>Assuming a range of dates, such as:
>
>9/26/2000
>2/13/2001
>9/11/2001
>2/12/2002
>9/16/2002
>2/3/2003
>2/4/2003
>9/29/2003
>9/30/2003
>2/24/2004
>2/25/2004
>6/7/2004
>6/8/2004
>10/4/2004
>10/5/2004
>12/22/2004
>2/15/2005
>2/16/2005
>
>Anyone have an algorithym that might predict future recurring events?

Oh, so you give us a RANGE of dates, and ask us to PREDICT THE FUTURE?? Not asking for much, are you? Now, if you'd given us a DATE SERIES and asked us to EXTRAPOLATE we might have been able to do something ;)

All kidding aside, it doesn't look very predictable:

Date       Days Since Last One
20000926        N/A
20010213        140
20010911        210
20020212        154
20020916        216
20030203        140
20030204          1
20030929        237
20030930          1
20040224        147
20040225          1
20040607        103
20040608          1
20041004        118
20041005          1
20041222         78
20050215         55
20050216          1

No doubt you could run a best-fit least-squares line through this, FWIW. You might also be able to calculate the chance of the next event happening the very next day (again, FWIW).