The only progress I track or pay attention to is my accounts level in a casino. When there are bonuses for level-up, and if there is little left to level-up, I would gamble to get more wager. But I never track the amount of bets I have placed, won or lost. Nor the amount I have bet. As I find it useless. I cant imagine situation when such statistics could help me. For example I know that I have won 11 and lost 25 bets during current session. That does not help me anyhow when I think if I should place next bet or no.
That still is a record and as long as the user benefited that then it is good. Well in that one sample you can truly cannot track anything, but in series of long records, you can identify patterns. And in terms of gambling I believe analyzing patterns of behavior will be very helpful For instance, journal your emotion every bet or day. Then analyze, which emotions should you avoid gambling especially when you lose. And you can probably bet more on your records of your emotion where you have greater probability of positive result. That's just one of many ways where you can use your track record. For it to be relevant and effective, the bigger the sample size then the better it would be.
What patterns I can identify and build if all rounds are random? If there was at least one constant variable, then it would be possible to try to build a pattern. What can behavior analysis also give, if when I win I am happy, and when I lose I am sad. In addition to this, all gambling sessions happen on random times. For me it is a complete random and I dont know how usage of such info might help me in future.