This is the week one project of my DreamHIGH. It is an introduction to R.

Activity 3 of Prof. Torres Aragon's DATA 101 class.

A reproducible "Kaggle-style" predictive analytics project testing whether Dota 2 draft-composition and lane-risk features can explain match outcomes. The report compares logistic regression and a tree-based model, evaluates performance against a baseline, and discusses limitations and next steps for lane-versus-lane modeling.

Three problems — one chi-square goodness-of-fit, one chi-square test of independence, and one one-way ANOVA.

Completed activity 3 from DATA 101

Backup version

SI/L ranking vs EB_HSM/L ranking

In a companion note, we examine how contract typology determines V&V exposure across EPC and EPCM engagements: Why 70–90% of EPC/EPCM Projects Fail — and What Contract Structure Actually Controls. This note extends that argument to the capital dimension.