2026 Gymnastics Worlds Qualification Stats: MAG by the Numbers

This article looks at the men’s artistic gymnastics qualification in Rotterdam through the numbers. It covers:

• Average and median scores on each apparatus
  
• How D and E scores spread, and which one decided each apparatus
  
• The most consistent all-arounders
  
• Total D and total E across the six apparatus in the AA
  
• Team depth: how much each team lost behind its top three
  
• Each team’s best and worst apparatus, ranked against the other teams
  
• Stuck landing bonus ratios per country
  

1. Averages & Medians

For each apparatus, we calculated two “typical” scores for the whole field. The average adds up all the scores and divides by the number of routines. The median is the score in the middle: if you line up every routine from lowest to highest, half the gymnasts scored above it and half below. We did the same for the D score and the E score.

Score, D and E – average vs median per apparatus

Median minus average score, by apparatus, points

Difficulty and Evolution scores, median minus, by apparatus, points

On five of six apparatus, execution scores trail the median more than difficulty scores. The wider the gap, the more the average understates the real level of the field.

What it shows:
For each apparatus, it shows the typical level of the field, measured in two ways. The average is the sum of all scores divided by the number of routines. The median is the middle score once all routines are sorted, so half the field scored above it and half below.

How it’s calculated:
The population is every routine performed in men’s artistic gymnastics qualification, with DNS entries excluded: 142 on FX, 149 on PH, 141 on SR, 146 on PB and 142 on HB. When the number of routines is even, the median is the average of the two middle values, rounded to three decimals. The “Avg – Median” column is the difference between the two measures.

How to read it:
The average is sensitive to extreme values. A single 3.00 penalty or a fall can pull it down noticeably, while the median barely moves. A negative “Avg – Median” means a tail of low scores is dragging the average below the typical score. The larger the gap, the more the average understates the real level of the field. On every apparatus here, the median is the more representative figure.

Limitations:
2026 Gymnastics Worlds Qualification stats have their limitations: For vault, D and E averages and medians cover 121 vaults, not 145. Twenty-four team members did a single vault that is published only as a total, without a D/E breakdown. Comparing averages between apparatus, such as rings at 12.442 against parallel bars at 13.010, describes different scoring scales, not which apparatus the field “performed better” on.

Example:
On parallel bars, the average is 13.010 and the median is 13.133. The gap of 0.123 is the largest of the six apparatus, driven by a handful of very low scores at the bottom of the field.

2. Difficulty & Execution Spread

This analysis asks a simple question: on each apparatus, are gymnasts separated mainly by the difficulty of their routines or by the quality of their execution?

Correlation of D and E with the final score, all routines

• Min, max and range. The range is the highest value minus the lowest, for D and for E.
• Standard deviation (SD). This is the typical distance of a value from the field’s average. It’s more robust than the range, which depends on only two extreme routines.
• Correlation with the final score. These are Pearson coefficients, between −1 and 1, measuring how closely D tracks the final score and how closely E tracks it. A value of 1 would mean perfect alignment.

Difficulty & Execution score, standard deviation points

• Min, max and range. The range is the highest value minus the lowest, for D and for E.
• Standard deviation (SD). This is the typical distance of a value from the field’s average. It’s more robust than the range, which depends on only two extreme routines.
• Correlation with the final score. These are Pearson coefficients, between −1 and 1, measuring how closely D tracks the final score and how closely E tracks it. A value of 1 would mean perfect alignment.

What it shows:
For each apparatus, it shows whether final scores are separated mainly by difficulty or by execution.

How it’s calculated:
It is computed for two populations: all routines, and the 102 all-around finishers only. For each apparatus, the tab gives:
– Min, max and range: The range is the highest value minus the lowest, for D and for E.
– Standard deviation (SD): This is the typical distance of a value from the field’s average. It’s more robust than the range, which depends on only two extreme routines.
– Correlation with the final score: These are Pearson coefficients, between −1 and 1, measuring how closely D tracks the final score and how closely E tracks it. A value of 1 would mean perfect alignment.
– “Score more tied to”: This column names whichever of the two correlations is higher.

How to read it:
A wide D spread combined with a narrow E spread means gymnasts execute at broadly similar levels and are separated mostly by what they attempt. That is the “difficulty-decided” profile. The reverse means execution errors create the gaps. The SDs and correlations should be read together.

Limitations:
Correlation describes association, not cause. When the two correlations are very close, as on pommel horse (0.84 vs 0.83), the label in the last column is technically correct but not meaningful, and the apparatus should be considered balanced. Bonus and penalties are part of the final score but not of D or E, so a 3.00 penalty weakens both correlations. Falls lower E sharply and can increase the E spread on their own. For vault, the all-routines population is again 121 vaults.

3. All-Around Consistency

Most consistent All-Arounders

• Best and worst. These are the gymnast’s highest and lowest of the six apparatus ranks.
• Spread. This is the worst rank minus the best.
• Average rank. This is the mean of the six ranks.
• “Best on” and “Worst on”. These name the apparatus concerned.

