An update on Garciaparra: a doctor who examined him on Monday has recommended surgery. According to Dan Duquette, surgery would sideline him from 10 weeks (about 60-65 games) to 16 weeks (about 100 games).

An update on Garciaparra: a doctor who examined him on Monday has recommended surgery. According to Dan Duquette, surgery would sideline him from 10 weeks (about 60-65 games) to 16 weeks (about 100 games).

An update on Garciaparra: a doctor who examined him on Monday has recommended surgery. According to Dan Duquette, surgery would sideline him from 10 weeks (about 60-65 games) to 16 weeks (about 100 games).

An update on Garciaparra: a doctor who examined him on Monday has recommended surgery. According to Dan Duquette, surgery would sideline him from 10 weeks (about 60-65 games) to 16 weeks (about 100 games).

I have to believe the Diamond Mind people ran their simulations under the assumption that everyone would be healthy.

I have to believe the Diamond Mind people ran their simulations under the assumption that everyone would be healthy.

I have to believe the Diamond Mind people ran their simulations under the assumption that everyone would be healthy.

I have to believe the Diamond Mind people ran their simulations under the assumption that everyone would be healthy.

Man, that took me a long time to read the whole article, but it was worth it. I must say that I am surprised that Diamond Mind has the Mariners winning only 91 games. I thought my projection for Seattle, 95 wins, was on the low end.

Perhaps it's the toughest division to predict in all of baseball this year, but I have quite a different outcome for the NL West. Diamond Mind has the Padres in last place, I have them in first!

Diamond Mind's NL West:

SFG 88-74

Jack,

Thanks very much for the link. I read it found the following fact the most interesting: over the last two years, Diamond Mind's average simulation was not as good as most of the other predictors. In fact, they ranked only 15th out of 22. That's not very good for a system which bills itself as so sophisticated. When Peter Gammons can wipe you out, you are necessarily a great prognosticator.

Again, my belief is that garbage in results in garbage out. I don't care if you run it 50 times or 1000. I have a feeling that the reason Diamond Mind has done so poorly is that they are not yet very good at projecting the trajectories of individual players. I think Ron Shandler is a better source for that. If they then used his projections, they would probably do much better with their average simulation.

For what it's worth, I did extremely well predicting the 2000 season, beating everyone listed in the ESPN.com poll. However, last year I was only fair, though still much better than Diamond Mind.

Jack,

Thanks very much for the link. I read it and found the following fact the most interesting: over the last two years, Diamond Mind's average simulation was not as good as most of the other predictors. In fact, they ranked only 15th out of 22. That's not very good for a system which bills itself as so sophisticated. When Peter Gammons can wipe you out, you are necessarily a great prognosticator.

Again, my belief is that garbage in results in garbage out. I don't care if you run it 50 times or 1000. I have a feeling that the reason Diamond Mind has done so poorly is that they are not yet very good at projecting the trajectories of individual players. I think Ron Shandler is a better source for that. If they then used his projections, they would probably do much better with their average simulation.

For what it's worth, I did extremely well predicting the 2000 season, beating everyone listed in the ESPN.com poll. However, last year I was only fair, though still much better than Diamond Mind.

I simmed 50 seasons using High Heat 2003 and I'm happy to report that the Royals won the AL Central 6.2 times and they won the wildcard 3.76 times. Carlos Febles broke the AL single-season homerun record, hitting 65, three times.

Woo hoo!! My Angels have a 6% chance of reaching the post-season. When do playoff tickets go on sale?

What do you mean the Tigers don't have a chance to win the World Series? What, a line-up with 10 DH's and 4 catchers doesn't strike fear in the hearts of anyone? C'mon we have Jose Lima for God's sake! Hey, the man can dance.

C'mon we have Jose Lima for God's sake! Hey, the man can dance.

There's dancing the Go-Go, then there's dancing the Do-Do. I'll leave it as an exercise for the student, which one Lima Time and the Tigers will be performing in the "summer of two." ;-) ...

--------------

on DMB, their primary problem seems to be in their "manager profiles" in which, frankly, their virtual managers tend to be a lot smarter about platoon advantages, bullpen use, valuing offense over defense, etc. than real-life managers. Also, a lot of it comes down to how well they predict playing time. They do also seem to under-project young players, but I haven't ever looked at that systematically.

