First, we thank our three reviewers.

The paper is an empirical study arising from three questions (in the abstract
and introduction).

There was a question about the "Knuth-shuffle" not being defined or explained. There is a citation
to it [2], i.e. CACM algorithm 235. It is also called the "Fisher-Yates shuffle" circa 1938. We will 
add text to better explain this.

"... summarize main findings" 
We believe this has been done in the conclusion.

"Other ways to morph" and "originality of Algorithms 1 and 2". 
We make no claims for novelty (see references [5] and [8]), but describe algorithms in the interests 
of reproducibility. We use another technique for large graphs [1]

"In Figure 2 ... on the right it only shows model A. Why?" 
On the left we show  contours for two different models with fixed n. On the right we have one model and vary n. 
If we had models A and B and vary n, the graphs becomes too cluttered and the observations not so obvious. We 
will add text to clarify this.

"... Figure 3 is surprising. Please, try to justify ...". Indeed, it is surprising! We have, so
far failed to come up with an explanation for this. Nevertheless, we believe that this phenomena is interesting 
enough that it should be reported even though we cannot explain it (but have given some report of what it is 
NOT due to, page 6, first paragraph).

First paragraph, page 6, "For even n all instances are stable for p = 0 ...". This should read 
"For all instances ...". When p = 0 it does not matter if n is odd or even as there will be a 
stable matching where all agents are happy to be alone.

"... paper should be written ... on black and white printed ...". We will address this and we apologise 
to the reader.