Time of day is an interesting metric, too. Doesn't seem like a random distribution to me... Maybe coupled with another metric or too it could help navigate the day better...?
Here are two I tried: One algo was similar to an algo used by a gentleman on ET day trading options. It uses a point system going back a few days based on the behavior of price movement, up day, down day, flat day... option movements... Based on the scores, going long, short, etc. Another I invented based on up, down, flat, pre market, post market, news.... option movements.... Both good brain exercises to stave off dementia but not good enough to earn a living day trading. Anyway, staring at the screen and click based on gut feels produced better outcome. i.e., "random" trades, but with strict risk management actually worked better.
From that info, one can deduce that price will break either yesterday's high or yesterday's low (or both) on 90% of occasions.
One way to use these stats - Line 1... O < C is an UP day (47% of days..bit surprising given this crazy bull run...but I don't know your sample size of course). So...24% of the day's daily range is likely to be BELOW the open and 76% ABOVE. If we use say ADR(20) for an average daily range, we can aim for BUYS after price falls roughly 1/4 of ADR(20) and aim for a full ADR(20) to the upside (taking profit at stages). BUT..of course...we don't know in advance that the Close will be above the Open. So we need to combine with other stats in the table.