Today's Fantasy Rx

As previously mentioned, please know that I don't intend to directly regurgitate the information presented by the speakers, since we don't want to reduce the value of the conference for the attendees or discourage anyone from attending in the future.

The Saturday morning sessions opened with Joe Sheehan, formerly of Baseball Prospectus, discussing the integration of performance and skills analysis. He began by distributing a handout depicting an approximately inverse linear relationship between the value of performance analysis and tools analysis from amateur ball through the minors to Major League baseball.

In high school and Latin America, the presence of aluminum bats and variable competition makes data worthless. College ball still has aluminum bats, though the competition is generally better. "Reliable workload data is gold" when discussing college, and he referenced Boyd Nation's website, which I've discussed in past articles. Most damningly, there's no motivation for high school or college coaches to protect their pitchers as winning is usually the only thing that helps coaches keep their coaching jobs.

Plate discipline numbers have value throughout professional baseball, although not many other stats matter at the lowest level of the minors as everyone's adjusting to wood bats. Stolen base percentage remains a good indicator of overall baseball skill. Middle infield prospects also need to stay in the middle as long as possible as players rarely move to the right of the defensive spectrum(towards CF, 2B, SS, or C). Strikeout rate is vital for projecting pitching prospects in the high minors, and almost no one at age 24 or above can be considered a top prospect. The percentage of extra-base hits among hits allowed is a decent proxy for G-F rates for minor league pitchers. Upon reaching the majors, more technical info is important, especially with the existence of advance scouts focused only upon a player's actions in the last week or two.

Perhaps the most intriguing item Joe noted was that he's aware of some recent studies that indicate improvement by players repeating leagues can be treated as skill development and not merely better familiarity with that league. We'll keep that in mind when evaluating players in the future, although we'd definitely like to view that research ourselves.

One of the last points he mentioned was that leagues like the AFL and Cape Cod League aren't good for evaluating stats due to the very small sample size. Again, we're stuck looking at pitcher K rates and hitter's power numbers to quickly identify the best prospects.

Anyone interested in Joe's recent work should contact him here to subscribe to his off-season newsletter for a reasonable fee.

Following Joe, BBHQ's Doug Dennis gave his first Symposium presentation. We're rather fond of Doug since he chauffeured us around Phoenix at our first Symposium(when we were sans rental) back in 2000. So we were quite pleased with his talk about bullpens and relievers, even if we were very familiar with everything he discussed.

He highlighted the ARP numbers at Baseball Prospectus as the best source of objective evaluation of reliever performance. Included in a rather lengthy handout were lists of relievers' skill ratios, as well as a detailed look at Detroit's bullpen. For those attendees without a deep background in bullpen science, he provided some welcome tools for predicting future closers.

Mat Olkin from Sports Weekly followed Doug, and he presented the same intriguing mix of player evaluations and sabermetric tools as in past years. Using MLEs, Mat identified Victor Martinez, Brandon Phillips, and Josh Phelps as prepared to succeed in the majors while Drew Henson shows no indication of every hitting at the upper levels of baseball.

He discussed the difference between actual ERA and predicted ERA, the latter calculated by multiplying a pitchers' OOBP by OSLG by the standard constant of 31(OOBPxOSLGx31). Among pitchers who should find additional success in 2003 are Miguel Batista, Dan Wright, Ryan Rupe, Mike Koplove, and Cliff Politte, while Elmer Dessens, Tomo Ohka, Casey Fossum, and Jorge Julio should struggle.

Mat coined perhaps the funniest phrase of the Symposium when discussing Grant Roberts. He suggested that the Sports Weekly' headline writers were disappointed that Roberts never had the chance to hit against young Braves' lefty Jung Bong, depriving the world of the potential banner of "Roberts giddy after hit off Bong".

Every year he details a new bunch of overworked pitchers, and about half the pitchers Mat names each fall wind up injured or very ineffective next season. Mark Prior and the young Padres, mainly Oliver Perez and Jake Peavy) ranked among the most worrisome youngsters, although C.C. Sabathia and Mark Buehrle have appeared here two years straight; we strongly recommend against acquiring either at market value. Focus instead on pitchers like Kip Wells and Aaron Harang.