What it shows:
It shows which of the leading All-Arounders performed at an even level across the six apparatus in the men’s artistic gymnastics qualifications, and which relied on strong events to offset weak ones.

How it’s calculated:
On each apparatus, the gymnasts are ranked by their apparatus score among the 102 gymnasts who completed the All-Around. This is not their official apparatus rank, which also includes specialists. Ties share the best rank.
The table shows the 33 gymnasts in the top part of the all-around ranking, in their qualification order: the 24 qualifiers for the All-Around Final, the five gymnasts excluded by the two-per-country rule (NR), and the four reserves (R1 to R4).

How to read it:
A small spread means a gymnast’s level hardly changes from one apparatus to the next. Ranks are used instead of raw scores because a 13.5 on rings and a 13.5 on vault mean very different things relative to the field. Ranks put every apparatus on the same scale. The spread should be read together with the average rank. Two gymnasts can have the same spread at very different levels, and the average rank shows where that consistency sits.

Limitations:
Ranks magnify small score differences in a crowded field, where 0.033 can mean ten places. A single weak apparatus dominates the spread by design, so a gymnast with five excellent events and one poor one will look less consistent than his overall level suggests.

Example:
Oka Shinnosuke (JPN) is the most consistent of the 33, ranking between 2nd and 15th on every apparatus (spread 13). Hashimoto Daiki (JPN) follows with a spread of 15 and the best average rank of all (6.3), thanks to his 1st place on pommel horse among All-Arounders. Among the gymnasts who missed the final, Kawakami Shohei (JPN) stands out: he has the fourth-best average rank (11.0), but the two-per-country rule kept him out. At the other end, Tang Chia-hung (TPE), the first reserve, has the largest spread (90), ranking 2nd on high bar but 92nd on parallel bars.

4. All-Around Total Difficulty & Execution

Total Difficulty and Total Execution Across 6 Apparatuses

• Best and worst. These are his highest and lowest of the six apparatus ranks.
• Spread. This is the worst rank minus the best rank.
• Average rank. This is the mean of the six ranks.
• “Best on” and “Worst on”. These name the apparatus concerned.

What it shows:
It shows the total difficulty and total execution of the leading All-Arounders across the six apparatus, with a rank for each.

How it’s calculated:
For each gymnast, total D is the sum of his six D scores and total E the sum of his six E scores. The vault is the one counted in the All-Around. Total bonus and total penalties are also shown. Together they add up exactly to the all-around total: Total D + Total E + Total bonus – Total penalties = AA total. The D and E ranks are calculated among the 102 gymnasts who completed the All-Around, with 1 the highest and ties shared.

How to read it:
Total D describes the overall value of a gymnast’s programme, and total E how cleanly it was performed. Summing across apparatus is accepted practice because every gymnast is measured over the same six events, which isn’t true when you compare a single FX D with a single PH D. The gap between a gymnast’s D rank and E rank is often the story. A high D rank with a low E rank points to an ambitious but error-prone programme, and the reverse to a clean but more conservative one.

Limitations:
One fall costs a full point of E on a single apparatus and can move an E rank dramatically, so total E reflects the errors of the day as much as general execution quality. Total D includes vault difficulty, which is set by the vault table and follows a different logic from the other five apparatus. That’s normal, but worth remembering when comparing gymnasts with a strong vault.

Examples:
Hashimoto Daiki (JPN) and Angel Barajas (COL) share the highest total D (33.2), with Illia Kovtun (CRO) just behind (33.1). Their execution tells two different stories: Daiki ranks 11th in E, Angel only 37th. Oka has the best total E of the 33 (52.998), which explains how he topped qualification with only the 7th-highest D. The widest contrasts are Jermain Gruenberg (NED), with the lowest D of the group (27.0, 65th) but the 5th-best E, and Shane Wiskus (USA), 16th in D but 72nd in E, the lowest E of the 33.

5. Team Depth (counting-score efficiency):
How big was the safety net?

In team qualification, each country can send five gymnasts. On each apparatus, up to four gymnasts per team compete, and only the three best scores count. The figure is the difference between the third-best score, the last one that counted, and the fourth-best score, the one that was dropped.

Here is an example: Switzerland on parallel bars (PB column):

14.166 − 14.100 = 0.066. That’s the figure you see for Switzerland in the PB column.

Counting-score Efficiency (depth)

What it shows:
For each team and apparatus, it shows how much cushion the team had behind its three counting scores. A small gap means the fourth gymnast was almost as good as the third. The team had a real alternative and was well protected. A large gap means the fourth score was far behind. Either the team depended on its top three, or the fourth routine went wrong, and the format did its job by dropping it.

The biggest gaps

Three perfect ties: On three occasions, the third and fourth scores were exactly the same: France on pommel horse, Canada on rings and Spain on pommel horse. The fourth gymnast could not have been closer to counting.

Japan and Switzerland had the full package. Both teams put four gymnasts up on all six apparatus and kept their average gap low (0.373 and 0.406). Among teams with four up on at least five apparatus, only Ukraine (0.340) and Spain (0.350) did slightly better.

