The distribution of hash function outputs
For ‘real-world’ inputs
What this page is
This page has a number of tables with the results of a test
on several well-known hash functions. It accompanies
this article.
There's a summary at the bottom if
you're just interested in seeing which function scores ‘best’
Longer explanation
In a set of values that are uniformly distributed over the
total output space, we can state that one half of the
output values should have 1 as its first bit. One half of
those values should have 1 as its second bit as well,
and again one half should have 1 as its third.
A hash function should, insofar possible, generate for any
set of inputs, a set of outputs that is uniformly distributed
over its output space. If we count the frequency of the
number of '1' bits in its outputs, we should get a nice,
clear binomial distribution
In theory, there is no way a hash function can
provide a uniformly distributed set of outputs for
any set of inputs. No hash can be perfect. But how
do they fare on large, everyday data sets?
This page
has tests the uniformity of the outputs of several common
hash functions, applied to text files containing a large set
of ‘things’.
This property is interesting for Flajolet-Martin (PDF) sketches; as explained here, it can be used to make a set of inputs ‘countable’
How to read these tables
Every column states, for each hash function, the frequency
of the number of values starting with (at least) n
‘1’ bits. In between parentheses is the p–value
resulting from the binomial
test.
Interpretation of this last value is up to you, but for convenience,
p < 0.20 is highlighted in pink, and p < 0.05
is highlighted in red.
Remember that none of these hashes are perfect. Some of them are just even more imperfect than others.
Caution
Even though most of the hashes used in this test are commonly used in cryptography, this test uses a property of hash functions that is only tangentially related to security issues. Don't draw any conclusion about the security or insecurity of these functions based on these tables.
Dataset 1: the four million address blacklist
This list was collected by Cor Gest, who blacklists people for looking at his mail server in a funny way. It contains about four million unique ip and email addresses in ASCII format.
| 1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 | 9 | 10 | 11 | 12 | 13 | 14 | 15 | 16 | 17 | 18 | 19 | 20 | 21 | 22 | |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Passthrough | 0 (0.000) | 0 (0.000) | 0 (0.000) | 0 (0.000) | 0 (0.000) | 0 (0.000) | 0 (0.000) | 0 (0.000) | 0 (0.000) | 0 (0.000) | 0 (0.000) | 0 (0.000) | 0 (0.000) | 0 (0.000) | 0 (0.000) | 0 (0.000) | 0 (0.000) | 0 (0.000) | 0 (0.001) | 0 (0.038) | 0 (0.275) | 0 (1.000) |
