Addiction, Diablo 3, and Portal 2

In The Tipping Point, Malcolm Gladwell talks about nicotine addiction amongst smokers. What you might not know is that nicotine's power varies quite a bit depending on the, uh, victim.

Of all the teenagers who experiment with cigarettes, only about a third ever go on to smoke regularly. Nicotine may be highly addictive, but it is only addictive in some people, some of the time. More important, it turns out that even among those who smoke regularly, there are enormous differences in the stickiness of their habit. Smoking experts used to think 90 to 95 percent of all those who smoked were regular smokers. But several years ago, the smoking questions on the federal government's national health survey were made more specific, and researchers discovered, to their astonishment, that a fifth of all smokers don't smoke every day. There are millions of Americans, in other words, who manage to smoke regularly and not be hooked—people for whom smoking is contagious but not sticky.

Gladwell goes on to call these sometimes-smokers "chippers." While chippers never feel the need to go beyond a certain level of their drug, true addicts escalate their drug usage over time. I was surprised when I first read that, as I would have imagined that everyone would be caught on the slippery slope toward needing more and more of a drug, but apparently not.

You might ask what separates chippers from more hardcore addicts. In Gladwell's summary of the situation, he says "probably genetic factors." One piece of support for this is a (kind of scary) study where mice were given toxic levels of nicotine. At some point, it's poisonous enough to cause a seizure. Is that point about the same in all mice? Turns out, it's not. While some mice had seizures at X amount of nicotine, other mice could tolerate two or three times that amount. There seems to be a genetic difference here. Further, that range of "toxic to some, but others can tolerate two or three times as much" is the same range for alcohol.

I've never personally been interested in smoking (or drinking alcohol, for that matter), but I drink coffee. It's something I originally did for practical reasons when I needed a bit of a boost to do some work, even though I didn't want the coffee at all, but now I certainly have some sort of chemical addition to it. That said, it's only at the level of a "chipper." I have basically never had more than one coffee in a day, while I know others who have escalated to four, five, six cups, etc.

Oh, and another thing about those mice. The experimenters wondered whether there was a correlation between how much nicotine a mice could tolerate (a genetic factor) and how much nicotine the mouse would *voluntarily* consume (behavioral factor). It turns out the correlation was almost perfect, and that the more a given mouse could tolerate, the more it voluntarily consumed. So I'm willing to be my own personal choice of having some coffee, but not nearly as much as some other people I know, is just my luck of the draw with genetics.

Wanting vs Liking

If you only read one thing in this post, this should be it: "wanting" and "liking" are governed by different circuits in the brain, based on different chemicals. (The wanting-circuit uses dopamine while the liking-circuit uses opioids.) It's actually super important for you to know that wanting and liking are so different in our brains. Things that you want are not necessarily things you will like. This illuminates how bad of an idea it is to expend huge amounts of effort to attain things that you will ultimately not like. The most obvious one here is money. Mountains of research say that

Click to read more ...

Diablo 3's Ability System

Diablo 3 comes out in a couple weeks. I'm giving it the coveted award for "Biggest Comeback In System Design." Diablo 2's ability system was so bad that it's almost unbelievable, while the way Diablo 3 handles ability customization is one of the very best systems I've seen.

Diablo 2

Diablo 2 had talent trees where you spend points to unlock new abilities, very similar to how talent trees work in World of Warcraft. Also, you could allocate stat points into various different stats however you wanted as you leveled up. At first glance, these seem like ok things, but let's look at just how deeply problematic they really are.

Don't Use Points!

First, the best way to play Diablo 2 is have this big red "+" button on your screen almost the entire time, the one that says you have extra points to spend. The reason that will be on you screen for weeks is that you'd be a sucker to actually spend the points as you get them. You counter-intuitively (and unfunly) should stock up on those and spend them much later on. So the simple and fun thing to do (spend points as you get them) is just a trap for noobs.

Next, the whole system of allocating points in the first place didn't really customize anything. It was just a giant test of if you did your web research enough to know the only reasonable way to spend those stat points. You don't have to take my word for it either, let's see what Jay Wilson has to say. He was an avid Diablo 2 player, and the Game Director of Diablo 3 for the last 5 or 6 years.

Here's a written transcript of the relevant part, in case you don't want to watch the vid:

"You usually take as much strength as you need to get the armor that you're targeting, and that's usually around 120 or 220, depending on what type of armor. You take 75 dexterity because that's the amount you generally  need for good block percentages. You take NO energy at all unless…there's like one type of build you can make on a sorceress that uses energy shield. And then you put everything else in vitality. That's a shitty customization system. That's just not a good system." --Jay Wilson, Diablo 3 Game Director on Diablo 2

Talent Trees

Next, there's the talent trees. There's two problems here, one medium sized and the other is one of the most mind-blowing fumbles in design out there. First the medium problem: it's pretty hard to make talent trees that give any real choice. They sure seem to allow choice,

Click to read more ...

