Jump to content


Yo! You're not logged in. Why am I seeing this ad?

Photo

Derek Carty Q and A


This topic has been archived. This means that you cannot reply to this topic.
34 replies to this topic

#1 Frisbetarian


  • ♫ ♫ ♫ ♫ ♫ ♫


  • 4632 posts

Posted 04 February 2009 - 04:24 PM

Derek will answer all your questions in this thread. This may take a day or so as Derek is unexpectedly extremely busy in real life right now, so please hold off on any follow-ups until he has completed his responses. I appreciate your patience.

Thanks for all the great questions, and thanks to Derek for agreeing to 'chat" with us.

#2 Frisbetarian


  • ♫ ♫ ♫ ♫ ♫ ♫


  • 4632 posts

Posted 05 February 2009 - 07:27 AM

Derek will answer all your questions in this thread. This may take a day or so as Derek is unexpectedly extremely busy in real life right now, so please hold off on any follow-ups until he has completed his responses. I appreciate your patience.

Thanks for all the great questions, and thanks to Derek for agreeing to 'chat" with us.


I heard from Derek late last night and he assures me he will be here today to post his responses. Unfortunately, he has had unanticipated issues arise that had to be dealt with before he could answer your questions. We both appreciate your patience.

Also, in case you haven't seen it, Davis Gassko has posted his always great Hardball Times Season Preview 2009 article.

#3 Derek Carty

  • 30 posts

Posted 05 February 2009 - 05:28 PM

Hey guys,
I want to quickly apologize for the delay in answering these. Some unexpected things have come up over the past couple of days that I needed to take care of.

That being said, I want to thank the guys running Sons of Sam Horn and all of you who asked questions or who are reading this for having me here. I had a lot of fun answering these questions, and if you guys have any follow-up questions, please feel free to post once I'm done posting all original responses.

Thanks!

Edited by Derek Carty, 05 February 2009 - 05:32 PM.


#4 Derek Carty

  • 30 posts

Posted 05 February 2009 - 05:31 PM

Nice work on the BABIP article, Derek. I see that you used Baseball Prospectus data for the study, and wonder what your thoughts are on how different the results may have looked if you could have used Baseball Info Solutions numbers or, better still, Hit F/X numbers. What I really want to know is if you think Hit F/X will become a reality (and when), and how much it will differ from the BP and BIS data.

This is actually something I'm looking into. I'm not completely familiar with how BP's batted ball data is derived, but I don't imagine their predictive ability would differ very much from BIS. If we were to use HITf/x derived batted ball data, though, our accuracy would definitely improve. This, of course, depends upon what data HITf/x provides, but if it gives us launch angle, we would be able to systematically classify batted balls as either ground balls, line drives, fly balls, etc. As it stands right now, it's up to a scorer to make a subjective judgment. Eliminating that element of subjectivity would definitely make the accuracy of the numbers better.

What I'm most excited about with HITf/x, though, is speed off bat data. Line drives tell us the angle at which a ball is hit, not the speed - the closer to the ground an airball is, the more likely it is that it hits the ground (and becomes a hit) before a fielder gets to it. With speed off bat data, though, we could tell how hard balls are being hit - definitely more valuable. I believe this will be the key to BABIP and will dramatically increase our ability to predict it.

As far as when HITf/x will be a reality, word is that it will be this year, at least the initial stages of it. Baseball Digest Daily (http://www.baseballd...hit-fx-in-2009/) first broke the news, and there was a good thread at the Book Blog (http://www.insidethe...tchf_x_in_2009/) about it.

Edited by Derek Carty, 05 February 2009 - 05:32 PM.


#5 Derek Carty

  • 30 posts

Posted 05 February 2009 - 05:37 PM

1) What do you think of MLBAM's attempt to "own" baseball statistics and how do you think they ever came to the conclusion that this was a smart thing to try?
2) Forecasting for things that fantasy baseball guys care about seems to have some differences, some subtle and some obvious, than projecting for performance evaluation. Do you think the tools are largely the same or are there some ways you think different projection tools would be applied for strictly fantasy baseball?
3) Say you are in a keeper league with minor leaguers. We know some general rules of thumb about how power develops but what should people really look for in their prospects in these cases?


1) Honestly, I'm not sure how they came to that conclusion. I think it was a poorly conceived idea, especially since the fantasy industry has really aided the baseball industry itself. It has rejuvenated interest in baseball among a lot of people and adds revenue for MLB and its clubs. I have no idea why they would do anything to try and halt this.

2) Well, fantasy baseball obviously ignores things like defense and baserunning, so right there is a big difference. It also ignores some important aspects of offensive value. As far as the tools themselves that are used to project talent, I don't think there is a big difference at all. If I'm trying to project a pitcher's ERA for my fantasy team, I'm not going to be looking at anything different or using any different methods than someone who is trying to project that pitcher's ERA for any other purpose. A major league club will have more resources than I do, but aside from this, I don't do anything systematically differently.

I believe that we can do our best work evaluating players when we combine proper use of stats with proper use of scouting, and I believe that true scouting is used very infrequently in the fantasy world. That's not to say it can't be used, just that those who know what they're doing are more likely to be doing it for a club or for "real" baseball purposes.

3) I think one of the big things people - or a lot of fantasy players, at least - ignore is context. Did a player hit 20 HRs in Double A or Triple A? In the PCL or the International League? Major League Equivalencies (MLEs) account for league quality and ballpark effects, and this is a very good starting point. Maybe a little obvious, but it can also help to look at doubles and triples. I do advise being wary of looking at these for several reasons but mostly because they are balls in play, which are prone to wide swings in luck and also because many doubles and triples are actually ground balls or line drives that don't actually represent power. Still, for minor leaguers, they are at least worth a look. Outfield fly rate is good to look at (although that doesn't necessarily represent raw power). A hitter with a high rate of flies will hit more HRs than one with a low rate, all other things equal. Aside from that, there's very little other data that we have access to that can be of use. I think if we can ever get a system like HITf/x or HitTracker installed in minor league parks, that will be a big step towards projecting future power potential.

