Research
See what it takes to reach the top across 17 PvP training drills. Explore how the wider player base is distributed in every mode.
Published August 2, 202635 min readTopic: TRAINING
Leaderboards are good at showing who is first. They are less helpful when you just want to know whether your own result is typical, strong, or genuinely rare.
We looked at 582,425 qualified training runs completed between January 1 and August 1, 2026. The final article covers 17 drills and 10,584 qualifying player-mode samples. That is not a unique-player count because the same person can qualify in more than one mode.
Personal bests are not used as the main benchmark. Each player is represented by the median of their latest 10 qualified runs in a mode. That gives a steadier picture of current performance and keeps one unusually good or bad run from deciding the result.
You can compare the benchmarks below with the current training leaderboards. A leaderboard rank and a population percentile answer different questions, so both are useful.
There is no single bell curve that describes every drill. Aim Tracking, Small Flicking, several mace drills, and First Hit are fairly balanced around their centers. The time-based drills are much more lopsided because a small group of slow results can stretch far beyond the main cluster.
Speed Anchor is the clearest example. Its median is 566 ms, while its mean is 759 ms. That gap is why the detailed 100 ms view is more useful than a chart that runs all the way to the most extreme stored time.
First Hit has the largest sample, with 2,216 qualifying players and 166,665 qualified runs. Shield Break and Speed Anchor also have large samples, which makes their overall shapes especially stable.
Safe Anchor Destruction is the only limited sample, with 57 qualifying players. The numbers are still useful as a reference, but they should not carry the same weight as the larger modes.
Related guides based on the learning path, explicit relationships, and matching topic metadata.
Every bar shows the share of qualifying players inside a performance range. The value used for each player is the median of their latest 10 qualified runs.
For KPS, accuracy, successful actions, and score, higher is better. For time in milliseconds, lower is better. A top-10% benchmark is therefore a high number in some modes and a low number in others.
The time charts focus on the central player range. Values outside the displayed window are not deleted. They still count in every median, percentile, threshold, and sample total. The note above each trimmed chart says exactly how many players sit outside the visible range.
Visible values are rounded to one decimal place or to whole milliseconds. All calculations use the original unrounded data.
| Mode | Players | Runs | Median | Middle 50% | Top 10% | Shape |
|---|---|---|---|---|---|---|
| Aim Flicking | 422 | 17,973 | 3.3 KPS | 2.9 KPS to 3.6 KPS | 4.1 KPS or higher | mildly right-skewed |
| Aim Tracking | 189 | 9,383 | 56.8% | 46% to 71.3% | 84.1% or higher | approximately symmetric |
| Small Flicking | 296 | 12,797 | 3.4 KPS | 2.8 KPS to 3.8 KPS | 4.3 KPS or higher | approximately symmetric |
| Anchor Destruction | 259 | 9,331 | 8,275 ms | 6,528 ms to 10,726 ms | 4,876 ms or faster | mildly right-skewed |
| Safe Anchor Destruction | 57 | 2,156 | 12,722 ms | 10,302 ms to 16,540 ms | 8,440 ms or faster | mildly right-skewed |
| Safe Speed Anchor | 599 | 32,227 | 876 ms | 675 ms to 1,200 ms | 500 ms or faster | strongly right-skewed |
| Speed Anchor | 1,429 | 115,854 | 566 ms | 328 ms to 874 ms | 224 ms or faster | strongly right-skewed |
| Totem Reaction | 729 | 28,923 | 1,554 ms | 1,217 ms to 1,960 ms | 1,045 ms or faster | strongly right-skewed |
| Totem Speed | 539 | 24,130 | 1,286 ms | 1,024 ms to 1,680 ms | 837 ms or faster | strongly right-skewed |
| Elytra Mace | 459 | 14,098 | 11 |
Player totals are per mode. Safe Anchor Destruction is based on a limited sample.
The aim drills are the closest group here to a familiar centered distribution. Even so, they reward different things. Flicking is mostly about pace, while tracking puts more weight on sustained control.
Aim Flicking has a clear center in the low 3 KPS range. Most established players sit close to that cluster, with a smaller group stretching the right side of the chart.
The sample includes 422 players and 17,973 qualified runs. The median is 3.3 KPS. The middle half of players fall between 2.9 KPS to 3.6 KPS. The top-10% benchmark is 4.1 KPS or higher.
422 qualifying players. Each player contributes the median of their latest 10 qualified runs.
Aim Tracking is one of the most even distributions in the dataset. There is still a wide gap between the lower quartile and the strongest players, but no single tail dominates the chart.
The sample includes 189 players and 9,383 qualified runs. The median is 56.8%. The middle half of players fall between 46% to 71.3%. The top-10% benchmark is 84.1% or higher.