| Adler32 | 2009900 (0.264) | 1003713 (0.039) | 502641 (0.864) | 251309 (0.889) | 125701 (0.971) | 47157 (0.000) | 19945 (0.000) | 7653 (0.000) | 4 (0.000) | 0 (0.000) | 0 (0.000) | 0 (0.000) | 0 (0.000) | 0 (0.000) | 0 (0.000) | 0 (0.000) | 0 (0.000) | 0 (0.000) | 0 (0.001) | 0 (0.038) | 0 (0.275) | 0 (1.000) |
| CRC32 | 2011349 (0.743) | 1004816 (0.425) | 501638 (0.092) | 250318 (0.029) | 125254 (0.214) | 62635 (0.402) | 31289 (0.453) | 15791 (0.522) | 7821 (0.701) | 3961 (0.593) | 1998 (0.436) | 981 (1.000) | 520 (0.191) | 252 (0.678) | 127 (0.684) | 60 (0.949) | 30 (1.000) | 13 (0.700) | 5 (0.467) | 1 (0.199) | 1 (1.000) | 1 (0.617) |
| MD5 | 2010200 (0.414) | 1003929 (0.069) | 502061 (0.296) | 250454 (0.057) | 125164 (0.133) | 62490 (0.155) | 31251 (0.334) | 15506 (0.101) | 7720 (0.127) | 3836 (0.144) | 1955 (0.857) | 977 (0.898) | 456 (0.119) | 237 (0.610) | 115 (0.527) | 62 (0.898) | 26 (0.469) | 13 (0.700) | 4 (0.273) | 3 (1.000) | 1 (1.000) | 0 (1.000) |
| SHA224 | 2012108 (0.278) | 1007283 (0.041) | 503842 (0.101) | 252319 (0.053) | 126287 (0.087) | 63311 (0.061) | 31652 (0.193) | 15821 (0.379) | 7882 (0.765) | 3915 (0.848) | 1937 (0.557) | 930 (0.100) | 455 (0.109) | 213 (0.038) | 120 (0.857) | 55 (0.482) | 25 (0.365) | 12 (0.521) | 3 (0.102) | 2 (0.602) | 2 (0.718) | 1 (0.617) |
| SHA512 | 2008519 (0.013) | 1004287 (0.159) | 501354 (0.035) | 251028 (0.472) | 125413 (0.431) | 62535 (0.214) | 31237 (0.296) | 15508 (0.105) | 7713 (0.109) | 3877 (0.425) | 1937 (0.557) | 1014 (0.307) | 515 (0.279) | 257 (0.463) | 127 (0.684) | 71 (0.225) | 31 (0.928) | 14 (0.898) | 7 (1.000) | 2 (0.602) | 2 (0.718) | 2 (0.249) |
| Whirlpool | 2011387 (0.714) | 1005639 (0.882) | 503111 (0.591) | 251730 (0.468) | 125905 (0.535) | 62890 (0.853) | 31328 (0.596) | 15885 (0.164) | 7874 (0.834) | 3993 (0.296) | 2000 (0.410) | 1006 (0.444) | 500 (0.685) | 251 (0.725) | 120 (0.857) | 62 (0.898) | 30 (1.000) | 16 (0.798) | 7 (1.000) | 2 (0.602) | 1 (1.000) | 0 (1.000) |
| ECDSA+SHA1 | 2011376 (0.723) | 1004509 (0.249) | 502106 (0.328) | 250887 (0.313) | 125841 (0.662) | 63215 (0.136) | 31666 (0.168) | 15857 (0.243) | 7889 (0.705) | 3925 (0.975) | 1947 (0.718) | 978 (0.924) | 462 (0.198) | 224 (0.180) | 127 (0.684) | 67 (0.444) | 37 (0.240) | 23 (0.055) | 8 (0.855) | 4 (0.797) | 1 (1.000) | 0 (1.000) |
| MD4 | 2011580 (0.577) | 1005899 (0.654) | 503254 (0.452) | 251404 (0.956) | 126127 (0.209) | 63305 (0.064) | 31736 (0.076) | 15774 (0.615) | 7892 (0.680) | 3937 (0.879) | 1995 (0.477) | 988 (0.836) | 472 (0.404) | 237 (0.610) | 113 (0.416) | 59 (0.848) | 28 (0.718) | 15 (1.000) | 7 (1.000) | 2 (0.602) | 2 (0.718) | 2 (0.249) |
| SHA1 | 2011376 (0.723) | 1004509 (0.249) | 502106 (0.328) | 250887 (0.313) | 125841 (0.662) | 63215 (0.136) | 31666 (0.168) | 15857 (0.243) | 7889 (0.705) | 3925 (0.975) | 1947 (0.718) | 978 (0.924) | 462 (0.198) | 224 (0.180) | 127 (0.684) | 67 (0.444) | 37 (0.240) | 23 (0.055) | 8 (0.855) | 4 (0.797) | 1 (1.000) | 0 (1.000) |
Dataset 2: cracklib
About 50.000 common English words and proper names, in ASCII.
