The trends over the past few years have shown that mining is 10+ times less efficient than simply buying coins outright, while offering very little protection if the coins go to zero due to all the hardware depreciation and electricity costs involved with mining. Not to mention babysitting miners to restart them if they crash, if the power goes out, etc., and replacing burned out PSUs/GPUs/etc.
I did a back of the envelope analysis in another thread where Silver claimed that if you bought something like 8 high-end GPUs at the beginning of the year that it would have produced 50k LTC. Assuming typical USA electric prices, the cost of the GPUs, CPUs, mobos, RAM, etc. + the power costs would have totaled about $7.5k for 2013. 50k LTC for $7.5k sounds pretty good, until you consider that if you had simply bought $7.5k worth of LTC at the beginning of the year, you would have 714k LTC. Or put another way, rather than spending $7.5k in mining and power costs and dealing with mining headaches, you could have gotten similar results by simply buying $750 worth of LTC at the beginning of the year.
"But it's riskier to buy coins than to mine" some people claim. That's not necessarily true and in fact it's often false. If litecoins had gone to zero, the guy who bought $750 in January 2013 would lose $750. The miner would have lost between $500 and $5500 depending on exactly when litecoin went to zero. $500 loss = if the coin went to zero value almost immediately, so that the miner doesn't really lose anything on electricity costs, but then has to return or resell all that equipment, probably for a loss (let's say $500 in losses; if the coin crashes, so does demand for mining equipment). $5500 loss = if LTC crashed at the end of the year, after 12 months of power bills and equipment depreciation. The breakeven point would probably be around 2-3 weeks of mining--at that point, both the guy who mined and the guy who bought $750 worth, are both going to lose about $750. Anything beyond that, and the miner would lose more money.
In conclusion, if you think a coin won't go to zero in the next 3 weeks and might go up, then buy it instead of mining it, because you will likely get at least 10 times more profit that way (this has held true for every year that BTC has existed--it's always been more profitable to buy coins directly than to mine). If you think that a coin will go to zero within 3 weeks, then mining is better than buying the coin directly, but in such a case, why are you even buying/mining at all? Just stay away from it. If you can justify mining some other way, like if you were going to buy a GPU anyway or whatever, that lessens the risk of mining because you were going to game on the card anyway. But once you start to buy more GPU power than you would normally buy, you are going to start losing out to the people who simply bought coins directly.
And always remember, all of this stuff is extremely risky so treat it like gambling and don't ever put more money into it than you can afford to lose.
P.S. for the people responding, I already covered your issues in this post except for coin-hopping. Re: hopping from coin to coin, all coins go up and down depending on what bitcoin does. Like precious metals being led around by gold (if gold goes up or down it tends to drag silver with it), there is a clear leader and that is bitcoin. If bitcoin goes up, it drags everything else up, and if goes down to zero, all the rest of the currencies will go to near-zero or zero as well.