Steve-o wrote:Philliebuster wrote:Steve-o wrote:Philliebuster wrote:I'm glad you posted this because I just received the same results for Huff, his 29 hr and 100+ RBI seem inflated, and do not correspond to the PECOTA card. None of the projections really exactly correspond to the PECOTA cards, but the Huff projection is way off.
That's because the PECOTA cards used the weighted mean, whereas the PFM tries to incorporate the actual playing time of the ball player.
Can you please elaborate? I need a idiots guide to these Pecota cards. So the Pecota cards don't necessarily take into consideration playing time? Nor the Five-year forecast? Hmm, they're pretty high on the Huffster then. I may have gotten in a bit over my head with these PECOTAs.
I am happy to share what I know, and then someone can correct me if I am wrong.
It is my understanding the the PECOTA cards are similar to what you see in the actual book (if you have gotten it in year's past). These are the weighted means, which a fairly conservative when it comes to playing time. For example, the weighted mean for Fielder has him getting 575 PA. That's fairly low for an everday young first baseman. When you look at his numbers through the PFM, which is geared towards fantasy, it shows 600 PA. That's a pretty significant difference.
It's not that the card doesn't take PT into consideration, it's just the the PECOTA card forecasts for PT are conservative. Same goes for the 5 year forecast.
Frankly, if you were to do a fantasy draft off PECOTA projections from the book or the cards, I don't believe you would do very well because all you counting stats would be skewed low. I think, in order to really take advantge of the PECTOA system, you need to either combine it with another forecasting system that predicts playing time (I have heard Fantastics does a good job with this) or you need to adjust PECOTA forecasts on your own for PT.
Random note: I read that PECOTA forecasts are done on Excel and take 4 straight days of CPU time. You think there would be a better program for that.
This is more or less correct. I use PECOTA and Forecaster to judge players, I look at the other systems for players that they disagree on greatly and then look at those players more in depth. I use a combination of forecaster and common sense to project actual AB's and IP's.