| 12 |
|
from $W$-decays, which is reconstructed as \met in the |
| 13 |
|
detector. |
| 14 |
|
|
| 15 |
< |
in 30 pb$^{-1}$ we expect $\approx$ 1 SM event in |
| 15 |
> |
In 30 pb$^{-1}$ we expect $\approx$ 1 SM event in |
| 16 |
|
the signal region. The expectations from the LMO |
| 17 |
< |
and LM1 SUSY benchmark points are {\color{red} XX} and |
| 18 |
< |
{\color{red} XX} events respectively. |
| 17 |
> |
and LM1 SUSY benchmark points are 5.6 and |
| 18 |
> |
2.2 events respectively. |
| 19 |
> |
%{\color{red} I took these |
| 20 |
> |
%numbers from the twiki, rescaling from 11.06 to 30/pb. |
| 21 |
> |
%They seem too large...are they really right?} |
| 22 |
|
|
| 23 |
|
|
| 24 |
|
\subsection{ABCD method} |
| 41 |
|
|
| 42 |
|
\begin{figure}[bt] |
| 43 |
|
\begin{center} |
| 44 |
< |
\includegraphics[width=0.75\linewidth]{abcdMC.jpg} |
| 44 |
> |
\includegraphics[width=0.5\linewidth, angle=90]{abcdMC.pdf} |
| 45 |
|
\caption{\label{fig:abcdMC}\protect Distributions of SumJetPt |
| 46 |
|
vs. MET$/\sqrt{\rm SumJetPt}$ for SM Monte Carlo. Here we also |
| 47 |
< |
show our choice of ABCD regions. {\color{red} We need a better |
| 45 |
< |
picture with the letters A-B-C-D and with the numerical values |
| 46 |
< |
of the boundaries clearly indicated.}} |
| 47 |
> |
show our choice of ABCD regions.} |
| 48 |
|
\end{center} |
| 49 |
|
\end{figure} |
| 50 |
|
|
| 54 |
|
in the four regions for the SM Monte Carlo, as well as the BG |
| 55 |
|
prediction AC/B are given in Table~\ref{tab:abcdMC} for an integrated |
| 56 |
|
luminosity of 30 pb$^{-1}$. The ABCD method is accurate |
| 57 |
< |
to about 10\%. |
| 57 |
> |
to about 10\%. |
| 58 |
> |
%{\color{red} Avi wants some statement about stability |
| 59 |
> |
%wrt changes in regions. I am not sure that we have done it and |
| 60 |
> |
%I am not sure it is necessary (Claudio).} |
| 61 |
|
|
| 62 |
|
\begin{table}[htb] |
| 63 |
|
\begin{center} |
| 66 |
|
\begin{tabular}{|l|c|c|c|c||c|} |
| 67 |
|
\hline |
| 68 |
|
Sample & A & B & C & D & AC/D \\ \hline |
| 69 |
< |
ttdil & 6.4 & 28.4 & 4.2 & 1.0 & 0.9 \\ |
| 70 |
< |
Zjets & 0.0 & 1.3 & 0.2 & 0.0 & 0.0 \\ |
| 71 |
< |
Other SM & 0.6 & 2.1 & 0.2 & 0.1 & 0.0 \\ \hline |
| 72 |
< |
total MC & 7.0 & 31.8 & 4.5 & 1.1 & 1.0 \\ \hline |
| 69 |
> |
ttdil & 6.9 & 28.6 & 4.2 & 1.0 & 1.0 \\ |
| 70 |
> |
Zjets & 0.0 & 1.3 & 0.1 & 0.1 & 0.0 \\ |
| 71 |
> |
Other SM & 0.5 & 2.0 & 0.1 & 0.1 & 0.0 \\ \hline |
| 72 |
> |
total MC & 7.4 & 31.9 & 4.4 & 1.2 & 1.0 \\ \hline |
| 73 |
|
\end{tabular} |
| 74 |
|
\end{center} |
| 75 |
|
\end{table} |
| 99 |
|
|
| 100 |
|
|
| 101 |
|
Monte Carlo studies give values of $K$ that are typically between 1.5 and 1.6, |
| 102 |
< |
depending on selection details. |
| 102 |
> |
depending on selection details. Given the integrated luminosity of the |
| 103 |
> |
present dataset, the determination of $K$ in data is severely statistics |
| 104 |
> |
limited. Thus, we take $K$ from $t\bar{t}$ Monte Carlo as |
| 105 |
> |
|
| 106 |
> |
\begin{center} |
| 107 |
> |
$ K_{MC} = \frac{\int_0^{\infty} {\cal N}(\met)~~d\met~}{\int_{50}^{\infty} {\cal N}(\met)~~d\met~}$ |
| 108 |
> |
\end{center} |
| 109 |
> |
|
| 110 |
> |
\noindent {\color{red} For the 11 pb result we have used $K$ from data.} |
| 111 |
|
|
| 112 |
|
There are several effects that spoil the correspondance between \met and |
| 113 |
|
$P_T(\ell\ell)$: |
| 143 |
|
