Tuesday, August 28, 2007

Projection Scoreboard

Last winter I published my CHONE projections, and Dave Appelman of fangraphs added the projections to his wonderful website, along with other projections systems like ZIPS, Marcel, and from the Bill James handbook. In addition, I have projections from the Hardball Times Season preview, where I wrote the Angels' team section but had nothing to do with the projections, and PECOTA from Baseball Prospectus. A full review of projections will wait till after the season, as who knows what will happen in September, but I couldn't wait and decided to see who projected the Angel players best. The Angels I'm looking at are the 8 who have played the most games at each position in the field, plus I'll count Reggie Willits as the DH even though Hillenbrand still has the most games there this year, because I want to forget he ever played for us, and finally, I'll throw in supersub Mighty Maicer Izturis. For pitchers I'll go with the 5 current starters, Shields, K-Rod, and Speier.

For pitchers the winner is whoever gets ERA the closest. For hitters I've decided to not use OPS, but instead the difference in OBA, and in Slugging. In other words, if you project a .350 OBA and a .450 SLG, and the hitter hits that, your error is zero, but if he has a .400 OBA and a .400 SLG, you may have hit the OPS but you are off 50 points in each measure, so your error is 100. By doing this, I'm not asking "how good is this guy" but more, "what are his stats going to look like?".

And the winners:

C Mike Napoli: Bill James and crew nailed his current slugging percentage at .459. For OBA, they are off by one. Of course this could change dramatically given the Napster's low PA total, once he finally comes off the DL.

1B Kotchman: Thanks to his lost 2006 season, Kotchman's projections are terrible, so the winner will be whoever was most optimistic for him. That would be THT, they projected a 276/345/429 line out of Casey. He has played much better than that, but all the other systems were even lower.

2B Kendrick: THT again, with a slight edge over Marcel the Monkey. They had Howie at 283/323/448

SS Cabrera: CHONE wins this one, projecting a 287/341/411 line. Cabrera has beaten them all so far.

3B Figgins: CHONE does not know Chone. At least I didn't project a .256 average like PECOTA, but this one goes to Marcel the Rally Monkey at 283/347/400. Nobody saw his high average coming. I'm sure most will predict it to go away though. Lets just enjoy it while it lasts.

UT Izturis: Marcel wins again: 276/346/410. Actual numbers: 278/332/415. Believe in the power of the Rally Monkey.

RF Guerrero: Hardball Times takes this one. All are very close. What are the odds, everyone predicts Vlad to have a great year, and he does!

CF Matthews Jr.: ZIPS has him at 274/342/433 against an actual of 264/328/436. Considering Matthews has played a little under the projections so far, and has still been a valuable contributor thanks to offensive versatility, speed, and mostly good defense (remember to catch before you throw), lets just say Angel fans badly overreacted when the deal was first announced.

LF Anderson: Bill James takes this one at 280/317/445.

DH/OF Willits: ZIPS wins at 272/352/346. The slugging was dead on, the actual OBA is 50 points higher, but the other systems had Reggie far below that level of play, except Marcel, which, being only a monkey, didn't know Reggie had no power.

Pitchers later. So far the score is: THT 3, Bill James, Zips, Marcel 2, Chone 1, Pecota 0

Saturday, August 18, 2007

K-Rod and rest

Last night's game, where Vlad bailed out K-Rod, was not a long rest situation, since he pitched the day before in Toronto. The Toronto game, however, was. K-Rod seems to lose his release point when he has a long layoff between outings. Here are his career results by day of rest:

Days ERA BB%
0-1 2.15 10.1
2-3 1.73 11.3
4+ 4.21 9.3

He actually walks fewer batters when he's on long rest, but the ERA doubles. My theory is that the walks don't tell the whole story, if his control is off he has to throw more fastballs and less of the nasty sliders, and becomes much more hittable.

Strikeout rate:

0-1 32.9
2-3 34.6
4+ 29.3

Not a huge difference.

Homeruns per at bat:
0-1 1.6%
2-3 2.0
4+ 6.6

There's the problem. Shaky control means hitters can sit on a get-me-over pitch instead of not knowing when he'll break out the killer slider.

Still, I don't know what to do about it. Giving him the 9th of a 10-2 game instead of Greg Jones just to get some work in may not be the answer. It sure wasn't for Troy Percival.

Friday, August 17, 2007

What a stinker

Playing against the team I dislike even more than Oakland, John Lackey gives up 6 runs in the first inning. That was pretty much the game, as the Angels outplayed the Red Sox for the rest of the game.

To salvage the doubleheader, Ervin Santana returns from his long walk into the cursed earth of Salt Lake to face a dangerous lineup, on the road in a park where he has pitched horribly the only two times he has started there. It has the makings of a massacre, as he'll be facing Josh Beckett, who attempts to do what Lackey could not: win his 16th game of the year. As I write this Beckett is ahead of Lackey among Cy Young contenders, though both have to be behind ERA leader Dan Haren, Our own Kelvim Escobar, who has seriously stepped up to ace status this year, whiff-o-maniac Eric Bedard, and everyyear candidate Johan Santana.

Back to Ervin though, while him facing Beckett looks like a complete mismatch, just one year ago they were comparable pitchers, and in fact, Santana was better. Both pitched 204 innings, and won 16 games. Santana allowed 181 hits, Beckett 191. Beckett threw 36 gopherballs, 15 more than Ervin, and walked 4 more (74). Beckett's only advantage was in strikeouts, but even there its not a huge difference (158 to 141). If you're dogmatic about your DIPS, then Santana was a little hit lucky compared to Beckett. Those 17 fewer strikeouts say he should have allowed 5 more hits than Beckett. That means Santana "should have" allowed 20 more hits in play, but 15 fewer homers and 4 fewer walks while recording the same number of outs. Even if you assume half of those hits in play are doubles, Santana was a full win better than Beckett last year.

Unfortunately, this is 2007 not 2006. I hope we can turn back the clock and get a good performance out of 2006 Ervin while Vlad hits a pair of bombs off 2006 Beckett, but if not, at least I won't have to watch the disaster unfold as I'll be bidding up the price of LaDanian Tomlinson for my fantasy football draft.

Tuesday, August 07, 2007

Hey man, nice shot

Barry Bonds is the new Home run king with a bomb off Mike Bacsik. Everyone expects that one day A-Rod will pass Bonds. Pujols may eventually beat A-Rod. The same night that Bonds breaks the record, 19 year old Justin Upton hits his first (as well as a double and triple).

Some day that might be significant.