Skip to content
About
Main content
Law

Digest collisions (the birthday bound)

At ten billion rows a 64-bit collision is not a risk. It is the expected result.

Rows
rows hashed
Digest bits
bits in the digest

Recurring principle: Independence is assumed

How it behaves

One input on the horizontal axis, the others held fixed.

64-bit digest96-bit digest
Data table
Chance of at least one collision, by rows hashed
Rows nseriesCollision probability
1M64-bit digest0%
1.19M64-bit digest0%
1.41M64-bit digest0%
1.68M64-bit digest0%
2M64-bit digest0%
2.37M64-bit digest0%
2.82M64-bit digest0%
3.35M64-bit digest0%
3.98M64-bit digest0%
4.73M64-bit digest0%
5.62M64-bit digest0%
6.68M64-bit digest0%
7.94M64-bit digest0%
9.44M64-bit digest0%
11.2M64-bit digest0%
13.3M64-bit digest0%
15.8M64-bit digest0.001%
18.8M64-bit digest0.001%
22.4M64-bit digest0.001%
26.6M64-bit digest0.002%
31.6M64-bit digest0.003%
37.6M64-bit digest0.004%
44.7M64-bit digest0.005%
53.1M64-bit digest0.008%
63.1M64-bit digest0.011%
75M64-bit digest0.015%
89.1M64-bit digest0.022%
106M64-bit digest0.03%
126M64-bit digest0.043%
150M64-bit digest0.061%
178M64-bit digest0.086%
211M64-bit digest0.121%
251M64-bit digest0.171%
299M64-bit digest0.241%
355M64-bit digest0.341%
422M64-bit digest0.481%
501M64-bit digest0.679%
596M64-bit digest0.957%
708M64-bit digest1.35%
841M64-bit digest1.9%
1B64-bit digest2.67%
1.19B64-bit digest3.76%
1.41B64-bit digest5.26%
1.68B64-bit digest7.35%
2B64-bit digest10.2%
2.37B64-bit digest14.1%
2.82B64-bit digest19.4%
3.35B64-bit digest26.2%
3.98B64-bit digest34.9%
4.73B64-bit digest45.5%
5.62B64-bit digest57.6%
6.68B64-bit digest70.2%
7.94B64-bit digest81.9%
9.44B64-bit digest91.1%
11.2B64-bit digest96.7%
13.3B64-bit digest99.2%
15.8B64-bit digest99.9%
18.8B64-bit digest100%
22.4B64-bit digest100%
26.6B64-bit digest100%
31.6B64-bit digest100%
37.6B64-bit digest100%
44.7B64-bit digest100%
53.1B64-bit digest100%
63.1B64-bit digest100%
75B64-bit digest100%
89.1B64-bit digest100%
106B64-bit digest100%
126B64-bit digest100%
150B64-bit digest100%
178B64-bit digest100%
211B64-bit digest100%
251B64-bit digest100%
299B64-bit digest100%
355B64-bit digest100%
422B64-bit digest100%
501B64-bit digest100%
596B64-bit digest100%
708B64-bit digest100%
841B64-bit digest100%
1,000B64-bit digest100%
3.35M96-bit digest0%
3.98M96-bit digest0%
4.73M96-bit digest0%
5.62M96-bit digest0%
6.68M96-bit digest0%
7.94M96-bit digest0%
9.44M96-bit digest0%
11.2M96-bit digest0%
13.3M96-bit digest0%
15.8M96-bit digest0%
18.8M96-bit digest0%
22.4M96-bit digest0%
26.6M96-bit digest0%
31.6M96-bit digest0%
37.6M96-bit digest0%
44.7M96-bit digest0%
53.1M96-bit digest0%
63.1M96-bit digest0%
75M96-bit digest0%
89.1M96-bit digest0%
106M96-bit digest0%
126M96-bit digest0%
150M96-bit digest0%
178M96-bit digest0%
211M96-bit digest0%
251M96-bit digest0%
299M96-bit digest0%
355M96-bit digest0%
422M96-bit digest0%
501M96-bit digest0%
596M96-bit digest0%
708M96-bit digest0%
841M96-bit digest0%
1B96-bit digest0%
1.19B96-bit digest0%
1.41B96-bit digest0%
1.68B96-bit digest0%
2B96-bit digest0%
2.37B96-bit digest0%
2.82B96-bit digest0%
3.35B96-bit digest0%
3.98B96-bit digest0%
4.73B96-bit digest0%
5.62B96-bit digest0%
6.68B96-bit digest0%
7.94B96-bit digest0%
9.44B96-bit digest0%
11.2B96-bit digest0%
13.3B96-bit digest0%
15.8B96-bit digest0%
18.8B96-bit digest0%
22.4B96-bit digest0%
26.6B96-bit digest0%
31.6B96-bit digest0%
37.6B96-bit digest0%
44.7B96-bit digest0%
53.1B96-bit digest0%
63.1B96-bit digest0%
75B96-bit digest0%
89.1B96-bit digest0%
106B96-bit digest0%
126B96-bit digest0%
150B96-bit digest0%
178B96-bit digest0%
211B96-bit digest0%
251B96-bit digest0%
299B96-bit digest0%
355B96-bit digest0%
422B96-bit digest0%
501B96-bit digest0%
596B96-bit digest0%
708B96-bit digest0%
841B96-bit digest0%
1,000B96-bit digest0.001%

Worked examples, recomputed

Every number the mechanisms quote for this relationship is recomputed from it, so none of them disagrees with it.

Where it appears

Each mechanism states it in its own names. The general symbol each name stands for is shown with it.

  • Query Processing

    Why can one question be answered by plans whose costs differ by a factor of ten thousand, and what does each operator force the engine to do?

    Read it in place →