I'm finding difficulty in what to really look for with drafting pitchers.

I usually look at OAV#, ERC# and WHIP# first and compare to OAV, ERC and WHIP.

Anything else I should target? I know ballpark and opposing batters play a huge factor, but I draft guys that have great # numbers that are all lower than their regular numbers and they suck. Case and point, Brandon League '06 has an era of 7.47 after over 15 IP. Smaller sample, but that's bad.

Thanks.

Overlooking HR/9 (+/#) is a big mistake. HRs are killers at all levels.

BB/9# is also big. As you get to higher caps, it is even more important with all the high OBP guys out there.

As you said, though, it is a very small sample. One or two bad outings can really skew the stats. Give it a little time.

I look at HR and BB/9 too.

I'm just always surprised when a guy with lower normalized stats performs horribly.

Thanks all

The # stats are not used in the at-bat calculations.  They are calculated in part by using historical averages from all seasons.  They can be a useful reference for knowing how a player's performance will be affected on average against competition from a wide range of historical seasons, but in the individual at-bat calculations, only raw AVG and AVG+ are used on the hitter's side, and raw OAV and OAV+ on the pitcher's side.  While the hitter's season (+ stats) and pitcher's season (+ stats) are calculated against each other, this excludes all the other seasons which are weighted equally by the # stats.  So # can be misleading.  If a pitcher has "lower" normalized stats (or higher, in the case of batters), that does not automatically mean a permanent boost in that player's performance.  

Over a sample size that small, bad luck is more often the culprit, but even over longer periods of time, # stats and whether they go "up" or "down" for a player are not the standard by which performance should be judged. 

skunk, y do u say # can be misleading?  With a large enough sample size (and facing the exact same pitching), a .250# avg hitter will have a lower batting average than a .260# avg hitter.  I don't look at the raw or + stats when evaluating players (just the # stats).

# stats can be misleading because it would probably take a very large sample size and fairly even distribution of opponent pitcher or hitter seasons and talent levels to definitively "prove" a .250# hitter will have a lower batting average than a .260# hitter.  Or that a RL .245 with a .250# will hit better than a RL .245 with a .245#.  But we know from countless "underperforming" complaint threads and the subsequent statistical analyses by minds greater than mine that even a full 162 game season is a very small sample size for a statistical simulation and results will vary widely.  You can look at the end season results of every single team in every single league you play and there will be plenty of examples of lesser # stats outperforming better # stats, or players with better # stats than RL stats performing worse than expected and guys with worse # stats than RL stats performing better than expected.  

I think a lot of owners think of players with better # stats than RL stats having some kind of built-in performance boost and thus have expectations that those players should always perform "better" somehow because of it, and that's just not true.  Many players who have worse # stats than RL stats can still be very effective in the small sample size of a 162 game season because the actual at-bat calculations only consider the opponent's raw stats and their value relative to their own individual season (represented by the + stat figures), and not the historical averages used to calculate the # stats.

As I said before, the # stats are a good reference for what would occur over an extremely large sample size (probably hundreds of full seasons) with an even distribution of opponent seasons to accurately reflect the historical average.  But that's not how the SIM works.