This also ties into what I was saying in question two about scouting. For minor leaguers, scouting can be important because stats can't always do the best job of illustrating upside. If you can see that a player has room to add muscle, has changed his swing to add more loft to the ball, or is hitting the ball a very long way, these things can indicate future power potential. It's less precise than stats, but the raw stats we currently have leave some portions of the equation out for minor leaguers.

#6 Derek Carty

  • 30 posts

Posted 05 February 2009 - 05:41 PM

I was wondering: From a fantasy perspective, who are some of the undervalued players going into this season? Who are the sleepers for 2009?


I like Nelson Cruz a lot, although the word seems to be out on him. Raul Ibanez might be undervalued. Mike Napoli makes a good pick too, although he looks to be turning into a Nelson Cruz case where people are picking up on him. After Round 15, I like guys like Ricky Weeks, Elijah Dukes, Adam Lind, Kenji Johjima, Aaron Harang, Max Scherzer, and Frank Francisco, to name a few.

Of the more well-known players who could out-produce their draft positions, I like Jose Reyes, Robinson Cano, James Loney, Dan Haren, and Javier Vazquez, among others.

I wrote articles about Cano, Loney, Vazquez, and Haren (in order) here:
http://www.hardballt...-robinson-cano/
http://www.hardballt...ty-james-loney/
http://www.hardballt...s-for-pitchers/
http://www.hardballt...d-park-factors/

I'm also in the process of writing up why injured/injury risk players, as a group, might be undervalued:
http://www.hardballt...rewards-part-2/

#7 Derek Carty

  • 30 posts

Posted 05 February 2009 - 05:44 PM

Welcome, Derek, and thanks for participating.

Do you consult any outside projection systems in determining your draft day targets and valuations? If so, do you find that some are more appropriate for keeper leagues vs. straight drafts?

If so, what do you believe are the critical differences? Or, if not, are the differences too minor and convoluted to gauge such suitability?


Well, up until this season I've never created my own projection system. I'm working on one now to enter into Tom Tango's Forecaster's Challenge, but as tests have shown that the differences in the major projection systems are minimal, I had never bothered to create my own. So, yes, I do consult outside systems. I'll look at CHONE, PECOTA, THT, Bill James, Oliver, and even Marcels.

As far as the keeper versus non-keeper question, this kind of depends on the situation of the keeper league. In the first year of a league like this, I won't put too much emphasis on getting young players, getting a few but mostly trying to build a good team to compete this year. If you're in a clear rebuilding mode, though, you'll want to go for young players (this is also dependent on the league setup). For this, you'll want to avoid Marcels as it definitely lags behind for young players. It doesn't use minor league stats, so anyone with less than three full years of experience will be assumed to be closer to the mean than they should be. For this, you'll definitely want CHONE or THT or PECOTA or Oliver instead. Aside from that, though, I wouldn't say there are any critical differences as to keeper versus redraft leagues.

#8 Derek Carty

  • 30 posts

Posted 05 February 2009 - 05:46 PM

Derek, thanks for stopping by.

It's nearing the time to set keepers in my league. What do you expect from the Upton brothers in 2009?


The Upton's are a couple if interesting players. I'd first recommend reading this FanGraphs article about B.J.: http://www.fangraphs...-the-long-ball/

So I'd definitely be expecting more than 9 HRs from B.J. I like Justin's raw power and think he could hit more HRs than his brother as soon as 2009. Both have trouble making contact (though B.J. made legitimate strides this year) and shouldn't be counted on for their batting averages. If they stay healthy, B.J. should hit around .270 with 20+ HRs (could get as high as 30) and a lot of steals. For Justin, if he gets 550 at-bats, he could get as high as 25 HRs. His average might only be .250-.260, though, with only a handful of steals. A lot of his value will hinge on where he bats in the order.

Whether or not you should keep them depends upon league setup, other options, expectations for your opponents, and other factors like that.

#9 Derek Carty

  • 30 posts

Posted 05 February 2009 - 05:49 PM

Hi Derek. Thanks for stoping by and giving us some of your time. Was wondering if you could chime in on your projections for Wieters for this upcoming season? When does he get the job all to his own, and do you see him being a perennial top 5 catcher in the immediate future?


Well, as said in a previous questions, I don't have my own systematic projection system completed yet. That being said, it's very hard to argue against Wieters being very good, very soon. Greg Zaun is currently the starting catcher in Baltimore, but I could see Wieters taking the job by Opening Day with a good Spring Training. Perhaps a bit more likely will be a May call-up (or a June call-up, if the O's are worried about arbitration and such, which would be wise since they won't be competing. There's a good explanation of the process here, if you're interested: http://www.brewcrewb...ation-clock-and).

I'm not sure I'd call him a top five catcher just yet because there are a lot of good catchers at the top now (Martin, McCann, Mauer, Soto, Doumit, Napoli, and maybe Victor Martinez and Iannetta), but assuming 2009 goes reasonably well, I think he should be in the top 5 by 2010. There was an interesting quote at Fantasy Ball Junkie (http://www.fantasyba...nkie.com/?p=257) that noted Keith Law's comparison to Mark Teixeira. If Teixeira were a catcher, he would be a consideration for the #1 fantasy pick. That's how good Wieters could be. That's his upside.

