What scrims are and why they are hard to read
A scrim is a practice match between teams, played outside official competition. Teams use scrims to test strategies, rehearse set plays, try new players and get used to particular opponents. Because nothing official is at stake, the results are less meaningful than they look. Teams may hide their best ideas, experiment with odd picks, play on one map for hours, or relax after a comfortable lead.
That makes scrim analysis valuable but fragile. The aim is to learn about your own team’s habits and plans, and not to forecast how you will do in a real tournament.
Decide what you are testing
Start each block with a stated purpose. Examples: “try the new opening on this map”, “practise our response to the early aggressive style”, “see how the fifth player fits in.” Without a purpose, the notes become a pile of results that answer nothing.
Scrims also often have different goals on each side. The opponent may be working on something unrelated to what you are testing. Keep that in mind before reading too much into a loss.
What to record
A simple, consistent record is better than an elaborate one nobody fills in. A basic row for each game might include:
- date, opponent and map
- the draft or pick and ban choices, and who started on which side
- the strategy or plan you meant to test
- result
- two or three notes on why it went the way it did
Beyond that, depending on the game, you might log round or objective timings, economy decisions, key fights and which planned calls succeeded. Link entries to saved replays so you can look back. Assign one person to keep the log, since shared responsibility often means no one does it.
Tracking is easier if the structure stays small, in the same spirit as stat tracking for individuals.
Comparing without over reading
Scrim data tends to be thin. A team might play a handful of games with a given setup, and the results are affected by who the opponent was, what they were practising and how motivated each side was. A few cautions:
Small differences are noise. Winning 6 of 10 scrims with one strategy and 5 of 10 with another tells you almost nothing. Ten games per setup is a small sample, and the gap is well within normal variation. Notice also that differences of one or two games are easy to produce by chance alone.
Opponent quality varies. A win against a weak team and a win against a strong one are not equal. If you compare strategies, check that they were tried against similar opposition.
Hidden information. Opponents may not show their real preparation, so a successful scrim strategy may not work once rivals use their full plans. Your own team may hold back as well.
Fatigue and conditions. Late-night blocks, back-to-back games and roster stand-ins change results. Note them in the log.
Selection bias. If you only analyse games you lost, you may think everything is broken. If you only discuss games you won, you miss real problems. Review a mix.
Looking for patterns, not scores
Rather than focusing on win totals, look for behaviours that repeat across games:
- Do the same mistakes appear in the first few minutes of each loss?
- Does a particular call or rotation usually work?
- Where does communication break down? This ties into team communication, which often explains more than individual mechanics.
- Which map or composition gives the team trouble? The wider picture of map pools and results applies at the scrim level too.
- How does the team respond after falling behind?
Behaviours, such as how often a planned trap was executed, can be counted over fewer games than wins, so they are a more reliable guide than results. When you want to know whether a plan is better, you still need a reasonable sample size, and the honest answer is often that you cannot tell yet.
Using scrims to prepare for real opponents
Scrims against a team you will face in tournament are a particular case. You can learn about their style and tendencies, but remember they can also learn about yours. Record the draft choices they make, such as in the pick and ban phase, and what they appear to rely on, while treating it as partial. Official head-to-head records give additional evidence but carry their own caveats.
For opponent habits that you notice, note how often they showed up. A single occurrence is an anecdote, while the same pattern across several games is worth planning around.
Running the review
A review routine might look like this:
- Spend a few minutes after each block with a quick debrief while memories are fresh.
- Update the log within a day, including the stated purpose and result.
- Once a week, read through the log for repeated patterns and pick one or two items for practice.
- Keep a running list of open questions that more games could answer.
Keep sessions short and constructive. Long, blame-heavy reviews tend to produce defensiveness, not changes.
A realistic stance
Scrim results are a rough guide to how a team is developing, and no more than that. Use them to test ideas and find habits, hold conclusions loosely, and wait for repeated evidence before changing a core plan. Teams that treat a scrim record like a tournament forecast are likely to be surprised.