CSS Button No Image Css3Menu.com
New! Search comments:
(NOTE: Relevance, Author, and Article are not applicable for comment searches)
I've seen this notion posted more than once that the currently available one year projections are sufficient for those playing in non-keeper leagues. As someone who does not play in keeper leagues, the player cards are still very important to my fantasy valuation.
For example, let's say I don't believe Nate Mclouth's weighted mean PECOTA projection. In this case, I might be inclined to use his 30th percentile projection for my own valuation. Without the player cards, there is no way for me to do this. The player cards are important even for those of us that are not looking for multi-year forecasts.
I'm really interested in seeing how much each of these teams has spent on the draft + international signings over the past three years. I would expect a strong correlation!
More important to me than tiers or rankings are your assessment of how realistic some of PECOTA's more surprising predictions are. For example, you mentioned in the 2B rankings that you don't believe Brian Roberts is going to steal 45 bases this year. That's helpful analysis that the PFM is not going to provide.
I would like to see this sort of analysis highlighted more than it is. Either (1) a special article devoted to people PECOTA loves and hates; or (2) some special syntax highlighting a paragraph in your positional ranking articles where you discuss a player that PECOTA loves/hates. Of these, I would prefer (2). I imagine (2) would also be less work for you.
Tim Alderson in 2009, according to Kevin: "Timetable: Alderson will move up to Double-A as a 20-year-old in 2009, and along with the addition of Bumgarner he should turn the Giants' rotation into one of the most feared in the game within three years."
One year later and Alderson is a two-star prospect. Did his velocity drop even more? Did he have a bad statistical year? What's going on?
Along these lines, how does Sabean do in the post-Bonds era?
I'd like to see each player's age presented inside the parenthesis next to what level they're at.
E.g., Tim Melville, RHP, Royals (Age 19, Low-A Burlington)
Few people could have guessed Dunn would be one of the top hitters in the league this year. PECOTA expected a WARP of 1.1 for Anderson and 2.9 for Dunn in 2009. Given the difference in their salaries and the fact that the Braves added quality starting pitching over the offseason, I don't see the evidence that the Braves' owners "just don't give a damn".
The fascinating thing to me about the Giants this season isn't just that the team is a collection of studs and duds, but how wildly underpaid the studs are, and conversely how wildy overpaid the duds are. Randy Winn is making $8.25M this year, sheesh!
One of the nice things about PECOTA is that it provides a distribution of possible outcomes for each player every year. Setting a strict limit of 10 points in EqA doesn't take into account that PECOTA sometimes predicts higher variance for some players versus others. I think a better way to do this would be to identify "swings and misses" as players who ended the season with an EqA below PECOTA's 10th percentile prediction or above PECOTA's 90th percentile prediction. Doing this would provide some context for the numbers in the first table: if 50% of players are finishing the season above the 90th percentile or below the 10th percentile of PECOTA's predictions, then (I would think) something is going awry in the predictions.
Also, I think it's important to throw out all players with a similarity score below a given threshold. You basically state that we shouldn't be surprised if PECOTA can't predict what Ichiro is going to do, so including these types of players in the analysis is only going to add noise to the analysis.
Like everyone else, I enjoyed reading this article. However, the "Multi-Positional Fatigue Comparison" chart is a bit alarming to me. Taken at face value, it suggests that the DH should be receiving the same pattern of rest as the catcher. The paragraph following the chart offers a few potential explanations for why such a pattern of rest may be appropriate for the DH.
However, I look at the way the DH line wiggles up and down like crazy as a function of consecutive days played, and I look at how the 1B line drops like a rock at Day 8 (and then returns to normal at Day 9), and I wonder if we are seeing statistically significant trends in the data. It would be nice to know (1) the sample size in each bin (we got a little bit of this information for catchers, but not enough for this particular chart) and (2) the rms variation or confidence interval associated with the data in each bin (for example, do catchers on Day 8 have similar OPS+ differentials -- i.e. a very tight distribution; or is there a wide range of OPS+ differential for catchers on Day 8?). I think this analysis would also be useful for the catcher OPS+ differential by month chart as well.