My diagram is too big.

- The problem is the red line, which is going below 50.
In some old books I found a nice work around for this. Their the area between the main part and the outlier is left out, and replaced with a zickzack line dividing the diagram.
Can someone help me doing something like that? Maybe it is even possible to show the outlier in a magnifying glass like here!
- My second problem is the x-scale which should be higher, below the last point at ~20.
My code so far:
\documentclass{article}
\usepackage{tikz}
\begin{document}
\begin{tikzpicture}[scale=.6, xscale=0.65, yscale=0.25]
% grid
\draw[gray!40, thin, step=5, dotted] (0,50) grid (24,115); % Grid
% axis
\draw[->] (0,0) -- (25,0);
\draw[->] (0,0) -- (0,116);
% scale
\foreach \x in {0,2,...,24} \draw (\x,0.05) -- (\x,-0.05) node[below] {\tiny\x};
\foreach \y in {0,10,...,115} \draw (-0.05,\y) -- (0.05,\y) node[left] {\tiny\y};
% lines
\draw[blue!80!black,opacity=0.6, dashed] (1,76.11) -- (2,114.29) -- (3,62.73) -- (4,65.75) -- (5,79.09) -- (6,76.11) -- (7,95.62) -- (8,93.83) -- (9,74.9) -- (10,85.81) -- (11,93.7) -- (12,79.75) -- (13,79.14) -- (14,87.52) -- (15,83.71) -- (16,76.74) -- (17,85.76) -- (18,81.35) -- (19,69.59) -- (20,81.74) -- (21,68.7) -- (22,74.21) -- (23,70.17) -- (24,83.33);
\draw[red!80!black] (1,55.56) -- (2,76.02) -- (3,87.25) -- (4,96.97) -- (5,90.28) -- (6,88.64) -- (7,78.28) -- (8,80.39) -- (9,77.01) -- (10,78.13) -- (11,80.91) -- (12,85.42) -- (13,74.55) -- (14,79.91) -- (15,78.78) -- (16,77.31) -- (17,84.75) -- (18,95.84) -- (19,83.94) -- (20,86.79) -- (21,76.44) -- (22,47.76) -- (23,66.06) -- (24,24.06);
\end{tikzpicture}
\end{document}
By the way, I really had fun drawing the picture! Be happy as well :)


pgfplotsinstead of doing them directly intikz. Also, please convert your code snippet into a fully compilable MWE including\documentclassand the appropriate packages that sets up the problem. While solving problems can be fun, setting them up is not. Then, those trying to help can simply cut and paste your MWE and get started on solving the problem. – Peter Grill Oct 29 '14 at 19:55