SPRT test not accepted with large LLR

[vondele]

https://tests.stockfishchess.org/html/live_elo.html?61a9d4bf9e8855bba1a35c4f

was stopped by hand after reaching 3.11

maybe related to the fact that there are many purged workers ?

[ppigazzini]

Weird, green residual purged too. Do we know if this was purged by hand?

[vdbergh]

This test had auto purge on. So perhaps it just kept on purging....

https://tests.stockfishchess.org/actions?count=1&before=1638681200.308

[vdbergh]

The fact that green workers were purged is weird. Green workers can in principle be purged if the p value is very low. But if all workers are green then you would not expect the p value to be very low.

[vdbergh]

The fact that green workers were purged is weird. Green workers can in principle be purged if the p value is very low. But if all workers are green then you would not expect the p value to be very low.

Ah but that's precisely what's happening. The p-value is still 0.00 despite all the workers now being green. Unfortunately it is tricky right now to audit the chi2 computation.

There should be a summary table per user, with an option of csv export.

[vdbergh]

It might be interesting to know how much more purging would be necessary for the p-value to go up.

[vondele]

you want me to push purge more on the test?

[vdbergh]

We could. But perhaps @ppigazzini can transfer the db to DEV?

[vondele]

let me purge now a couple of times, doesn't really make a difference

[vondele]

btw, we should make you an approver on fishtest..

[vondele]

I pressed purge once more (IIRC), and the run got LLR of 2.71 and started running again. p value 0.01%