Jesse Davis
@jessejdavis.bsky.social
Professor @ KU Leuven researching AI, machine learning, and sports. Co founder of runeasi
Burger time! With a very fitting flipper thanks to @janvanhaaren.be and Giels
September 8, 2025 at 5:25 PM
Burger time! With a very fitting flipper thanks to @janvanhaaren.be and Giels
At OHL for women’s champions league qualifiers with the whole family
August 27, 2025 at 5:58 PM
At OHL for women’s champions league qualifiers with the whole family
This is one meal that makes both kids very happy
June 10, 2025 at 3:27 PM
This is one meal that makes both kids very happy
Spain - Belgium #FlameTime with a very happy daughter
May 30, 2025 at 7:26 PM
Spain - Belgium #FlameTime with a very happy daughter
All possession value frameworks establish a time window over which the effect of an action sequence should be measured. How this window is defined leads to 4 key differences. Here’s what we found 👇
December 12, 2024 at 12:03 PM
All possession value frameworks establish a time window over which the effect of an action sequence should be measured. How this window is defined leads to 4 key differences. Here’s what we found 👇
One can try to predict the probability that a goal will be scored from a possession or estimate the xG of a possession. Which one is better?
According to our experiments, using xG as a label (soft classification) converges faster.
According to our experiments, using xG as a label (soft classification) converges faster.
November 21, 2024 at 5:40 PM
One can try to predict the probability that a goal will be scored from a possession or estimate the xG of a possession. Which one is better?
According to our experiments, using xG as a label (soft classification) converges faster.
According to our experiments, using xG as a label (soft classification) converges faster.