Mr81.xing

Ph.D candidate in Basketball Performance Analysis #inefmadrid 🇨🇳🇪🇸 Disciple of Brad Stevens 🍀

[Mr81.xing @XingW_INEF]

Thanks! So glad we got to build this together.
Here’s a little preview 👇 — more to come at #SSAC26! yellow indicates higher value movement actions.

[Kostas Pelechrinis @kpelechrinis]

If you are attending #SSAC26 make sure to check our work on HoopEval. Led by @XingW_INEF !

[Mr81.xing @XingW_INEF]

@Anderson_MCI There is no predictive metric that is completely accurate.

[Mr81.xing @XingW_INEF]

Version 2.0 Incorporating contrastive learning into a multi-task self-supervised pre-training framework can enhance the representation of players' collective trajectories.

Hoop-MSSL: Multitask self-supervised representation learning on basket... https://t.co/E5y2GSw5SM

[Mr81.xing @XingW_INEF]

@M_Robertson100 and the shoot is a pass to the rim

[Mr81.xing @XingW_INEF]

Can dribbling be considered a pass to the ball handler himself?

[Mr81.xing @XingW_INEF]

@Vvirgilicky @shirley_MHSE Congrats👏

[Mr81.xing @XingW_INEF]

@keisuke_fj Thanks, by "CQL-like" do you mean tuning the CQL regularization term so that low-frequency actions are penalized less and high-frequency actions are penalized more?

[Mr81.xing @XingW_INEF]

RL in Sport 🟰 imbalanced multi-agents offline data😵‍💫

[Mr81.xing @XingW_INEF]

@JSportsSci can accept latex as an editable file?

[Mr81.xing @XingW_INEF]

Tracking data has revolutionized performance analysis (most on collective behavior level) —so what impact will skeleton data bring?

[AIT Lab @ait_eth]

Together with @FIFAWorldCup @FIFAcom, we introduce #WorldPose: a dataset designed to advance multi-person global pose estimation in the wild, featuring footage from the 2022 FIFA World Cup. Website: https://t.co/AKbv5Ae8Z5

[Mr81.xing @XingW_INEF]

@YuluPan_00 这个真能从视频里定位到指定的球员么…有没有这方面的验证？

[Mr81.xing @XingW_INEF]

@DTRadke7 @UWCheritonCS @AAMASconf Use the data from simulation environment to build some evaluate model than apply to real data？

[Mr81.xing @XingW_INEF]

@karun1710 I'm doing the same thing in basketball. Are you using RL for “multi-agent action valuation”?

[Mr81.xing @XingW_INEF]

@arbues6 @susanaferreras @arecasens @AleixMMartinez Is this in English or Catalan?

[Mr81.xing @XingW_INEF]

That why we need to value the off-ball action

[Chris Steed @csteed]

Basketball is played WITH the basketball 5% of the time.

One of my favorites quotes from Billy Donovan is:

‘The other 95% of the time most players spend trying to get the basketball’.

Players also can’t conceptualize the amount of time they have the ball in their hands in a game. They think it’s exponentially more than it actually is.

The ball is glorified, and rightfully so, but what are we doing with the other 95%?

Do be ELITE REQUIRES impacting the game without the basketball.

The video below is a MUST WATCH for all players and coaches 👇

🎥 @chicagobulls

[Mr81.xing @XingW_INEF]

@chrisleckey_ @BJSM_BMJ @EamonnDelahunt @NicolvanDyk @DohertyCailbhe @aonghuslawlor XGBoost alway performs best 🤣🤣🤣 cause we only have tabular data

[Mr81.xing @XingW_INEF]

It was a great time to work with @kpelechrinis ！！

[Kostas Pelechrinis @kpelechrinis]

I unfortunately could not attend #CASSIS2024 this past weekend but here is the poster from our work led by @XingW_INEF!

[Mr81.xing @XingW_INEF]

@JDeLeo @_NeilC_ @Wod_Society @alicesweeting @MeghanMHall mixed linear

[Mr81.xing @XingW_INEF]

@burakcankoc 🙌