The Braves didn’t want to shift Chipper to third base, where he’d block Andy Marte, but really had no choice. He couldn’t stay in the lineup in left field, and Mark DeRosa was killing the team with horrible all-around play. Moving Chipper to third improved the defense at two positions and replaced DeRosa’s bat with a good platoon. What it didn’t really do was energize Chipper, who didn’t improve as much as you think.
Chipper as a left fielder, 2004: .242/.395/.453.
Chipper as a third baseman, 2004: .261/.368/.501.
Not all that big of a difference, really, but his best work came after the switch. He was generally mediocre all year, except June, when he was awful, and August, when he was great. Yes, his second-half numbers were much improved as a whole, but that was on account of August. In September, he hit .222. And his post-All-Star numbers weren’t exactly MVP caliber either, .278/.391/.548. It’s mostly the batting average that’s down, though. His career OBP is .401, his career SLG .537. If he could get back to hitting .300 his numbers would look a whole lot better.
Chipper’s health is a concern; if healthy, you have to figure he’ll at least match what he did as a third baseman last season. He was able to stay in the lineup at third base, unlike left field. The sabermetric line is that Chipper’s defense will kill the Braves, but his range factors are pretty solid; I suppose there could be a statistical illusion there but I don’t see where it could come from… Chipper has slowed down a lot in the last three years and doesn’t run much anymore. Last year, he attempted only two stolen bases (he was successful both times).
Chipper’s MVP season days are probably over. But if he hits his averages then he is in the top 4 3b in the league. I always thought the move to LF was dumb. He looks lost out there. Even if he is a below average defensive 3b he is still more valuable there than in left. Because I’m confident that Jordan will suck and not so confident that Langerhans is replacement level I wonder what the Braves will do if Marte starts to blast AAA pitching. All descriptions say that he is a good defensive 3b and has never played the OF in his pro life. Ya think Chip would move back to LF again?
I’ve seen the BP and Dayne Perry stuff about Chipper being a defensive black hole but I watched him a lot when he moved back to third and just didn’t see it. I mean he ain’t no Brooks Robinson but my observations were that he was playing a pretty good defensive 3b.
Chip is my favorite player on the Braves. I hope that he can stay healthy becuase the team is going to need his bat more than ever this season.
I’ve read several reports saying that Chipper’s offensive numbers are trending down the last three years and will continue to head that way & that he is no longer a real impact player. “Trend” or not, his numbers were still pretty good in 2003, and he will have to stink this season before I believe 2004 was anything but a horrible aberration, attributable more than a little to his injuries (he still racked plenty of RBIs, though some stats freaks will probably give all the credit there to the guys who were on base when he came to the plate). This season will be pretty telling in answering the question whether Chipper’s career really is on the big decline. A lot of experts are writing him off. I hope (and suspect) they might be surprised.
A couple of things. I remember hearing ruminations a couple of years ago that Larry may have been on the juice too, hence his recent drop in weight. I wonder if his decline has anything to do with that. Also, what are the odds of him winding up his career in Atlanta and HOF chances. I think he is close to a lock for the Hall myself.
Chipper’s power numbers didn’t drop in 2004; in fact, his isolated power was up, from .209 in 2002 and .212 in 2003 to .237. It’s all batting average, and a lot of that is strikeouts. He struck out more last season than any time since his rookie season, despite setting career lows for games and at-bats. But that shouldn’t have caused a sixty-point drop in batting average all by itself, so some of it’s probably just luck.
Chipper should go into the Hall if he stays at third base.
Johnny, his BP defensive numbers aren’t even that bad. He’s about halfway between replacement level and average. Obviously, that’s not good, but it won’t kill you. He was very bad in 2001, but a lot better last year.
Claims of Chipper’s demise are premature. Most of his problems last year were injury related (though they could come back), not position related. Chipper is like a diesel engine. He takes a while to get going, but when he does he can carry you for a long time. Last year he kept getting interrupted with injuries before he ever got going. And they would be spaced out just enough to keep not let him play and not do much before getting hurt again. If healthy this season will be significantly better than his second half last season. I don’t think he’ll put up as many HR’s as he use to back in the day, but he’ll cut down on strikeouts and hit about .300 again.