It's also worth considering the point I made about injury risks at THT earlier this week (http://www.hardballt...rewards-part-2/). Wieters isn't necessarily an injury risk, but the premise can be applied here. If you're league setup has room for you to stash Wieters until he is recalled, he could be more valuable than some might otherwise think.

#10 Derek Carty

  • 30 posts

Posted 05 February 2009 - 05:50 PM

What are your thoughts on Kila Ka'aihue?

Simply put, Kila is a guy who has drawn the short end of the stick. He seems to have excellent, Major League-ready power, draws a ton of walks, and isn't supposed to be a liability on defense. Yet, despite already having 1B/DH candidates in Kila, Ryan Shealy, and Billy Butler, the Royals go out and get Mike Jacobs this off-season. Come on.

Kila's MLE batting average is fairly low, though his contact rate was 81% this year and the low average was driven by a .246 BABIP. Some scouts seem to believe that he'll strike out more at the major league level, but we can't know for sure. Even more so, we won't know for sure because he's blocked. While he could have some serious fantasy value hitting in the middle of the order for the Royals, he probably won't be much of a factor unless someone (or a couple someones) get hurt or traded.

#11 Derek Carty

  • 30 posts

Posted 05 February 2009 - 05:52 PM

I read your Pedroia article that focused on power, contact, and speed... how does Alexei Ramirez stack up and what are your expectations from him this season? He obviously has all the tools to be a stud, but he swings at pitches outside the zone at a rate nearly twice the Major League average [while making contact 70% of the time]. Plus his walk rate is 3%.


Well, I definitely like Pedroia better, but Alexei was a nice player this past year.

His xBABIP of .280 was lower than his actual BABIP of .296, which would drop his average to .277. If he gets a full-season of at-bats for the Sox, though, he could hit 22 HRs or so and steal maybe 18 or 20 bases. With Orlando Cabrera and Nick Swisher gone, the lead-off spot appears to be open, which would also boost Ramirez's value if given the opportunity (although the walks would limit his run opportunities). A line of 575 AB/.277 AVG/22 HR/18 SB/90 R/60 RBI would make Ramirez worth roughly $19 in a traditional 12-team mixed league and worthy of a pick in the fourth round.

I will say, though, that he's lucky to have Ozzie Guillen as a manager. On a lot of clubs, a 59% stolen base success rate just wouldn't fly. For Ozzie, he'll probably continue letting him run. Good aggressive baseball, am I right? Posted Image

#12 Derek Carty

  • 30 posts

Posted 05 February 2009 - 05:53 PM

Hi, I was wondering what your thoughts were re. Jason Heyward. He has been compared to Dave Winfield at times, and his impressive minor league debut has only raised expectations. He's an intriguing 6'4" RF/1B with multiple tools, even if his raw power has not yet translated; at Single-A Rome (he finished with a handful of AB at A+ Myrtle Beach), his BB rate approached 10% and his K rate was below 15%, in addition to a LD% of 19.

Everyone seems to love the guy -- do you have any unique insights into his future or skill set?


Everyone does seem to love Heyward, but because of how far away he is from the majors (he topped out at Advanced-A this year and batted just 22 times there), he hasn't exploded onto the fantasy radar yet. He's still a couple of years away, and as such, I probably don't know as much about Heyward as you're hoping for. He's still so far away from the majors it would be very difficult to project him with any reasonable degree of confidence. His numbers aren't eye-popping yet, so I'd temper expectations, at least until we see how he matures a little bit more.

Sorry I couldn't give you a better answer here.

#13 Derek Carty

  • 30 posts

Posted 05 February 2009 - 05:54 PM

Hi Derek,

I enjoyed the article on your approach to drafting late in R1 in 2009. You mentioned several players that, while they are being drafted there in mock drafts, you weren't interested in. Beyond the first round, are there any players that, given where they are being drafted/their perceived auction values during the pre-season, you would expect will not be on your teams in 2009? Is there a theme you see in these "overvalued" players?

Thanks!


One immediate theme is pitchers. In mixed leagues, I'll rarely take a pitcher before Round 8, sometimes before Round 12. Closers I may even wait until Round 18 or 19.

As far as non-pitchers go, how about Hanley Ramirez, Ian Kinsler, Justin Morneau, Adrian Gonzalez, and Chris Davis, if you're looking for some higher profile names. No trends there, really; I'm just not buying the skills for where they are being taken in drafts. For mid-round pitchers, Joe Saunders and Gavin Floyd immediately stick out as guys I'm not buying.

#14 Derek Carty

  • 30 posts

Posted 05 February 2009 - 05:54 PM

Derek,

What probability would you assign Varitek having an OPS of 730 or better in 2009?

Thanks.


As a rough estimate, I don't think 35% is unreasonable, if he catches a few breaks.

#15 Derek Carty

  • 30 posts

Posted 05 February 2009 - 05:58 PM

I have a few non-fantasy questions, Derek.

Which of the defensive evaluation systems do you feel is most accurate, and why? How much data do you think is necessary for a defensive metric to show a true talent level? What do you see as the future for defensive measurement, and will hit f/x play any part?

Thanks.


Well, as a fantasy guy, I pay less attention to defense than traditional baseball analysts do. I do keep up with what's going on, though, and when evaluating defense on my own, I like looking at MGL's UZR. I feel the methodology is the strongest (and have heard the same from people whom I trust and who are more familiar with defensive metrics than I am), and it's now readily available at FanGraphs.

