To Play or Observe?:
The best battle reports I have ever written were done as a third-party observer. As observer, you are able to devote your full attentions to the other players, and do not need to devote any concentration to strategy or playing. However, unless you are a hardcore writer, or you have some personal stake in the game, it is quite possible that you will become disinterested over time, and might miss important facts. Writing a battle report of a game you played provides for the possibility of more intense story, but it only tells half the story, unless the opponents have shared vision on throughout the game. Writing up a game you play in also results in a break in concentration as you try to divide your efforts between keeping track of your base and your screenshots and your tabs on other players. There are pros and cons to both approaches, but necause I prefer to write a battle report from the third-party perspective, the rest of this essay will deal with that.
When to Write:
I find that I do my best observing after I have played a couple short games. By playing myself, I am warmed up and in the "Starcraft groove". Too few games beforehand, and its hard to really get into the game I'm observing. Too many, and I have trouble staying focused since it's much more likely that I've overdosed on too much Starcraft beforehand. As for writing the report, do it immediately after the game if you can. The story will be fresh in your mind and easier to put together. Sit down before hand and sketch an outline of your story; its high points, and its direction. If you just start writing, your report will lack a sense of continuity and probably result in a rambling discourse that begins to drag long before you reach the end. Hit the high points and conflicts, as well as any out-of-the-ordinary moves, but focus on the exciting, not the mundane. For example, supposed a player puts two zealots on patrol around his base (in mid to late game). In the early game, almost everything is important, but in the mid and late game, this isn't particularly noteworthy. I would only mention it if it caught an opponent coming in, or he did something sneaky, like start the zealots at opposite ends so the fog was lifted twice as often.
The More the Merrier...But Too Many Cooks...:
By personal preference, I have limited myself to only writing up 1v1s, 3FFAs, or 2v2s. This saves some sanity and provides for a better, more coherent report. If more than 4 people are in the game (or even four people in a FFA) it becomes extremely difficult to keep track of the players and their strategies. The more people you have in your game, the less interesting detail you will be able to add to your report. A 6FFA on Meatgrinder could provide for some of the most intense battles ever played on battle.net, but a lone observer cannot possibly do justice to more than a few of the players. For any games other than the three types I mentioned in the beginning, I would highly recommend having two or more observers, each watching pre-assigned players, and then combining notes at the end. As a beginning reporter, I would recommend writing only 1v1s until you feel confident enough to have more players.
Always include the map, races, colours, teams, and start locations. If possible, include the players' race preferences and provide a link to the map for your readers. All Blizzard maps can be found in the Starcraft Compendium. Write as much of this material up in the chat room as you can, because the first few minutes of the game are going to require your full attention.
The Observer Mentality:
Much has been posted on forums concerning the role of the observer in a game. Essentially, I try to be completely invisible until the end of the game. If the players in the game forget that I am there until I say GG at the end, then I've done my job. The observer must be as unobtrusive as possible; if he or she was removed from the game, it should finish in exactly the same manner. Every comment you make can affect the balance of the game or the players' concentrations so it is best to say nothing at all. Try not to request a game pause, or ask players to "hold up for a second" until you can get ready to take notes, or ask a player what he is building towards. A good observer should be able to work around the players, not through them. Take a look at the map before the game. Mentally note the start locations and be prepared to write them down quickly when the game commences. If you are on a big map, you have time to spare, but if you are near a player, quickly lift off your CC and clone your SCVs to kill each other. Fly your CC away from the players to a corner that will not likely see any action and do the same with the last SCV. The players in the game should never see your CC or SCV in transit; this breaks their concentration and can lead to false reconnaissance.
I always take notes. Period. My notes are handwritten on unlined paper. Before the game, I will make columns for each player and list their race preferences and any other information. In the early game, I write down exact build orders the best I can, and then I take note of any actions that are significant or out of the ordinary.
In the very beginning, you have a few seconds grace before you need to take notes. Take care of your CC and hotkey locations as soon as the players share vision with you. Position your screen over the area where their main base will likely be and then hit Shift+F2. Now, hitting F2 will always jump back to this location. You only have 3 location keys (F2, F3, F4) so decide what to mark wisely. In 1v1s, I tend to mark the two mains, and leave the last one to be dynamically allocated to wherever most of the action is. In games with more players, I forgoe location keys entirely and rely on minimap clicking to jump to bases. If you hotkey 3 of 4 bases, the last base will not get as much coverage, simply because it's off the beaten path. Make every base equally hard to jump to and you will report on them all equally. After this is taken care of, I write down early builds and scouting techniques. This information is very interesting to some readers, who like to know what types of builds win games and what types good players use. If it is a standard build, you can always label it as such, i.e. "10/9 Hatch" instead of "Extractor, Drone, Cancel, Hatch". Take a screenshot of every base at regular intervals (but not too often). This will help immensely when you try to recreate the game later.
