Is Practice the Best Way to Learn?
No. Playing a lot of hands is a very slow way to learn. Certainly not the best way. It's just trial and error. The best way, by far, is to read books and watch videos.
One example is AI computers. Some have been running 24 hours a day for years and are nowhere near a solution. They are based on neural networks, a fancy name for trial and error. It's how the human mind learns. A human is a million times slower, and with an imperfect memory. I would put my money on the computer.
This is not to say that practice is not required. It's just not efficient.
Humans have learned that the better way is with a teacher, books, videotaped instruction. We all know this.
For poker, a teacher, or a coach is very expensive. Perhaps the best way but not always practical. Video tapes are good, but generally available only on poker coaching websites. I like RedChip.com but there is a $50 / month membership fee. Books are relatively cheap, easy to get, and very useful. You can read them over and over again. Highlight them, make notes, whatever.
I believe that books are the most practical way to learn to play Hold'em well. One of my favorite authors are Ed Miller, who wrote “The Course”. His books are generally easy to understand.
Alexander Fitzgerald has also written a great book. “The Myth of Poker Talent”. Not as easy to read, but very good.
Will Tipton is a PHD Computer Scientist that has written two books, for Heads Up Holdem, Game Theory. He has even made his source GTO code open source. Harder to read but worth the effort.
Matthew Janda is another. His second book is perhaps the best Hold'em book written so far. If you really want to get deep into poker theory he is the best.
To study and experiment on your own there are some excellent applications available. Holdem Manager and Flopzilla are the two that I have used the most.
The shortcoming of even the best software is that once you have used it to develop a Strategy, you must test that strategy by risking real money, and spending many hours at the table, before you know the result. Even then, you can't know if you actually followed that strategy 100% at the table.
PeakHoldem offers a good alternative, not a replacement. You can test your strategy by editing a strategy element, such as a preflop range, then run 100,000 simulated hands, full table, against a variety of opponent types. No risk, no long hours, and many tools to help you perfect your game. It's free.
PeakHoldem can dramatically improve your win rate.
Every future release of PeakHoldem will increase the accuracy of opponent play. Unique enough to be patent pending. 62/732,519.