This is true RL. I am sure there are several reasons that happens.
Going all in on one digit, pair, trip, or set is one reason.
Running tight filters is another reason.
Setting the the wrong filter first is a big reason in my case. The first filter, call it the base filter, should be one that produces the largest amount of possible outcomes.
The more I research filters the more I realize I don't know squat about filtering techniques. I'm always learning and most of the time it is the same lesson from a different standpoint. This is a problem most everyone faces with filters.
Let's use some of the filters in this thread to demonstrate why a single filter may produce a loser before you buy a ticket.
To do this we'll use the Texas 2Step 4/35 and I'll quote stats frm the entire matrix.
One of the most popular filters is the Even/Odd Totals. It is simply the total of the even and odd numbers. Its considered ubiquitous. Its used in every program and game.
There are 2380 sets containg 4Even/0Odd in Texas 2Step. They make up 4.5% of the total sets.
There are 12240 sets containing 3Even/1Odd. They make up 23.5% of the total sets.
There are 20808 sets containing 2Even/2Odd that make up 39.7% of all combinations.
There are 13872 sets containing 1Even/3Odd that make up 26.5% of all combinations.
And there are 3060 sets of 0Even/4Odd that make up 5.8% of all combinations.
Knowing that, it makes sense to jump on 2Even/2Odd with the expectation that almost 4 of the next ten drawing will match the E/O patterns. Combined with 1E/3O almost 7 of the next ten drawings could be expected to match those patterns.
But the lottery rarely plays by its statistical rules.
A good example of this is last ear when I was watching the sum patterns od Pick3. he sum went to 70 draws out, the longest it had ever been and a length that, for a mid sum, was unusual. I immediately set the trap playing all combinations totalling 15 with the belief that I would cash in quick.
Big fail. I spent $187 to win $250. Certainly not the big score I anticipated. So what happened?
Somewhere along the way I forgot to check the trends. And those current trends have a way of disrupting the statistical process. We fool ourselves by calling it an abberration of the norm, but the norm is whatever is happing at the time. Current trends can push aside digits, sums, E/O patterns, E/O totals, and every other filter we apply.
Statistically the drawing can follow the matrix perfectly overall, but the more important question is is the game following the matrix NOW and in recent week or months?
IMHO the biggest reason filters fail is we apply them by what we know overall, without considering what the trends are in recent history.
All games have the propensity to go "crazy" and times; producing eerie combinations that make us go, "Huh?"
Pardon me, if I've missed your point, seems I've meandered around and deluded mine as well.