Monday, March 31, 2008
Somebody's Gonna Be Wrong
Comparing preseason predictions from various sources is a lot of fun, and also a pretty good way to get a feel for what's expected from teams from various sections of the media. What I've done here is take predictions from a few different places- 5 ESPN analysts in their season preview, three Yahoo! guys, the SI staff, Joe Sheehan (AL, NL), and PECOTA- and find the biggest discrepancies bewteen win totals for each team. The first largest differences are below, followed by a discussion of why there's such a lack of consensus, and who looks to be correct.
Tampa Bay Rays
Average: 77.3
High: PECOTA, 88
Low: Steve Henson (Yahoo!), 72
The thing you have to love about PECOTA is that it's 100% unbiased. When it runs the numbers and comes up with 88 wins for a team that's never won 70, it doesn't adjust that to something that seems a little more reasonable. This paid off with the White Sox prediction last year; considering its history of success (not limited to that one example, obviously), the extreme predictions for Seattle and Tampa are hard to ignore.
I don't really know who this Steve Henson fellow is, but that's okay- he's got some wacky predictions, which are always appreciated. Here is his analysis on the Rays:
Combine that with the addition of Matt Garza, and the progress of Kazmir, Shields, Sonnanstine and Co., and it's easy to see that their run prevention will be much improved. PECOTA has a team that allowed 944 runs last year decreasing that by a whopping 226 runs. Without looking it up, I'm going to go ahead and assume that that'd be the largest reduction in the history of baseball; that's about three months worth of runs for the Giants' offense.
Henson's prediction of 72 wins for the Rays is insanely low; 88 is high, but not that high. It's hard to both see and quantify these internal improvements- switching up defensive alignments, young players improving, old ones regressing- which is why PECOTA is so far off from the general consensus.
Somebody's Gonna Be Wrong
Comparing preseason predictions from various sources is a lot of fun, and also a pretty good way to get a feel for what's expected from teams from various sections of the media. What I've done here is take predictions from a few different places- 5 ESPN analysts in their season preview, three Yahoo! guys, the SI staff, Joe Sheehan (AL, NL), and PECOTA- and find the biggest discrepancies bewteen win totals for each team. The first largest differences are below, followed by a discussion of why there's such a lack of consensus, and who looks to be correct.
Tampa Bay Rays
Average: 77.3
High: PECOTA, 88
Low: Steve Henson (Yahoo!), 72
The thing you have to love about PECOTA is that it's 100% unbiased. When it runs the numbers and comes up with 88 wins for a team that's never won 70, it doesn't adjust that to something that seems a little more reasonable. This paid off with the White Sox prediction last year; considering its history of success (not limited to that one example, obviously), the extreme predictions for Seattle and Tampa are hard to ignore.
I don't really know who this Steve Henson fellow is, but that's okay- he's got some wacky predictions, which are always appreciated. Here is his analysis on the Rays:
"The Rays are improving but are still middle-school level to the Red Sox graduate students."
This is a little over the top, but I think that's the mainstream consensus. Personally, I have no idea how many games this team is going to win (although I'd certainly take the over on 72). There's no big Pythag gap here- last year their expected record was 67-95, and their actual record was 66-96. Three things are causing the huge expected jump- a vastly improved defense, additions to the bullpen, and the development of young players. They were a horrible fielding team last year, but PECOTA expects them to be a little above average this season. The biggest upgrade is going from Brendan Harris (-19 in Dewan's system) embarrassing himself at short to Jason Barlett's +18 glove. The also have Upton finally spending a full year in center, and the (eventual) addition of Longoria to the lineup will allow Iwamura to slide over to second.Combine that with the addition of Matt Garza, and the progress of Kazmir, Shields, Sonnanstine and Co., and it's easy to see that their run prevention will be much improved. PECOTA has a team that allowed 944 runs last year decreasing that by a whopping 226 runs. Without looking it up, I'm going to go ahead and assume that that'd be the largest reduction in the history of baseball; that's about three months worth of runs for the Giants' offense.
Henson's prediction of 72 wins for the Rays is insanely low; 88 is high, but not that high. It's hard to both see and quantify these internal improvements- switching up defensive alignments, young players improving, old ones regressing- which is why PECOTA is so far off from the general consensus.