Anyway, just finished my first run of the 2002 season with me managing the Cubs and DMB handling everything else. Very interesting and fun results:

The AL went as planned, with NY, Chicago, Seattle, and Oakland as the wild card. The Yanks won easily. Baltimore stood at a surprising 60-66 ... then promptly lost 16 straight and won just 7 of their last 36 games. Still, they easily beat out the D-Rays, who needed 3 straight wins at the end of the season to break 50!

Cheer up Angels fans, your team won an impressive 97 games! Even the Rangers finished 500 (the AL West had a winning percentage a bit above 667 vs. the other divisions). The A's were really the best team, underperforming pythagoras by 9 wins to finish at "only" 102 wins (Seattle was 4 wins above expected).

The Central was the interesting race. The Twins were up 10 games with about 2 months to go, but were down 1 going into the final weekend against the White Sox. But the Sox won 2 out of 3 to eke out the title. The Royals surprised a bit by finishing 3rd (the Indians really stunk)

The NL was a hoot. The Cards won the Central pretty easy (6 games) but that was partly due to being 3 wins better than pythagoras. Going into the final weekend, after late season surges, the Braves and Mets were 3 games behind the Marlins. The Marlins cooperated by getting swept by Philly while the Braves swept the Mets, forcing a one-game playoff which, alas, the Braves won.

The wild-card race was good too. Going into that final weekend, the Cubs held a 3 game lead on Cincy and Houston and 2 game lead on Atlanta and New York. The Cubs split their first two against the Pirates and needed to win the last one to clinch it. 2 outs, bottom of the ninth, tying run on 2nd ... and Bobby Hill of all people goes deep to win it. Cubs take the wild card.

And the NL West? Going into the final weekend, all 5 teams were still mathematically alive! The Giants took it with an 83-79 record, beating Colorado by 1 game. Arizona was the real over-achiever, reaching 500 despite being outscored by over 40 runs.

Every team in the NL finished between .400 and .600. Other than the Brewers and Cards, every team was between .450 and .550.

The Cubs finished with the 2nd best record with 87 wins. The Cubs were also the best-scoring non-Coors team in NL, and even outscored all AL teams but Texas. Needless to say, I put a better hitting lineup in the field than Baylor is likely to (lots of Bellhorn at SS and Brown in CF). Sammy led the NL with 62 HR, despite missing 21 games due to injury. Being a Cub fan, I was happy to see that this wasn't really a season where everything went right. I played the first month without Mueller and the first 4 months without Gordon. The team finished with the 4th worst pitching in the NL. McGriff tore it up with 25 HR's in the first half, then managed just 3 over the second half and slumped horribly to a below-800 OPS for the season. Hundley had a nice comeback (30 HR's, OPS about 850) and Rosie Brown hit the ball well (OPS about 850). Had a nasty stretch where Patterson, Brown, Lewis, followed by Sosa on Patterson's return were all injured, and I was reduced to playing guys like Kimera Bartee, Angel Echevarria, and Turner Ward (signed off the FA list). Jon Lieber had an ERA over 5, though still better than Juan Cruz (eventually replaced by Mike Meyers). Other than Hundley, the only out-of-whack perfromances were 157 gp from Alou and the 140 AB's of 950 OPS I got out of Choi and Zuleta when I started giving up on McGriff. And the starters were all pretty healthy.

The playoffs weren't that interesting. The A's did finally conquer the Yankees, the M's took the Sox, the Cards beat the Giants, but the Cubs crapped out against the Braves pitching, all in 4 games. The A's over the M's and the Braves over the Cards, both in 5. Then the A's dropped the first 3 to the Braves, did manage to get the series back to Oakland, but went down in 6.

I'm looking to join a second DMB league next season. Is anybody in a league with an opening?

Make that two openings.

One more thing before I move on to the other league. ESPN.com's story about the Gold Glove selections included this comment by an unnamed AL coach: "I voted for Adam Kennedy because he made some great plays against us and I happened to catch Bret when he made a couple of errors." We have no way of knowing whether this is typical of the amount of thought that goes into the voting, but it wouldn't surprise me if it is.

I would be willing to wager a fair amount of money that the aforementioned coach was a member of the Twins staff.

Any other year and we'd have quite a few, but for some weird reason our turnover this winter has been nil. We even have some great owners on the waiting list. So it goes. I wonder if other leagues are experiencing this as well...

