In evaluating minor leaguers, I look at stats, age relative to level of ball and 'scouting reports' from reputable sources, such as BA. For stats, I like to use OPS and adjust that up or down based on the player's age and level. I like to use a slightly more liberal age per level than most probably do. This is partly because players drafted out of college (virtually all of whom are at least in their age-21 year) almost always are initially assigned to a minor league level (Lo-A or below) that many or most probably would consider too low for a prospect or that age. And the age I like to use is the age the player will be on 12/31 of the year in question. This seems like a good method to me because this is how people are assigned to grades in school (people born on 1/1 and 12/31 of the same year are assigned to the same grade). For me, here are the appropriate ages per level for a bona fide prospect:
R - 18-19
SS - 19-20
A-Lo - 20-21
A-Hi - 21-22
AA - 22-23
AAA - 23-24
A BB/K ratio of about 3/5 or 60% is about average. So I consider anything close to that to be acceptable. A low ratio is not good, but only very low rates raise red flags for me. For example, Reggie Abercrombie's BB/K was so low (121/793 or 15.3%!) that I didnt see how he could be successful in MLB. And this year I think he's performed as could be expected, IOW not very good. So Brandon Wood's ratio is not good (114/308 or 37.0%) but it's nowhere near Abercrombie territory. There are some guys that K a lot but also walk a lot. Like Troy Glaus and Adam Dunn. Guys like that who also have exceptional power can still be top prospects. And some guys have excellent BB/K ratios but dont walk much, like Alberto Callaspo. In these cases, I think the BB/K ratio is a little misleading. These guys still can be good prospects, but probably are slap hitters who make excellent contact but w/o much power. So, they may have to be very good defensively in order to be top prospects.
For pitchers, I like to see a K/IP close to 1.0, low HR/IP (considerably less than 0.1), BB/IP less than .33 and that old standby, ERA. I also like OPS against, but I dont know of a place currently that lists it. BA had this stat last year, but they dont this year. Groundball/Flyball rate is nice too.
Of course, you also want to consider whether the league is a hitter's or pitcher's league or not, altho I think a little too much emphasis is placed on this.
After looking at and analyzing the stats, you need to read some scouting reports, particularly for the defensive reputations of the players. Lots of guys have impressive stats but arent considered bona fide prospects. Like Travis Denker and Kila Kaaihue from last year. Their stats indicated that they should be pretty good prospects. But neither one was considered a very good fielder and it was questionable where they would play and they werent rated highly by the reputable prospect mavens. Also, this year Denker is not hitting nearly as well as last and Kaaihue's year is atrocious (dont confuse him w his brother).
It's nice to actually see the players play in person, and I do go to a handful of minor league games. But I dont find this helps me evaluate prospects much. First of all, I'm not a trained scout so I really dont know what to look for. And I dont have a scout's tools, like a radar gun. But even if I was a trained scout w all the tools, I think I would have to see a player several times before I could accurately evalute him. I saw Erick Aybar play a game in 2004. If hadnt known beforehand that he was a top prospect, I couldnt have found this out by seeing this game. I think he went 0-5, struck out twice, hit into a DP and made an error. But it was one of the few poor games he had that year.