Classification of the major league pitchers was based upon the average WAR (from Fangraphs) for all starting pitchers from 2007-2009.

Sonnanstine was the 11th pitcher in the BORP list. Thanks for catching the mistake.

The standard deviations for the short-season leagues shown in the bottom tables are fairly large due to smaller sample sizes and relatively few innings pitched. Most of the pitchers used in the study pitched in 10+ games in each of the full season leagues, so those values are much more reliable as an indicator of the group. The lower tables were used more to get a sense of how the various groups of pitchers progressed through the minor leagues than to drive teh conclusions of the study.

Anderson – 104.5 – TORP
Cahill – 103.4 – TORP
Mazzaro – 85.1 – MORP
Simmons – 70.5 – Long Reliever

I was restricted to looking at major league pitchers whose minor league careers began within the past 7-10 years. Of the 20 or so pitchers outside the study that I used for validation, 15 were accurately categorized. Most of the outliers were pitchers whose minor league careers were shortened due to injuries.

In regard to Tillman’s GB/FB ratio, it is worth noting that according to FanGraphs, only 6 of the 35 most valuable starting pitchers in baseball this year have GB/FB rates that are at least as low as Tillman’s career minor league ratio of 1.05. The six fly ball pitchers (Verlander, Jackson, Braden, Weaver, Santana, and Cain) certainly support your suggestion that a fly ball pitcher can have success in the major leagues. The counterpoint, of course, is the other 29 pitchers on the most valuable list whose average GB/FB ratio is 1.47.

GB/FB ratio is one of the major components of tRA, for instance. And I’m pretty sure Graham would slap you if you mentioned errors to him.

I agree Ks are more important, but a really high GB rate, even if there are more errors behind you, (and at the major league level that should be less of a problem than in say A ball), you also get more DPs. But an interesting point.

With large enough samples sets, I suspect that it would be possible to create decent classifiers for pitching prospects at each minor league level. The accuracy of the classifiers would likely be pretty good at the upper levels and rather poor at the lower levels. With the resources that I have available, I am forced to build classifiers with relatively small sample sets and then to ensure their relative accuracy with test sets.

As noted above, I was limited to pitchers who have pitched long enough in the major leagues to establish their performance level and who are young enough to have relatively complete minor league statistical records. I restricted the analysis to 34 pitchers due to my own time constraints and to ensure that I had enough pitchers for testing the classifier that resulted from the initial study.

Based on the results of the validation study, I am confident that the TORP/MORP/BORP classifier is reasonably accurate for pitching prospects who have become major league starters in the past ten years. I can’t speak to the accuracy of the classifier for pitchers who predated that time nor can I speak to its accuracy for pitchers who failed to become major league starters.

95 major league pitchers tossed enough innings in 2 of the past 3 seasons to qualify. Of those, I eliminated pitchers whose performances had significantly regressed after decent careers (e.g. Zito, Arroyo, Moyer) or who were still too early in their careers to properly classify as TORPs, MORPs, or BORPs. The pitchers who were left comprised the sample sets that were used for the study. After completing the analysis, I tested the conclusions by determining if the career minor stats of 20 or so pitchers matched how I would classify them and found the method to be very accurate.

If you are interested, please provide the names of 10 or so major league pitchers whose minor league careers began in the past ten years. I’ll run their numbers and you can see if you agree with the classification that results.

they are just getting lucky haha

His minor league career suggests that he can be successful despite his high FB rate. The fact that his K/9 remained outstanding in both AA and AAA makes him unlike most of the BORPs that were considered in this study. When taking his career path into account, Hanson does appear to be a guy who could be an exception to the rule.

His minor league statistical profile sits right between that of the TORPs and BORPs used in the study. His career ERA (2.80), K/9 (9.7), and BB/9 (2.0) all fit very nicely with the TORPs. His career 1.18 GB/FB ratio (0.84 in the upper minors) coincides with the group of BORPs. Based solely on the numbers, I think that Hellickson is more likely to slot as a #4 or #5 pitcher in the major leagues.

TORP score = 86.4
In a small sample size, Withrow’s strong strikeout rate (10.4 K/9) is over-ridden by his high walk rate (4.8 BB/9) and average GB/FB rate (1.24). Improved command would certainly help, but his limited statistical profile doesn’t look like that of a TORP.

Chris Tillman-itis.

If this analysis properly weighed K%, BB% and GB%, both of those guys’s stocks would skyrocket.

As noted above, the fact that Hanson maintained a strong strikeout rate through AAA separates him from most of the BORPs considered in the study. That suggests that he could develop into a Santana/Verlander type pitcher who excels despite a sub-optimal batted ball profile. Worth noting in regards to Hanson’s early major league success, his ERA (3.05) is significantly better than his peripherals (4.08 FIP and 4.50 tRA) would suggest. A lot of that appears to be due to an unsustainably high LOB% (82.3%) and an unsustainably low HR/FB rate (7.8%). As those numbers move toward what would be expected of an above average major league pitcher, his ERA will move into the 4.00-4.50 range unless he can improve his strikeout and/or walk rates.