Comments for combinatorial backgrounds in the tracks

As noted in the Introduction, SciFi tracker is basically 2D device and it provides two independent measurement on X-Z and Y-Z plane. So, in the case that two same lenth tracks exist in one event like Fig. 2, the tracker makes 4 combinations and in principle, there is no way to separate real two real tracks out of 4 tracks with 100% efficiency.

1. The event in Fig. 2 has two real tracks(Trk1 and Trk2) and 4 fragments.

   a. Trk1 has Fragment1(X projection) and Fragment3(Y projection)

   b. Trk2 has Fragment2(X projection) and Fragment4(Y projection)

2. In the reconstruction, tracker makes 4 combinations.

   a. Fragment1 + Fragment3 : Comb1

   b. Fragment1 + Fragment4 : Comb2

   c. Fragment2 + Fragment3 : Comb3

   d. Fragment2 + Fragment4 : Comb4

3. Out of 4 combinations, only two are real(Comb1 and Comb4).

   a. If we use other detector information(VETO/LG/MUC),
   $\Longrightarrow$ We don't lose any real track $\longrightarrow$ Maximize efficiency.
   $\Longrightarrow$ We have many tracks(including bad combinations) before we use other detector information. $\longrightarrow$ Looks noisy.

   b. If we select two combinations out of 4 using only SciFi information,
   $\Longrightarrow$ In some cases, we lose real tracks and pick up the bad ones. $\longrightarrow$ We lose Efficiency.
   $\Longrightarrow$ We have the cleaner event display(What does it mean???) and easy to understand.

4. We choose option 3.a rather than 3.b in order to KEEP EFFICIENCY as high as possible.
5. Miscellaneous : Users(Serious user) can optimize the selection criteria for their purpose, for themselves.

Figure 2: Event example which has two real tracks and 4 fragments.