| 1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 | 9 | 10 | 11 | 12 | 13 | 14 | 15 | 16 | 17 | 18 | |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Passthrough | 0 (0.000) | 0 (0.000) | 0 (0.000) | 0 (0.000) | 0 (0.000) | 0 (0.000) | 0 (0.000) | 0 (0.000) | 0 (0.000) | 0 (0.000) | 0 (0.000) | 0 (0.000) | 0 (0.003) | 0 (0.086) | 0 (0.419) | 0 (1.000) | 0 (1.000) | 0 (1.000) |
| Adler32 | 26297 (0.271) | 13115 (0.332) | 6566 (0.603) | 3261 (0.456) | 1630 (0.600) | 844 (0.516) | 465 (0.011) | 235 (0.051) | 28 (0.000) | 19 (0.000) | 0 (0.000) | 0 (0.000) | 0 (0.003) | 0 (0.086) | 0 (0.419) | 0 (1.000) | 0 (1.000) | 0 (1.000) |
| CRC32 | 26262 (0.160) | 13221 (0.928) | 6623 (0.823) | 3342 (0.483) | 1618 (0.409) | 803 (0.440) | 386 (0.190) | 181 (0.081) | 86 (0.094) | 43 (0.264) | 24 (0.844) | 14 (0.676) | 6 (1.000) | 3 (1.000) | 2 (0.679) | 0 (1.000) | 0 (1.000) | 0 (1.000) |
| MD5 | 26536 (0.332) | 13402 (0.057) | 6777 (0.025) | 3466 (0.004) | 1703 (0.198) | 877 (0.074) | 441 (0.166) | 224 (0.222) | 123 (0.055) | 56 (0.530) | 31 (0.279) | 14 (0.676) | 6 (1.000) | 3 (1.000) | 1 (1.000) | 1 (0.554) | 1 (0.332) | 1 (0.183) |
| SHA224 | 26445 (0.858) | 13141 (0.479) | 6583 (0.767) | 3278 (0.660) | 1638 (0.745) | 795 (0.293) | 401 (0.587) | 205 (0.972) | 103 (1.000) | 53 (0.834) | 25 (1.000) | 8 (0.209) | 4 (0.432) | 1 (0.392) | 1 (1.000) | 0 (1.000) | 0 (1.000) | 0 (1.000) |
| SHA512 | 26567 (0.215) | 13321 (0.274) | 6656 (0.511) | 3303 (1.000) | 1630 (0.600) | 824 (0.972) | 431 (0.374) | 213 (0.625) | 109 (0.554) | 54 (0.727) | 24 (0.844) | 14 (0.676) | 5 (0.842) | 1 (0.392) | 1 (1.000) | 0 (1.000) | 0 (1.000) | 0 (1.000) |
| Whirlpool | 26499 (0.517) | 13215 (0.976) | 6627 (0.782) | 3308 (0.928) | 1647 (0.920) | 831 (0.847) | 424 (0.570) | 208 (0.889) | 98 (0.657) | 44 (0.329) | 24 (0.844) | 10 (0.575) | 6 (1.000) | 2 (0.778) | 1 (1.000) | 1 (0.554) | 0 (1.000) | 0 (1.000) |
| ECDSA+SHA1 | 26296 (0.267) | 13152 (0.550) | 6552 (0.482) | 3257 (0.414) | 1609 (0.294) | 806 (0.505) | 376 (0.071) | 189 (0.236) | 79 (0.016) | 40 (0.109) | 19 (0.200) | 13 (0.889) | 8 (0.549) | 3 (1.000) | 1 (1.000) | 0 (1.000) | 0 (1.000) | 0 (1.000) |
| MD4 | 26566 (0.218) | 13390 (0.075) | 6675 (0.364) | 3415 (0.045) | 1705 (0.181) | 878 (0.068) | 443 (0.138) | 223 (0.250) | 117 (0.183) | 61 (0.186) | 35 (0.076) | 19 (0.093) | 6 (1.000) | 3 (1.000) | 2 (0.679) | 1 (0.554) | 0 (1.000) | 0 (1.000) |
| SHA1 | 26296 (0.267) | 13152 (0.550) | 6552 (0.482) | 3257 (0.414) | 1609 (0.294) | 806 (0.505) | 376 (0.071) | 189 (0.236) | 79 (0.016) | 40 (0.109) | 19 (0.200) | 13 (0.889) | 8 (0.549) | 3 (1.000) | 1 (1.000) | 0 (1.000) | 0 (1.000) | 0 (1.000) |
Dataset 3: A big Apache log file
This is an Apache error log file, sifted through sort(1) and uniq(1) to obtain only unique lines. This still left about 50 megabytes worth of log data. Since most lines are very similar (i.e. vary only in the address or time stamp logged), this is probably the most stringent test employed here.
