Back To Main Page

I am still puzzled by the background estimates method (p 3 l 25 ... of the 12/9/2012 version).
I want to know the ratio of two quantities,(i) the number of 3-jet ttbar events that have b-tags (in MC+simulation), scaled to the luminosity used, (ii) the observed number in 3-jet events with b tags in the data. This question relates to what constitutes "negligible").

Although I don't have the number quoted here, we can have a sense of the value of the ratio from page 6, 2nd paragraph of internal note: The estimated ttbar presence in 4-6 jets is about 5% to 10%, so the ratio for 3-jet events should be significantly lower.

Also, I do not understand the response to question by another GP. Question was is there an assumption that relative contributions of different background processes are the same in 4-jet events as in 3-jet events? For example, if there were just 2 background processes, and one had zero 3-jet events and many 4-jet events, and the other had both 3-jet and 4-jet events,would the procedure still work?
Restated,the 3-jet events yield some distribution that is an average over several processes,and the averaging would be different for the 4-jet events.

Since the construction of the b-tag matrix does not treat varies of background processes differently, then the bkgd estimation does not carry the assumption stated above.