How to Get Better at Competitive Games: A Drill-Based Roadmap

The short answer

You get better at competitive games through short, structured practice with a specific target, not through hours of unfocused ranked. The pattern coaches and aim-training research converge on is simple: 15–30 minutes of isolated mechanical drills daily, followed immediately by in-game application, then one focused review of one loss. Distributed short sessions beat marathon grinding for motor learning, and four to six focused hours will move you further than twelve unfocused ones.

Everything below is that idea, broken into things you can actually do tonight: lock your settings, run a 30-minute drill block, review one match with three questions, and stop tilting. The numbers are the ones the competitive community and coaching literature actually use — not vibes.

Step 1 — Lock your settings, then stop touching them

Settings are a build, not a mood. The single most damaging habit I see in plateaued players is changing sensitivity every week, which resets muscle memory before it can form. Coaches are unanimous on this: pick a value, then lock it for at least 30 days.

Sensitivity, with real numbers

For tactical shooters — Valorant, CS2 played tactically — the professional cluster is 200–400 eDPI, which works out to roughly 30–45 cm of mousepad per 360° turn. Some CS2 pros run higher, up to 600–1200 eDPI, for a more dynamic peeking style, but the majority still sit low. Wider surveys of pro FPS players put the range at 200–800 eDPI, or about 25–50 cm/360.

eDPI is just mouse DPI × in-game sensitivity. If you’re on 800 DPI, a 0.4 in-game sens gives you 320 eDPI — squarely in the pro range. Start there, adjust by ±10% if it feels genuinely wrong, then commit.

Input latency: what’s worth doing

• 144 Hz is the competitive minimum in 2026. The jump from 60 Hz is the single biggest hardware upgrade a competitive FPS player can make. Beyond that, take as high a refresh rate as you can drive with stable frame times.
• Enable Nvidia Reflex (or your game’s low-latency mode) where it exists. It shortens the render queue and cuts end-to-end latency measurably.
• Cap FPS just below your refresh rate — 141 on a 144 Hz panel. This keeps frame times consistent and avoids the latency spike from running uncapped against a sync limit.
• 1000 Hz polling is the accepted baseline. Newer mice offer 2000 Hz and above; stability matters more than the number.
• Turn off: motion blur, film grain, heavy post-processing, and “Enhance pointer precision” in Windows mouse settings. Set Windows pointer speed to 6/11.

On a TV rather than a monitor, the equivalent checklist is Game Mode on the specific HDMI port, motion smoothing off, and VRR enabled on both TV and console — our 2026 gaming TV guide covers the input-lag numbers worth caring about.

Controller baseline

Start with a 0.08–0.12 stick deadzone, a linear response curve, and ADS sensitivity around 0.8× your hip-fire value. On PS5 and Xbox, enable 120 Hz output in system settings and the game’s performance mode — people routinely enable one and not the other.

Audio

Drop music to 0–20%, raise effects, and switch voice chat to push-to-talk. On PS5 specifically, enable and personalise Tempest 3D Audio; positional accuracy improves more from that two-minute test than from most headset upgrades. Our PS5 headset guide covers why connection type matters more than driver size, and why Bluetooth is a latency trap.

Step 2 — The 30-minute drill block

I wasted months “warming up” with five chaotic deathmatches. Swapping to a structured block moved the needle inside two weeks — which matches what aim-training research reports: measurable gains from 20–30 minutes of focused daily practice within one to two weeks.

Three rules make it work:

• One skill per set. Speed or accuracy or timing. Not all three.
• External focus. Think about the crosshair and the target, never about your hand or wrist. This is a well-established motor-learning finding and it changes results immediately.
• Bridge to the game. Aim trainers build mouse control. They do not build crosshair placement, peeking, recoil or decision-making. Always finish with in-game application or the training doesn’t transfer.

FPS, mouse and keyboard (30 min)

• 0–5 min: warm-up on large, slow targets. Don’t start cold at max difficulty.
• 5–13 min: tracking — follow a smoothly moving target, prioritise stability over speed.
• 13–21 min: flicking — snap to random targets, clean single movements, no correction wobble.
• 21–25 min: target switching — multiple targets in sequence, fast but accurate.
• 25–30 min: in-game. Movement mechanics (counter-strafing, jiggle peeks, wide swings) then a headshot-only deathmatch.

FPS, controller (30 min)

• 10 min micro-adjustment: strafe a dummy and keep the reticle locked on its head.
• 10 min flick-to-centre drills.
• 10 min bot deathmatch with strict crosshair discipline — head height, pre-aimed at the next angle.

Soulslike / action

Pick one enemy and run 3×10 reps of perfect dodge → punish, or guard-counter. Reset between reps so the timing stays crisp rather than drifting into muscle-memory mush.

Rocket League

• 10 min free play, car control only, no ball.
• 10 min custom training: ground shots plus one wall pack.
• 10 min 1v1 focused solely on boost management.

If a drill tilts you: lower the difficulty, finish the set, move on. Consistency beats heroics, and quitting a set mid-way teaches your brain that frustration ends practice.

Step 3 — VOD review that takes fifteen minutes, not your night

The breakthrough for me was capping review at one match and three questions. Long review sessions feel productive and change nothing.

Right after your session, pull up one loss and answer:

• Where did the first mistake of the losing sequence happen? Not the death — the decision two steps before it.
• What information did I have and ignore? A minimap ping, a sound cue, a teammate’s call.
• What single rule would have prevented it?