I’ve never seen his defense as particularly bad. I’d say it’s about average. He’s capable of making some tough plays and hardly ever bungles routine stuff.
I think Chipper will go down as one of the 5 best 3B to ever play. He is already in the top 15 and climbing. All he needs for the Hall is solid (not spectacular) years from here to 35 to add to his counting stats and he is a shoe-in.
If he plays ’til he’s 40, he would need to average about 165 hits and a touch under 25 homers to get 3000 and 500. If he does that, he is inner-circle HOF 3b with Matthews, Schmidt, Brett and Boggs.
For the next year or three I expect .290-.300 with 25-35 homers, and 80-100 walks. In other words, your basic all star level play, but not superstar level play.
But for what its worth, I have always hated his defense. He refuses to get infront of balls, instead chosing to play them off to the side while making a swiping motion to try to catch the ball cleanly. Likewise, he doesn’t seem to move more than a step or so and never dives. Probably the tradeoff in longterm health is good. But it is frustrating to watch balls go just under his glove.
I was doing some projections and I had Chipper at .315/35/105. His 2003 was better than people tend to remember and he never got on track in 2004. I have only found one website that already has their projections up. Is there a site that has a good method for projections? Thanks in advance.
The PECOTA (or something or another)system on Baseball Prospectus projects Chipper with a .277 BA, .382 OBP, .490 slugging, and 22 HR. I’m thinking those numbers, though not terrible, are a bit low. Count one more group of stat freaks as being down on Chipper.
Re the BBall Prospectus projected numbers for Chipper, it looks like they basically split the difference in his 2003 and 2004 seasons in most categories and then really knocked down his HR totals.
It’s actually pretty close to his numbers as a 3B last year. I’m not wild about PECOTA, to be honest; I think it’s trying to do too much and introduces too many variables.
“… I think it’s trying to do too much and introduces too many variables.”
Amen.
Baseball Think Factory’s “ZiPS” has Chipper at .278 BA .388 OBA and .471 SLG with 27 homers and 92 ribbies.
Back when I played rotisserie ball, I used a system much like Tangotiger’s “Marcel.” I’m at work now and can’t get into his, but the link to the entire league’s projections is here:
http://www.tangotiger.net/marcel/index.html
Tango’s is weighted 5 for the most recent year, 4 for the prior year, 3 for two years removed, and 2 for league average then adjusts for age. My unpublished one was 2 parts career (per 150 games) average, 1.5 parts most recent year, 1 part prior year and then a subjective fudge factor based on age, injury, and other non-statistical information.
Before subjective factors, I have Chipper at .289 with 28 doubles 29 homers, 89 walks, a .392 OBA and .512 SLG. For Chipper, I have those subjective factors largely at a wash — he is about to turn 33, an age at which the vast majority of players are declining quickly, but on the other hand, his numbers last year were significantly below expectation due, in part, to an injury to an otherwise durable player.
The probem with the MARCEL system seems to be that it mispredicts the number of plate appearances a young player will have. For example, it has Marcus getting 480 PAs. This would meant that he plays in just over 100 games and it predicts that based on two years of part time duty and one injury riddled year. It undervalues how many games Marcus played when he was a healthy regular.
BP usually breaks down PECOTA at 70th 80th 90th percentile projections. Hank at what percentile were those projections? They seem kind of low.
If Chipper continues to play 3b at or near his career level for the next 2 to 3 years and then doesn’t experience a free fall in his late 30’s he is a Hall of Famer. With he exception of Brooks Robinson third basemen don’t get into the HOF based on their defensive prowess.
The Sporting News:
28hr/101rbi/.292avg
The Fantasy Baseball Guide:
27/97/.290
The Baldbrave Player Predictor:
33/118/.319
I agree Larry Jr. only needs a few more above average years to be a HOFer. As a 33 yr old huge fan of Dale, I need someone from my era to get in there. Lord knows the Veterans Committee won’t vote in Glenn Hubbard or Bruce Benedict.
