## Saturday, August 06, 2016 ... /////

### ISIS claims responsibility for the diphoton excess

Anniversary: Exactly 25 years ago, on August 6th, 1991, Tim Berners-Lee posted a modest USENET post in which he announced the World Wide Web to the world.
Dr Abu Bakr al-Baghdadi, Nude Socialist is honored that you agreed to give us this exclusive interview. After Hillary Clinton, you became another leader of global importance – one who fights against the discrimination of Muslims – who talks to us. Why was there a diphoton excess near $750\GeV$?

We realized that some of our warriors can't even get to a disco club anymore. We were thinking about ways to harm the Western infidels most severely. Airplanes? Trains? Trucks? Guns? Axes? Machetes? Safety matches? Fingernails? Finally, Allah gave me a better idea. We would create a new God for the infidels – a God many of them consider more important than Allah and Mohammed (PBUH) – and then we would kill him.

How did you do that?

In Fall 2015, we have used a machete and cut the heads of 1 experimenter from ATLAS and 1 experimenter from CMS. Surprisingly, their replacements have agreed to inject some 4 sigma of signal into the ATLAS and CMS 2015 datasets. A new fake God, the $750\GeV$ diphoton resonance, was born.

And in 2016, it was enough to do nothing.

Did you really harm the Western civilization?

I am sure that this terrorist attack was more effective than Al-Qaeda's 9/11. You can see it if you read comments like
Jaswin Kasi: I feel sad for my friend who had to work really hard to work on a theory which explained $750\GeV$ excess.
and numerous gloomy articles in the media and the blogosphere, you know that these people were harmed more than after the 3,000 deaths on 9/11.

Wasn't it sensible for the physicists to believe that the bump was a statistical fluke, anyway?

Many of them did, others were uncertain. After I resign as the leader of ISIS, I want to recommend a great idea to the LHC physicists. They should express the "global significance" in terms of "a false positive per $X$ papers".

For example, a bump of the local significance of 4 sigma may be weakened to the global significance of 3 sigma because of some 30 bins in the same graph. 3 sigma means that a false positive appears in "1 in 300" situations. The number 300 should be reduced by another factor $N$ equal to the effective number of "similar but independent graphs" in the present paper where an anomaly could emerge, e.g. to 100.

And this number 100 should be expressed by the sentence "a deviation of the same global significance appears once in 100 papers" and the number of similar papers that ATLAS or CMS have published so far (and those that they plan to publish in the ongoing run) should be written for a comparison.