Re: DMB's evaluation, I'm a little surprised that Andruw Jones got a VG. Yes, Bill James still loves him, and Jones got a ton of defensive win shares...but the question remains, can Jones really have such a horrible ZR but an above-average UZR (which is probably as close to DMB ratings as it gets until we get some DRAs for 2003)? I understand that ZR doesn't correlate incredibly well with UZR, but when you take into account the things that ZR doesn't handle well, I'd expect significant differences...just not this significant. MGL, an inquiring mind wants to know - is this a case of DMB giving out fluffed-up ratings (and did this happen with Jeter in previous years? I can't recall now...), or is my skepticism re: the huge difference between ZR and the DMB ratings unwarranted?

MGL, an inquiring mind wants to know - is this a case of DMB giving out fluffed-up ratings (and did this happen with Jeter in previous years? I can't recall now...), or is my skepticism re: the huge difference between ZR and the DMB ratings unwarranted?

I don't understand your question. A. Jones UZR rating in very good again this year (+16 per 162 I think) after a poor rating last year. These kinds of year to year fluctuations is one reason why I don't even like to "look at" year by year ratings - I like to look at multi-year ratings in order to estimate a player's true defenensive value or to project their future defensive value or performance.

Of course, a Gold Glove is an award for performance and not for true talent, so if a player has an "unlucky" or "fluke" year like Andruw may have had last year, he doesn't deserve to get a GG award. This year his UZR was very good, which means he caught a lot more balls than the average CF'er would have caught given similar opportunities (in both number and "kind"), so he deserves consideration for the GG award.

Since DMB's defensive ratings use similar methodology to mine (UZR) (I assume), it would be surprising if our ratings were very far apart for any given player. Yes, it is true that regular ZR correlates very well with UZR and DMB's ratings. However, that doesn't mean that they can't be far apart for any given player. Same thing for UZR and DMB. They will tend to be very close, but again, it doesn't mean that for any given player, once in a while they can't be very far apart.

Since, ZR and UZR are very similar methodologies, with UZR being much more precise, I would not "combine" the ratings in any way shape or form. UZR should trump ZR, since it essentially a ZR++. Kind of like BA and OPS. One is an extension of the other. If a player had an OPS of .850 but a BA of only .230, you wouldn't "combine" the two metrics and say that this player is not as good as the .850 OPS suggests because he has a low BA. OPS trumps BA in the same way that UZR (or any other good PBP or pseudo-PBP metric) trumps ZR.

OTOH, when two methodologies are similar, like UZR and DMB (again, I am assuming, since as far as I know they have never released their complete methodology), but they are on the same "plane" (i.e., one is not "better" or more precise than the other, like BA and OPS are), then you can "combine" the two or use one to complement the other. If UZR had a player rated very high and DMB had a player rated very low, it would be fair to give them an in-between (average) rating.

Similarly, if UZR and DMB both have a player rated very well, like with A. Jones, you can have a higher degreee of confidence that A. Jones, did indeed have a very good defensive year, regardless of what his simple ZR was.

That is one reason why I wrote in another thread that DMB's contention that they consider range factor and ZR in their GG assessments is silly. I think they are just being politcially correct in not saying that they ignore these vastly inferior defensive metrics. But the truth is that if you have a simpe ZR, you can completely ignore range factor, and if you have a UZR or a DMB rating (or a DRA or any good metric that relies on PBP data or "faux" data), whatever a DMB rating is, you can completely ignore simple ZR...

I think in the end I was surprised, as much as anything, at the difference between Andruw's ZR and DMB rating. I've seen a lack of correlation before, but rarely if ever this extreme. So my question was - is this EXTREME lack of correlation common for ZR? Can you point to other players with similar/greater differences between the two?

Going to go out on a limb here...

Pac Bell makes good rightfielders great, average rightfielders good, and poor rightfielders average. Cruz won't have the same numbers in Tropicana, just like Sanders suffered in PNC. I'm not fluent with defensive stats, so that's nothing more than a guess.

Will,

I belong to 2 leagues that have openings. I'll email you the info.

Dave, I didn't mean that literally anyway. The BA might "inform" the OPS, since the walks and hits in OPS are not given the proper weightings, as you correctly point out.

As far as large differences between ZR and DMB or UZR, I don't know how prevalent that is. It shouldn't be THAT prevalent or course, otherwise there wouldn't be a high correlation, which there is. I don't think it should significiantly reduce your confidence in the DMB rating, for the reasons given above, and especialy since UZR agrees with DMB...

for all you DMB players, how much time commitment is required to put up a decent effort, and how is it distributed throughout the year?

I get pretty busy at times, can you let your team run on autopilot for long stretches of time? (I know you might be at a disadvantage for aquisitions, or are there daily/weekly decisions that must be made?)

