People who compile dictionaries would have us equate a "streak" to that like a layer of fat in bacon, etc. A better description applicable to gambling would be a "period." That could then be a recurring phenomenon or a series of events with a limit or end of a cycle, etc. Personally, I like to call it a continuum of similar decisions, which is more apropos for a gambler.
A lot of betting decisions are based on the streak phenomenon. It entails getting on board a series of repetitive event decisions and riding it out until it inevitably exhausts itself. It can be ascribed to the "momentum" of a ball team, when for brief moments; they seem to get it all together. That’s like healthy key personnel, and coordinated execution of offense and defense simultaneously, etc.
It can also be an inevitable aberration in probabilities. Probability tells us heads are a 50 percent occurrence in a coin flip. But, we know it isn’t a series of single chops HTHTHT, etc. which qualifies as 50/50 as well, but we can observe eight or 10 heads in a row at times, etc.
Such longer streaks are inevitable though infrequent. The longer the streak persists, the less likely it is to continue. So there is also a probability on the length of streaks. Players of casino games, especially of the 1/1 or "even" payoffs are very cognizant of streaks and base their wagering systems on such.
A streak is characterized as any number of like decisions finally terminated by an unlike or opposite decision. Thus, it takes two decisions to identify a streak of one. A head followed by a tail constitutes a streak of one head. (It also kicks off a streak of tails of undetermined duration until a head occurs, etc.)
Thus, even though we know a head can be a .50 (50 percent) probability, a streak of one head only occurs .25 or 25 percent of the time! That’s figured: .50 x .50 = .25. The first .50 is the probability of flipping a head, and the second .50 is the probability of flipping a tail, and thus terminating and defining a streak of one head.
Probably the most streak-prone "side" in the casino, is the bank wager in baccarat (exclude the "push" ties). Over a period of time, a bank decision will be dealt about 50.68 percent of the time. That also means that player or losing decision will be dealt about 49.32 percent of the time.
That means, that a bank streak of one will occur .5068 x .4932 = .24995, or about 24.995 percent of the time! Even through the bank will show a bit over half the time, a single bank streak will occur less than a quarter of the time.
Ironically, figuring a streak of one for the player decision, we multiply the same two numbers, so a player’s one streak also occurs 24.995 percent of the time! Quite a paradox considering that player’s win decision is dealt less than half the time!
However, when we get off the single streak, probability begins to assert itself, as we look at longer streaks. A streak of two for the bank is figured .5068 x .5068 x .4932 or 12.668 percent. While a players streak of two is .4932 x .4932 x .5068 or 12.328 percent, slightly less. The difference between the two decisions becomes more pronounced the longer the streak, etc.
Streak Bank Player
Size Percent Percent
1 24.995 24.995
2 12.668 12.328
3 6.420 6.080
4 3.254 2.999
5 1.649 1.479
6 0.836 0.729
7 0.424 0.360
etc. etc. etc.
So we see, it becomes easier to hit a bank streak of seven than a player streak of seven and so on.
If we carried the streaks out to ridiculous infinity, we could add up all the band and player streak percentages, and they would total 50.68 and 49.32 percent respectively, which gives us our true probability.
If you are going to gamble on streaks, it’s nice to get to know them.