ranking.jags

model {
	for (i in 1:nplayers) {
		player_quality[i] ~ dnorm(0, 1)
	}
	for (k in 1:N) {
	    wins[k] ~ dbern(p[k])
	    logit(p[k]) <- player_quality[winnerids[k]] - player_quality[loserids[k]]
	}

}