189 qualifying players. Each player contributes the median of their latest 10 qualified runs.
Small Flicking is also fairly balanced. The middle half covers almost one full KPS, so there is meaningful separation even among players with an established sample.
The sample includes 296 players and 12,797 qualified runs. The median is 3.4 KPS. The middle half of players fall between 2.8 KPS to 3.8 KPS. The top-10% benchmark is 4.3 KPS or higher.
296 qualifying players. Each player contributes the median of their latest 10 qualified runs.
Every anchor result is measured in milliseconds, so lower is better. The charts focus on the central player range because a few very slow results can otherwise flatten the part of the distribution that matters to most players.
Anchor Destruction has a noticeable slow-time tail. The 500 ms chart keeps the main player range readable without changing the median, percentiles, or top-decile benchmark.
The sample includes 259 players and 9,331 qualified runs. The median is 8,275 ms. The middle half of players fall between 6,528 ms to 10,726 ms. The top-10% benchmark is 4,876 ms or faster.
Chart range: 3,000 to 18,500 ms in 500 ms steps. The chart shows 255 of 259 players. The remaining 4 players, or 1.5%, are outside the visible range but remain included in every statistic.
259 qualifying players. 500 ms buckets show 255 players in the central range. 4 outside the display remain in all statistics.
Safe Anchor Destruction has the smallest published sample. It is useful as an early reference, but it should not be treated with the same confidence as the larger modes.
The sample includes 57 players and 2,156 qualified runs. The median is 12,722 ms. The middle half of players fall between 10,302 ms to 16,540 ms. The top-10% benchmark is 8,440 ms or faster.
Chart range: 6,000 to 24,500 ms in 500 ms steps. The chart shows 55 of 57 players. The remaining 2 players, or 3.5%, are outside the visible range but remain included in every statistic.
57 qualifying players. 500 ms buckets show 55 players in the central range. 2 outside the display remain in all statistics.
Safe Speed Anchor is strongly right-skewed. Most stable times sit in a compact band, while a small number of slower results extend well beyond it. The 100 ms view makes that central band clear.
The sample includes 599 players and 32,227 qualified runs. The median is 876 ms. The middle half of players fall between 675 ms to 1,200 ms. The top-10% benchmark is 500 ms or faster.
Chart range: 100 to 2,900 ms in 100 ms steps. The chart shows 593 of 599 players. The remaining 6 players, or 1.0%, are outside the visible range but remain included in every statistic.
599 qualifying players. 100 ms buckets show 593 players in the central range. 6 outside the display remain in all statistics.
Speed Anchor has the longest slow-time tail in the dataset. A 50 ms view would need 50 bars, which is more than the article renderer supports, so this chart uses 100 ms buckets across P01 to P99.
The sample includes 1,429 players and 115,854 qualified runs. The median is 566 ms. The middle half of players fall between 328 ms to 874 ms. The top-10% benchmark is 224 ms or faster.
Chart range: 100 to 2,600 ms in 100 ms steps. The chart shows 1,400 of 1,429 players. The remaining 29 players, or 2.0%, are outside the visible range but remain included in every statistic.
1,429 qualifying players. 100 ms buckets show 1,400 players in the central range. 29 outside the display remain in all statistics.
The reaction drills have a dense middle and a longer slow tail. The detailed charts use 100 ms steps, which makes the difference between common and strong times much easier to see.
Totem Reaction has the familiar shape of a reaction-time metric. Most players gather around a central band, while slower performances create a long tail. The 100 ms chart avoids letting a handful of extreme times flatten the useful part of the graph.
The sample includes 729 players and 28,923 qualified runs. The median is 1,554 ms. The middle half of players fall between 1,217 ms to 1,960 ms. The top-10% benchmark is 1,045 ms or faster.
Chart range: 700 to 3,700 ms in 100 ms steps. The chart shows 717 of 729 players. The remaining 12 players, or 1.6%, are outside the visible range but remain included in every statistic.
729 qualifying players. 100 ms buckets show 717 players in the central range. 12 outside the display remain in all statistics.
Totem Speed follows the same broad pattern. The median stays close to the main cluster, while the average is pulled upward by the slower side of the distribution.
The sample includes 539 players and 24,130 qualified runs. The median is 1,286 ms. The middle half of players fall between 1,024 ms to 1,680 ms. The top-10% benchmark is 837 ms or faster.
Chart range: 600 to 3,600 ms in 100 ms steps. The chart shows 529 of 539 players. The remaining 10 players, or 1.9%, are outside the visible range but remain included in every statistic.
539 qualifying players. 100 ms buckets show 529 players in the central range. 10 outside the display remain in all statistics.
The mace drills count successful actions. Most of these distributions are fairly balanced, but the distance from the median to the top 10% varies a lot between drills.
