Weight of evidence and information value calculator
For automated credit decisions, check a characteristic for your scorecard. Enter its bins with goods and bads, or add raw data to see each bin’s WOE and information value in a minute.
Bin a file of raw rows
The file is read in this browser and never uploaded. Only the bin counts go into the link and, if you request it, the email.
Information value
Across 6 bins with 6,400 goods and 550 bads, the information value is 0.3627. The bin >=96 adds the most to it.
Weight of evidence and bad rate by bin
Bars show each bin’s weight of evidence, and the line shows its bad rate on the right-hand scale.
Bin results
| Bin | Goods | Bads | Share of goods | Share of bads | Bad rate | WOE | IV part |
|---|---|---|---|---|---|---|---|
| <6 | 520 | 110 | 8.1% | 20.0% | 17.5% | −0.901 | 0.1070 |
| [6,12) | 610 | 95 | 9.5% | 17.3% | 13.5% | −0.595 | 0.0460 |
| [12,24) | 980 | 120 | 15.3% | 21.8% | 10.9% | −0.354 | 0.0230 |
| [24,48) | 1,400 | 105 | 21.9% | 19.1% | 7.0% | 0.136 | 0.0038 |
| [48,96) | 1,650 | 80 | 25.8% | 14.5% | 4.6% | 0.572 | 0.0643 |
| >=96 | 1,240 | 40 | 19.4% | 7.3% | 3.1% | 0.980 | 0.1186 |
| Total | 6,400 | 550 | 100.0% | 100.0% | 7.9% | 0.3627 |
The bins paste into a decision table as two columns, with the bin as the condition and its WOE as the result.
Get the result by email
The scorecard is the next part of the credit decision, with its bins as a table and each change tested on past applications. Start free
Show formulas
Share of goods = Goods ÷ All goods Share of bads = Bads ÷ All bads WOE = ln(Share of goods ÷ Share of bads) IV part = (Share of goods − Share of bads) × WOE Information value = Σ IV part A bin with no goods or no bads gets 0.5 added to both its counts before the shares are taken.
This looks at one characteristic on its own. It says nothing about how characteristics overlap in a scorecard, and the strength bands are a rule of thumb, not a test.
Calculate weight of evidence and information value
Weight of evidence shows how the goods and bads are distributed in each bin of a characteristic. Information value adds up each bin’s contribution, calculated from its share of goods, share of bads and weight of evidence.
These measures help you review a characteristic for a scorecard. Use the bin results to compare how goods and bads are distributed, and check whether the bad rate rises or falls across the bins.
Common questions
How do I calculate weight of evidence and information value?
For each bin, calculate the share of all goods and the share of all bads. Weight of evidence is the natural logarithm of the goods share divided by the bads share, and information value is the sum of each bin’s contribution.
What does information value tell me about a scorecard characteristic?
It sums the contributions from each bin, based on the difference between its goods and bads shares and its weight of evidence. The calculator shows its strength using a common rule of thumb.
How should I bin a characteristic for a scorecard?
Enter bins with their goods and bads, or provide raw rows and choose the value and good or bad columns. The calculator groups numbers into ranges and text values into bins.
What should I do with a bin that has no bads?
The calculator adds 0.5 to both the goods and bads counts before calculating the shares. This keeps the bin’s weight of evidence finite.
Why should the bad rate move one way across bins?
The calculator checks whether the bad rate rises or falls steadily across the bins, or turns. That pattern is one part of reviewing a characteristic, alongside its weight of evidence and information value.