Mat concluded by returning to the G-F analyzation that we so enjoyed last year. Among the hitters he identified to see a nice jump in homers in 2003 were Eric Chavez, Brian Giles, Carlos Lee, Frank Thomas, and Nomar Garciaparra. Other power sources that we can rely upon include Joe Crede, Mark Bellhorn, and Carlos Pena. Lastly he looked at pitchers with both great G-F rates and solid strikeout rates, identifying Carlos Zambrano as perhaps the most intriguing target for 2003.

You should consider one minor note when researching G-F numbers. The G-F data at ESPN is from Stats, Inc., but the numbers at MLB.com are from Elias, and each company uses a different method of categorizing hits into groundball or fly balls.

The remainder of the morning flew by rather quickly since we were already running behind schedule. Ron Shandler gave his annual lecture, discussing bias in choosing fantasy players. His list of "dangers of perception" included preconceived notions, long and short memory(Frank Thomas, Francisco Rodriguez), media influence(Blalock & Beckett in 2002), local bias, personal bias, and statistical bias. As a group, we discussed the recent stat histories of five players without knowing their names to project their probable futures, and we identified Greg Colbrunn as someone who could make a significant contribution to fantasy teams in the near future depending on what team he joins.

For the last morning sessions, the attendees divided into three groups: Challenge players with Gene McCaffrey, Scoresheet players with Jeff Barton, and Roto owners with Alex Patton. John Hunt was supposed to join Patton but he was unable to attend the Symposium. Patton led a discussion on perceived values vs. actual valuation as a prelude to a new experts' keeper league that was drafting on Sunday night, and among the "surprises" for most people was that Soriano earned more 4x4 value than ARod, Gagne earned more than Randy, and Chris Hammond finished very close to Matt Morris.

Perhaps the most vital information was that, based on the projected 2002 dollar values of Patton and Hunt vs. the actual performance of players, the conventional wisdom to spread your money and avoid $30+ players wasn't supported by the data. The most expensive players were generally more likely to return a small profit than anyone in the $20-$30 range, however we'd need to see a few years of similar data before changing our belief on spreading around cap money in most leagues.

The afternoon game, featuring Grand Canyon at Phoenix, was a complete disappointment for several reasons. They only played seven innings due to the lack of available pitchers, we'd already seen both teams in action, and no impressive prospects pitched for either team. Also, the concession stand ran out of hot dogs and wouldn't let us pay for the food and then wait until the new batch was ready; we were supposed to go to the end of the 15-person line and wait for another three innings.

We have no idea if the Florida Marlins were responsible for the concession operations at Phoenix Municipal Stadium.

As most of the players in the AFL were in their tenth month of baseball this year, few players hustled, and even clocking players running to first didn't appear reliable. About the only good scouting data we tabulated was that Philadelphia outfield prospect Jay Sitzman is extremely fast; we're confident in clocking him at a major league-average 4.3 seconds down the line to first base. Tampa lefty Mark Malaska also owns a very consistent release point.

While we weren't too pleased with the Saturday afternoon game and didn't learn as much as we'd hoped during the morning sessions, Sunday looked brighter as the first two speakers were former Houston Assistant Scouting Director David Rawnsley, followed by AAA Tucson and Scottsdale manager Al Pedrique.

I'll continue tomorrow with a recap of that fourth day of the Symposium. While David Rawnsley's Scouting Seminar actually began Saturday night, I'm going to wait to discuss it until this coming Monday so I can first conclude the regular sessions and AFL games.

Today's Fantasy Rx: If anyone reading this is either currently or was recently involved in high school baseball, I'd love to know if anyone kept pitch count data. Any team considering investing a very high pick in a high school pitcher should find some way to tabulate the pitch count data from his senior season, especially since we're not aware of anyone keeping track of that data. We'd think that teams could at least convince players' girlfriends or parents to track pitch counts since they stand to directly benefit from keeping a top prospect healthy.