Elytra Mace is fairly balanced, with the middle half running from the high single digits into the low teens. The upper tail is visible without taking over the chart.
The sample includes 459 players and 14,098 qualified runs. The median is 11. The middle half of players fall between 7 to 14. The top-10% benchmark is 17.1 or higher.
459 qualifying players. Each player contributes the median of their latest 10 qualified runs.
Elytra Mace Stun Slam is centered around five successful actions. The distribution has a mild right lean, with the stronger group extending into double figures.
The sample includes 207 players and 7,033 qualified runs. The median is 5. The middle half of players fall between 3 to 7.5. The top-10% benchmark is 10.5 or higher.
207 qualifying players. Each player contributes the median of their latest 10 qualified runs.
Lunge Mace has one of the tighter middle ranges among the mace drills. Half of the qualifying players sit within only a few successful actions of the median.
The sample includes 749 players and 27,975 qualified runs. The median is 9. The middle half of players fall between 8 to 10.5. The top-10% benchmark is 11.6 or higher.
749 qualifying players. Each player contributes the median of their latest 10 qualified runs.
Mace Pearl Catch has a mild right tail. Most players sit in the mid-single digits, while stronger performers extend the distribution into the teens.
The sample includes 242 players and 7,772 qualified runs. The median is 7.5. The middle half of players fall between 6 to 9.5. The top-10% benchmark is 12 or higher.
242 qualifying players. Each player contributes the median of their latest 10 qualified runs.
Mace Stun Slam has a wider spread than several other mace drills. The distance from the median to the top 10% leaves a clearly separated upper group.
The sample includes 428 players and 14,240 qualified runs. The median is 6. The middle half of players fall between 2.5 to 11.1. The top-10% benchmark is 15 or higher.
428 qualifying players. Each player contributes the median of their latest 10 qualified runs.
Spear Mace is fairly balanced overall. Results are concentrated around the middle single digits, with a smaller group reaching into double figures.
The sample includes 253 players and 8,208 qualified runs. The median is 5.5. The middle half of players fall between 2 to 8.5. The top-10% benchmark is 10.5 or higher.
253 qualifying players. Each player contributes the median of their latest 10 qualified runs.
The melee scores come from the largest samples in the article. First Hit is fairly balanced, while Shield Break has many players grouped toward the high end.
First Hit has the largest qualifying sample in the article. Its distribution is broad but balanced enough that the median gives a reliable picture of the center.
The sample includes 2,216 players and 166,665 qualified runs. The median is 57. The middle half of players fall between 50.2 to 62.7. The top-10% benchmark is 70.4 or higher.
2,216 qualifying players. Each player contributes the median of their latest 10 qualified runs.
Shield Break looks very different from the time-based modes. Many players cluster near the high end of the score range, while lower scores form the longer tail on the left.
The sample includes 1,511 players and 83,660 qualified runs. The median is 73.7. The middle half of players fall between 67.3 to 79.8. The top-10% benchmark is 88.9 or higher.
1,511 qualifying players. Each player contributes the median of their latest 10 qualified runs.
For a balanced mode, the mean and median tell a similar story. For a skewed mode, the mean can be pulled toward a long tail. That is why the median is the main benchmark throughout this article.
These thresholds describe the McTrainer.Club population during this reporting period. They will move as more people play, improve, and qualify. They are best used as a practical goal, not as a permanent title.
A mode with several hundred or several thousand qualifying players gives a steadier picture than one with a few dozen. The limited sample is marked clearly for that reason.
Milliseconds, KPS, percentages, successful actions, and score all behave differently. A wide chart does not automatically mean a drill is harder. The useful comparison is where a result sits inside its own mode.
The extraction found no missing profile UUIDs, no duplicate profile UUID groups, no unknown stored mode keys, no unknown metric directions, and no duplicate run IDs in the published data.
The analysis skipped 4,103 runs with invalid timestamps and 48,164 runs without the required primary metric in the published modes. It also found 85 unusually distant values. Those values were kept whenever they were otherwise valid.
Another 14,167 player-mode candidates had at least one qualified run but fewer than the required 10, so they did not contribute to the distributions.
One chart-format check rejected a 50 ms Speed Anchor view because it would require 50 bars, while the article renderer supports at most 40. The published chart uses the valid 100 ms P01-to-P99 view instead. This changes only the display, not the statistics.
Stored histories have per-mode limits. Some highly active profiles reached those limits, which means lifetime run totals can be lower bounds. The main benchmark only uses the latest 10 qualified runs, so older-history limits have much less effect on the player-level median.
The published files contain aggregate mode-level statistics only. They do not include player names, UUIDs, internal database IDs, run IDs, IP addresses, or individual training histories. The extraction used read-only access and made no database changes.