Starcraft's Essence in Card Form?

A Puzzle Strike player named BT mentioned that he thought Puzzle Strike captures what Starcraft is about better / more elegantly than the Starcraft board game. I thought it was an interesting statement, so I'll explain what he meant.

First, I should say that I have not played the Starcraft board game, but a quick look at it shows that it comes with about a thousand pieces, and looks complicated and long to play. My guess is that it tries hard to capture literally what's going on in the computer game, but that is generally a dangerous approach. Computers are good at keeping track of all sorts of numbers and resources that would be tedious (and yeah, "inelegant") in a physical game. Sorry if my impression of the board game is unfair though, I stress again that I have not played it, but BT said this is part of what he meant.

Meanwhile in Puzzle Strike, you have choices that basically amount to "expand," "tech," and "army."

Expanding

In Starcraft, you ideally would like to invest as much as possible in your economy, as a way of being weaker now but very strong later. In Puzzle Strike, this means spending your money to buy more gem chips for your deck. Gem chips basically are money, so buying them will make your economy much stronger later, but at the expense of not building "tech" or "army" now. In both games, you have to keep an eye on how much the opponent is threatening you with his army to know how much you can safely invest in your own economy.

Tech

In Starcraft, investing in tech gives you the potential to do powerful things. For example, building a Templar Archives gives you the ability to build High Templars and researching Psionic Storm gives your Templars access to that powerful spell. In Puzzle Strike, the analog is buying what players call "engine chips." These are chips that all work together to produce powerful combinations. For example, chips that give you more actions and chips that let you spend those actions drawing more chips. Building an engine in your deck is sort of like teching up in Starcraft, as it gives you access to powerful turns, but it's not the "tech" itself that wins--that's what your army is for.

Army

In Starcraft, your army is your set of attack and support units. It's your army--not your economy and not your tech buildings--that actually apply force to the enemy and win the game for you. In Puzzle Strike, your "army" is your set of purple chips, the ones capable of combining gems in your gem pile and crashing them so they leave your pile and go to your enemy's. Filling up his gem pile to 10 is how you win, so these purple chips are what allow you to directly attack the enemy and to defend against his purple assaults.

Putting it all together

It would be nice if you could just sit back and build economy, but if you take too long to build any tech or army, you're going to lose before you get to use all that money. Having just a bit of army early can let you hold off incoming attacks long enough to let your economy kick in. How much tech and when to build it is also a hard question. It's possible to completely overwhelm other players if you build a solid tech engine, but you could very well be overwhelmed by an opponent's army while you're trying to get that together.

Asymmetry, Build Orders, and Maps

In addition to the expand vs. tech. vs army concept, Puzzle Strike also has asymmetric starts (3 races in Starcraft; 10 characters in Puzzle Strike) and it has the concept of build orders and maps. A build order in Starcraft is a combination of moves that results in a certain level of expansion / tech / army and a certain composition and timing of that army. Doing a Zealot / Stalker rush is a very different build than putting up some static defense and going for air units like Void Rays. Likewise, trying to clog up everyone's deck with useless wound chips while yours stays tight and efficient is a very different "build" than a draw engine or a mono-purple rush.

In Starcraft, your choice of builds depends partly on the map you're playing on. While any given map allows for many viable builds, some builds become stronger or weaker--or even possible / not possible--on certain maps. In Puzzle Strike, the "map" is set of bank chips you can buy for your deck in the current game. There are 24 types of these chips total, but each "map" consists of a set of 10 of these, so that there are millions of possible starting conditions. Your build depends a lot on which of the millions of possible maps you're playing on.

Conclusion

Puzzle Strike certainly isn't the same game as Starcraft, and I'm sure you can easily think of differences, but BT's point is that it's striking how many core similarities there are. None of it was even intentional except for the inspiration of using 4-gems to fill a similar role to Protoss Carriers that I mentioned in this article. Anyway, I hope you enjoy the game, it's got a lot of really interesting dynamics.

• Quick game overview
• Read an article about Puzzle Strike's design
• Buying options for Puzzle Strike

Analyzing Starcraft 2's Ranking System

Updated on Sunday, July 25, 2010 at 3:03PM by Sirlin

Updated on Sunday, July 25, 2010 at 4:33PM by Sirlin

Let's talk about Starcraft 2's ranking system, specifically the bonus pool system, the focus on ranked matches, and the division system. Before we get into all that, I'll give some background info from three years ago when I talked with Rob Pardo (VP of Game Design at Blizzard) about me possibly taking on the role of ranking-and-tournament systems designer at Blizzard. I ended up having to back out of that race because the Street Fighter HD Remix project suddenly became a reality, and I couldn't pass that up.

