7+d Upset losses against 5-d


LOSSES

Sooah Kim 7d KR below top group 7:3
- Main Round 2, Zihao He 4d GB below top group 6:4
- Rapid Round 5, top group, Spiegel Lothar 5d AT top group 3:5

Wada Hirohide 7d JP top group 5:5
- Main Round 2, Shu Xin 5d CN top group 5:5
- Main Round 4, Ougier Guillaume 5d FR below top group 6:2
- Senior Round 5, top group, Habu Koichiro 5d PL top group 5:0

van Zeijst Rob 7d NL top group 5:5
- Main Round 3, Dobranis Denis 5d RO below top group 2:3
- Main Round 10, Mantle Ben 5d CA below top group 6:4
- Senior Round 4, top group, Habu Koichiro 5d PL top group 5:0

Cho Daehee 7d KR below top group 6:4
- Main Round 2, Ben Malek Farid 5d FR below top group 4:4
- Main Round 9, Zhang Shukai 5d DE below top group 7:3
- Rapid Round 2, top group, Zihao He 4d GB below top group 5:3
- Rapid Round 4, top group, Shu Xin 5d CN top group 4:4

Chang Hyukkoo 7d KR below top group 5:5
- Main Round 1, Ougier Ariane 4d FR below top group 3:2
- Main Round 6, Côme Théo 4d FR below top group 6:4
- Main Round 8, Ulbricht David 4d DE top group 4:6
- Main Round 10, Lidor Jonathan 5d IL top group 5:5
- Rapid Round 4, top group, Dobranis Denis 5d RO top group 3:1
- Rapid Round 5, top group, Cobbold Scott 4d GB below top group 5:3
- Rapid Round 7, top group, Jasiek Robert 5d DE top group 4:3
- Rapid Round 8, top group, Barcza Bende 3d HU below top group 5:3

Debarre Thomas 7d FR top group 4:1
- Main Round 2, Labouret Florent 5d FR top group 2:2

Frejlak Stanislaw 7d PL top group 3:2
- Main Round 1, Labouret Florent 5d FR top group 2:2

Yang Seamo 7d KR below top group 2:2
- Main Round 4, Pittner Arved 5d DE below top group 6:4

Sato Sotaro 7d JP top group 2:3
- Weekend Round 5, top group, Pittner Arved 5d DE below top group 4:0


5-d PLAYERS WITH TWO UPSET WINS

- Zihao He 4d
- Habu Koichiro 5d
- Shu Xin 5d
- Dobranis Denis 5d
- Labouret Florent 5d
- Pittner Arved 5d


Chang Hyukkoo 7d REMARKS

According to Ulbricht David 4d about the games against himself and Welticke Jonas 6d, Chang's opening might be weak. In my Main game against Chang, I played a few too passive moves in the opening and lost easily. In my Rapid game against him, I did not recognise his face or recall his name, so played with unhindered motivation but the expectation of having no chance against a 7d, however, he played a few too passive moves in the opening, after which I used positional judgement to establish a slight winning position and then played good endgame to win by 11.5 points; according to Mateusz Surma 8d, I played well while Chang should have played more aggressively during the opening, as (my comment) younger Korean 7d would do. In the last Rapid round 8 on his second-last move, Chang was ahead by a few points, had some 12 minutes remaining time, knew his very young 3d opponent was down to the last seconds or the 1 second Fisher overtime, saw his opponent's great desire to win, self-ataried five stones while perfectly pretending carelessness but surely did lose this way intentionally in this game. Chang may have played like a 7d in former times but his well intermediate age may have reduced his actual strength to weak European 6d level, although he still plays occasional 7d moves.


CONCLUSION

During most previous congresses, wins by 5-d against 7+d was mostly unthinkable. It was a rare exception that a 5d beat a 7d in Rapid (or a 7d from Taiwan was ranked wrongly). The many upsets in this 2024 congress are a new phenomenon. Is it a consequence of AI study that 5-d players are catching up?

Possibly the organizers didn't push back as much as they would in the past on Asian high dan claims? Nowadays there are more players claiming 7d and pro ranks in Europe. Fewer players seem to shy away from declaring 7d like they would have in the past. If the requirement isn't to be able to kill the top group at EGC and there are more Europen 7d now, then maybe it is only fair that there are more Asian 7d.

I'd look more closely at the rating differences and the predicted win rates before drawing conclusions. The first and only result I checked was Rob van Zeijst Rob vs Denis Dobranis and it is hardly an upset on 21.7% predicted win rate in the GoR system. I thought the predicted win rate would be more extreme when trailing -126 points at this level but this is what was reported on EGD webpage.

I imagine other "upsets" could be explained in the same way but I don't think the results are in EGD yet. Also it might not indicate much for non-European players who only have been to a few European tournaments so for.

The fact that he lost against so many 3-5d shows that his level is not really 7d.



Just before the Congres Denis Dobranis had GoR 2546, and he was 2553 in June.
Just before the Congress, Florent Labouret had GoR 2539, and was over 2550 about 2 years ago.

As explained, the 3d loss must not be taken seriously. Otherwise, for his current strength, you are right.



I am not a believer in ratings, especially below the very top. However, please explain what these ratings mean compared to 7d ratings and what 2550 is supposed to mean compared to ranks!

In the EGF scale, a 6d rank corresponds to ratings between 2550 and 2650 and a 7d rank corresponds to ratings between 2650 and 2750. So both Dobranis and Labouret are about at the limit between 5d and 6d, so are approximately 1 stone weaker than a "weak" 7d.

Over the past 15 years, the EGD has recorded 235 wins out of 1789 even games (14.1%) between 5d and 7d, and the EGD also expects a 15% winrate for even games between a nominal 5d (GoR 2500) and a nominal 7d (GoR 2700).
Those percentages are similar to what is observed and expected for even games between 5k and 1k.
For both of those cases (5d vs 7d and 5k vs 1k) the observed and expected winning chances are not really as close to 0 as people may think. All players have their good days and their bad days, and consistency may vary between players of the same overall level (some 15 years ago I regularly beat 6d players, but I could also easily lose against 1k players).

To evaluate whether these EGC results are anomalous, we'd need to know how many EGC games in total were played between 5d and 7d.

Maybe those relatively few Korean European-ranked 7d at earlier congresses were among the strongest Korean amateurs fighting for Korean promotion (to Korean 7d?) by winning the European Go Congress main tournament. Only Alexandr Dinerchtein used to have chances to beat (some of) them. Some of those 7d continue to dominate side tournaments, such as lighting, 13x13, 9x9. (Except that this year a 9d (I heard 9p but I have not checked this) participated and won the lightning.)

When a strong 7d player from Korea scores consistently better than expected for 7d EGF (like a lot more than 85% wins against European 5d players), their rating in the EGD will gradually increase.
This has happened with some strong Korean amateurs who lived in Europe for many years and participated in many tournaments. Some of them (usually former insei in Korea when the insei competition was very stiff there and they failed to become pro before they were too old) are now rated 8d EGF (like Hwang In-Seong, Kim Young-Sam, Kim Seong-Jin, Cho Seok-Bin).

The EGD expects 8d EGF (GoR 2800) to win 88% against 6d EGF (GoR 2600) and 95% against 5d EGF (GoR 2500), so if they do worse than that, their rating will decrease, while if they do better than that, their rating will increase.