\begin{center} |
| 144 |
|
\caption{\label{tab:victorybad} Test of the data driven method in Monte Carlo |
| 145 |
|
under different assumptions. See text for details.} |
| 146 |
< |
\begin{tabular}{|l|c|c|c|c|c|c|c|} |
| 146 |
> |
\begin{tabular}{|l|c|c|c|c|c|c|c|c|} |
| 147 |
|
\hline |
| 148 |
< |
& True $t\bar{t}$ dilepton & $t\to W\to\tau$& other SM & GEN or & Lepton $P_T$ & \met $>$ 50& obs/pred \\ |
| 149 |
< |
& included & included & included & RECOSIM & and $\eta$ cuts & & \\ \hline |
| 150 |
< |
1&Y & N & N & GEN & N & N & \\ |
| 151 |
< |
2&Y & N & N & GEN & Y & N & \\ |
| 152 |
< |
3&Y & N & N & GEN & Y & Y & \\ |
| 153 |
< |
4&Y & N & N & RECOSIM & Y & Y & \\ |
| 154 |
< |
5&Y & Y & N & RECOSIM & Y & Y & \\ |
| 155 |
< |
6&Y & Y & Y & RECOSIM & Y & Y & \\ |
| 148 |
> |
& True $t\bar{t}$ dilepton & $t\to W\to\tau$& other SM & GEN or & Lepton $P_T$ & Z veto & \met $>$ 50& obs/pred \\ |
| 149 |
> |
& included & included & included & RECOSIM & and $\eta$ cuts & & & \\ \hline |
| 150 |
> |
1&Y & N & N & GEN & N & N & N & 1.90 \\ |
| 151 |
> |
2&Y & N & N & GEN & Y & N & N & 1.64 \\ |
| 152 |
> |
3&Y & N & N & GEN & Y & Y & N & 1.59 \\ |
| 153 |
> |
4&Y & N & N & GEN & Y & Y & Y & 1.55 \\ |
| 154 |
> |
5&Y & N & N & RECOSIM & Y & Y & Y & 1.51 \\ |
| 155 |
> |
6&Y & Y & N & RECOSIM & Y & Y & Y & 1.58 \\ |
| 156 |
> |
7&Y & Y & Y & RECOSIM & Y & Y & Y & 1.18 \\ |
| 157 |
|
\hline |
| 158 |
|
\end{tabular} |
| 159 |
|
\end{center} |
| 165 |
|
cuts. We have verified that this effect is due to the polarization of |
| 166 |
|
the $W$ (we remove the polarization by reweighting the events and we get |
| 167 |
|
good agreement between prediction and observation). The kinematical |
| 168 |
< |
requirements (lines 2 and 3) do not have a significant additional effect. |
| 169 |
< |
Going from GEN to RECOSIM there is a significant change in observed/predicted. |
| 170 |
< |
We have tracked this down to the fact that tcMET underestimates the true \met |
| 171 |
< |
by $\approx 4\%$\footnote{We find that observed/predicted changes by roughly 0.1 |
| 172 |
< |
for each 1.5\% change in \met response.}. Finally, contamination from non $t\bar{t}$ |
| 168 |
> |
requirements (lines 2,3,4) compensate somewhat for the effect of W polarization. |
| 169 |
> |
Going from GEN to RECOSIM, the change in observed/predicted is small. |
| 170 |
> |
% We have tracked this down to the fact that tcMET underestimates the true \met |
| 171 |
> |
% by $\approx 4\%$\footnote{We find that observed/predicted changes by roughly 0.1 |
| 172 |
> |
%for each 1.5\% change in \met response.}. |
| 173 |
> |
Finally, contamination from non $t\bar{t}$ |
| 174 |
|
events can have a significant impact on the BG prediction. The changes between |
| 175 |
< |
lines 5 and 6 of Table~\ref{tab:victorybad} is driven by only {\color{red} 3} |
| 176 |
< |
Drell Yan events that pass the \met selection. |
| 175 |
> |
lines 6 and 7 of Table~\ref{tab:victorybad} is driven by 3 |
| 176 |
> |
Drell Yan events that pass the \met selection in Monte Carlo (thus the effect |
| 177 |
> |
is statistically not well quantified). |
| 178 |
|
|
| 179 |
|
An additional source of concern is that the CMS Madgraph $t\bar{t}$ MC does |
| 180 |
|
not include effects of spin correlations between the two top quarks. |
| 181 |
|
We have studied this effect at the generator level using Alpgen. We find |
| 182 |
< |
