Too Many Minerals #8: Protoss Placement 2/5

[ ?posts_id=4593994&dest=81606]

In this episode, MikeXIV examines a placement match between Ender (protoss) and Dreaper (zerg). Thanks to Ender for supplying replays of all his 1v1 placement games!

Trading Bases

The enemy army is pushing up your ramp, and your army is out of position, halfway across the map. Time to make a decision: run your army home to defend your base, or attack his base and wipe him out before he can do the same to you. Making the right choice often means the difference between winning and losing the game on the spot, so it is important to get it right.

Good reasons to start a base trade

  • If you can’t possibly kill his attacking units
  • If both armies are about the same size, and your workers are already dying en masse
  • If you think you can stop or greatly slow down the enemy army just by building a few more units

Of all the tactics in Starcraft II, trading bases is among the most drastic. If you let your opponent run wild in your base and don’t win the game shortly thereafter, he probably will. Here are a few tricks to make sure that you come out ahead in the base-killing race, divided into two handy lists.

How to kill a base

  • Kill his workers first to 1) stop his income and 2) stop him from running off to build another base
  • Against Protoss, if you can shut down buildings easially by killing just a few pylons, do that
  • Kill unit production buildings next, to stop him from building reinforcements. This includes hatcheries for the Zerg.
  • Leave geysers, supply depots, Zerg tech structures,  and other buildings that don’t do anything for last.

How to slow down someone killing your base

  • Keep building units to slow down the attacking units
  • If you can, move workers near the front back to another base
  • If that is impossible, send a few workers out to start new bases
  • If even that is impossible, have your workers attack the incoming army rather than die without a struggle
  • Cancel half-finished upgrades – you might need the resources to start a new base
  • If you’re going to lift off Terran buildings to hide at the edge of the map, do so before they go into the red

A base race is one of the more dramatic an exciting ways for a game of Starcraft II to end. Win or lose, you’ll have a better idea if you can start a base trade in similar situations in the future.


Too Many Minerals #7: Protoss Placement 1/5

[ ?posts_id=4569022&dest=81606]

In this episode, MikeXIV examines a placement match between Ender (protoss) and JimTreeLivin (terran). Thanks to Ender for supplying replays of all his 1v1 placement games!

Back in the Saddle Again

I know I haven’t put anything up since approximately thanksgiving. Sorry about that. There is good stuff coming.

  • I got a new monitor, so from now on all my videos will be in WIDE SCREEN format. Which is awesome.
  • Too Many Minerals reader Ender sent me all 5 of his 1v1 placement matches. I recorded commentary of the first one last night, and will be uploading it ASAP. The rest will follow.
  • I’m working on an article about base trades, look for that before the end of this week.


A Personal Milestone

If you’ve been checking Too Many Minerals in the last week and a half, I owe you an apology for not posting anything. Fact is, I’ve been busy playing a whole lot of Starcraft II instead of making new content for this site, and it has payed off:

I probably shouldn’t put up this post just to brag about making it into silver, so I’m going to try to offer some advice regarding promotion. It isn’t a big deal. The matchmaking system doesn’t restrict you to playing within your own leage – in fact I’ve been playing against a mix of Bronze, Silver, and Gold league opponents for the last month or so. (tip: to check your ladder opponents rating after a game, left click on their name in the score screen and click “view profile”)
Good matchmaking ensures that players lose half the time. This can be very frustrating, and I find that the best way to blast past that frustration is to have a goal with a broader scope than a single game. Promotion is completely out of your hands, and as such it is a pretty terrible goal. A month ago I wanted to be promoted to Silver, but I had no idea how close I was. Hoping to get promoted left me discouraged. In the last few weeks, I’ve been focusing on a tangible goal – getting my bonus pool points down. Games have been more rewarding because every time I win I’ve seen exactly how much progress I made towards my goal. You might set a goal of climbing up in your division, or just reaching a certain ranking, or just to play a certain number of games each week. At the end of the day, might reward you with a promotion or an achievement for reaching some number of wins, and if it does, that will be the icing on the cake.

Too Many Minerals #6: Silver Terran vs. Zerg

[ ?posts_id=4475480&dest=81606]

MikeXIV examines a Silver league game on Steppes of War between Bronzite (Terran) and Skelpig (Zerg). Thanks to Bronzite for submitting this replay!

Matchmaker, Matchmaker…

I’m pretty bad at Starcraft II. Sc2Ranks pulls data from, and it tells me that there are about fifty thousand people better than me, and that is just in the Bronze league. There are nearly one million player accounts in North America, and more than two million worldwide. And yet despite this, the matchmaking system always finds me a fair fight.

Why aren’t I being overrun?

In short, because of good math. Every Starcraft player has a score, called his Matchmaking Score. Blizzard keeps the exact formula secret, but it works something like this:

  • A new player plays 5 games (placement) against random opponents who are already in the system.
  • The Matchmaking Score of those 5 opponents and the results of those games determines the new player’s Matchmaking Score
  • The matchmaking system tries to pair up players with similar Matchmaking Scores for games
  • After every game, each player’s Matchmaking Score is adjusted – the winner’s goes up and the looser’s goes down.
  • It’s that simple.

The real magic is in how much each score changes after a game. If the players had the same Matchmaking Score, then both players scores will change by some “average” amount. If a player with a very low Matchmaking Score wins against a player with a very high score, the winning player will get a big boost, and the loosing player will get a very big penalty. If, however, a player with a high score beats a player with a very low score, the winner’s boost will be very small, and the looser’s penalty will also be very small. In this way, the secret Matchmaking Score gets more and more accurate over time.

Don’t leagues and the visible rankings matter?

For the purposes of determining who plays who on the ladder, no.  A player’s league, division, division rank, and league points do not matter to matchmaking. If someone would like to compare his skill level to that of another Starcraft II player, they aren’t a bad place to start, especially since the real Matchmaking Score is kept hidden by Blizzard. In the end, leagues, divisions, and the bonus pool are all just clever mind games that plays with us, but that is a matter for another post.


TeamLiquid forum threads used as sources used for this post are here and here.