Phygas: strengthening the results for publication

https://zenodo.org/records/23680281
Phygas has finished tracing the Gaia DR3 stars back against young clusters, and the first comparison with chance-alignment control runs shows a clear excess of slow, cluster-linked candidates over what chance alone produces. We are now preparing these results for a peer-reviewed journal (Physics and Astronomy Reports, Istanbul University), and a published claim needs more than one control run.

What is changing in the work you receive
Over the coming days and weeks Phygas tasks will come in two extra kinds. You do not need to do anything: they run exactly like the tasks you already know, with the same application and the same credit.

• Control runs (20 of them): the whole star catalogue is screened again after each star has been given the motion of a similar star nearby in the sky and at a similar distance. This breaks the link between a star and its cluster while keeping the Galaxy's large-scale motions. Repeating it with 20 different random draws tells us how much the chance level itself varies.
  
• Measurement-error runs (20 of them): the real catalogue is screened again after every star's parallax, proper motion and radial velocity are re-drawn within their published Gaia uncertainties. A real cluster link should survive this; a link that only exists because of noise will not.
  

Why this matters
A candidate list is only worth publishing if we can show how often the same list appears by chance. With 20 control draws and 20 error draws we can give every cluster an honest probability, correct for the fact that we looked at 663 clusters at once, and state clearly which excesses are real.

Every task is checked twice
Each of these tasks is sent to two different volunteers and only counts when both results agree, so the published numbers rest on independently verified computations. Because every task goes to two people, there is plenty of work for everyone who wants to take part.

What you will see
Your normal Phygas credit and badges keep counting. The tasks are the same size as before. All results, real and control, will be published openly on Zenodo with the paper, and everyone who contributed computing time is credited by their BOINC user name.

Thank you for every task you have returned so far: the controls are the part that turns a long candidate list into a result other astronomers can trust.

---

I see today Phygas and Nostos have no unsent tasks. Will there be a new batch?

In order to get results before 20th November i recommend to send at least 1 extra task per WU... And then to save CPU times, cancel unstarted tasks on server for completed Wus...

---

I crunch for Ukraine

There are still a lot of tasks in the queue. I slowed down the producer; file compression can't keep up with it.

---

In reply to Alperen Yavuz's message of 2 Nov 2026:

There are still a lot of tasks in the queue. I slowed down the producer; file compression can't keep up with it.

ah. ok!

i turned on "receive tasks from other projects ...." but will turn it off now again to get more Nostos and Phygas in cache...

---

I crunch for Ukraine