| 1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 | 9 | 10 | 11 | 12 | 13 | 14 | 15 | 16 | 17 | 18 | |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Passthrough | 0 (0.000) | 0 (0.000) | 0 (0.000) | 0 (0.000) | 0 (0.000) | 0 (0.000) | 0 (0.000) | 0 (0.000) | 0 (0.000) | 0 (0.000) | 0 (0.000) | 0 (0.000) | 0 (0.000) | 0 (0.000) | 0 (0.000) | 0 (0.006) | 0 (0.129) | 0 (0.413) |
| Adler32 | 190619 (0.021) | 95269 (0.236) | 47554 (0.704) | 23666 (0.632) | 11845 (0.827) | 5647 (0.000) | 2738 (0.000) | 1180 (0.000) | 712 (0.278) | 346 (0.203) | 276 (0.000) | 205 (0.000) | 193 (0.000) | 5 (0.000) | 0 (0.000) | 0 (0.006) | 0 (0.129) | 0 (0.413) |
| CRC32 | 189662 (0.430) | 94960 (0.979) | 47463 (0.949) | 23688 (0.740) | 11859 (0.929) | 5870 (0.402) | 2925 (0.444) | 1488 (0.907) | 772 (0.270) | 378 (0.697) | 189 (0.769) | 91 (0.917) | 45 (0.941) | 23 (1.000) | 10 (0.769) | 6 (0.834) | 4 (0.545) | 2 (0.660) |
| MD5 | 189158 (0.015) | 94875 (0.772) | 47284 (0.346) | 23673 (0.665) | 11968 (0.356) | 6015 (0.292) | 3063 (0.078) | 1515 (0.413) | 759 (0.520) | 387 (0.406) | 199 (0.321) | 100 (0.436) | 63 (0.018) | 29 (0.213) | 20 (0.026) | 10 (0.091) | 4 (0.545) | 4 (0.059) |
| SHA224 | 189971 (0.833) | 94800 (0.568) | 47259 (0.287) | 23651 (0.562) | 11707 (0.132) | 5776 (0.038) | 2881 (0.113) | 1433 (0.193) | 712 (0.278) | 359 (0.568) | 176 (0.509) | 83 (0.349) | 43 (0.713) | 22 (0.917) | 12 (0.882) | 3 (0.399) | 3 (0.768) | 1 (1.000) |
| SHA512 | 189887 (0.953) | 94949 (0.991) | 47322 (0.450) | 23538 (0.180) | 11757 (0.298) | 5878 (0.464) | 2972 (0.927) | 1460 (0.550) | 753 (0.673) | 370 (1.000) | 189 (0.769) | 96 (0.716) | 47 (0.883) | 19 (0.466) | 7 (0.236) | 5 (1.000) | 3 (0.768) | 1 (1.000) |
| Whirlpool | 189189 (0.020) | 94544 (0.126) | 47158 (0.119) | 23577 (0.282) | 11731 (0.200) | 5810 (0.105) | 2953 (0.804) | 1455 (0.466) | 745 (0.898) | 375 (0.815) | 181 (0.797) | 101 (0.377) | 55 (0.211) | 26 (0.532) | 16 (0.185) | 7 (0.531) | 2 (1.000) | 2 (0.660) |
| ECDSA+SHA1 | 190455 (0.075) | 95201 (0.353) | 47517 (0.842) | 23748 (0.947) | 11944 (0.484) | 6086 (0.048) | 3064 (0.075) | 1539 (0.149) | 754 (0.646) | 367 (0.876) | 184 (0.971) | 90 (0.835) | 45 (0.941) | 20 (0.603) | 10 (0.769) | 3 (0.399) | 0 (0.129) | 0 (0.413) |
