NIJ's Recidivism Forecasting Challenge aims to improve the ability to forecast recidivism using person- and place-based variables with the goal of improving outcomes for those serving a community supervision sentence. We are pleased to post official results below in Student, Small Team, Large Team, and Accounting for Racial Bias categories. Download results spreadsheet (.xlsx).
Total Submissions by Year/Category
| Category |
Year 1 |
Year 2 |
Year 3 |
| Student |
1 |
0 |
1 |
| Small Team |
41 |
38 |
39 |
| Large Team |
15 |
17 |
20 |
| Totals |
57 |
55 |
60 |
Student Category
We received only one student entry for year one, none for year two, and one for year three from a different student than year one.
Year 1, Student: Shah
| Population |
Score |
| Male Parolees |
0.389 |
| Female Parolees |
0.236 |
| Average Accuracy |
0.312 |
Year 3, Student: Hanson
| Population |
Score |
| Male Parolees |
0.308 |
| Female Parolees |
0.267 |
| Average Accuracy |
0.288 |
Small Team Category
Small Team Year 1
Year 1, Male Parolees
| Place |
Team Name |
Brier Score |
| 1st |
IdleSpeculation |
0.1916 |
| 2nd |
SRLLC |
0.1919 |
| 3rd |
Sevigny |
0.192 |
| 4th |
Team Smith |
0.1922 |
Year 1, Female Parolees
| Place |
Team Name |
Brief Score |
| 1st |
TeamKlus |
0.1542 |
| 2nd |
TeamSherrill |
0.15539 |
| 3rd |
IdleSpeculation |
0.15544 |
| 4th |
Sevigny |
0.1555 |
Year 1, Average Accuracy
| Place |
Team Name |
Brier Score |
| 1st |
TeamKlus |
0.1733 |
| 2nd |
IdleSpeculation |
0.1735 |
| 3rd |
Sevigny |
0.1738 |
| 4th |
JianyeGe |
0.174 |
Small Team Year 2
Year 2, Male Parolees
| Place |
Team Name |
Brier Score |
| 1st |
MCHawks |
0.1658 |
| 2nd |
Oracle |
0.167 |
| 3rd |
CategOracles |
0.1679 |
| 4th |
VT-ISE |
0.1685 |
Year 2, Female Parolees
| Place |
Team Name |
Brief Score |
| 1st |
Oracle |
0.1233 |
| 2nd |
MCHawks |
0.1242 |
| 3rd |
VT-ISE |
0.126 |
| 4th |
DEAP |
0.1263 |
Year 2, Average Accuracy
| Place |
Team Name |
Brier Score |
| 1st |
MCHawks |
0.145 |
| 2nd |
Oracle |
0.1451 |
| 3rd |
VT-ISE |
0.1472 |
| 4th |
DEAP |
0.1481 |
Small Team Year 3
Year 3, Male Parolees
| Place |
Team Name |
Brier Score |
| 1st |
DEAP |
0.15237 |
| 2nd |
EconCGU |
0.152425 |
| 3rd |
CategOracles |
0.1526 |
| 4th |
Duddon Research |
0.1534 |
Year 3, Female Parolees
| Place |
Team Name |
Brief Score |
| 1st |
TEAMSherrill |
0.11614 |
| 2nd |
Aurors |
0.1179 |
| 3rd |
MNLB |
0.11815 |
| 4th |
Oracle |
0.1182 |
Year 3, Average Accuracy
| Place |
Team Name |
Brier Score |
| 1st |
EconCGU |
0.13577 |
| 2nd |
DEAP |
0.1358 |
| 3rd |
CategOracles |
0.1359 |
| 4th |
MNLB |
0.1363 |
Large Team Category
Large Team Year 1
Year 1, Male Parolees
| Place |
Team Name |
Brier Score |
| 1st |
CrimeFree |
0.19 |
| 2nd |
MattMarifelSora |
0.1907 |
| 3rd |
CrescentStar |
0.191 |
| 4th |
KMG_BQR |
0.1913 |
| 5th |
PASDA |
0.1914 |
Year 1, Female Parolees
| Place |
Team Name |
Brief Score |
| 1st |
CrimeFree |
0.1538 |
| 2nd |
