How confident can I be that I have an edge if these results are true?

So I developed a simple method to evaluate college basketball game lines. The method gives a spread and a total. The betting rules are simple. If the line is off by 4 points or more, bet it. If the total is off by 4 points or more bet it. There is absolutely no discretion involved. Even if half the players on a team are suspended and the numbers say bet it then bet it. This way there is no data interpretation or curve fitting, the numbers are true. I waited 15 games into the season to get good hard data to start paper betting it, but there's no bias there, simply decided before running any numbers to get 15 games in.

I am good at basic math, but my higher math skills are lacking and I have never had a firm grasp on standard deviation.

With these results, how sure can I be that I have an actual edge and not a statistical outlier?

Results vs spread: 264-242 52.2% This barely, if at all, beats the vig, there's no edge here.

Results vs Totals: 314-208 60.2%

Totals Breakdown:
Results taking the over: 215-137 61.1%
Results taking the under: 99-71 58.2%

So this methods seems to not be much better than break even against the spread but seems to do well on picking the totals, especially the over.

The question is, and it doesn't matter if you believe the results or not, it could even be theoretical, if the results are correct, what is the likelihood I have a true edge? My gut says the totals certainly bear a full season of betting.

I haven't run any numbers on the "strength" of the play (ie my total is 15 points off the vegas total instead of 4 or 5 creating a theoretical stronger play) it's just straight "if it's 4 points or more off, bet it".

Super question you ask. I would ask you what happens when you are off exactly 4, 5, 6... etc.... 3, 2, 1.... It is possible you have an advantage worth exploiting at 3.... It is possible you have a disadvantage at 10 or more (maybe it only happens with major injury or suspension). But whatever.... at 215-137 I'd start betting it!

Start posting picks each day.

Its still a relatively small sample size but I agree with HowMany, start posting picks and lets see for ourselves.

I think your great question gets to the heart of sports betting, which is its mysteriousness, its fascinating aspects and its frustrations. I think even if you had a sample of 5,000 results you could not be sure that you had an edge. Whatever factors that caused you to have a winning % on the first 5,000 bets might disappear over the next 5,000 bets. I think the real answer is that you can never be totally confident that you have an edge except on those very rare occasions where you might have an arbitrage play or some kind of unusual mathematical type of play. IMO sports betting is an art, not a science. The best thing to be confident about are your analytical abilities; not your edge.

Pick up a copy of conquering risk by Elihu fuestel. It talks about sports betting models. Comparing how often you beat the closing line to what your model projected line will be a strong indicator of long term success. For example if you are consistently laying 3 points and the market closes minus 4.5 you will be a winner in the long run.

Ugh, I don't know if I want to commit to posting picks. I usually run them at night since there's no discretion involved, knowing the result is irrelevant. And I am not asking the question to sell a system or even get people to believe it's working. I was just asking about the possibilities of the results showing a true bias and not a statistical outlier.

Thanks for the book recommendation, I bought it.