| MD4 | 190187 (0.362) | 94852 (0.706) | 47496 (0.924) | 23866 (0.391) | 11929 (0.576) | 5846 (0.250) | 2988 (0.699) | 1451 (0.405) | 715 (0.339) | 346 (0.203) | 160 (0.061) | 82 (0.298) | 37 (0.186) | 17 (0.251) | 7 (0.236) | 3 (0.399) | 1 (0.383) | 1 (1.000) |
| SHA1 | 190455 (0.075) | 95201 (0.353) | 47517 (0.842) | 23748 (0.947) | 11944 (0.484) | 6086 (0.048) | 3064 (0.075) | 1539 (0.149) | 754 (0.646) | 367 (0.876) | 184 (0.971) | 90 (0.835) | 45 (0.941) | 20 (0.603) | 10 (0.769) | 3 (0.399) | 0 (0.129) | 0 (0.413) |
Dataset 4: A sequence of numbers, ASCII
This set contains numbers ranging from 00001 to 65536, written as plain ASCII.
| 1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 | 9 | 10 | 11 | 12 | 13 | 14 | 15 | 16 | 17 | 18 | |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Passthrough | 0 (0.000) | 0 (0.000) | 0 (0.000) | 0 (0.000) | 0 (0.000) | 0 (0.000) | 0 (0.000) | 0 (0.000) | 0 (0.000) | 0 (0.000) | 0 (0.000) | 0 (0.000) | 0 (0.001) | 0 (0.040) | 0 (0.278) | 0 (0.632) | 0 (1.000) | 0 (1.000) |
| Adler32 | 32767 (0.997) | 16383 (0.996) | 8195 (0.972) | 4447 (0.000) | 200 (0.000) | 70 (0.000) | 70 (0.000) | 70 (0.000) | 0 (0.000) | 0 (0.000) | 0 (0.000) | 0 (0.000) | 0 (0.001) | 0 (0.040) | 0 (0.278) | 0 (0.632) | 0 (1.000) | 0 (1.000) |
| CRC32 | 32768 (1.000) | 16384 (1.000) | 8192 (1.000) | 4073 (0.717) | 2034 (0.762) | 1017 (0.838) | 509 (0.912) | 253 (0.876) | 125 (0.825) | 57 (0.416) | 28 (0.536) | 15 (0.901) | 9 (0.721) | 6 (0.306) | 6 (0.017) | 4 (0.019) | 2 (0.090) | 0 (1.000) |
| MD5 | 32787 (0.885) | 16362 (0.846) | 8203 (0.897) | 4072 (0.705) | 2063 (0.736) | 1023 (0.987) | 500 (0.610) | 259 (0.851) | 127 (0.965) | 61 (0.755) | 33 (0.859) | 13 (0.532) | 6 (0.597) | 2 (0.453) | 1 (0.729) | 0 (0.632) | 0 (1.000) | 0 (1.000) |
| SHA224 | 32593 (0.173) | 16383 (0.996) | 8168 (0.781) | 4128 (0.606) | 2061 (0.770) | 1063 (0.219) | 554 (0.066) | 280 (0.133) | 143 (0.184) | 75 (0.169) | 37 (0.375) | 13 (0.532) | 3 (0.077) | 2 (0.453) | 0 (0.278) | 0 (0.632) | 0 (1.000) | 0 (1.000) |
| SHA512 | 32864 (0.456) | 16395 (0.921) | 8212 (0.813) | 4163 (0.280) | 2081 (0.459) | 1032 (0.801) | 521 (0.690) | 250 (0.731) | 119 (0.452) | 30 (0.000) | 11 (0.000) | 9 (0.080) | 5 (0.375) | 3 (0.805) | 1 (0.729) | 0 (0.632) | 0 (1.000) | 0 (1.000) |
| Whirlpool | 32906 (0.283) | 16430 (0.678) | 8152 (0.641) | 4080 (0.802) | 1979 (0.124) | 996 (0.386) | 480 (0.162) | 251 (0.778) | 132 (0.723) | 63 (0.950) | 24 (0.184) | 10 (0.167) | 6 (0.597) | 1 (0.202) | 0 (0.278) | 0 (0.632) | 0 (1.000) | 0 (1.000) |
