sheesh, this team has on-base problems. Five projected starters--Cabrera, McPherson, Garret, Kendrick & Napoli--are below the .330 mark. The pitching, as one would expect, looks terrific though.

This team changed its name again?

Someone please give Rivera a full year, full time job

Free Reggie Willits!

Someone please give Rivera a full year, full time job

Not until he returns Derek Jeter's glove.

take anderson out back and shoot him, for everyone's sake. what a terrible contract (and one that primer groupthink actually got right - hip hip hooray!).

dan, how well-informed is ZIPS of mcpherson's injury history?

That projection for Erstad looks too high, 705 OPS??? He's bettered that mark just once (2004) in the last 6 years.
.691, .702, .642, .746, .696, .605...

Marcel probably brings in Erstad at about... OK Brock2 puts him at .671 (which still looks too high to me for a 33 year old in an apparent freefall..)

I've got to love any pitching staff projected to have Bartolo Colon as the 6th best starter.

sheesh, this team has on-base problems. Five projected starters--Cabrera, McPherson, Garret, Kendrick & Napoli--are below the .330 mark.

Very true. This team needs another bat, but with the OBP challenged hitters they have, someone like Soriano is not the answer.

Someone please give Rivera a full year, full time job

He has one. Rivera was injured part of last year. When he returned he was an everyday player for the rest of the year.

That projection for Erstad looks too high, 705 OPS???

1. Beware of trends. They are not predictive.

2. The .605 and .642 come in years where Erstad had very few plate appearances. The other years (which at first glance seem pretty close to a .705 average) are seasons where Erstad played pretty much every day.

3. A .705 OPS isn't good enough to get him a starting job anyway.

Can Erstad still play an above-average centerfield?

Can Izturis play a good SS? Because those numbers would be mighty fine if he could.

2. The .605 and .642 come in years where Erstad had very few plate appearances. The other years (which at first glance seem pretty close to a .705 average) are seasons where Erstad played pretty much every day.

3. A .705 OPS isn't good enough to get him a starting job anyway.

2001 711 PA OPS+ of 78
2002 663 PA OPS+ of 88
2004 543 PA OPS+ of 95
2005 663 PA OPS+ of 89 (as a 1B)

What happen dto him anyway?
After his age 24 season Brock 2 had him peaking at a .319 average, 104 rbi and 31 homers
2490 career hits, .290 average
after his age 25 season it downgraded him tremendously: 1607 hits, .276-186-770
after his age 26 season he was back on track, 2573 hits, .297-266-1161
Now? it has him with 150 ab left in his career (I had to lower the "floor" to get it to project him with ANY 2007+ playing time)

was it really just one nagging injury after another?

Can Erstad still play an above-average centerfield?

Yes, for about 3 games. Then he'll hurt himself and miss 2-3 months.

dan, how well-informed is ZIPS of mcpherson's injury history?

Yup, he looks good in Dial's numbers (i still have the spreadsheet to finish cleaning up for upload). I have a version of Dial's stuff with Chone's park factors and have Erstad at +4 in 2006 (+25! per 150 defensive games)

What happened to Erstad?

I wish I knew. I've watched 90% of Angels games since 2002 ( when I first got the baseball cable package). But I don't have the slightest clue. Injuries haven't helped, but a lot of players get banged up. I can't think of any who showed as much early promise as a hitter and produced so little, other than old player skill guys like Ben Grieve or players who didn't work or hustle enough (which certainly does not describe Erstad).

I call over on Weaver's ERA. If gets under 3.30 in a full season I'll buy a Rally Monkey and put it in my car's back window for a year.

Cabrera, McPherson, Garret, Kendrick & Napoli

Napoli is certainly a different sort of low-OBP hitter than the other four; he doesn't lack for on-base skills, he's just projected to hit .222. On second thought, I guess you could say he severely lacks the primary on-base skill.

Vlad is projected to have his worst season since 1997 and he's still by far the best hitter on the team. His walks really took a nosedive when he joined the Angels, and for no apparent gain.

"His walks really took a nosedive when he joined the Angels, and for no apparent gain."

Hatcherball!

Just eyeballing it, those look like some very low projected BABIPs. No?

dan, i have a feeling you weren't responding to my comment about mcpherson's injury history there... either that, or i'm a retard.

