ERRATA CORRIGE







Someone write me few comments on counting fully matched patterns. Following his suggestions, I found a mistake in my reasoning. Indeed, the 5 choices, rejected as an average to assemble the 7 Tangram pieces, are mainly due to matching onto shared sides. But I have already taken into account this fact in my counting! For instance, in the third step of the example I have explicitly subtracted the two shared sides from the number of possible choices on the two assembled pieces. As a matter of fact, I considered twice that phenomenon. Therefore, the division by 25 has not to be done. On the other hand, considering at each step the free sides of the settled pieces introduces multiple counting of the same pattern. To properly take into account this effect, I have considered the tree structure of matching. So doing, I found a lower bound of nearly 70 millions. Interested people can ask me the proof. I am not going to change the current title of the site, but it should be now

FEW TENS OF MILLIONS OF TANGRAM SILHOUETTES.

Prova