As for how much data is necessary to show true talent level, the answer is infinite Posted Image. We can never know a player's true talent level with absolute certainty. We can only estimate it, and the more data we have, the better that estimate will be. I imagine what you're looking for is when it starts to stabilize, though. I've never run or seen any studies on this (although I'd imagine they're out there), but anecdotally, I think it would take at least two full years worth, if not more (which is definitely possible). It just seems that a lot of players see such wide swings from year to year for it to be any less. Take a look at Robinson Cano's FanGraphs page and check out his UZR/150. It's all over the map.

The future for defensive measurements, in my opinion, is more detailed data - HITf/x or FIELDf/x. As of right now, we're basing a lot of these metrics, in part, on the subjective judgments of scorers. MGL's UZR, for example, has been calculated in exactly the same way using both Baseball Info Solutions and STATS Inc. data, yet for some players the two different UZR's are conflicting. Check out this article at the Book Blog for some talk about that: http://www.insidethe...ts/suzr_v_buzr/

They are, however, better than anything else we have and are clearly better than using any single person's eyes (that players like Derek Jeter and Michael Young win Gold Gloves can attest to this). Also, if you look at With or Without You Studies (made famous by Tom Tango), these largely confirm (the ones I've seen, at least) the findings of systems like UZR, RZR, OPA!, or whatever other advanced system you'd like to use.

Once we get something like a FIELDf/x where we can see where fielders are positioned, exactly where the balls are being hit, the speed off bat of all balls, and things of that nature, we will be able to do a much better job of evaluating defense.

#16 Derek Carty

  • 30 posts

Posted 05 February 2009 - 05:58 PM

What is your opinion of a player who is showing doubles power as a young major leaguer developing home run power in the future?


I discussed this a bit in a previous question. I think there are inherent flaws to using doubles/triples to project power, and for major leaguers, I prefer using something like HitTracker where we can see how far batters are hitting balls and control for things like ballpark and weather. Of course, HitTracker doesn't yet track all fly balls, but I still think we can get a reasonably good picture of a hitter's power.

Doubles and triples are worth a look, but I wouldn't be making any definitive judgments based on them.

#17 Derek Carty

  • 30 posts

Posted 05 February 2009 - 06:00 PM

It's now conventional wisdom, at least in stat-savvy circles, that workload is a special concern for young pitchers. At what age do you think a pitcher ceases to be "young" for this purpose? (I have Jon Lester specifically in mind, though I realize his situation is not typical.)


Honestly, I don't fully buy into this idea. First, while we stat-savvy fans love to quantify everything, we need to remember that we're dealing with human beings, and every human being has their own unique physiology. There are also plenty of other factors (mechanics - of which there are hundreds of components - health history, pitching history, types/mix of pitches thrown, etc.) that are unique to each pitcher. Even if this was true, though, there wouldn't be one age where pitchers magically transform into being less injury-prone. It may follow some sort of individualized slope, but again, I don't quite buy it.

Evidence has been inconclusive, at best, that this type of thing exists. Some evidence (http://www.hardballt...r-after-effect/) has actually shown the opposite.

This is a topic that I really think needs to be researched in more detail before we can make any definitive judgments. I would love to get my hands on an injury database, as we could do so much with this, but for now I don't have a real opinion one way or the other.

#18 Derek Carty

  • 30 posts

Posted 05 February 2009 - 06:02 PM

Obviously the biggest debate on SoSH this offseason was the never-ending Varitek "negotiations." Within this endless debate there were essentially two sides expressed:

1) Varitek is done as an offensive player. His 2008 was really the continuation of 2006, his decent offensive season of 2007 was not only a "dead-cat bounce", it may not be accurate as a good full season due to his second half numbers (decreased BA/increased walk (especially over historical average)).
2) Yeah, Varitek had a really bad season. But... he was apparently ill/injured in the beginning of the year and the injuries/sickness lingered. And, then there was the homelife going down the tubes (the divorce). These things had to have an effect... and he should do better in 2009.

My question is not the easy one... which would be which side do you support (although if you wish to answer that, be my guest). My question is this:

Do you/can you project seasons based on how a previous season may have been affected by outside influences, such as Varitek's divorce? I would figure the general lack of knowledge on most players personal lives would be a limiting factor in attempting to make a model... but I do think we know of cases where it appears that outside influences did affect a player on the field (Tony Clark with Boston in '02 comes to mind... as he was also involved with union negotiations). Furthermore, if such a model exists, would it even be useful for a player like Varitek (who is not young and is a catcher... which is not historically a good combination).


Unfortunately, we don't have a way of systematically projecting players based on these kind of factors. They undoubtedly have an effect on performance, but it's difficult to know how large those effects are/potentially can be, plus they will vary for every player. Every person is different and things affect people in different (sometimes even opposite) ways. And as you noted, we're surely missing a lot of information.

Baseball is, in part, a game of information for front office execs, and there are advantages to keeping a lot of information a secret. There are also advantages to feeding inaccurate information, which further complicates things for us (see: J.P. Ricciardi and B.J. Ryan, 2007). As a member of the public, we'll never have all the information we need or be able to fully trust all the information we have. I suppose if someone put a database together, we could construct a crude model, although I'm not sure how valuable/accurate it would be.

If a model did exist, we would be fine to use it on Varitek because we could control for things like age, although we'd need to understand that it would be very, very crude and probably not super accurate if it was a model that would be generalized to all players (which it would need to be without extensive psychological reports and other such information for each player).

#19 Derek Carty

  • 30 posts

Posted 05 February 2009 - 06:06 PM

Derek, I noticed in the Rotoworld Fantasy Season Preview, you picked the Red Sox to win the World Series, Beckett to win the Cy Young, and Pedroia to repeat as MVP. Care to elaborate on your feelings about the AL East, or were you just being toungue in cheek with your fellow writers? Also, what are your thoughts on the new Pirates management team, what they have done so far, and an ETA when you think that team may be in the hunt for the wild card?