Write the rule as one line and put it somewhere you’ll see it next session. “If we lose site control, stop re-peeking — set up a crossfire.” That’s it. One rule per review, carried into the next session as your only decision goal.

For PvE, the equivalent is noting the tell before each death — the audio cue, the wind-up animation, the phase trigger — then practising that specific window for five minutes somewhere safe.

Step 4 — Communication that actually lands

My comms used to be noise. A template fixed it: info + number + plan.

• “Two A Main, I’m low — play retake.”
• “Storm closing in 30, rotate north.”

Bind push-to-talk somewhere natural (Mouse 5 or Caps Lock). On console, learn the ping wheel properly — fewer, clearer pings beat spam every time. And after a mistake, call the fix rather than apologising into the void: “My bad, I’ll anchor next.” It resets team tilt faster than silence.

Step 5 — Tilt-proofing

Recovery is part of the training, not a break from it. Motor consolidation happens during sleep, and every coaching framework worth reading treats rest and hydration as part of the plan rather than an afterthought.

• Two-loss rule: after two consecutive ranked losses, stand up for five minutes. Breathe slowly, hydrate. Resume only when your hands feel calm.
• Two goals per session, maximum: one mechanical (“keep crosshair head-height”), one decision-based (“trade, don’t duel”).
• Cool down: five minutes of low-stakes play so you don’t end the night tilted. Jot one thing that worked and one to fix.

Step 6 — A weekly plan that fits a real life

You don’t need to grind thirty hours. The coaching consensus is that four to six focused hours beats twelve unfocused ones, and five deliberate hours a week is enough to climb.

• Mon / Wed: 30-minute drill block + two matches (60–90 min total).
• Fri: 30-minute drill block + one VOD review.
• Sat or Sun: longer session — warm-up, ranked, team play, cool down.
• Track exactly two metrics for two weeks. Headshot percentage and deaths per round, say, or boss attempts and healing items used. If both improve, keep the routine. If not, change one variable.

Two metrics is deliberate. Track five and you’ll learn nothing from any of them.

Step 7 — Patch-day adaptation

I used to ignore patch notes and wonder why my timing felt off for a week. Now it’s a 30-minute routine:

• Read the changes and write down one experiment to run.
• 15 minutes in a custom game or range testing the specific change — recoil, ability timing, boss pattern.
• Two unranked games focused only on that change.
• Decide: adopt, adapt, or skip. Don’t force a nerfed playstyle when another one suits you better.

Troubleshooting

• Aim feels floaty: Windows → Bluetooth & devices → Mouse → Additional mouse settings → Pointer Options → uncheck “Enhance pointer precision.”
• Over-aiming and overshooting: reduce sensitivity 5–10% and commit for two weeks. Predictable beats snappy.
• Random frame drops: cap FPS below max refresh, close overlays, and check whether a background updater is stealing cycles.
• Input lag on a TV: Game Mode on that specific port, motion smoothing off, console set to 120 Hz. If HDR adds delay, disable it for competitive play.
• Wrist or shoulder pain: arm at roughly 90°, monitor at eye level, clear mouse surface. Lowering my desk by 2 cm eliminated a wrist ache I’d assumed was permanent.

FAQ

Do aim trainers actually transfer to real matches?

Partially, and that’s the honest answer. Aim Lab and KovaaK’s reliably improve raw mouse control — flicking, tracking, target switching — and you can expect noticeable gains within one to two weeks of daily 20–30 minute sessions. What they don’t teach is crosshair placement, peeking discipline, recoil control or decision-making. Treat them as mechanical isolation, always followed by in-game application. Aim training alone will make you a better mouse operator and not necessarily a better player.

What eDPI should I use for Valorant?

Start around 320–400 eDPI, which is inside the 200–400 range most Valorant pros use and gives you roughly 30–45 cm per 360° turn. Adjust by ±10% if it feels genuinely off, then lock it for 30 days. Valorant rewards precise crosshair placement over fast turning, which is why the pro range skews low.

How long should I warm up?

15 minutes minimum, 30 as an upper bound. Split it into tracking, flicking and target switching, then finish in-game. Longer warm-ups mostly eat the session you were warming up for.

How many hours a day should I play to improve?

Four to six focused hours is what coaching frameworks recommend for players training seriously, and it consistently outperforms twelve unfocused ones. If you have five hours a week rather than a day, structure matters more, not less — the drill block and single-loss review are exactly what make limited time count.

Should I upgrade my hardware to climb?

Only two upgrades reliably pay off: getting to 144 Hz if you’re on 60, and a wired connection if you’re on unstable Wi-Fi. Beyond that, a better mouse or headset is comfort, not rank. If you are building or replacing a machine, our 2026 PC build guide covers what’s actually worth the money this year.

What to expect in 2–4 weeks

• Week 1: settings stop feeling variable. Fewer wild whiffs, clearer audio, fewer “I never saw him” deaths.
• Week 2: aim curves smooth out. Comms get shorter and more useful. Dodge and ability timing tightens.
• Weeks 3–4: decision-making sharpens. You trade rather than chase. Ranked feels less random because your floor is higher, not because your ceiling moved.

You don’t need superhuman reflexes — you need a system you trust and the discipline to leave your sensitivity alone. Lock the settings, drill with intent for half an hour, review one loss with three questions, adapt on patch day, and protect your mental. That’s the whole thing.