Subset or Head to Head

[rfr3sh]

Team 1 -117
Team 2 + 300
Draw +250

If I had head 2 head results of Team 1. vs. Team 2 , a small sample of say 80 games, I would assume this is too small of a sample size to approximate the % of times Team 1 wins, so my question is , is there any relevance to looking at away teams valued at [-110,-120] and seeing what percentage of time they won?

[sharpcat]

It is most likely going to end up between 52% and 55% over a large sample and it would have to stray awfully far away from that % in order to have any significant value before you consider investing on something of this nature without taking player and team statistics into consideration.

[Justin7]

One problem with looking at the previous 80 match-ups is relevancy. How much have the teams changed during those 80 match-ups?