I think Dan is tired of repeating his disclaimer that Zips doesn't predict playing time- and may not have realized that the question had more to do with the potential impact an injury like McPherson's has on rate production

Just eyeballing it, those look like some very low projected BABIPs. No?

Not really, no. There are two pitchers below .280 (Santana and Weaver) and 4 above .300 (Gregg, Moseley, Arredondo, Bulger). Everyone else is between. The Angels have had good defenses, but not as good as a few years ago.

right that was my question - surely a back injury affects more than just playing time.

Not really, no. There are two pitchers below .280 (Santana and Weaver) and 4 above .300 (Gregg, Moseley, Arredondo, Bulger). Everyone else is between. The Angels have had good defenses, but not as good as a few years ago.

Huh, guess not. Probably just the strong K rates up and down the staff, though Carrasco still looks low.

dan, i have a feeling you weren't responding to my comment about mcpherson's injury history there... either that, or i'm a retard.

Kyle, I cut-and-pasted the wrong line.

ZiPS doesn't know anything about McPherson's injury history other than changes in patterns of his at-bats.

I'm trying to keep human input out of ZiPS as much as possible. I'm not so much trying to make the best projection system as I am trying to make the best computer projection system, if this makes any sense.

I feel that when we have human inputs within projection systems, it makes the projections less useful to people even if they are slightly less accurate. This might sound counter-intuitive, but I feel that when certain non-statistical things are included, it makes it harder to analyze players individually.

Take for example, an injured player who's very big and muscular that hit 230/300/320 in some year. If the computer spits out a 260/320/370 projection, I know that it's based on statistics and that the computer is basing none of its judgment of a player's injury or the fact that he's a big guy that should add power.

But now, what if I know that the 260/320/370 projection instead contains something about his injury and the player's physical attributes? I have no idea what part of that projection is stats and what part of that projection is other. No additional knowledge about a player's injury or physical attributes can now be used to enhance the quality of that projection when predicting the future.

I think my goals are a little different than other poeple doing projections. I'm not trying to make the best projections possible, I'm trying to make the best computer projections possible.

Perhaps someone can explain this better than I - I don't think I've ever really been able to express my goals completely accurately.

I don't think I've ever really been able to express my goals completely accurately.

a lot of time has been spent on this. Think the right answer is 42

Dan, I completely get what you're going for, and think that it's a very worthwhile goal. Keep up the good work!

Agreed Dan. Projections have to be taken with a grain of salt and shouldn't be viewed an an end-all to evaluation of players. Well done.

I'm trying to keep human input out of ZiPS as much as possible. I'm not so much trying to make the best projection system as I am trying to make the best computer projection system, if this makes any sense.

Actually, it does. It's nice to be able to compare your purely computer-based system with, say, Sickel's purely human approach. Both have their merits.

thanks for the response, dan. i didn't mean to be glib or anything; sorry about that.

i think i take your meaning, too. the computer does the computer part, the part that is tedious to do by hand (age adjusting, park adjusting, 3/2/1 weights, etc).

thanks for clarifying.

I've done a few projections so far for the CHONE system, no entire teams yet. I'll try and post them sometime before the season starts. This year I'm doing my projections side by side with the Marcel formula, since its easy enough to do.

I'm kind of interested in where there are significant differences between CHONE and ZIPS. So far, there doesn't seem to be a whole lot of difference.

Dan, thanks for the explanation. Your argument is compelling. It allows us to make individual assessments without having to crunch any numbers and without having to what mysteries are included in the formula. Since I don't know how to crunch the numbers but I do like making individual assessments I love ZIPS.

Dallas McPherson* 263/323/503 with 136 K in 380 AB

No
Freakin'
Way

Not as a mean projection anyway. To do this, McPherson has to hit 410/783 on contact. That's certainly not impossible, but that can't be his expected performance. For his career in the majors he's 372/695. To his credit, his MLB K-rate is much lower than his AAA K-rate.

I'm surprised no one has weighed in on Brandon Wood yet. He could be interesting data point for what excessive strikeouts in the minors means for MLB performance...

Brandon Wood is not going to be in the majors in 2007, at least not until late in the year.

