⌖ Open the aim trainer
Online Aim Trainer / Guides / Research
Aim research

Never queue cold: what 7,500 warm-ups say about your first fight

Your first minutes of aiming are your worst. Here is how big the gap is, why it happens, and how little it takes to close it.

By Black Prince, founder of Online Aim Trainer · October 11, 2026 · 7 min read

Everyone has played the first round of the night and whiffed a shot they would hit nine times out of ten an hour later. Most players shrug it off as bad luck. It is not luck. It is predictable, it is measurable, and it costs the most at exactly the wrong moment: the opening duel of a ranked game.

We measured it in our own players, then checked what independent research says. They agree.

What our players' data shows

Online Aim Trainer records a skill score for every run. To isolate warm-up from everything else, we looked only at the same player playing the same drill twice in the same session, so player, drill and sensitivity are all held fixed. That gave us 7,584 pairs from 1,899 players, played a median of 8 minutes apart.

+9%higher score on the second attempt at the same drill (95% CI 8.3–9.4%)
+3.0points of skill score on a 0–100 scale (95% CI 2.7–3.3)
54%of players improved on their second try. 37% got worse.

The difficulty level did not change between the two attempts (it moved by 0.02 levels on average), so this is not the game going easier on anyone. Same player, same drill, same settings, a few minutes apart: the second run is simply better.

How much each attempt at the same drill improves on the one before
Difficulty-weighted score, change from the previous attempt in the same session
1st → 2nd attempt
+8.4%
2nd → 3rd attempt
+3.2%

The first repeat does most of the work, then gains shrink. Skill score (accuracy against the target bar) tells the same story: 71.9 on the first attempt, 74.9 on the second, 74.5 on the third.

Show the numbers
GroupValueSample
1st → 2nd attempt+8.4%3,082 sessions, 955 players
2nd → 3rd attempt+3.2%same sessions
Online Aim Trainer production data, 25 Jul – 10 Oct 2026. Sessions where a player ran the same drill at least three times.

Some drills need it more than others

The warm-up effect was largest on the drills that demand continuous control: tracking, multi-target and moving-target precision.

Score gain from the 1st to the 2nd attempt, by drill
Difficulty-weighted score, 2nd attempt vs 1st
Track
+13.0%
Multishot
+12.8%
Motionshot Precision
+12.3%
Flick
+10.1%
Micro Adjust
+5.7%
Motionshot Speed
+4.5%
Headshot
+4.0%

Tracking is the most cold-sensitive skill we measured, which is why our routine starts with it.

Show the numbers
GroupValueSample
Track+13.0%1,274 pairs
Multishot+12.8%477 pairs
Motionshot Precision+12.3%435 pairs
Flick+10.1%2,718 pairs
Micro Adjust+5.7%1,481 pairs
Motionshot Speed+4.5%462 pairs
Headshot+4.0%737 pairs
Online Aim Trainer production data. Reflexshot excluded (its skill score floors at zero for many new players).

Newer players lose the most to a cold start

The cold penalty shrinks with experience, but it never disappears. Players in their first 10 runs gained 12.7% on the second attempt; players with 11 to 50 runs gained 6.7%; players with more than 50 runs still gained 3.1%. Part of the beginners' jump is plain learning, but even experienced players start every session below their level.

And the gain does not carry over

Here is the clearest sign this is warm-up, not skill. We compared the last attempt at a drill in one session with the first attempt at it in the player's next session. On average, players started the next session 0.6 points below where they finished (95% CI −0.86 to −0.28), whether the gap was two hours or two days. The skill you built mostly stayed; the warm-up did not.

What the research says

Motor-learning researchers have studied this since the 1950s under the name warm-up decrement: the drop in performance on a motor skill after a break, which recovers within a few attempts.1 Modern models describe two processes running at once: slow, lasting learning, and a fast, temporary "warm-up" component that resets between sessions and degrades performance "at the beginning of the next practice session."2 That is exactly the saw-tooth pattern in our data.

It shows up wherever precision matters:

Not every study agrees. A randomized trial of trainees doing real robotic surgery found no significant benefit from warming up,7 and a small 2025 study of 16 CS2 and Valorant players found that physical warm-ups (jogging-style, reaction games) did not change KovaaK's scores.8 The pattern across all of this: warm-ups help most when they practise the exact skill, and they will not make you better than you are. They let you play at your level from the first round.

Warm up with aim, not with "a game"

The best real-world evidence on Valorant is a 2026 analysis of about 175,000 players' ranked histories by the stats site OTPStats.9 Taken together, "any warm-up" before ranked made no difference to win rate. But when players warmed up with Deathmatch or Team Deathmatch, their ranked win rate rose by 0.9 percentage points (95% CI 0.47–1.34) compared with their own cold starts. Unrated and custom games showed nothing. Timing mattered too: a warm-up within 15 minutes of queueing beat one an hour or two earlier.