I have never looked into this, but it sounds like fun.

One more thing before I move on to the other league. ESPN.com's story about the Gold Glove selections included this comment by an unnamed AL coach: "I voted for Adam Kennedy because he made some great plays against us and I happened to catch Bret when he made a couple of errors." We have no way of knowing whether this is typical of the amount of thought that goes into the voting, but it wouldn't surprise me if it is.

I would be willing to wager a fair amount of money that the aforementioned coach was a member of the Twins staff.

Actually, I'm almost positive I've seen that quote attributed to Buck Showalter, could be wrong though.

Homepage is a link to another DMB league with openings...

So how do these Diamond Mind baseball leagues work, anyway? I've been looking for something more fulfilling than the Yahoo! leagues (where half the managers always seem to quit paying attention two months into the season -- two years ago, one guy kept Darryl Kile in his starting rotation from his death to the end of the season). I'd love to join a league with some serious baseball people, and this DM thing always sounded good.

So I guess if anyone has an opening, I'd be interested as well. I've got a couple of months to figure out what to do...

I was curious to note there was no mention of holding runners/pickoff moves in the pitchers' review, but throwing out baserunners was quite prominent in the evaluation of catchers.

Does that not strike anyone else as a fairly ridiculous omission? I personally consider holding runners/pickoff moves a pretty significant part of the "fielding" value of a pitcher. . . and maybe the most overlooked.

spb

I was curious to note there was no mention of holding runners/pickoff moves in the pitchers' review, but throwing out baserunners was quite prominent in the evaluation of catchers.

Does that not strike anyone else as a fairly ridiculous omission? I personally consider holding runners/pickoff moves a pretty significant part of the "fielding" value of a pitcher. . . and maybe the most overlooked.<i>

From the article:

<i>Kenny Rogers made more plays, both in absolute terms and relative to the number of balls hit his way, REALLY shut down the running game (only 4 stolen bases allowed all year, 3 pickoffs)

and

But my vote goes to Kirk Rueter. He handled 43 chances without an error in 2003. In fact, he hasn't made an error since 1999, successfully completing 209 plays in the last four years. Rueter had a hand in 5 double plays, one shy of the league lead. And he continues to be nearly impossible to run on

Seems to me that holding runners on was actually a fairly important part of what went into Tom Tippett's selection.

Aw crap, let me try to fix that

Testing, testing 1, 2,</i> 3!

OOTP is way more fun then DMB. Check out the homepage for the league i'm a tri-commissioner of.

Just to expand on what Jhat said and answer a few other questions:

Participation in a DMB league typically requires two types of responsibility. The first is roster construction; you're responsible for putting together your roster through an annual draft/auction/whatever, making trades, acquiring free agents and otherwise assembling your team.

The other requirement of running a DMB team is participating in the season. For most leagues, this means setting up and maintaining a "Manager Profile" (MP), which is a compilation of starting lineups vs. LHP/RHP, bench roles, starting rotation, bullpen roles and manager tendencies (steal, bunt, use closers, etc) to be used when you're not actively running games. In some leagues, maintaining your MP is all you need to do as the commissioner sims everything; in others, the home team is given the opportunity to manually sim the games and manage his own team hands-on while using the visiting team's MP.

As far as time commitments go, the off-season player acquisition is usually the most time-intensive portion of participating in a league, especially if it's a real-time auction: not only do you need a strategy as to what holes you need to fill, who's available to fill those holes and how much you can afford for each, but you have to pay attention when you're outbid for a player and adjust your current approach dynamically. You can put as much or as little energy into trading as you see fit, but it certainly can pay off.

Actually running the season isn't particularly time-intensive. Unless you're constantly shuttling people on and off the 25-man roster, you'll find that your manager profile remains fairly constant throughout the season; I'd say it rarely takes me more than five minutes to scan my MP and make any changes needed. Simming a series takes well under an hour for me; if you're in a hurry, it can be done in far less.

Hi! Will,

True. "No" mention overstated my case inexcusably. But it reads more like an afterthought in the pitchers' section (I don't think it got mentioned in the first 3-4 paragraphs) while it reads as a primary consideration in the catchers' segment (I think it was in the first or second sentence that he referred to Pudge's caught stealing prowess).

But maybe it is just my personal bias that flavored the article for me. If so, I apologize.

