Faith in theory and curve-fitting, at least…
Tommaso Dorigo reports some new results, which are based on a figure that could be titled “Why I Am Not a Particle Physicist #729”:
“What’s the problem?,” you ask, “There’s a nice big peak there, looking a little like a black-body spectrum.”
Ah, but that’s not the signal. The green shaded region in the big plot is all background. The signal is in the tiny little gap between the green background line and the blue data points on the right-hand side of the enormous background peak.
Now, this is actually a pretty solid result, as you can see from the inset, which shows the difference between the data and the background, with error bars. There’s a clear peak, at least 5-6 times larger than the error bars, and it’s right about where they expect it (which I believe is what’s indicated by the little red shaded peak in the bigger plot). I’m not saying this is a bogus result because I can’t immediately see the signal in the raw data, or anything like that.
I’m just saying that I’m glad I don’t work in a field where I’m trying to tease a signal that small out of a background that large. You really need to have some confidence in your measurements of the background level, and do some serious statistics to confirm that it’s a real signal. I just don’t have the right temperment for that sort of work…