As soon as conflicts begin, I put down my pencil and take a massive numbr of screenshots. Count the units. See which ones are being micromanaged. Hotkey the battle, and jump back to the bases. Often, the heat of the battle is when the players will do something sneaky back home. Battles make reports exciting (hence the term "Battle" Report) but often times I read the conclusion of the battle only to find that one player suddenly has a spire and 12 mutas flying around the map. Flip back to the battle long enough to take screenshots, but keep an eye on everything else. You are not there to watch the battle; you are there to tell the whole story to everyone else.
If you haven't played a lot of Starcraft, this will be somewhat difficult. As the opponents build or expand, try to determine what they are going for. As you get better, you will be able to more accurately determine strategies. This provides coherence to a story, since you can see an end result from the early battle plans. Attempting to predict what a good player will do allows you to take notes in a down time instead of while the player is acting, and also helps you to become a better player, since you can learn much more from a single recon trip.
Mid and Late Game:
As the battle progresses, I take notes mostly in a macro sense. I record buildings that are produced only if they are important to the tech tree, and focus more on the troops involved. It is important that you not become bored at this point; try to maintain the same focus and intensity or else readers will notice the change in your report. If you don't find the report interesting when you write it, your readers won't find it interesting to read. I made up the distinction between early, mid, and late game, simply because many games seem to fit this mold. I define early game to include everything up to the last initial attack, when both sides switch their focus from head-to-head fighting to outmaneuvering. This switch is hard to observe for new observers, but experience will give you a handle on it. Mid game is the bulk of the game and usually includes power moves and sneaky strategies. I define late game as the segment where it looks obvious who the winner is, there are no resources left on the map, or one or more players is building a "grand victory" army or fleet. Many games never get this far, so I also classify late game to include any attack which results in a game ending.
Screenshots and Editing:
A picture can make a battle report interesting, but it can also destroy one. Some of the best battle reports I've ever read were written on the B.net Forum where pictures cannot be posted. A good battle report is good with or without pictures. If the report relys on pictures too much, then you should work on your writing style and note taking. The ultimate goal is to have a report which is an excellent read without pictures, and then use pictures to enhance the story. Because I use screenshots as time markers in a game as well, I tend to have nearly a hundred shots per game. It would be easy to put them all in and call it a battle report, but it would be a poor report. Instead, I select a few that are interesting or pivotal to include. If the reader cannot enjoy my report without pictures, then I haven't done a good enough job writing it. It is very easy to crop pictures and make them pretty, but they will never make up for poor writing style.
Writing style really cannot be taught; it is acquired through much experience or intuition. You need to know when to expound and when to cut out whole paragraphs. Keep in mind while writing, that you are trying to convey the full feeling of the battle to the reader who wasn't there. This requires the use of suspense, facts, details, and maybe a little bit of humour...all dependent on your style and pacing. The best way to work on your style is to read some excellent reports and then trying to emulate them. Gradually, your own style will develop as a result.
Who Do You Like?:
The purpose of a battle report is not to please the players; it's to please your audience. A battle report should not be an ego trip for the player (unless of course the writer is a player). The job of the reporter is to tell the story of a battle and allow the reader to draw his or her own conclusions and possibly learn new tricks. Never slander or blame any player or write in a condescending manner about any person in the game. Try to write as neutrally as possible; you should not be rooting for a side, even mentally because it will show up in your report. Of course, you can say that a player had a nice strategy or played well, but never allow your report to become skewed in the favour of or the detriment of someone in the game.
Report of the Day:
One excellent report on Saturday is better than 7 mediocre ones all week. Quality is always key over quantity. Do observe many games and take many notes, but don't write the battle report just because you feel obligated to. Be selective about the ones that make it to the board and they will be of much higher quality. The most I ever wrote was for the 3-game series with DisgruntledApe and TerranFlavor. I made sure to take a break afterwards for a couple days. If you write too often, your style becomes cramped and repetitive and cramped and repetitive. If you write up every single game you observe, your style becomes cramped and repetitive. Sometimes after a game, I will look at my notes and realize that it isn't the best game to have a report written about. If this is the case, I will just throw away the notes and try again another day. If you write a report because you took the notes, it will be mediocre. If you write a report because it was an interesting game, you will enjoy writing it and your readers will enjoy reading it.
Don't worry about formatting beyond paragraphs and spacing. The more you focus on formatting, the less you focus on storytelling. I've found a fairly simple table layout to write my reports and stick to it because it's simply and effective.
How Long Does It Take?:
The average battle report takes me about 1.5 hours to prepare and write. I spend a half hour sorting pictures and deciding what to write about, and then an hour actually writing the report. This may seem long, but a report that you spend more time on will not seem rushed, and will probably make for a better read. You have to want to write the report. A half-effort report not only wastes your time, it also wastes your readers time.
The best way to become a better report writer is through feedback from the players and readers. Be open to criticisms and comments. Don't be afraid if no one comments within the first day, allow readers to comment at their leisure. It's a good feeling to work on a battle report and find that your readers enjoy it; but you will always have a few that aren't quite as good as the rest. Don't worry, just keep writing!