China shows why the dropped score matters. China has by far the largest gaps among the top teams (2.572 on average), but this is not a lack of depth. On several apparatus, the fourth routine produced a very low score, including two that carried a 3.00 penalty (Tian on rings, Lan on pommel horse). Thanks to the format, none of those scores counted. Without the safety net, China would have lost far more.

Vault is the tightest apparatus. The average gap on vault is only 0.430, against 0.746 to 0.912 on the other apparatus. Vault scores sit close together, so a team’s fourth vault is usually a credible replacement.

Some teams took the risk. Israel and Cyprus competed with only three gymnasts on five apparatus, and Kazakhstan on three. On those apparatus, every routine counted, and there was no margin for error.

Switzerland and the USA have gaps of just 0.066 on parallel bars, and Japan 0.067 on vault, which is near-perfect depth. China‘s average gap is 2.572, the largest among the top teams, because two falls became its dropped scores.
On vault, Japan‘s last counting score was Doi’s 14.033, and the dropped score was Oka’s 13.966. The gap is only 0.067, so Japan had a near-identical vault in reserve.

How it’s calculated:
For each team and apparatus, the gymnasts’ scores are sorted, giving the 3rd score (the last one that counted) and the 4th (the dropped one, under the 5-4-3 format). The gap is the difference between the two. When only three gymnasts competed on an apparatus, there is no dropped score and the gap is shown as “–”. The average gap is calculated only over the apparatus where four gymnasts competed.

How to read it:
The gap is the team’s insurance. It’s the most the team could have lost if a counting gymnast had a bad routine, because the 4th score would have replaced it. A small gap means the team had a strong fourth score and was well protected. A large gap means it depended on its top three.

Limitations:
A large gap doesn’t always mean a lack of depth. Often it means the fourth gymnast fell and his score was simply dropped. China’s large gaps on pommel horse and rings come from exactly that, not from a weak squad. A gap of 0.000 means the 3rd and 4th scores were identical. Apparatus with only three gymnasts up have no insurance at all, which is itself a strategic choice worth noting.

6. Team Apparatus Ranks

Each Team’s Rank on Every Apparatus (24 Teams)

What it shows:
It shows each team’s position on every apparatus, among all 24 teams.

How it’s calculated:
On each apparatus, the 24 team scores (the sum of the three counting scores) are ranked from highest to lowest, and tied scores share a rank. For each team, the tab shows its six ranks, its best rank and the apparatus where it was achieved, and its worst rank and apparatus. The data comes from the full 24-team ranking, including Switzerland, Kazakhstan, Israel and Germany, verified against the official live page. Green rows are Team Final qualifiers.

How to read it:
This is the fair way to name a team’s best event. Comparing a team’s own scores across apparatus would simply favour apparatus with higher scoring scales. Ranking each team against the others on the same apparatus avoids that, so “4th-best team score on PH” is meaningful in a way that “their highest score was on PB” is not.

Limitations:
Ranks hide margins. A team can be 2nd by 0.001 or by 2 points, so it’s worth checking the team scores themselves when a rank is used in a headline. Ties occur, such as Ukraine and Korea sharing 17th on FX, Hungary and France 14th on SR, and Germany and Hungary 13th on VT, and the following rank is skipped.

Example:
Japan has the best team score on both parallel bars and high bar. Kazakhstan, only 12th overall, had the 2nd-best team score on pommel horse. Switzerland’s best apparatus are PH and PB, ranking 4th on each.

7. Stick Leaders – Nations & Apparatus

Stik Rate Per Team

Stick Rate Per Apparatus

What it shows:
It shows the total sticks per NOC, and in a second table the stick rate per apparatus.

How it’s calculated:
It includes all bonuses obtained by every gymnast of each NOC, whether team members or individuals. Only countries with at least one stick are listed. The per-apparatus table divides sticks by routines with a known bonus on each apparatus.

How to read it:
The total favours large delegations, while the rate measures the habit of sticking regardless of team size. The per-apparatus stick rate is one of the few valid cross-apparatus comparisons, because a percentage of routines doesn’t depend on each apparatus’s scoring scale.

Limitations:
Rates from small samples are unstable. Mongolia‘s 60% comes from one gymnast’s 5 routines, and a single extra or missing stick would change it by 20 points. Vault sticks are undercounted for nations whose team vaulters only appear as totals. Also, the 2026 Gymnastics Worlds Qualification stats show there were fewer sticks on vault in this competition (for example, only 1 out of 16 vaults for the gymnasts qualified for the final).

Examples:
Japan has 10 sticks from 20 routines (50%), ahead of Great Britain and the USA (6 each, 30%). By apparatus, high bar has the highest rate (35 of 142, 25%) and vault the lowest (8 of 148, 5%).

More: 2026 Gymnastics Worlds Qualifications: Hits, Misses & Surprises

2026 Gymnastics Worlds Qualification Stats 2026 Gymnastics Worlds Qualification Stats 2026 Gymnastics Worlds Qualification Stats