of the High Frontier Colonization late game - because the game had been so one sided. I had failed every prospect I attempted except the last (and where prospecting was automatically successful), he had succeeded at every prospect no matter how unlikely. But what sticks with me, is no matter how far behind I was, there was a possibility at almost every turn which I could leverage, if I looked hard enough. The game felt like playing chess, but where I could find cracks between his bishops and rooks with which to move my remaining pawns.

And this is why High Frontier Colonization rewards the time you invest in it. There is no game winning tactic, easy synergy or cheap trick: and yet when you look hard enough, every move you make takes advantage of your knowledge and understanding of the interlocking systems in the game.

Thursday 2 January 2014

10-100

I'm not going to get the runner up details together for a little while, but Darren Grey made the intriguing suggestion that the roguelikes which get between 10-100 votes are the most interesting. With that in mind, and with the assistance of Joseph Bradshaw who has converted the results to a spreadsheet for further analysis (TBD by another volunteer if you want to help out), I'd like to present 'The most interesting roguelikes of the year 2013' as a unicorn chaser.

Wayward99
Sil98
Dungeonmans92
Infra Arcana84
NEOScavenger71
Prospector67
Steam Marines50
One Way Heroics46
Pixel Piracy45
Red Rogue39
WazHack39
868-HACK38
Bionic Dues38
Diablo, the Roguelike37
Dungeon of the Endless35
Legend of Dungeon35
Triangle Wizard31
Eldritch27
Delver26
Hoplite26
Hack, Slash, Loot25
UnBrogue25
NPPAngband24
8685652723
KeeperRL23
Forays into Norrendrin20
HyperRogue20
Aurora18
Nuclear Throne18
PRIME18
307917
Paranautical Activity17
The Veins of the Earth16
Hydra Slayer15
308915
Drakefire Chasm14
Tower of Guns14
Kerkerkruip13
Ending12
Voyage to Farland12
Dungeon Dashers11
DynaHack11
Mosaic11
PonyRL11
Rogue's Tale10
8.68-HACK10
Occult Chronicles10

Wednesday 1 January 2014

Winner of the Ascii Dreams Roguelike of the Year 2013: Noxico

You probably haven't heard of Noxico. It is one of the sub genre of adult roguelikes, that I only was made aware of when I ran this poll last year. The link I'm not going to include in this post has an adults only warning on the front page. The site beyond it is NSFW but in reality is only barely titillating. I'm bemused to see that the first bullet point feature the game has is 'Reasonable Unicode Support'.

Nonetheless, it garnered more votes than any other for the roguelike of the year, and I wish to congratulate the developer for his or her persistent fan base.