Thanks.


Nope, nothing tongue-in-cheek about it. I think the Red Sox are the best team in the baseball. I especially liked the signings of Smoltz, Saito, Penny, Fernando Cabrera, and Baldelli this off-season. Low-risk signings that could (and likely will) pay big rewards. I really wanted my Mets to get Smoltz and Saito.

Also, while I don't think Pedroia deserved the MVP, I do think he'll win it again. As I wrote in this article that was quoted a few times in the original thread, I think he's actually in for a bit of an improvement this year. Unless someone like A-Rod or David Ortiz or Miguel Cabrera puts up a ridiculous year, I don't see why Pedroia wouldn't win the MVP again if he actually plays better. That was my logic with the pick, anyway.

Beckett has a very good arsenal pitches, terrific peripherals, and was very unlucky this past season. As long as he stays healthy, I think he's the best pitcher in the AL, so he gets my pre-season Cy Young vote.

As far as Pirates management goes, it's certainly better than the previous regime. I really would like to see Paul DePodesta get a GM job again, but Huntington seems open-minded about combining stats and scouting. I loved seeing them bring on Dan Fox.

As far as when they'll be in the wild card hunt, I think it's going to depend on how well their younger players develop and how willing they are to spend a little money to push themselves over the top. It'll probably be at least a couple of years, but they have a good core in Maholm, McLouth, Doumit, and Capps. After that, Adam LaRoche is solid, Ian Snell could bounce back, and the development of Andy LaRoche, Andrew McCutchen, Pedro Alvarez, Steve Pearce, Ross Ohlendorf, Jose Tabata, and a few others will be very important. Honestly, I wouldn't be surprised if they're competing in 2011 if some things break right and they show that they know what they're doing when it comes to finding undervalued players.

#20 Derek Carty

  • 30 posts

Posted 05 February 2009 - 06:08 PM

To play off of Robinson Checo's question a little, I'll ask another question involving Dustin Pedroia. I'll start by referencing your article about him from December 8th where you said two things that I found interesting. First, you said you didn't think he had any business winning the MVP. I'm curious about who you thought was more deserving and why.

Second, you also said you think that not only is Pedroia's 2008 performance sustainable going forward, but also that you think he will improve on it. One of the reasons you cite, if I read you correctly, is that you don't believe Pedroia will be as susceptible to negative fluctuations in his BABIP. You specifically mention the Green Monster and the 28 of his 35 Fenway doubles that were plunked off it. Are there other variables driving that opinion or is it mostly just that he has the Monster there for 81 games a season?

Thanks for taking the time to answer these questions for us.


1) Factoring everything in (offense, defense, baserunning), I think Grady Sizemore deserved it. You could also make a case for Joe Mauer. Pedroia wasn't an absurd pick, but I don't think he was the correct pick.

2) Well, I first want to note that I mentioned at the time that it was just speculation on my part. I haven't run any studies or seen any numbers that show that this is true. This being said, that was the only variable driving that opinion, and the effect of it (if existent) wouldn't be enormous. Given that Pedroia is hitting balls at a wall (and doesn't seem to have trouble reaching it) that will automatically become hits, I think that may count for a little more than for other hitters who will be hitting their balls at fielders where anything can happen.

#21 Derek Carty

  • 30 posts

Posted 05 February 2009 - 06:10 PM

Derek-Non-fantasy question here. I was wondering which metrics, and combination of metrics, that people use for player values and/or projections are not really that good to use and which combination of ones you think are best. Thanks


Well, you probably know by now that batting average doesn't capture a player's full very well. OBP also has it's flaws, though, as does slugging percentage and OPS. OPS+ and Win Shares are more sabermetric-y stats that are also pretty flawed. Win Shares likely wouldn't have lasted this long without Bill James's name attached to it (and please don't take that as a jab at James, because it's not). VORP is another sabermetric stat that is actually getting a lot of mainstream press (albeit from guys poking fun at it) that is flawed. Just about all of the mainstream defensive metrics are flawed (errors, fielding percentage, etc). And stolen base success rate is flawed in the sense that some people use it as a proxy for baserunning, which is a poor assumption.

For pitchers, ERA is no good, WHIP is no good, BAA (batting average against) is no good, wins are very, very bad. As far as more sabermetric-y pitching stats go, Quality Starts are no good, FIP has some flaws, ERC is flawed, even K/BB is flawed, and that gets a lot of play. Lots of "sabermetric" guys use K/BB without taking the time to think about the problems with it.

As far as what is good to use, personally, I like Tom Tango's wOBA (which is based on linear weights) for offense, MGL's UZR for defense, and Dan Fox's work on baserunning. For pitchers, I like David Gassko's LIPS ERA (though defense, ballpark, etc also need to be taken into consideration).

#22 Derek Carty

  • 30 posts

Posted 05 February 2009 - 06:14 PM

Linear weights or WPA/LI?

Also, many, including MGL and David Gassko, have made the argument that Manny, because of his defensive shortcomings, is not worth much more (about 10 -15 runs, iirc) than a league average left fielder. What are your views on this?


Linear weights. WPA/LI is a cool stat, but as a fantasy guy, I'm more about isolating true talent and forward-looking thinking. WPA/LI is based on how hitters perform in different situations, and as we know, the evidence is inconclusive at best that things like "clutch" and "situational hitting" actually exist to a significant enough degree. I'm much more comfortable simply taking a player's raw line and seeing the situation-neutral run value of his contributions.

