granted, but Magglio Ordonez is definitely a better second half producer. IN the last 3 seasons, his average jumped an average of 30 points in the second half each season. 3 seasons is a pretty large sample and consistent pattern.
An unanswered question is better than an unquestioned answer
shuffcard wrote:halladay got off to a horrible start last year before lighting it up the rest of the year.
Go take a look at who he was pitching against in April and May. This is another thing you have to take into account with splits. With an unbalanced schedule and some divisions weaker or stronger than others, a pitcher's numbers may tell you more about the competition he faced that month than how well he pitched.
New Zealand Fan wrote: Guys that do repeatedly heat up after the break are
Jeff Bagwell Miguel Tejada Eric Chavez Kelvin Escobar
Only a small list though isn't it?
Tejada does have slightly better numbers after ASB than before (263/320/447/767 versus 277/343/474/817), but his second best month for hitting is May. While he was better after the ASB in 2002 and 2003, he was better before in 2001.
In 2002, Chavez was better pre ASB than post ASB. Lifetime, there's literally no difference in his numbers in April, June, July, and September. The only monthly variation is bad Mays and great Augusts.
Similarly, early in his career Bagwell was a noted "early starter". Lately, he's been slow to start. Pattern or random? It's just too hard to tell.
There is so much random variation in player's performance that it is really dangerous to make decisions based on these data.
Kelly Gruber wrote:granted, but Magglio Ordonez is definitely a better second half producer. IN the last 3 seasons, his average jumped an average of 30 points in the second half each season. 3 seasons is a pretty large sample and consistent pattern.
Lifetime, Magglio Ordonez
Pre ASB
306/368/536/904
Post ASB
309/362/517/879
An even larger sample says it's not a consistent pattern.
Yes, but what is more significant in predicting his future? Recent trends or trends from his early carreer? Obviously, his last 3 seasons are much more of an indication of how he'll do next year than his rookie season.
Join my league, draft Magglio and offer me a trade around the all-star break. I'll happily accept it (assuming it's fair).
An unanswered question is better than an unquestioned answer
Kelly Gruber wrote:I take it you don't like many stats (or overuse of stats). IN that case, which are indicative of future success, or which are meaningful?
No, anyone around here can tell you that I love stats and use them almost exclusively in my decision-making. What I'm talking about is the misuse of stats in making decisions.
If you are talking about during the course of a season, the best predictor of a player's performance in any given month is his overall past performance, appropriately adjusted for age. In other words, long-term past performance is the best predictor of future performance. My best guess of Magglio's performance in the first half of the season would be some version og a weighted average of his overall three-year trend for the full season, adjusted for age.
Since I think a minimum of 3 years of data is necessary to get a good prediction, a minimum of 6 years of half-season data would be necessary to predict half-season performance. In addition, you'd have to do some work to remove ballpark and unbalanced schedule effects in that half-season data, which taint the results.
Cafe Home
Fantasy Football
Fantasy Basketball
Fantasy Hockey
by shuffcard » Mon Mar 15, 2004 11:23 pm 