| ECDSA+SHA1 | 32894 (0.327) | 16477 (0.401) | 8338 (0.085) | 4193 (0.118) | 2108 (0.178) | 1100 (0.017) | 555 (0.059) | 277 (0.188) | 155 (0.019) | 74 (0.211) | 30 (0.791) | 19 (0.451) | 6 (0.597) | 2 (0.453) | 2 (1.000) | 1 (1.000) | 1 (0.393) | 1 (0.221) |
| MD4 | 32897 (0.315) | 16440 (0.613) | 8184 (0.929) | 4152 (0.366) | 2036 (0.796) | 1004 (0.539) | 505 (0.773) | 245 (0.511) | 120 (0.507) | 59 (0.574) | 33 (0.859) | 16 (1.000) | 7 (0.860) | 4 (1.000) | 0 (0.278) | 0 (0.632) | 0 (1.000) | 0 (1.000) |
| SHA1 | 32894 (0.327) | 16477 (0.401) | 8338 (0.085) | 4193 (0.118) | 2108 (0.178) | 1100 (0.017) | 555 (0.059) | 277 (0.188) | 155 (0.019) | 74 (0.211) | 30 (0.791) | 19 (0.451) | 6 (0.597) | 2 (0.453) | 2 (1.000) | 1 (1.000) | 1 (0.393) | 1 (0.221) |
Dataset 5: A sequence of numbers, binary
This set contains all values ranging from 0 to 65535, encoded in two bytes. This input set is special; it has exactly the property we expect of the output set, which is why the Passthrough ‘test’hash, which does nothing more than pass through the input string unaltered, gets a perfect score.
Interestingly, CRC32 also receives a perfect score; and even Adler32, notable for its under-performance in the other tests, fares pretty well.
| 1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 | 9 | 10 | 11 | 12 | 13 | 14 | 15 | 16 | 17 | 18 | |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Passthrough | 32768 (1.000) | 16384 (1.000) | 8192 (1.000) | 4096 (1.000) | 2048 (1.000) | 1024 (1.000) | 512 (1.000) | 256 (1.000) | 128 (1.000) | 64 (1.000) | 32 (1.000) | 16 (1.000) | 8 (1.000) | 4 (1.000) | 2 (1.000) | 1 (1.000) | 0 (1.000) | 0 (1.000) |
| Adler32 | 32768 (1.000) | 16384 (1.000) | 8192 (1.000) | 4096 (1.000) | 2048 (1.000) | 1024 (1.000) | 512 (1.000) | 256 (1.000) | 1 (0.000) | 0 (0.000) | 0 (0.000) | 0 (0.000) | 0 (0.001) | 0 (0.040) | 0 (0.278) | 0 (0.632) | 0 (1.000) | 0 (1.000) |
| CRC32 | 32768 (1.000) | 16384 (1.000) | 8192 (1.000) | 4096 (1.000) | 2048 (1.000) | 1024 (1.000) | 512 (1.000) | 256 (1.000) | 128 (1.000) | 64 (1.000) | 32 (1.000) | 16 (1.000) | 8 (1.000) | 4 (1.000) | 2 (1.000) | 2 (0.264) | 0 (1.000) | 0 (1.000) |
