Two national publications came out at the end of last week and said that the Tigers could be good this year. Now they didn’t say the Tigers would be good, just that they could be good.
First Baseball Prospectus’ Tigers notebook had this to say:
But is the ridicule justified this year? The AL Central isn
Billfer-
A PECOTA question for you: does anyone do a 2 year PECOTA projection? I don’t recall ever seeing it, but it seems like it could be done by plugging in the 1st year projected data as the actual statline.
i would love to know if, using 2 year PECOTA projctions, it would’ve predicted the same ’06 breakout years for Bonderman, Maroth and Roberston last offseason, or if their ’05 performances did something that resulted in the breakout projections.
Zimm,
If you’re a BP subscriber, you can actually see 5 year projections. Looking at the projections for Bonderman, last year they projected Bonderman’s breakout in 07/08. So last year he did enough to shift his comparables to expedite the process.
In the case of Maroth and Robertson, neither had breakouts indicated in their previous 5 year projections.
I’ll take it.
thanks Billfer! i’m not a BP subsriber. i rely on guys like you and Lee to keep me informed. =)
interesting about the influence of ’05 on the projections. i guess my follow-up would be to wonder if by increasing their chances of an ’06 breakout through their ’05 performances, Maroth and Roberston also decreased their chances of an ’06 collapse (which is something that i would worry about for a soft-tosser like Maroth and a sometimes overexposed 2 trick (pitch) pony like Roberston…)
What is PECOTA? From the sound of your conversations it sounds like a great tool. How does it work (in general terms)?
It’s a prediction tool that works by creating a list of comparable players for an individual and then projecting that individuals performance based on how his comparable players performed. It also uses those comparable players to create a predicted distribution of performance for the player, which is where the breakout/collapse figures come in. If a player has a 30% breakout rate, it means roughly 30% of his comparable players performed significantly better than their previous baseline performance.
i was once skeptical of PECOTA, but became a believer in its relevance when it accurately prediicted Willy Mo Pena’s breakout year (the projections seemed absurd at the time)
You have to be carefull when it makes projections based on low minors data or on guys with little to no track record, but that should be self evident. It’s really pretty good for general projections, though I’d like to see someone actually take a look at whether the distributions actually work out. It would be tough to do though, because players that do bad (fall in the bottom 10% of their projections) tend to stop getting playing time, while players that perform good (top 10% of their projections) tend to get more playing time.
Health (Guillen, Ordonez) and older players fighting off decline ( Rogers, Jones, Pudge). These are the huge keys to the season. Those and improvement by young pitchers.
I expect both Chicago and Cleveland to come back to the pack a little. Minnesota should be better. The Tigers MUST put up a decent record against these teams to have a shot (19-37 last year).
Leyland in the Freep today says in essence that Young, Monroe, Logan and Granderson will all get their ABs. They really need Granderson in there full-time, and they need for him to learn to hit lefties as well. Doesn’t sound like he will get the chance.
I really hope Logan doesn’t get 100 starts this year. Ok, I really hope he doesn’t get any starts this year, but since I know that isn’t going to happen, I’d settle for him being limited to playing only occasionally against righties, and as a defensive sub/pinch runner.