Moreover, I am out of line for not acknowledging that I am nitpicking a very good and interesting article. (hell - SERIES)

One more thing before I move on to the other league. ESPN.com's story about the Gold Glove selections included this comment by an unnamed AL coach: "I voted for Adam Kennedy because he made some great plays against us and I happened to catch Bret when he made a couple of errors." We have no way of knowing whether this is typical of the amount of thought that goes into the voting, but it wouldn't surprise me if it is.

I'm not sure what's so terrible about this. The coach has rated Kennedy and Boone based on how they played against his team, which, except for a couple of games on TV, is probably the only time he saw either of them play a full game. What's he supposed to base his vote on, SportsCenter highlights? UZR? If you're asking a coach for his opinions on the best fielders, don't you expect him to base that on how well he saw them field?

KM, even if a players' true defensive ability stayed the same every year (as if he were a robot or a computer) and even if UZR or any other metric were 100% reliable and accurate, the measurement is based on binomial results around a certain mean, that mean being the player's true defensive value, same as flipping a coin (where the 50/50 represents an average defensive player). So no matter what, you aer going to see random fluctuations in a player's year-to-year defensive measures. Any variation is possible of course. If you know just a little about stats and the properties of normal courves, you know that 68% of those (random) variations will be withing one standard deviation of a player's true value, 95% within 2 standard deviations, etc.

Your assertion that we "should see no or very little variation in a player's year to year defensive ratings," is like saying "if we flipped a coin 2 or 300 times each year, we should see very little variation in h/t % from year to year. All of these measures (BA, UZR, ZR, range factor, fielding %, OBA) are binomial and are exactly like the coin flipping analogy. That is ONE reason (and the major reason) why you see year to year fluctuations in any of a player's rate stats. The other three reasons are environment (the true value or mean changes), talent (agaion, the true value changes), and different kinds of sample error (most of these measures are not "true" binomials, like a coin flip). As I said, by far and away, the most predominant influence in the year to year fliuctuations is "randomness", like the coin, which is why we use multiple years (larger sample reduces the sample error) for estimating a player's true talent and for projections....

Just to give you more of an idea as to the random nature of a defensive metric, even an excellent one, and why you can and will see all kinds of year to year differences, even if the underlying talent has not changed at all....

Everyone seems to know and understand how an offensive metric can vary and flutuate from player to player and from time period to time period regardless of the underlying talent. Player A can hit 30 scoreched line drives and have 3 of them caught in one year and 6 of them caught the next year (or player A can have 3 of them caught while player B can have 5 of them caught). Player A can have 20 bloop hits in one season and only 15 the next season, without there being a change in his talent. Player A and B might hit 20 380 foot fly balls to left center field, but 10 of those might be for a home run for player A and only 7 for player B (due to the different stadiums they were hit in). These and many, many other examples are the "reasons" behind the random fluctuations in offensive metrics.

For some reason, some people have a hard time understanding that the same types of things occur in defensive metrics, only in reverse of course. Player A might get 10 scorched balls hit right at him and he catches all 10, while player B might get 5 of those scorched balls hit just to his left or hust to his right, out of his reach. Almost all defensive metrics, even granular PBP metrics, like UZR or ZR, will probably have all 10 of those balls as "the same", such that player A will "look like" the much better fielder. Rather than player A and player B, of course, it could be player A one year and player A the next year (hence, "random" year to year fluctuations in the defensive metric). Even in the best zone-type system, the best we can do is to say, for example, that in such and such small area of the field, an average player catches 40% (or 20%, or 60%, etc.) of the medium height fly balls. Even if we have two average players (or one average player in 2 separate years), it might so happen that player A gets 10 medium fly balls balls hit into that zone, but that 7 of them are in the "catchable" portion of that zone, while player B also gets 10 balls hit into that zone, but only 4 of them are in the "catchable" portion. Even though they are both average defenders, one will "appear to be" better than the other when we look at their defensive metrics. Same thing for one player in two different years. Again, there are slews of examples of how the random nature of the distribution of the balls hit can affect a player's sample defensive metric without the underlying talent changing, just as there is with offensive metrics.

Hope that helps....

Silas,

Funny you mention this - I just poked around the EBA website two days ago. The thing that struck me right away - you're moving to weekly reporting. For a hands-on league, that has to be huge - both in the sense that the owners are more immersed in the season, and that they are forced to get those games in at a pretty quick tempo. I'd love to see it in the SSL, but we have too many owners with busy schedules, and in the end I value the community (slow owners and all) more than I would the slightly enhanced realism.

Regardless, good luck!