Presentation is loading. Please wait.

Presentation is loading. Please wait.

0.27% expected error New predictive rational method.

Similar presentations


Presentation on theme: "0.27% expected error New predictive rational method."— Presentation transcript:

1 0.27% expected error New predictive rational method

2 2 Who needs: everyone, especially politicians, financiers, entrepreneurs, CEO, executives and sales managers. Why it is useful: for any kind of forecast, from Stock Exchange to Sales In writing the book Prevedere per Decidere, Marco Galleri considered the general and particular aspects of rational predictive methods and he conceived the Crowdshang idea, where the binary surveys, previously submitted to thousands of people, are perfected by a group of trained experts. The idea is quite simple: the worth of the Shang method is greatly enhanced by the wisdom of a huge crowd; the common sense of the analysts and the coordinator plays any differences that could improve this method. On the other hand: in a huge crowd, you can find large numbers of specialists and thus obtain correct surveys. MARCO GALLERI Teoria e Pratica d’Azienda. www.marcogalleri.it IN SHORT

3 A large crowd responds to repeated binary questions. Average Estimate An experienced group in face-to-face interviews verifies the estimate and perfections it. Estimate completed by the group Average historical error 2.5% Average expected error 0.9% Average expected error 0.3% Free estimate of the crowd CROWDSHANG Greater or less than a central value? MARCO GALLERI Teoria e Pratica d’Azienda. www.marcogalleri.it CROWDSHANG METHOD

4 Free estimate of the crowd: Central value 0 Average Value Ask for the minimum (m) and the maximum (M) By averaging you obtain: the minimum 1 (m1), the maximum 1 (M1), and the central value (C1) Ask if the value is greater or less than C1 (or before or after) If the most frequent (modal) response is… Greater than C1: C1 it becomes m2 and we average C1 and M1 to determine C2 Lower than C1: C1 becomes M2 and we average c1 and m1 to determine C2 Ask if greater or less than C2 (as above C3 is obtained) 1000 m1 = 980 ; M1 = 1060; C1 = 1020 Major: m2 = 1020 C2 = 1040 Major: m2 = 1020 C2 = 1040 Minor: M2 = 1020 C2 = 1000 Minor: M2 = 1020 C2 = 1000 NUMERICAL EXAMPLE Once C4 is obtained, it is proposed, as an average, to the expert group (C4 … C7) C7 concludes the estimation process You can determine the range of change also with: max of the max, minimum of the minimum, minimum of the max, average of the max, average of the minimum, median of the max, median of the minimum Summary The free estimate of the crowd is based on the Wisdom of Crowds and produces an historical average error of 2.5%. After three of his consultations with the Shang method the expected error is 0.9% The group adopts the same procedure with the Pfizer method : the expected error ( C7 ) is 0.3%

5 MARCO GALLERI strategia organizzazione comunicazione marketing MARCO GALLERI strategia organizzazione comunicazione marketing Il Poggio 58036 Sassofortino (GR) tel. & fax 0564.567.118 mobile 333.2456.338 www.marcogalleri.it marco@marcogalleri.it For more information: www.crowdshang.com www.marcogalleri.it 5


Download ppt "0.27% expected error New predictive rational method."

Similar presentations


Ads by Google