qbert72: no really good resources, like a biography, but if you search on "steve dalkowski" you'll find a fair number of mini-bios and write-ups. Baseball Library has a nice collection of anecdotes by players who saw him. Grum: another good post from the gruminator! While I agree with others that Sinins is a little too grumpy about no-hitter hype (where's the fun in that?!), I think Grum summarizes his basic thesis well- which is also the basic thesis of sabermetric analysis, that we shouldn't get caught up in gaudy but secondary stats like RBI, which are largely influenced by context, and instead should focus on numbers that reflect more accurately the genuine value of a player. As his Edgar stat line shows, the same player had nearly identical years, but the difference was entirely in how many people got on base in front of him, and how many hit well behind him. While those factors are in a GM's control, in terms of asking whether Edgar did his job, or how much he's worth (both in terms of debates like MVP or HOF and in actual hard salary negotiations), it's only valuable to look at the numbers Edgar alone can control. Unlike the other major US sports- basketball, football, hockey, etc- baseball is the only team sport that so thoroughly isolates a player- pitcher or hitter- into very discrete quanta that can be measured. It's why baseball is the most statistical sport: it can be. The whole point of sabermetric exercises is to devise a way to scientifically define a player's performance. This means that- to pass the scientific test- we can somewhat accurately predict future performance, as well as retroactively explain their prior performance. The value of this, as Billy Beane has shown, is that if you can subtract out the hype stats- like RBI- and get a "real" picture of a player, you can find the solid performers who because they didn't post big secondary numbers come cheaper than the "stars", yet when you put a whole lineup of these performers together they all have, happily enough, great secondary numbers. Here's a fun stat/formula most fans aren't aware of: For statistical reasons that make intuitive sense when you think of it, if you take the total runs scored (RS) and the total runs allowed (RA) by a team in the season, you can plug those into the thumbnail "Pythagorean formula" invented by Bill James like so: RS^2 / (RA^2 + RS^2) (runs scored squared, divided by the sum of runs scored squared and runs allowed quared) ... to generate an expected winning percentage, which is remarkably accurate. For example, looking at last year's results, the 2002 Seattle Mariners scored 814 runs and allowed 699. Plug that into the formula, and you get a winning percentage of .5755, which translates to an expected 93-69 record. The actual record for the 2002 Mariners? 93-69. I can tell you from watching this stat for a few years that 2002 was surprisingly turbulent; most of the time, the vast majority of teams fall within 1-2 games of their expected wins, such as in 2001, with only a couple of outliers like the 2001 M's (who won 116 but performed at a level of winning 111 wins) or Colorado Rockies who won an unusual 10 less than expected. As those links show, by the way, ESPN.com conveniently has in their standings page for MLB an "Expanded" option that will show you the Pythagorean expectations through the course of the season. While the season is still very young, and thus these numbers are subject to the curse of a small sample size, teams that through say May/June remain well under or over their statistical expectations (the Cubs -3 and Giants +4, for example) are either killing/dying in one-run games, or due for a "regression to the mean" (just like a hitter who hits .400 in April in May is almost certain to come waaaaay down to earth for the rest of the season). The 2002 Red Sox, for example, *should* have won 101 games with their pitching and hitting, but were also-rans. This tells us that the Sox didn't need to significantly retool, and that they should probably have a season, for better or worse, much closer to expected than was last years. The point of this, and it's a *very* cool thing to impress your less statistical baseball-loving friends, is that if we can predict a team's runs scored from their lineup, and their runs allowed from their rotation, bullpen, and defense, we can get a reasonable estimation of how good they'll be. Too many factors prevent us from knowing exactly how good they'll be, and players are known to have horrible seasons, get injured, or bust out a career year like they've never had, but in general when building a team- or just watching your own team- you can ask these questions. You score runs by getting on base and driving in base runners; combining great leadoff mean with .400+ OBP with great hitting machines like Edgar and their .400/.500 numbers, means the runs pour in. Forget how many RBI that guy has had; if you put the right people before and after him, and he will drive the leadoff guys in and then come around to score later in the inning. If you as a team consistently get runners on, you may occasionally strand the bases loaded, but over a long season those runners will score. And those runs scoring... well, you saw the Pythagorean Theorem yourself. This was the genius of the Billy Beane A's: focus on OBP when other teams didn't even know it was a stat; they stressed walks and plate discipline, even requiring a certain number of walks per month before one could advance to the next level, and the result was a team of mashers who came cheap and scored a ton of runs. Baseballprospectus.com (who publish a print version every year) has running stats that blow the doors of what most "experts" think about baseball, and show how to more accurately evaluate a player based on his actual contributions, and thus determine how much he really contributed to his team winning. It helps us better find those diamonds in the rough who, because of lousy supporting casts, don't get the attention they deserve, and overhyped players who put up great numbers thanks to the quietly consistent efforts of those around them, but aren't really worth the money they'll demand at negotiating time. Beane, and other teams that have succeeded, realized this and would quickly dump an overhyped player making too much money for an equally good one who doesn't have the Sportscenter hype to propel his name.

posted by Hal Incandenza at 11:57 PM - reply - flag: Flag: (Choose)FunnyNewsworthyOffensiveAbusive