Greg Starek

Equally passionate about football and dataframes? Endlessly curious? Motivated by team success? We’re hiring a Football Data Analyst with the Colts, and I’d love to hear from you. teamworkonline.com/football-jobs/…

The Indianapolis Colts are hiring for the next batch of Harriet P. Irsay fellows! Analysts are especially encouraged to apply. Reach out to me with any questions. Learn more and apply here: colts.com/news/harriet-i…

Let’s gooooo!

Totality

Heartbroken to hear of the loss of Bill Catterall today. Thank you, Bill, for inspiring me to continue asking questions.

When all I could see was my project come to an end, Bill only saw new questions. This was the beauty of science: new findings lead to new questions which lead to new findings. He didn’t have answers, he insisted, but evidence that needed to interrogated, *especially* with theory.

At a tiny conference in Italy in 2011, a buzz started going around one afternoon: Bill Catterall had solved a voltage gated sodium channel structure, and he was going to share the structure publicly for the first time. (1/x)

@EstebanNG_ @BenJedlovec @ZelusAnalytics Incredible hire!

Now that it's officially official, I couldn't be more excited to welcome @BenJedlovec to @ZelusAnalytics🥳 Ben will help us (among many other things) to develop our data engineering as a service offering. If you're trying to squeeze wins out of data, we'd love to talk to you!

I wonder to what extent the growing fetishization of translational and applied research is just an end-of-history illusion: the inability to imagine that there are things we don't understand leading to the idea that we simply need to try harder to apply what we think we know

The Indianapolis Colts have hired a new director of football analytics:

FINALLY

1/7 of humanity uses Ezoic every month 🤯

So, so, so awesome.

Women’s NCAA basketball championship let’s gooooooo

@notshibo SO GOOD

another ke huy quan win 🥹🥹🥹

@johncole1 I worry about the long-term stability of these tools for this very reason. As more and more online content inevitably becomes AI-generated, if we are unable to reliably determine artificial information, AI will be producing its own training set, thereby baking in its own biases.

Generative Pre-trained Transformers GPT3+ Bard are pulverising all data, then reforming it to look like new data. That's not necessarily bad, but without validation and verification, it's going to be tough to know what's real and what's not. What do you think?