Why was it an anomaly? Because they lost? What about the last few weeks? Should we exclude those too? Were those anomalies?
You can't back-fit data, it makes the results worthless.
"The point of keeping records is to give an indication what happens if you play these." Which is precisely why you don't cherry pick weeks. I can guarantee you, if week 1 went 4-1 you wouldn't have excluded that week from your records. Saying its week 1 is a convenient excuse to back-fit data.
Its already been shown that the only anomaly (I'm assuming that's what you meant) was last year, that was the only thing abnormal. As I and others have pointed out, it is not a longterm profitable method for playing Overs. Its been posted previously in this thread what the records long term have been.
Same thing for PT Overs. This year is an aberration, the best in the last 10 years. Last year 28-21. Last 12 years, including this year, 266-241. This is for any game that started at 7pm EST or later. There are many "systems" that are good short term, but for any system to be meaningful you need consistent year to year wins. Neither of these have done that. Now, the last 2 years have been good, so perhaps it will continue. That's the only thing "we" know.
As for Thursday home team, next week over? Again, whats the long term record? I'm not gonna dig that up, but I would doubt that it isn't anything more than a recent phenomenon.
You trumpeted in your thread title how your overs went 80% last year, now you are backing off and telling people to essentially not to play them. Often times people get excited over what they think is a great angle, and sometimes they are. But it takes more than a 1 year, or any other short time period, to make them reliable.