that the bias is a the few percent level. |
| 182 |
> |
that the bias is at the few percent level. |
| 183 |
|
|
| 184 |
|
Based on the results of Table~\ref{tab:victorybad}, we conclude that the |
| 185 |
< |
naive data driven background estimate based on $P_T{\ell\ell)}$ needs to |
| 186 |
< |
be corrected by a factor of {\color{red} $1.4 \pm 0.3$ (We need to |
| 187 |
< |
decide what this number should be)}. The quoted |
| 185 |
> |
naive data driven background estimate based on $P_T{(\ell\ell)}$ needs to |
| 186 |
> |
be corrected by a factor of {\color{red} $ K_{\rm{fudge}} =1.2 \pm 0.3$ |
| 187 |
> |
(We still need to settle on thie exact value of this. |
| 188 |
> |
For the 11 pb analysis it is taken as =1.)} . The quoted |
| 189 |
|
uncertainty is based on the stability of the Monte Carlo tests under |
| 190 |
|
variations of event selections, choices of \met algorithm, etc. |
| 191 |
+ |
For example, we find that observed/predicted changes by roughly 0.1 |
| 192 |
+ |
for each 1.5\% change in the average \met response. |
| 193 |
+ |
|
| 194 |
|
|
| 195 |
|
|
| 196 |
|
\subsection{Signal Contamination} |
| 197 |
|
\label{sec:sigcont} |
| 198 |
|
|
| 199 |
< |
All data-driven methods are principle subject to signal contaminations |
| 199 |
> |
All data-driven methods are in principle subject to signal contaminations |
| 200 |
|
in the control regions, and the methods described in |
| 201 |
|
Sections~\ref{sec:abcd} and~\ref{sec:victory} are not exceptions. |
| 202 |
|
Signal contamination tends to dilute the significance of a signal |
| 209 |
|
in the different control regions for the two methods. |
| 210 |
|
For example, in the extreme case of a |
| 211 |
|
new physics signal |
| 212 |
< |
with $P_T(\ell \ell) = \met$, an excess of ev ents would be seen |
| 212 |
> |
with $P_T(\ell \ell) = \met$, an excess of events would be seen |
| 213 |
|
in the ABCD method but not in the $P_T(\ell \ell)$ method. |
| 214 |
|
|
| 215 |
+ |
|
| 216 |
|
The LM points are benchmarks for SUSY analyses at CMS. The effects |
| 217 |
|
of signal contaminations for a couple such points are summarized |
| 218 |
|
in Table~\ref{tab:sigcontABCD} and~\ref{tab:sigcontPT}. |
| 229 |
|
are normalized to 30 pb$^{-1}$.} |
| 230 |
|
\begin{tabular}{|c||c|c||c|c|} |
| 231 |
|
\hline |
| 232 |
< |
SM & LM0 & BG Prediction & LM1 & BG Prediction \\ |
| 232 |
> |
SM & SM$+$LM0 & BG Prediction & SM$+$LM1 & BG Prediction \\ |
| 233 |
|
Background & Contribution& Including LM0 & Contribution & Including LM1 \\ \hline |
| 234 |
< |
x & x & x & x & x \\ |
| 234 |
> |
1.2 & 6.8 & 3.7 & 3.4 & 1.3 \\ |
| 235 |
|
\hline |
| 236 |
|
\end{tabular} |
| 237 |
|
\end{center} |
| 242 |
|
\caption{\label{tab:sigcontPT} Effects of signal contamination |
| 243 |
|
for the background predictions of the $P_T(\ell\ell)$ method including LM0 or |
| 244 |
|
LM1. Results |
| 245 |
< |
are normalized to 30 pb$^{-1}$.} |
| 245 |
> |
are normalized to 30 pb$^{-1}$.} |
| 246 |
|
\begin{tabular}{|c||c|c||c|c|} |
| 247 |
|
\hline |
| 248 |
< |
SM & LM0 & BG Prediction & LM1 & BG Prediction \\ |
| 248 |
> |
SM & SM$+$LM0 & BG Prediction & SM$+$LM1 & BG Prediction \\ |
| 249 |
|
Background & Contribution& Including LM0 & Contribution & Including LM1 \\ \hline |
| 250 |
< |
x & x & x & x & x \\ |
| 250 |
> |
1.2 & 6.8 & 2.2 & 3.4 & 1.5 \\ |
| 251 |
|
\hline |
| 252 |
|
\end{tabular} |
| 253 |
|
\end{center} |