The report covers runs completed from January 1 through August 1, 2026, using UTC completion time. Sandbox runs and voided runs were excluded. A valid run needed a usable completion timestamp and the configured primary metric.
Players needed at least 10 qualified runs in a mode. The latest 10 were selected, then the median primary-metric value became that player's stable-performance result.
Percentiles use the R7 method. Standard deviation is calculated across the qualifying player-level results. Full-range histograms use the Freedman-Diaconis rule with readability limits. The detailed time charts use aligned fixed-width buckets around the central range.
The observed histogram is always the main evidence. The article does not force a normal curve onto modes that are clearly skewed.
The table below keeps the main descriptive statistics in one place. Time values are shown in milliseconds. Other values use the mode's normal metric.
| Mode | N | Mean | Median | SD | P10 | P25 | P75 | P90 | Skewness |
|---|---|---|---|---|---|---|---|---|---|
| Aim Flicking | 422 | 3.3 | 3.3 | 0.6 | 2.6 | 2.9 | 3.6 | 4.1 | 0.6 |
| Aim Tracking | 189 | 57.8 | 56.8 | 17.7 | 35.2 | 46 | 71.3 | 84.1 | 0.1 |
| Small Flicking | 296 | 3.4 | 3.4 | 0.7 | 2.5 | 2.8 | 3.8 | 4.3 | 0.3 |
| Anchor Destruction | 259 | 8,864 | 8,275 | 3,500 | 4,876 | 6,528 | 10,726 | 13,686 | 0.9 |
| Safe Anchor Destruction | 57 | 13,492 | 12,722 | 4,114 | 8,440 | 10,302 | 16,540 | 18,166 | 0.6 |
| Safe Speed Anchor | 599 | 980 | 876 | 483 | 500 | 675 | 1,200 | 1,525 | 2.0 |
| Speed Anchor | 1,429 | 759 | 566 | 1,416 | 224 | 328 | 874 | 1,224 | 12.2 |
| Totem Reaction | 729 | 1,678 | 1,554 |
The useful question is not whether one score is good in the abstract. It is where that score sits inside the distribution for that drill. The median shows the center, the middle 50% shows the common range, and the top-10% line gives strong players a realistic next target.
Check the training leaderboards for current rankings, then use these distributions to understand what those numbers actually mean.
| 7 to 14 |
| 17.1 or higher |
| approximately symmetric |
| Elytra Mace Stun Slam | 207 | 7,033 | 5 | 3 to 7.5 | 10.5 or higher | mildly right-skewed |
| Lunge Mace | 749 | 27,975 | 9 | 8 to 10.5 | 11.6 or higher | approximately symmetric |
| Mace Pearl Catch | 242 | 7,772 | 7.5 | 6 to 9.5 | 12 or higher | mildly right-skewed |
| Mace Stun Slam | 428 | 14,240 | 6 | 2.5 to 11.1 | 15 or higher | mildly right-skewed |
| Spear Mace | 253 | 8,208 | 5.5 | 2 to 8.5 | 10.5 or higher | approximately symmetric |
| First Hit | 2,216 | 166,665 | 57 | 50.2 to 62.7 | 70.4 or higher | approximately symmetric |
| Shield Break | 1,511 | 83,660 | 73.7 | 67.3 to 79.8 | 88.9 or higher | strongly left-skewed |
| 641 |
| 1,045 |
| 1,217 |
| 1,960 |
| 2,449 |
| 1.8 |
| Totem Speed | 539 | 1,417 | 1,286 | 588 | 837 | 1,024 | 1,680 | 2,100 | 1.9 |
| Elytra Mace | 459 | 10.8 | 11 | 4.9 | 4.5 | 7 | 14 | 17.1 | 0.4 |
| Elytra Mace Stun Slam | 207 | 5.5 | 5 | 3.6 | 1 | 3 | 7.5 | 10.5 | 0.5 |
| Lunge Mace | 749 | 9.2 | 9 | 2 | 6.5 | 8 | 10.5 | 11.6 | -0.2 |
| Mace Pearl Catch | 242 | 7.9 | 7.5 | 3.1 | 4 | 6 | 9.5 | 12 | 0.8 |
| Mace Stun Slam | 428 | 7 | 6 | 5.4 | 0.5 | 2.5 | 11.1 | 15 | 0.5 |
| Spear Mace | 253 | 5.4 | 5.5 | 3.9 | 0 | 2 | 8.5 | 10.5 | 0.3 |
| First Hit | 2,216 | 56.6 | 57 | 11.6 | 43.3 | 50.2 | 62.7 | 70.4 | -0.4 |
| Shield Break | 1,511 | 71.8 | 73.7 | 16.4 | 54.2 | 67.3 | 79.8 | 88.9 | -1.6 |
