Beginner miners always learn from losses. I also did not listen to the advice of experienced miners before, and mine shitcoins. The price of shitcoins may fall by more than 85% and not rise for many years, or the project may be closed.
I don't think there's anything wrong with dedicating a rig for speculation that aren't listed on exchanges, as long as you accept making no revenue with that rig.
Other than that, the lesson I learned was to buy much more video cards when the price is below MSRP, and buy much less during a gold rush. I should mine through a bear market, sell enough coins to cover operating expenses, then HODL the rest in anticipation of a gold rush. If there's a profit during that bear market, I should contribute the cash value to a Solo 401(k) on a Roth basis. When the bull market arrives again, I can put the profit into the 401(k) on a pre-tax basis, then borrow against it when the bear cycle starts again in order to build more rigs.
Also, you should always mine the most profitable coin and sell it to buy the coins you like. This is why only half my rigs mine ETH at the moment. Minerstat has a good profit-switching feature (which is really a difficulty switcher). It's surprising how much difficulty for coins like ERG, FLUX, XEQ, or SERO can rise/fall every day. The 'averages' on WTM are irrelevant.