As far as Manny goes, I've read some things recently that show that Manny's defense may not be quite as bad as some make it out to be, that the Green Monster had a large effect on him. Still, he is undoubtedly a below-average defender, and he isn't worth anywhere near the $25 million for 4 years he's said to want. Even the 1-year offer the Dodgers made was probably too high. While I'm not sure if he's a league average left fielder, I'd think he's closer to league average than he is to a superstar. Keep in mind, though, that I'm much better at evaluating fantasy value than real-life value, at least when we're talking in precise terms.

Also, this article by HitTracker's Greg Rybarczyk is also a must read on Manny, as is Greg's article in the 2009 THT Annual. Manny is very well suited for the Dodgers, and if someone is going to offer him a big deal, it would make much more sense for the Dodgers to than for a team like the Mets.

#23 Derek Carty

  • 30 posts

Posted 05 February 2009 - 06:18 PM

How do you balance taking the best player available, to drafting by weaker positions? Do you have any strategy on when to take C, 2B, or SS? Also, I was wondering on your opinion of Jose Reyes. I have always been kind of puzzled by Reyes high ranking year in and out. He's a nice fantasy player and being a ss definitely helps him, but I don't see how he can be considered a top 10 fantasy player. Especially when Rollins is ranked a round or so lower and there is basically no difference between to the two.


Ah, now we're back on my turf Posted Image. To answer your first question, I use simple replacement level theory. I take the unadjusted value of every player and subtract the value of a replacement level player at the same position. So, as an example, if Brian McCann is worth 10 units, and the 25th most valuable catcher (e.x. Jarrod Saltalamacchia) is worth 4 units, you get McCann at a position-adjusted 6 units. Do this for every position and you can compare players directly, making subjective attempts at balancing best player available/weaker positions unnecessary.

While I'm not quite sure I'd call it a strategy, I do find myself taking catchers and (to a much smaller extent) middle infielders early because other owners don't value them properly. The top catcher is probably worth a second round pick, or at least a third, in 12-team mixed leagues. Taking him in the fourth or fifth round is capitalizing on a great opportunity. The position adjustment is smaller for middle infielders, but it's still there and can lead you to have one at the top of your board early on.

Just make sure you don't overcompensate for position scarcity, especially when it comes to middle infielders. Taking them too early is just as bad as ignoring position scarcity in the first place.

Reyes I like a lot this year. I worry that Citi Field will hurt his HR totals, but I still took him second overall (while Hanley and A-Rod were still on the board) in the recent experts mock draft over at Mock Draft Central (http://www.mockdraft...ert_draft_2.jsp). I'd probably take him #1 overall if I had to. Rollins is valuable as well and worth at least a top 7 or 8 pick, but there is a definite difference between Reyes's .295/15/55/65/115 and Rollins's .280/18/35/70/110. The big difference is the steals, and they're worth paying for. If you can trade down for Rollins it's not a bad idea, but if you're stuck in the top 5, Reyes is the man.

#24 Derek Carty

  • 30 posts

Posted 05 February 2009 - 06:19 PM

Thanks for your time Derek. My question has to do with catchers. It seems that once you get past the top 5 or so, there seems to be a drop off and a plateau. Is it your strategy to try and draft one of those top catchers or try and steal a high upside guy in the late rounds? Cheers!


This is pulling from my answer to the last question, but I definitely try to get a top catcher, if not two (in two-catcher leagues). I'll usually take my first in round four or five, and then take the second depending on what I know about my opponents and the catcher depth that year. In 2009 (after getting Martin, McCann, or Mauer), I'll likely wait out a guy like Ryan Doumit or Chris Iannetta (Round 9 or 10) for my second catcher instead of taking Victor Martinez or Geovany Soto in Round 5, 6, or 7. Mike Napoli is also a guy I like, and while his value is really starting to heat up, if I'm in a league where I can get him after Round 15 (or better yet, in an auction where I can see what he goes for before the other top catchers come out), I'm waiting on him. I think he could be a top five catcher this year, and I would take him in the top 10 or 12 rounds if necessary. If I somehow miss out on my second catcher, I'm waiting until the end of the draft to take Kenji Johjima.

#25 Derek Carty

  • 30 posts

Posted 05 February 2009 - 06:28 PM

My question has to do with career trends and splits. You tout Pedroia's consistancy and I don't think many Sox fans would argue that the quality is one that is appreciated here. However, over his two years (which I know isn't enough for a trend, but I'll get to it), he seems to have three months where he is an ~.800 OPS guy, two months where he is a ~.900 OPS guy, and one where he is below .700 (April). For Pedroia, luck seems to be a part of this and his BABIP is fairly low in that, but for some players it seems to be something to set your watch by.

Shea Hillenbrand is primarily who I am thinking of, and while it evened out by the end of his career, March/April was still his best month on the whole and he seemed to always get off to a roaring start and then cool down a couple months during his short time in Boston.

Early on in a draft, you seem to value consistancy, but if a player has a history of being a cold starter, or tiring by August, do you take any precautions for those past streaks? For instance, suppose Pedroia makes a cold April a trend and it turns out to be more than luck. A few years down the road, would you hold a 2B on your bench and play him until Pedroia becomes his usual self and then drop him, or do you weather the storm of a .260 BA, a HR or two, maybe seven or eight RBI and R, and a handful of steals?

(Side-Note: This is going under the assumption that it would be a consistantly set-up streak that happens, more or less, annually. Neither Pedroia nor Hillenbrand would qualify under my view, so it's mostly hypothetical).

Thanks for your time.


If we could prove that these kinds of things were a legitimate trend, then the projection would need to be adjusted a bit. In your hypothetical Pedroia case, I would combine a May-September Pedroia projection with 1/6 of the projection for whoever I'll be playing in April. It makes little sense to play Pedroia if his true April talent level is worse than a potential replacement. That's wasting value. In this situation, I might put a slightly higher priority on getting a quality 2B (i.e. better than 'April Pedroia') later in the draft with the intent to play him in April.