Dan, I don't get it. I think that all projections should be 100% computer-based, but that doesn't mean you cannot use more refined inputs, i.e. varying aging patterns or accounting for player size.

McPherson is not a projected starter.

And though we'll never find out, I don't think there's any way in hell Brandon Wood would hit that mean projection in 2007.

FACT: I will be beyond myself with joy if the starting pitching can match these projections.

I think that all projections should be 100% computer-based, but that doesn't mean you cannot use more refined inputs, i.e. varying aging patterns or accounting for player size.

Oh, I use varying aging patterns as well. I don't agree with the latter, though - I think that creates a bunch of noise due to the fact that specific accurate data is hard to come by and that height and weight have so many different configurations and even if we could get that for the past, we probably wouldn't then have enough in baseball history for it to help us much.

After all, I'm a little bit taller and a little bit heavier than LaDainian Tomlinson. I'm guessing he's a bit harder to tackle.

To elaborate on Wood, his MLE last year was 216/284/396, so hitting 248/303/444 would be a pretty big jump. I'd be stoked with an MLE of 230/300/420, actually, as that would show some progress. I guess that's not too from off from the projection.

Basically, I want to see him hit at least 300/370/520 at AAA.

***

Erstad:

My view on The Punter is that he can only hit a straight pitch right at the knees (either fast or change) and has no plate discipline. His swing is also top-hand dominant, which leads to 50 groundballs to second every year, some of which become PRODUCTIVE OUTS!. He just doesn't seem to hit the ball as hard as he did back when he was good, which may be an artifact of his throwing his body around like Paris Hilton at a sailors' convention.

Someone please give Rivera a full year, full time job
Can Izturis play a good SS?

Trader Jim Bowden sure knows his horseflesh.

I'm skeptical about Ztu's defense. He hasn't looked very good at third. Haven't really seen a lot of him at short due to The OC. I doubt he's a disaster, but I'd be surprised if he were very good.

I suspect Izturis' defense at short would be notably better than it is at third. He's said before he'd never taken more than a few grounders at third before being called up. From my personal experiences in high school and in fast pitch rec leagues, the adjustments from short to third base are surprisingly demanding. That said, I agree that Izturis' glove at short would probably be average at best.

I really hope Stoneman finds the Angels a bat. That offense looks pathetic.

Erstad hurt his legs and lost his power.

There's got to be more to it than that. Erstad last hit for power in 2000. In 2002 he played almost every and had one of the greatest defensive season in the history of play by play metrics, while stealing bases at a high percentage. His legs couldn't have been that bad.

Not until he returns Derek Jeter's glove.

Wrong Rivera... still funny though.

Hmmm... OBP challenged team with a glaring need at CF and a revolving DH spot... anyone else think Stoneman needs to get Boras on the phone about J.D. Drew right now?

God, I have a hard time believing Weaver's ERA will be a full run lower than Verlander's over the same workload.

I don't. Think about it... which one plays for the Tigers, and which one doesn't?

I think my goals are a little different than other poeple doing projections. I'm not trying to make the best projections possible, I'm trying to make the best computer projections possible.

What you're trying to do is come up with a reasonable set of "objective" projections. If you included more "subjective" information, it would be very difficult to know what the system was doing and what you were doing. By providing us (thanks, BTW) with the purely "objective" numbers, we can actually be as subjective when we interpret them as we want.

Granted, any system has elements of subjectivity as to what you include and what you don't include, but it sounds like the ZiPS system is as objective as one could reasonably expect. So people can take the ZiPS projection as your baseline and bias them from there with as much subjective information as you want to/are able to use.

I don't. Think about it... which one plays for the Tigers, and which one doesn't?

Really? You're taking the flyball pitcher on a team with a pretty abysmal outfield? Over the younger guy on the best defensive team in the AL?

Um.

OK.

Besides, I just think Verlander's the better pitcher:

Weaver's xFIP 4.66 (over 123 innings)

Verlander's xFIP 4.67 (over 186 innings)

And that includes Verlander's second half swoon.

Verlander's ZiPS seems about right. Weaver's looks insane. (CAN'T WAIT to see Liriano's... woulda, coulda, shoulda.)

Has anyone put forward any evidence that xFIP is any more reliable than FIP, DIPS, or plain ol' ERA? This is a serious inquiry.