| MD5 | 32732 (0.782) | 16472 (0.427) | 8116 (0.373) | 4017 (0.205) | 2044 (0.937) | 1008 (0.625) | 496 (0.492) | 240 (0.332) | 125 (0.825) | 64 (1.000) | 34 (0.723) | 14 (0.708) | 5 (0.375) | 4 (1.000) | 2 (1.000) | 1 (1.000) | 0 (1.000) | 0 (1.000) |
| SHA224 | 32853 (0.509) | 16279 (0.346) | 8271 (0.351) | 4238 (0.022) | 2143 (0.034) | 1056 (0.313) | 511 (0.982) | 256 (1.000) | 121 (0.565) | 63 (0.950) | 29 (0.659) | 12 (0.381) | 6 (0.597) | 4 (1.000) | 2 (1.000) | 0 (0.632) | 0 (1.000) | 0 (1.000) |
| SHA512 | 32693 (0.561) | 16353 (0.783) | 8243 (0.547) | 4096 (1.000) | 2037 (0.814) | 1050 (0.413) | 499 (0.579) | 231 (0.125) | 109 (0.101) | 62 (0.851) | 30 (0.791) | 14 (0.708) | 8 (1.000) | 2 (0.453) | 1 (0.729) | 0 (0.632) | 0 (1.000) | 0 (1.000) |
| Whirlpool | 32604 (0.201) | 16348 (0.749) | 8242 (0.555) | 4206 (0.077) | 2110 (0.164) | 1047 (0.469) | 510 (0.947) | 252 (0.827) | 114 (0.232) | 56 (0.348) | 25 (0.249) | 10 (0.167) | 6 (0.597) | 3 (0.805) | 1 (0.729) | 1 (1.000) | 1 (0.393) | 1 (0.221) |
| ECDSA+SHA1 | 32684 (0.514) | 16353 (0.783) | 8153 (0.649) | 4088 (0.904) | 2077 (0.515) | 1033 (0.777) | 508 (0.877) | 245 (0.511) | 124 (0.757) | 69 (0.531) | 30 (0.791) | 17 (0.802) | 10 (0.475) | 6 (0.306) | 3 (0.459) | 1 (1.000) | 0 (1.000) | 0 (1.000) |
| MD4 | 32777 (0.947) | 16271 (0.310) | 8122 (0.412) | 4029 (0.283) | 2053 (0.911) | 1012 (0.717) | 497 (0.520) | 255 (0.975) | 130 (0.859) | 62 (0.851) | 35 (0.595) | 17 (0.802) | 12 (0.154) | 6 (0.306) | 3 (0.459) | 1 (1.000) | 1 (0.393) | 0 (1.000) |
| SHA1 | 32684 (0.514) | 16353 (0.783) | 8153 (0.649) | 4088 (0.904) | 2077 (0.515) | 1033 (0.777) | 508 (0.877) | 245 (0.511) | 124 (0.757) | 69 (0.531) | 30 (0.791) | 17 (0.802) | 10 (0.475) | 6 (0.306) | 3 (0.459) | 1 (1.000) | 0 (1.000) | 0 (1.000) |
Summary: distribution of p-values
As a quick summary, perhaps it's interesting to present graphs of the computed p-values themselves. Bars to the left of these graphs represent low p-values. Bars nearer to the right edge represent high p-values.
CRC32 seems to score really well, but its graph is skewed by the results of Dataset 5 (binary numbers), which may or may not be too synthetic to be considered a fair benchmark. But even if you substract the results from that test, it does not fare significantly worse than other, cryptographic hash functions.
| Adler32 |
|
CRC32 |
|
MD5 |
|
|---|---|---|---|---|---|
| SHA-224 |
|
SHA512 |
|
Whirlpool |
|
| ECDSA+SHA1 |
|
MD4 |
|
SHA1 |
|