I’m a beginner in data analysis and I’m using data from the past 5 years NBA playoff games. I made a logistic regression model where the result of the game (0 or 1) is the dependent variable. I also used whether the team played at home (0 or 1), and several other variables like 3 point %, Offensive Rebound %, etc. Yet my output confuses me. Supposedly being at home isn’t even significant, which obviously isn’t true. And the beta coefficient of some variables is way too high for what I think log-odds should be (19, 20, etc). What did I do wrong? I made the home variable a factor btw.

Call:
glm(formula = Win ~ ., family = “binomial”, data = fivedata)

Coefficients:
Estimate Std. Error z value Pr(>|z|)
(Intercept) -34.428796 2.688105 -12.808 < 2e-16 ***
Home -0.008206 0.215093 -0.038 0.96957
threePCT 20.276459 1.861469 10.893 < 2e-16 ***
FTA 0.049225 0.016449 2.993 0.00277 **
FTpct 6.566342 1.152709 5.696 1.22e-08 ***
ORB 0.204625 0.032390 6.317 2.66e-10 ***
AST -0.022875 0.027131 -0.843 0.39916
STL 0.347226 0.042683 8.135 4.12e-16 ***
BLK 0.108892 0.047946 2.271 0.02314 *
TOV -0.295402 0.034487 -8.566 < 2e-16 ***
PF -0.063317 0.028169 -2.248 0.02459 *
DRB 0.336799 0.028429 11.847 < 2e-16 ***
TwoPCT 19.864978 1.963007 10.120 < 2e-16 ***
threeRate -0.315986 1.408751 -0.224 0.82252

Signif. codes: 0 ‘’ 0.001 ‘’ 0.01 ‘’ 0.05 ‘.’ 0.1 ‘ ’ 1

(Dispersion parameter for binomial family taken to be 1)

Null deviance: 1167.26  on 841  degrees of freedom

Residual deviance: 574.81 on 828 degrees of freedom
AIC: 602.81

Number of Fisher Scoring iterations: 6