Class is one of those things that plenty of people believe in but very few can define. A horse seems to either have the requisite class or they don't and it one of those things that most observe with the naked eye rather than with actual numbers.
My thought or observation on the matter is that class is essentially the type of race a certain horse fits into. It's a tough thing to nail down because horses continually compete at various class levels and achieve mixed results because of many different reasons that go well beyond class. For instance a horse may be a G-1 quality sprinter but only a G-3 quality miler. It's hard to gather enough data to accurately classify horses because of brief campaigns.
It's a commonly held notion that the Preakness unlike the Derby is often won by the best horse. A look over the list of past winners reveals that the vast majority of them were extremely classy individuals, true G-1 horses. It then stands to reason that a secret to unlocking the Preakness puzzle beforehand would be to identify those horses who are classy enough. The trick is finding the tools to do that accurately.
I've often seen average earnings used as a benchmark for class. If you were to take the lifetime earnings of every Preakness starter and divide it by their career starts then compare it against the field average you'd find that 69 horses failed to better the field's average earnings mark. Those 69 horses combined for a 69-2-7-6 in the Preakness from 2000 to the present. Since there were only 96 Preakness starters during that era it means the remaining 27 horses achieved a record of 7-2-3. Sadly betting all those 27 horses would not have yielded a profit since the highest paying winner was Curlin ($8.80). You'd have lost $10.80 over 9 years but you'd have had 7 winners.
A far more simplistic class measurement is looking at whether or not a horse has been able to win a Graded Stakes race recently. One could divide the Preakness field into those that have won a Graded Stakes in the current season and those that have not. The "have nots" combined for a very poor 52-0-4-4 since 2000 while the "haves" went 44-9-5-5. Sadly this angle doesn't show a flat win bet profit either. Betting the 44 graded stakes winners would cost $88 and would only return $85.40. In a race where short priced winners are so common it's rare to find any cut and dried angle like this that would show a profit.
As always judgment needs to be applied, but class is still a major key. However you define it, handicappers will need to sort out the class levels of this Preakness field to decide who the true G-1 horses are.
It's an odd year because in most renewals of the Preakness there are two or three horses that look really good and the rest are up and comers. This year one must decide between Rachel Alexandra, Mine That Bird, Pioneerof the Nile, Friesan Fire, Papa Clem, Musket Man and maybe even Big Drama. All of them have either hinted at or outright confirmed G-1 class but they can't all win.