The reason I want to tell you about the ranking stuff from back then is to illustrate two points of view--mine and Pardo's--and to explain how it took me three years to understand that his point of view is probably best after all.

TrueSkill vs. Monkeying With Rankings

He asked me to come up with a ranking system for a game, we'll just call it Game X. My first response was that this is easy because it's already a solved problem: Microsoft solved it with what they call TrueSkill. TrueSkill is a refinement of the well-known ELO system used in Chess. One of TrueSkill's main features is that it can apply to games with more than 2 players, while ELO can't. Also, TrueSkill uses a bell curve rather than a single point when referring to a player's skill level. As the system gets more information about the player, it becomes more "certain" that the skill ranking is accurate, and that player's personal bell curve shrinks.

More important than any of those details though, is Microsoft's philosophy about rankings. The premise of their whole system is that players will have the most fun if the ranking system can give matches as close to 50-50 as possible. Yeah it's fun to have to have a few matches that are easy and some that are hard (and you will because of the inevitable variance), but you really do want the ranking system to try to give you close matches. Consider a matchmaking system that simply gave you random opponents, and how from the perspective of a bad player, he just gets stomped repeatedly then probably quits. He is better served by getting evenly matched with other bad players. Over time, he might become a good player rather than quitting.

Microsoft makes another good point here that ONLY winning and losing can be allowed to affect these stats. You can't adjust the matchmaking stat by "experience points" or even by any skill-based stats such as headshots, number of kills, time to finish a lap in racing, etc. All those stats can be gamed, and you will end up trying to get more headshots or something instead of winning. Any formula that equates number of headshots (or any other stat besides wins/losses) with how likely you are to win or lose introduces a layer of imperfect simulation. If we want to know how likely you are to beat someone, we should only consider your wins and losses, and not any in-game stats.

Short version: Microsoft's philosophy is correct (or is it...?), our ranking system should only consider wins and losses and should maximize close matches using TrueSkill, or a close implementation.

But Pardo was not quite on board. Such a ranking system is harsh and while hardcores are ok with it, regular people just get crushed and discouraged. They don't get to see any real progress, he said.

Click to read more ...

UC Berkeley StarCraft Class, Week 14 (Final Week)

In the last class of the semester for UC Berkeley's StarCraft class, several students gave their presentations.

Zergling Rush

The first student's topic was the Zergling Rush. The point of this strategy is to take advantage of Zerg's window of dominance at the very beginning of the game. Zerglings are weak, but you can get them out so early that your opponents might very well have little or nothing to defend with.

There are two main varieties of the zergling rush: the 4-pool and the 6-pool. 4-pool means that you only have four drones (resource collectors) when you build your spawning pool (the building that lets you produce zerglings). So you start the game by gathering resources for a bit, not building any drones, and going straight to spawning pool and zerglings. This is an "all in" strategy though. If you don't win with it, you are so severely behind economically, that you will probably lose. The reason to choose this strategy over a slightly more conservative 6-pool rush is if you think getting the zerglings to the opponent a few seconds faster will allow you to win the game on the spot.

The 6-pool strategy is still risky (putting you behind economically if you fail), but you can recover from it if things don't go as planned. The build order is to first build two drones (getting you to six total), then mine a bit, then build a spawning pool as soon as you can, then build another drone to replace the one that you just turned into a spawning pool, then build an overlord. If you do this correctly, these things will happen simultaneously: 1) your overlord finishes building, 2) you have all three larva ready to build units, 3) you have 150 minerals, the exact cost of building 6 zerglings from your 3 larva.

The student said the first goal of a zergling rush is to end the game immediately if you can. As soon as you arrive at their base, you probably have a good idea if this is even possible. If it's not, your next goal is to disrupt their economy. Remember that your economy is already disrupted by you doing a zergling rush (rather than building up your own economy) so you'll have to disrupt theirs just to stay even. If you can't disrupt their economy much, the next priority would be to at least force them to change their strategy or force them into a certain strategy. For example, they might have to build sunken colonies or bunkers right away to survive. And failing all that, at the very least you should harass with your zerglings and force the enemy to spend a lot of clicks to deal with you. The student also pointed out that there is some value in getting the opponent angry or flustered here, too. Zergling rushes are considered "cheesy" by some, so that's great to

Click to read more ...