The Third Bulgaria Bus Bombing Experiment
Our third experiment used the same protocol as the first experiment
except that Bulgaria, בולגריה
had the alternative spelling with a Hey. The best table is shown below. It uses the full
spellilng of Iran in the From Iran keyword and has a p-value of 28.5/100,000.
This is the most compact table of the third experiment resulting when the expected number of ELSs for the axis key word was set to 100
and for the other ELSs was set to 800. The automatically determined cylinder size is 941.
The probability that a table this good would arise from
the ELS random placement monkey text population is 28.5/100,000. Click on the table to see a full
resolution image of the table.
Best table using axis protocol and the key word Bulgaria with the spelling
בולגריה.
The next best table is similar to the best table with the key word From Iran replaced by Iran.
Its p-value is 36.5/100,000.
This is the second most compact table of the third experiment resulting when the expected number of ELSs for the axis key word was set to 100
and for the other ELSs was set to 800. The automatically determined cylinder size is 941.
The probability that a table this good would arise from
the ELS random placement monkey text population is 36.5/100,000. Click on the table to see a full
resolution image of the table.
Second best table using axis protocol and the key word Bulgaria with the spelling
בולגריה.
The third best table is also similar to the first best table except that it the spelling of the
key word From Iran without the Aleph.
Its p-value is 39.5.5/100,000.
This is the third most compact table of the first experiment resulting when the expected number of ELSs for the axis key word was set to 100
and for the other ELSs was set to 800. The automatically determined cylinder size is 941.
The probability that a table this good would arise from
the ELS random placement monkey text population is 39.5/100,000. Click on the table to see a full
resolution image of the table.
Third best table using axis protocol and the key word Bulgaria with the spelling
בולגריה.
The fourth best table is similar to the first best table
only omitting the keyword (5)772.
Its p-value is 43.5/100,000.
This is the fourth most compact table of the third experiment resulting when the expected number of ELSs for the axis key word was set to 100
and for the other ELSs was set to 800. The automatically determined cylinder size is 941.
The probability that a table this good would arise from
the ELS random placement monkey text population is 43.5/100,000. Click on the table to see a full
resolution image of the table.
Fourth best table using axis protocol and the key word Bulgaria with the spelling
בולגריה.
The fifth best table is like the third best table with omissio of the key word (5)772.
Its p-value is 50.5/100,000.
This is the fifth best most compact table of the third experiment resulting when the expected number of ELSs for the axis key word was set to 100
and for the other ELSs was set to 800. The automatically determined cylinder size is 941.
The probability that a table this good would arise from
the ELS random placement monkey text population is 43.5/100,000. Click on the table to see a full
resolution image of the table.
Fifth best table using axis protocol and the key word Bulgaria with the spelling
בולגריה.
There are three more tables like these last with p-values less than 1/1000 that we do not show.
The Fourth Bulgaria Bus Bombing Experiment
Our fourth experiment uses the same key words as the third experiment, but we use In Burgas as the axis
key word. Key word sets had the possibility of containing the key word Bulgaria or not including
Bulgaria. With our protocol, any ELS of the axis key word must have a row skip of at least 1
and no more than 5 and a column absolute skip of no more than 5. So the ELS for In Bulgaria
that appears in the tables of the third experiment cannot appear in the tables of the fourth experiment
(since it had an row skip of 10). As a result the best table had a p-value of 30/10,000 and as we are showing
only tables with p-values no more than 1/1000 we are not showing any of the tables of the fourth experiment.