MattMarifelSora |
0.1543 |
| 3rd |
CrescentStar |
0.1548 |
| 4th |
DataRobot |
0.155 |
| 5th |
KMG_BQR |
0.1552 |
Year 1, Average Accuracy
| Place |
Team Name |
Brier Score |
| 1st |
CrimeFree |
0.1719 |
| 2nd |
MattMarifelSora |
0.1725 |
| 3rd |
CrescentStar |
0.1729 |
| 4th |
KMG_BQR |
0.1733 |
| 5th |
PASDA |
0.1734 |
Large Team Year 2
Year 2, Male Parolees
| Place |
Team Name |
Brier Score |
| 1st |
TrueFit |
0.1542 |
| 2nd |
IdleSpeculation |
0.1552 |
| 3rd |
EarlyStopping |
0.1619 |
| 4th |
KMG_BQR |
0.1631 |
| 5h |
PASDA |
0.1638 |
Year 2, Female Parolees
| Place |
Team Name |
Brief Score |
| 1st |
TrueFit |
0.1196 |
| 2nd |
IdleSpeculation |
0.1209 |
| 3rd |
EarlyStopping |
0.1222 |
| 4th |
KMG_BQR |
0.1224 |
| 5th |
Crescent Star |
0.1245 |
Year 2, Average Accuracy
| Place |
Team Name |
Brier Score |
| 1st |
TrueFit |
0.1369 |
| 2nd |
IdleSpeculation |
0.1381 |
| 3rd |
EarlyStopping |
0.142 |
| 4th |
KMG_BQR |
0.1428 |
| 5th |
PASDA |
0.1442 |
Large Team Year 3
Year 3, Male Parolees
| Place |
Team Name |
Brier Score |
| 1st |
IdleSpeculation |
0.1463 |
| 2nd |
TrueFit |
0.1472 |
| 3rd |
SAS Institute |
0.149 |
| 4th |
EarlyStopping |
0.1498 |
| 5h |
KMG BQR |
0.1507 |
Year 3, Female Parolees
| Place |
Team Name |
Brier Score |
| 1st |
TrueFit |
0.1139 |
| 2nd |
IdleSpeculation |
0.1147 |
| 3rd |
PASDA |
0.1165 |
| 4th |
SAS Institute |
0.1173 |
| 5th |
EarlyStopping |
0.11734 |
Year 3, Average Accuracy
| Place |
Team Name |
Brier Score |
| 1st |
IdleSpeculation |
0.13048 |
| 2nd |
TrueFit |
0.13052 |
| 3rd |
SAS Institute |
0.1332 |
| 4th |
EarlyStopping |
0.1336 |
| 5th |
PASDA |
0.1343 |
Accounting for Racial Bias Category
Bias Year 1
Year 1, Male Parolees
| Place |
Team Name |
Brier Score |
| 1st |
MattMarifelSora |
0.809 |
| 2nd |
SEVIGNY |
0.808 |
| 3rd |
MCHawks |
0.8069 |
| 4th |
SRLLC |
0.8066 |
| 5th |
Team_Smith |
0.8064 |
Year 1, Female Parolees
| Place |
Team Name |
Brier Score |
| 1st |
PASDA |
0.8446 |
| 2nd |
MCHawks |
0.8442 |
| 3rd |
EconCGU |
0.8432 |
| 4th |
IdleSpeculation |
0.8428 |
| 5th |
MengHuang |
0.841 |
Bias Year 2
Year 2, Male Parolees
| Place |
Team Name |
Brier Score |
| 1st |
TrueFit |
0.8397 |
| 2nd |
KMG_BQR |
0.8356 |
| 3rd |
EarlyStopping |
0.8354 |
| 4th |
SAS Institute |
0.8343 |
| 5th |
MCHawks |
0.8342 |
Year 2, Female Parolees
| Place |
Team Name |
Brier Score |
| 1st |
IdleSpeculation |
0.8771 |
| 2nd |
MCHawks |
0.8758 |
| 3rd |
PASDA |
0.8755 |
| 4th |
Oracle |
0.8754 |
| 5th |
SAS Institute |
0.8742 |
Bias Year 3
Year 3, Male Parolees
| Place |
Team Name |
Brier Score |
| 1st |
SAS Institute |
0.851 |
| 2nd |
IdleSpeculation |
0.8504 |
| 3rd |
TrueFit |
0.8502 |
| 4th |
EarlyStopping |
0.8501 |
| 5th |
KMG_BQR |
0.8481 |
Year 3, Female Parolees
| Place |
Team Name |
Brier Score |
| 1st |
IdleSpeculation |
0.8853 |
| 2nd |
PASDA |
0.8835 |
| 3rd |
SAS Institute |
0.88297 |
| 4th |
EarlyStopping |
0.882663 |
| 5th |
CategOracles |
0.8821 |
