Here is an exchange of mails with Kyle Cranmer (kyle.cranmer@nyu.edu)
[Posted here on February 10, 2018]
GdA-1 (2015-12-23 15:16) -------------------------------------------
Dear Kyle,
a friend has pointed me the article in NYT in which you are cited
http://www.nytimes.com/2015/12/16/science/physicists-in-europe-find-tantalizing-hints-of-a-mysterious-new-particle.html?_r=0
According to the journalist you state that
"the bump in the Atlas data had about a 1-in-93 chance of being a fluke",
THAT IS 92-in-93 of NOT being a fluke.
In other words, FAIR bet odds are 1 to 92, right?
If this is you opinion, you should be ready to accept the bet in either direction.
For my reasons, I choose to bet 10 CHF on Fluke, asking you to bet 920 CHF on non-Fluke.
To be more clear (its is a question of money!):
- I pay 10 CHF and you pay 920 CHF;
- if the present excess will result to be something a real new particle, you will get the 930 CHF;
- if the present excess will turn out to be just a fluke, I will get the 930 CHF.
If you accept, we can define a trusted person at CERN to hold the money,
waiting for the result of the new run.
Best regards,
Giulio
--------------------------------------------------------------------
--------------------------------------------------------------------
KC-1 (2015-12-23 19:08) --------------------------------------------
Hi Giulio
I understand the betting odds, but that wasn't my quote. I provided the p-value number and he wrote the part about being a "fluke".
That phrase is not precise and I can interpret either as a classic probability inversion (mistake) or as a colloquial way of saying "a bump at least this big assuming there is no signal" (i.e. a p-value.)
My odds are more like 1/3 that this is real.
I'll bet you 30CHF if you want.
Kyle
-----------------------------------------------------------
GdA-2 (2015-12-23 19:47) ---------------------------------
Thanks a lot for your prompt reply, Kyle!
This is want I wanted to hear, although I can ensure you that
in other cases similar statements have been provided _verbatim_ to
journalists by our colleagues, or they have been directly written by them[*]
(And also in this case, an Italian physicist of ATLAS has WRITTEN something similar,
so that he cannot blame the journalist:
http://www.lescienze.it/news/2015/12/19/news/qualcosa_di_nuovo_a_lhc_solo_il_tempo_lo_dira_-2900622/)
Anyway, I accept the bet you propose (10CHF Vs 30CHF),
and I do not think we need a kind of notary :-)
Best wishes,
Giulio
-----------------------------------------------------------
KC-2 (2015-12-23 22:38) ----------------------------------
I agree and appreciate your interest in these matters. I took an extended interview trying to break down these points of confusion.
I'll take the bet, and I agree, no notary is needed. I would hope that by this time next year it will be clear.
All the best,
Kyle
-----------------------------------------------------------
Back