The best simulations tend to reproduce somewhere in the neighborhood of about forty five % of the couples that we've sampled. So he said, i if this was just totally random, and you'd just scrambled these people up in the computer maybe you'd expect about three percent or so of the couples to get back together. The fact that we're already hitting a Forty Five% mark suggests we're doing a good job. But it also tells us if we are missing things in our models, we have a lot of work left to do too.