PS Really good essay regarding Dale and his fleeting chances at being enshrined in Cooperstown.
http://www.all-baseball.com/richbeat/archives/017111.html
It just blows my mind. Chipper missed a lot of games with the hamstring injury, and played in many more while injured, yet had the team high in RBIS and an above average August. I chalk up September to conditioning. He’ll be back at 100% this Spring, playing a position that is more comfortable to him, and put up the best numbers in the NL, if not the Majors.
He’s gonna miss hitting against Randy Johnson (.370 BA) and not look forward to Pedro (.138 BA).
“Hank at what percentile were those projections? They seem kind of low.”
Johnny, I wish I knew to tell you. I can’t figure out their spreadsheet, and I’m looking right at the damn thing.
I don’t see any percentile numbers on it in the ranges you’re speaking of. Maybe the BP braintrust has some map legend with the answer somewhere. I’ll try to see if I can SABER it out. (I knew I should have subscribed to Baseball America, instead.)
FYI, Tango’s 2005 Marcel for Chipper is .284/.386/.498 with about 25 homers. That seems pretty reasonable to me.
I think the quoted PECOTA forecast is probably either the 50 percentile projection or the weighted mean of all 10 percentile projections. It’s low because Chipper’s comparable players haven’t done so hot as they got older; let’s hope that he doesn’t turn out like them.
If you look closer at the projections, they are terrible. If his projections are correct, then no pitcher will win more than 14 games or pitch more than 200 innings next year. It also assumes that just about every player will have a signigificant decline.
Mathew: Are you referring to the Marcel or the Pecota projections? I’m looking at the hodgepodge of numbers on the Pecota spreadsheet for pitchers but cannot find a W or L column. Am I looking in the wrong place, or has the significance of wins and losses been assigned to the ash heap of history as far as the Baseball Prospectus guys are concerned? (I’m sure to some, those are just matters of pure luck.) I wonder, are there articles out there where someone has gone back and compared these highly specialized projections with what actually happened for the seasons they were charted for? Reading a lot of the articles on BP, with their ubiquitous charts and their recondite mysteries to reveal, I come away much of the time with the same reaction caused by listening to papers given at MLA conferences: “Uh, whatever.”
I am looking at the Marcel. He admits on the website that he didnt put much time into it and the results bare that out. What methods do other amateur sabrmetricians use for projections? I am playing with numbers but I am not happy with the results. Subjectivity seems to be an important aspect, especially for younger players.
The probem with the MARCEL system seems to be that it mispredicts the number of plate appearances a young player will have. For example, it has Marcus getting 480 PAs. This would meant that he plays in just over 100 games and it predicts that based on two years of part time duty and one injury riddled year. It undervalues how many games Marcus played when he was a healthy regular.
Its a simple system, but there is some truth to the statement that a player who has been injured in the past is more likely to be injured in the future. As a class, this is undoubtedly true. Of course some players will buck the trend. But if you frequently bet against the odds, your money goes to the house. Betting that a player like Andruw Jones will play 150+ is a much sounder bet than that Marcus Giles will.
For what little its worth, prior to subjective adjustments, my simplistic system has Giles at 490 ABs even though the heaviest weighting (44.4%)is given to an assumption of 150 games, a total he has never reached. For Giles, his figures are weighted 2×150 games, 1.5×102 games (2004 season) and 1×145 games (2003 season) or approximately 133 games played.
My other figures have Giles batting .304 with an OBA of .370, SLG of .473, 15 homers, 34 doubles and 51 walks. Subjectively, I bump the performance up a pretty good bit due to his age but I leave the playing time projection alone.
Giles is one of my favorite players. I would love to be wrong. But I can’t see a reasonable projection (as opposed to highly optimistic) that he plays more than his career high of 145 games.
490 ABs is different than 480 Plate Appearances. A typical 2 hole hitter in 140 games should have about 550 Plate Appearances.
