<![CDATA[The low cognitive learning outcomes in mathematics, particularly in fraction material, was the main problem identified in fifth grade at MI Cipamutih, with 60% of students failing to achieve the Minimum Mastery Criteria (KKM). This study aims to examine the effectiveness of the Game-Based Learning (GBL) scramble model in improving cognitive mathematics learning outcomes. The method used a quantitative approach with a quasi-experimental Nonequivalent Control Group Design. The subjects consisted of 54 fifth-grade students (27 experimental, 27 control). The learning achievement test instrument was tested for validity, reliability (0.887), discrimination power, and difficulty level before use. Data analysis used the Mann-Whitney U test, Wilcoxon signed-rank test, N-Gain, and Cohen's r effect size. The results showed no significant difference in the pre-test (Sig.=0.660; p>0.05), but a very significant difference in the post-test (Sig.=0.000; p<0.05) with the experimental class mean score (71.78) much higher than the control (29.37). The experimental class increased by 43.74 points (156%), while the control only increased by 1.18 points (4.2%). The N-Gain value of the experimental class was 60.88% (medium category) and the effect size r=0.799 (large effect category). The conclusion of this study is that the GBL scramble model is proven effective and has a large effect in improving students' cognitive mathematics learning outcomes. This model is recommended as an innovative, low-cost learning alternative that is adaptive for madrasahs with limited digital facilities.]]>
