Comparing results between rows based on column value

[OceanView]

I am collecting data from many different runs of software and need to do comparisons between them.

I need to know the average difference between the Run 2 or 3 as compared to base.  Expressed as an equation:

Example average (Case A = (97 - 50), Case B =(84-60), Case C = (88-65), Case D = (67 - 40) )= avg (47, 24,27)

Case	Run	Value
A	base	50
A	2	98
A	3	97
B	base	60
B	2	78
B	3	83
C	base	65
C	2	83
C	3	88
D	base	40
D	2	56
D	3	67

I am currently reading the data a second time and creating a second table linked by Case with the Value given a different name and when statement that only loads Run=base.  This seems very inefficient as I might want to compare Run 2 to Run 3. 

Is there a better way to structure the data?  Or, a better way to create a chart or pivot table with set analysis? Should I break it into 3 loads where I use the where statement to separate each Run into their own table?

Side note, since I keep getting asked this: I don't want (sum(run(3)) - sum(run(base))/sum(run((base)))

[Saravanan_Desingh]

Try this,

tab1:
LOAD * INLINE [
    Case, Run, Value
    A, base, 50
    A, 2, 98
    A, 3, 97
    B, base, 60
    B, 2, 78
    B, 3, 83
    C , base, 65
    C , 2, 83
    C , 3, 88
    D , base, 40
    D , 2, 56
    D , 3, 67
];

Left Join(tab1)
LOAD Case, Evaluate(FirstSortedValue(Value,Run=3)&'-'&FirstSortedValue(Value,Run='base')) As Key
Resident tab1
Group By Case;

[Saravanan_Desingh]

Output:

[OceanView]

Thank you for the fast response, and your suggestion does work for this specific case.  I was hoping there could be something done in the charts/pivot tables.

I regularly have multiple Values, which makes your solution more challenging than the one I am currently using:

load Case,

...

From [file] ();

Load Case,

      Value 1 as Value1_base,

      Value2 as Value2_base,

      Value3 as Value3_base

From [file] (...) where Run = 'base';

This allows me to attach all of the base Values in one extra load.