Another consideration would be to let someone else get Pedroia and then trade for him at the end of April. Of course, if this has happened long enough to call it a real trend, it's likely others have noticed it and won't give him up very easily. Same goes for drafting a player and trying to trade him in August if he's known for tiring out.

#26 Derek Carty

  • 30 posts

Posted 05 February 2009 - 06:29 PM

And that's all of them!

Thanks again to all of you for your questions. If you have any follow-ups, feel free to post them and I'll answer them as they come in.

#27 Frisbetarian


  • ♫ ♫ ♫ ♫ ♫ ♫


  • 4632 posts

Posted 05 February 2009 - 06:40 PM

And that's all of them!

Thanks again to all of you for your questions. If you have any follow-ups, feel free to post them and I'll answer them as they come in.


Excellent work, Derek. I especially liked the answer to the linear weights vs wpa/li question as that has been the subject of a few classes Sal Baxamusa and I have done together, and I'm always on the lwts side.

Thank you so much for taking the time to respond to all these questions. If anyone has any follow-ups, please feel free to post them here. Also, if any lurkers have follow-ups, feel free to PM them to me.

Remember, the Hardball Times Annual 2009 is currently available, and David Gassko's preview of the annual can be seen here.

#28 absintheofmalaise


  • too many flowers


  • 10773 posts

Posted 06 February 2009 - 08:58 AM

As far as what is good to use, personally, I like Tom Tango's wOBA (which is based on linear weights) for offense, MGL's UZR for defense, and Dan Fox's work on baserunning. For pitchers, I like David Gassko's LIPS ERA (though defense, ballpark, etc also need to be taken into consideration).


Thanks for taking the time to do this Derek. I don't put a lot of stock into any defensive metric either, but I'm curious about why you prefer UZR to Dewan's +/- system.

#29 Frisbetarian


  • ♫ ♫ ♫ ♫ ♫ ♫


  • 4632 posts

Posted 06 February 2009 - 09:24 AM

They are, however, better than anything else we have and are clearly better than using any single person's eyes (that players like Derek Jeter and Michael Young win Gold Gloves can attest to this). Also, if you look at With or Without You Studies (made famous by Tom Tango), these largely confirm (the ones I've seen, at least) the findings of systems like UZR, RZR, OPA!, or whatever other advanced system you'd like to use.

Once we get something like a FIELDf/x where we can see where fielders are positioned, exactly where the balls are being hit, the speed off bat of all balls, and things of that nature, we will be able to do a much better job of evaluating defense.


I think that depends somewhat on whose pair of eyes. I was involved with a sabermetric seminar last summer, and one of the speakers was a scout from the statistically inclined Arizona Diamondbacks who was in town to do a game at Fenway. He said something that really stuck with me. I'm paraphrasing here, but the gist of his comment was that the evolution of sabermetric methods was slowing, and he believed teams in the future may gain the largest benefit over their rivals by having scouts that could more accurately judge player's ability (we were talking defense at the time) on the field. It was an interesting take, but I do wonder how much "traditional" scouting has been overlooked (maybe too strong a word) in the sabermetric community, and how you feel about his comment.

#30 Derek Carty

  • 30 posts

Posted 06 February 2009 - 10:28 AM

Thanks for taking the time to do this Derek. I don't put a lot of stock into any defensive metric either, but I'm curious about why you prefer UZR to Dewan's +/- system.


From what I understand, the two are very, very similar, and I wouldn't begrudge anyone who uses +/-. MGL has said himself that the two are basically identical. I believe UZR is park-adjusted and +/- isn't, and +/- uses a different database (not saying that's good or bad). The big factor, for me, is that UZR is now more readily available at FanGraphs. Also, UZR is presented as runs above average, not plays above average, which makes a little more intuitive sense.

#31 DJnVa


  • Yes


  • PipPipPipPip
  • 17253 posts

Posted 06 February 2009 - 10:43 AM

I'm also in the process of writing up why injured/injury risk players, as a group, might be undervalued:
http://www.hardballt...rewards-part-2/


Apparently the Sox FO agrees as this describes the Sox off-season.

#32 Derek Carty

  • 30 posts

Posted 06 February 2009 - 10:50 AM

I think that depends somewhat on whose pair of eyes. I was involved with a sabermetric seminar last summer, and one of the speakers was a scout from the statistically inclined Arizona Diamondbacks who was in town to do a game at Fenway. He said something that really stuck with me. I'm paraphrasing here, but the gist of his comment was that the evolution of sabermetric methods was slowing, and he believed teams in the future may gain the largest benefit over their rivals by having scouts that could more accurately judge player's ability (we were talking defense at the time) on the field. It was an interesting take, but I do wonder how much "traditional" scouting has been overlooked (maybe too strong a word) in the sabermetric community, and how you feel about his comment.


I absolutely agree that a large part of the sabermetric community doesn't give enough credence to "traditional" scouting. I wrote an article discussing my thoughts on sabermetrics/scouting/player evaluation here: http://www.hardballt...ion-philosophy/ Tom Tango also chimed in at The Book Blog shortly after: http://www.insidethe...f_sabermetrics/

The analogy Tom mentions of scouting and statistics being two lenses of the same pair of glasses is a great one, and that those glasses are sabermetrics.