A one-point win rate gain sounds small, but matchmaking pushes everyone toward 50%, so small shifts are what real effects look like. That study is observational and not peer-reviewed, and it cannot see aim-trainer warm-ups at all. But its direction matches everything above: warm up the specific skill, close to when you play.

Why five minutes

No study has compared a 5-minute aim warm-up against other lengths, so here is what the evidence actually supports:

So: four or five one-minute runs, covering different skills, right before you queue. That is five minutes.

The 5-minute warm-up

  1. Set Online Aim Trainer to your exact in-game sensitivity (convert it here if you play something other than Valorant).
  2. Track, 60 seconds. The most cold-sensitive skill, so do it first.
  3. Flick, 60 seconds. Snap to targets and stop dead.
  4. Micro Adjust, 60 seconds. Small corrections onto the head.
  5. Multishot or Headshot, 60 seconds.
  6. One more run of whichever drill felt worst. Then queue straight away: the effect fades the longer you wait.

There is a fuller breakdown of each step in our 5-minute Valorant warm-up routine. And if you lose the feeling mid-match, see the 30-second reset.

Run the 5-minute warm-up. Five one-minute drills at your real sensitivity, free in your browser.

Start warming up →

How we measured this

Data. Every completed run on Online Aim Trainer between 25 July and 10 October 2026 on the production site: 49,908 runs by 4,916 mouse players after cleaning. We left out touch-screen runs, and runs where the player's computer could not hold a steady frame rate (achieved frame rate under 80% of their display's refresh rate, or software-only graphics), so that slow hardware does not pass for slow aim.

Sessions and attempts. A session is a run of play with no gap longer than 30 minutes. "Attempts" are the first, second and third time a player ran the same drill in that session. Confidence intervals are 95% bootstrap intervals over pairs.

Metrics. "Skill score" is the trainer's 0–100 performance measure for a run. "Difficulty-weighted score" multiplies the run's score by the difficulty it was played at, so players at different levels are comparable. The trainer adjusts difficulty between runs, which is why we checked that it did not change between paired attempts.

Limits. This is observational: players chose when to play and what to repeat, and people who repeat a drill may differ from those who do not. Part of the gain for brand-new players is ordinary learning, not warm-up. These are averages, and individuals vary widely.

Sources

  1. Adams, J. A. (1961). The second facet of forgetting: A review of warm-up decrement. Psychological Bulletin, 58(4), 257–273. doi.org/10.1037/h0044798
  2. Newell, K. M., Mayer-Kress, G., Hong, S. L., & Liu, Y.-T. (2009). Adaptation and learning: characteristic time scales of performance dynamics. Human Movement Science, 28(6), 655–687. doi.org/10.1016/j.humov.2009.07.001
  3. Wunderlich, F., Heuer, H., Furley, P., & Memmert, D. (2020). A serial-position curve in high-performance darts: The effect of visuomotor calibration on throwing accuracy. Psychological Research, 84, 2057–2064. doi.org/10.1007/s00426-019-01205-2
  4. Smith, D. J. C., Furley, P., Wunderlich, F., Heuer, H., & Memmert, D. (2025). The perils of the first try: experimental evidence for visuomotor calibration in darts and hammering. Psychological Research, 89, 126. pmc.ncbi.nlm.nih.gov/articles/PMC12283878
  5. Listman, J. B., Tsay, J. S., Kim, H. E., Mackey, W. E., & Heeger, D. J. (2021). Long-term motor learning in the "wild" with high volume video game data. Frontiers in Human Neuroscience, 15, 777779. frontiersin.org
  6. Lendvay, T. S., et al. (2013). Virtual reality robotic surgery warm-up improves task performance in a dry laboratory environment: a prospective randomized controlled study. Journal of the American College of Surgeons, 216(6), 1181–1192. doi.org/10.1016/j.jamcollsurg.2013.02.012
  7. Chen, C. C. G., et al. (2023). Effect of pre-operative warm-up on trainee intraoperative performance during robot-assisted hysterectomy: a randomized controlled trial. International Urogynecology Journal, 34(11), 2751–2758. doi.org/10.1007/s00192-023-05595-1
  8. Kulmakorpi, T. (2025). Effects of different physical warm-ups on the aiming performance of Counter Strike 2 & Valorant players (Master's thesis). University of Jyväskylä. urn.fi/URN:NBN:fi:jyu-202506245672
  9. OTPStats (2026). Does warming up before ranked actually help? VALORANT data study. otpstats.com/studies/warmup
  10. Tholl, C., Hansen, T., & Froböse, I. (2025). BMC Sports Science, Medicine and Rehabilitation, 17, 261. doi.org/10.1186/s13102-025-01305-0