The Marcel system seems to be slanted toward decline and the pitching projections are poorly done. (No pitcher with more than 14 wins i believe)
The Marcel system seems to be slanted toward decline
The system is designed around regression to the mean. A player above average will tend to move down towards average while a player below average will move up towards average. How far they are projected to move to that mean is driven by how much and how well they’ve played in the past.
(No pitcher with more than 14 wins i believe)
I assume you are referring to the Braves. But since last year our leaders were at 15 wins and topped out at 205 IP that doesn’t seem like an unreasonable projection. Our new aces are a pitcher coming off a 12 win, 188 IP season and a 38 year old who hasn’t had 200 IP in a season since before anyone knew who Monica Lewinski is.
Nope, he has Oswalt and Mulder at 15 wins. Then a bunch at 14. I would be willing to bet the house payment that there will be at least one 16 game winner next year.
I’ve never seen the use of trying to project games played. Injury is too dependent on external variables to accurately predict. My humble advice would be to concentrate on rate stat projections, and leave matters of chance to Fortuna (unless you speak to her directly — if so, please say hello for me).
There is very good data suggesting that a high regression level should be used for pitchers. At the start of last year, what ONE pitcher would you have bet the most money on to win 20? I would have chosen Mark Prior without hesitation and been dead wrong. I ask the rest of you to try that excercise, and don’t cheat.
Matthew, it’s certainly true that there will be a 16 game winner. That’s not the point – the difficult part of projecting pitchers is figuring out WHICH one will win 20 games. For example, in 2003 Esteban Loaiza came out of nowhere to almost lead the league in ERA and wins; his ’03 projection was nowhere near that. Should you have projected 20 wins and ~3.00 ERA for ’04? No – the proper method is to consider his prior history, stir in a large percentage of regression to mean, bake for 20 minutes, and then see what you get – and you still have little confidence. Even Halladay, who was consistently better than Loaiza, missed his ’04 projection big time as well.
Anyway, the Marcel data is most useful not for projecting player performance in the aggregate, but for projecting relative performance. I don’t really care how many wins the AL league leader ends up with, so long as I was able to identify WHO he would be before the season, and draft him for my fantasy team 🙂
I understand this stuff pretty well (as do a bunch of other posters here, obviously) so if you disagree with me or are still unclear, ask, and maybe we can do a better job of explaining our/tango’s reasoning.
I wanted to comment on a story that was posted on Atlantabraves.com on Thursday. It was commenting that Chipper was willing to redo his contract to make the Braves more competitive! He is willing to lower the money that he is making to get or keep talent in Atlanta, mainly he spoke of Tim Hudson.
Is Chipper a cool guy or what! He is the real throwback to the good old days of baseball, when guys just enjoyed playing!
I posted this elsewhere, and it is appropriate here. Perhaps I should it to to my website as well.
=====================
The comments I’ve read here seem to point to some lack of understanding of Marcel.
The Marcel the Monkey forecasting system is simply a weighted average of the last few years, regressed towards the mean based on sample size, and includes a modest aging factor. It represents what any person looking at the back of a baseball card what they think that player will hit in the coming year. Think of it as a poll for expectations.
It is fairly accurate.
***
It is not the best forecasting system, and it doesn’t pretend to be. But, any forecasting system out there should be ashamed of itself for not beating Marcel.
I ran a test a couple of years ago at my site, and Marcel did as well as other professional forecasting systems, and as well as the consensus among 150 fans.
***
If 1 million people play in a lottery where the chances of winning are 1 in a million, I will give the odds of each of those people exactly 1/1,000,000. The *sum* of all those odds will come out to exactly 1. Now, my forecast for any single person is virtually zero, but I still expect someone to win. I just don’t know who.
I forecast a handful of hitters to hit between 30 and 39 home runs. I expect someone (and more likely a few) to hit at least 40 HR. I just don’t have anyone that has a very good chance of doing it.
Same thing with wins, or any other stat.
***
As for RJ or any player that switched teams, this is not built into the forecasting engine. Then again, I said the same thing last year when Javy Vazquez was signed by the Yanks, and Marcel said he’d win 11 or 12 games.
***