I think that scouting largely gets ignored for a few reasons:
1) A lot of "sabermetricians" or statisticians or whatever you want to call them don't understand the truth of this analogy. They think since scouting can't be perfectly quantified and objectified it must be ignored completely. There is still some measure of reliability with a person's judgment, though, and when looking at things that stats can't capture, this has to suffice.
2) The ones that do understand the analogy may not know which scouts and scouting perspectives are trustworthy.
3) Very few people are capable of analyzing both stats and scouting by themselves, so they must rely on others for the other half.
4) There are very few scouts of any nature publishing their work on a regular basis to reference. Most work for major league clubs.
5) The proper way to combine them isn't completely understood by a lot of writers

Clubs like the Indians and A's and especially the Red Sox seem to understand the analogy, have been able to overcome some of the hurdles, and work to combine the two in an efficient manner. Others just seem to be clueless. And if MLB clubs are clueless, how can we expect the majority of internet writers and bloggers to be better? That's not to say that there are some great minds out there - there absolutely are quite a few - but there are others who, like you said, have some learning to do.

Also, in regard to the comment about gaining value from this in the future, teams are gaining value from it in the present. And there may still be more value to be gained down the road once more teams start adapting statistical analysis. There will likely always be an information lag, though, and some teams will always be ahead of the pack. The key is finding what is undervalued and exploiting it. Billy Beane has said that, at some point, speed will be undervalued and the A's will have a team full of 50 SB guys. I think the same can be said about scouting.

#33 Derek Carty

  • 30 posts

Posted 06 February 2009 - 10:52 AM

Apparently the Sox FO agrees as this describes the Sox off-season.


Yup. As I mentioned, I love the approach of the Sox this off-season. I said earlier on that it's one I would have wanted the Mets to take. They also probably need an impact player or two, but these kinds of signing are excellent investments.

#34 absintheofmalaise


  • too many flowers


  • 10773 posts

Posted 06 February 2009 - 12:47 PM

I absolutely agree that a large part of the sabermetric community doesn't give enough credence to "traditional" scouting. I wrote an article discussing my thoughts on sabermetrics/scouting/player evaluation here: http://www.hardballt...ion-philosophy/ Tom Tango also chimed in at The Book Blog shortly after: http://www.insidethe...f_sabermetrics/

The analogy Tom mentions of scouting and statistics being two lenses of the same pair of glasses is a great one, and that those glasses are sabermetrics.

I think that scouting largely gets ignored for a few reasons:
1) A lot of "sabermetricians" or statisticians or whatever you want to call them don't understand the truth of this analogy. They think since scouting can't be perfectly quantified and objectified it must be ignored completely. There is still some measure of reliability with a person's judgment, though, and when looking at things that stats can't capture, this has to suffice.
2) The ones that do understand the analogy may not know which scouts and scouting perspectives are trustworthy.
3) Very few people are capable of analyzing both stats and scouting by themselves, so they must rely on others for the other half.
4) There are very few scouts of any nature publishing their work on a regular basis to reference. Most work for major league clubs.
5) The proper way to combine them isn't completely understood by a lot of writers

Clubs like the Indians and A's and especially the Red Sox seem to understand the analogy, have been able to overcome some of the hurdles, and work to combine the two in an efficient manner. Others just seem to be clueless. And if MLB clubs are clueless, how can we expect the majority of internet writers and bloggers to be better? That's not to say that there are some great minds out there - there absolutely are quite a few - but there are others who, like you said, have some learning to do.

Also, in regard to the comment about gaining value from this in the future, teams are gaining value from it in the present. And there may still be more value to be gained down the road once more teams start adapting statistical analysis. There will likely always be an information lag, though, and some teams will always be ahead of the pack. The key is finding what is undervalued and exploiting it. Billy Beane has said that, at some point, speed will be undervalued and the A's will have a team full of 50 SB guys. I think the same can be said about scouting.

I'm in full agreement that you must combine traditional scouting along with advanced stats in order to properly evaluate players. You mentioned that the Indians, A's and Sox are ahead of the curve in that regard. Which other teams would you add to that group and which teams do you think are behind the curve when it comes to combining both aspects of player evaluation. Which teams do you think are still lagging behind in using saber stats. I'm guessing that the Braves are lagging behind still. I know they were when JS was the GM. Not as much now with Wren as the GM, based on some comments he made about Vasquez, he mentioned on a radio show here that they used park factors, defense and other factors in evaluating him. Maybe there is some light at the end of the tunnel for some teams.

#35 Derek Carty

  • 30 posts

Posted 06 February 2009 - 04:21 PM

I'm in full agreement that you must combine traditional scouting along with advanced stats in order to properly evaluate players. You mentioned that the Indians, A's and Sox are ahead of the curve in that regard. Which other teams would you add to that group and which teams do you think are behind the curve when it comes to combining both aspects of player evaluation. Which teams do you think are still lagging behind in using saber stats. I'm guessing that the Braves are lagging behind still. I know they were when JS was the GM. Not as much now with Wren as the GM, based on some comments he made about Vasquez, he mentioned on a radio show here that they used park factors, defense and other factors in evaluating him. Maybe there is some light at the end of the tunnel for some teams.


Well, I think the Rays and the D'Backs would be in that group (although why they let Randy Johnson go perplexes me). You might be able to make a case for the Jays or Padres, though some of their moves seem to indicate otherwise. The Pirates may join the group soon. Jury is still out there.

Lagging behind are teams like the Giants, Reds, Royals, Astros, and even successful teams like the White Sox, Dodgers, Angels, Mets, and Phillies. You'll notice the successful ones all play in pretty sizeable markets and have succeeded (at least in part) as a result of either having money to throw around at players who everyone knows is good (Beltan, Vlad, Manny, etc) or through good scouting and player development (Utley, Howard, Hamels, etc.). I think those are kind of the common denominator there.