Marc Saint-Jour

@elliboops I agree, I think there should be an xG cutoff for PK. An xG of 0.03 cannot be rewarded an xG of 0.8; the disproportion is too absurd. I would say a FK in the box (wherever the foul happened) for clear fouls with an insignificant xG below the cutoff and PK for clear goal chances.

Spain hasn’t registered a single shot on target against France, yet they are gifted an 80% xG penalty out of absolutely nothing. It is deciding the game on a non-scoring situation, not to mention a clear handball by Yamal during the buildup to gain control. This is exactly why football rules are broken. It’s too binary. We need a Graded Penalty System: 1️⃣ Standard Penalty (11m): ONLY if a foul denies a direct, active goal-scoring opportunity (attacker facing goal, in motion to shoot). 2️⃣ Distant Penalty (14-15m): For borderline/reckless fouls that don't deny a 100% clear chance. Gives the GK a fighting chance. 3️⃣ Edge-of-Box Free Kick: For soft fouls or back-to-goal situations. No more free gifts. And the ultimate balancer: The player fouled must take the kick. Stop deciding massive tournaments on technicalities. Let the players play!!! To be clear: Under my proposal, Spain's incident today is a clear Option 3 (Edge-of-Box Free Kick). Furthermore, we must fix the absurd handball rule—any contact that advances an attacker's advantage, even if accidental or close-body, should be penalized. The current binary penalty rule is completely illogical and should have been scrapped years ago. Just look at other sports: In basketball, if you foul a player in the paint, they get 2 free throws to potentially score 2 points. It’s proportional. They don’t get awarded 20 free throws just because the foul happened inside the key. Yet in football, a minor, zero-threat trip at the edge of the box is rewarded with a penalty kick—turning a 3% scoring chance into an 80% free gift. It’s the equivalent of giving a basketball team 15 free throws for a common foul. The punishment is completely disconnected from the crime. It’s time to bring mathematical proportion and logic back to football

@theMadridZone Rodri…

🤔 Who should be Real Madrid’s priority target — Rodri or Michael Olise?

😂 😝 😆

@shaunmmaguire @RoKhanna I got fooled once by commie Ro Khanna. I won’t be fooled again. You guys were wrong for selling this guy as a moderate. He says nothing that deserves credit; let it go, he is a commie.

Credit where credit is due (And I have been highly critical of @RoKhanna’s antics this week) I think he handled this segment well, and under pressure

This is a perfect analysis..

@RalphEFrancois Se sak fè Pedri pran ban ...paske-l FLOP?

⚽️ Men jwè ki flòp tande. Si w pa malad, ou an sante, ekip bezwen gòl yo retire w, ou pa priyoritè..

Fact check before the propaganda ..!

@ralphganthier Exactly, it was all fun when EDF was playing inferior teams but with a superior team who has the ball, DD should have done what he knows best.. Ugly football..

M kwè sa k rive DD jodi a se tèt koupe ak sa k te rive Mou an novanm 2010 lè li pran manita a. Yo rive nan match yo (vs Espay pou DD, vs Barça pou Mou) sou on dinamik pozitif kote y ap jwe on foutbòl ofansif, ekspansif e pran risk ki pwodui anpil gòl ak viktwa. Laprès ansanse yo.

💔

I think Real Madrid is going to have to reconsider Rodri .. He is back ..

Rodri + Fabian Ruiz —> Menace to society!

@nbraun Adrenaline bro 😂 ..

@MSaintjour Penal la te ase clair wi 🤷‍♂️

Very poor refereeing .. this is a bullshit Pk

Physician-builder takeaway: Precision medicine AI should not be judged only by AUC. The bar is whether it turns messy biology into a decision clinicians can trust, audit, and explain to patients. This is non-ambient AI medicine worth watching.

For physicians, the hard part is workflow. Can the test be reproduced across labs? Will the inputs be available fast enough? Does it change treatment selection, trial matching, or follow-up? And who owns the risk when predictions conflict with clinical judgment?

AI in medicine is not just notes. A new npj Digital Medicine paper points to another lane: multi-omics ML for ovarian cancer treatment-response prediction. The question is not "can AI read the chart?" It is: can it help make precision oncology more actionable?

@simonw Nope.. Hermes all day

It's been about six months since OpenClaw burst onto the scene - are you still using yours? Did it become a daily driver? Any interesting lessons or anecdotes you can share?