OK, we’ve got a few more bases covered. In the PFM, all the fielding data should be up-to-date, so Ryan Braun is back to being an exclusive left fielder. There was a pretty big problem with the players I added manually, as a temporary patch - Moyer, Uehara, and a few more. We got the programs to read their data, and so now they show up in the PFM and in the depth charts - trouble was, “WHIP” wasn’t one of the fields I initially entered, and when the PFM saw guys with 100 innings and a zero WHIP it gave them all Sabathia-sized contracts. That is fixed. These players still don’t have entries in all fields, though, and if your league uses one of the un-entered fields those players could still have strange results.
On the depth charts, the batting/onbase/slugging digits problems has been broken, fixed, re-broken, and re-fixed, and I think I know where I made the mistake that re-broke them so it should (fingers crossed) not happen again, unless I come up with a third way to break them. The quality start numbers have been largely repaired, but more in the sense of a field dressing than a proper fix; I’ll revisit them shortly.
What else - I’m going to redo the way saves are handled tonight, so expect the PFM pitcher results to look a lot better tomorrow morning. Braden Looper signed with the Brewers yesterday, so I had him displace some of the spot starters on their roster. I fixed a typo that had Josh Willingham drawing 5% of left field time instead of 75%. I’d like to get the PECOTA-based Playoff Odds report up and running in the next couple of days, and play around with some of the remaining free agents.
I got through a large portion of the to-do list. Three decimal places on BA/OBP/SLG, check. Another pass at lineup orders. Games started fixed; quality starts and saves are still wonky, as is pitcher’s upside. The Nationals list is now complete. Jason Botts is gone. The defense numbers will be added later today; they aren’t in yet, but at least I tracked down the what and where to get it fixed. I also made a series of player moves that came out yesterday, some of which I’ll revisit down the road - Jeff Weaver, Rich Aurilia, Chad Bradford, Edgar Gonzalez, Adam Kennedy, Brendan Donnelly, Andruw Jones, and Rick Porcello, along with various changes to accomodate them.
Some players are still “missing”. You can’t see me using air quotes there, but their statistics (Jamie Moyer, Koji Uehara) ARE being included in the team total lines and they ARE in the database; for a reason I haven’t figured out yet, they are just not being displayed on either the Depth Chart or PFM pages that you see.
That’s not to say that all is done, no no no. I’m going to be your executive guide this year, and I expect to follow an update schedule that will be pretty close to daily (I’m setting 9 PM eastern, as my target) throughout the fantasy draft season - allowing that there will be days when nothing significantly chart-changing happens. And we all know that the fantasy season doesn’t end with the draft, so I’m aiming to keep these updated throughout the season - dropping to weekly - where they will also serve as improved inputs to our playoff odds reports.
“Chart-changing events” are primarily going to be clear and obvious - a player signs, or gets cut, or gets injured. I am not going to be changing somebody’s playing time because he went 3-3 in a spring training game; that’s a severe over-reaction. I’m a fairly serious fantasy player, in contrast to some of the other BP writers, and I’m approaching this as exactly as I would one of my own drafts (I have my first one on 2/21, less than two weeks from now - got some open spots if you’re in the DC area. Play against someone who’s giving you his entire draft prep!).
A lot will depend on spring training games, though - I really don’t see a clear choice between Gaby Sanchez and Dallas McPherson in Florida, nor do I see any obvious answers as to how the Yankee center field situation will shake out. In situations like that I’ve often split the baby, knowing that one of them is likely to push close to 100% while the other drops down near zero. The numbers that I’ve put down are meant to be big averages - as if we had a large number of universes with their own baseball seasons, and in some of them they went with Melky Cabrera again, some with Nick Swisher, and some with Brett Gardner. I have _not_ tried to capture any effects of trading personnel or signing free agents - the depth charts are based on allocating reasonable amounts of time to the players they have now. Minor league callups, however, are included, albeit subject to big(ger than otherwise) error bars.
As I write this, the appearance of the cards is the same as last year’s. There are some changes coming, like getting comments onto the individual team cards. I’d like to be able to explain my thinking on specific cases, and I want to hear from people who, frankly, know more about most of these teams than I do. I am quite sure that the best way to do that is to contain the discussions to each team - else the Giant and Twin comments will have a hard time cutting through the expected New York/Boston flood. The downside is that it gives me 30 threads to read.
Most of the software that we have to run the depth charts was already in place, and I’m admittedly still learning to use it. I made several “fixes” over the weekend, only to realize that it was already covered and had to fix it right back. I’m seeing some of this output for the first time as I type this, and like a freshly painted wall, I can see several places where I need to touch up. The batting/onbase/slug numbers are only going to two significant digits - that looks like a copy/paste artifact, and at least is easy to fix. There are a few players who are in my data but aren’t showing up - Kenji Kawakama and Koji Uehara, I’m looking at you, but I’m not seeing you. You guys need to be added to our master registry. The games started figures are so far out of whack that I must have made a serious mistake somewhere, and the way the PFM hands out saves is something I’ve grumbled about before - I’ll be revisiting both of those topics. No, I don’t really think Mark DeRosa will bat ninth for the Indians; add “double-check all lineup orders” to the to-do list. There are a number of players, maybe a dozen, who did not have a PECOTA projection, and I subbed in what was basically their three-year average DT; those should get fixed when we get the next PECOTA run from Nate, possibly with a very different forecast.
So please, don’t tell me about those problems, I already know about them. Do let me know if you spot something else, particularly any players on a wrong team - I really hate it when I do that. Enjoy, debate, use, criticize; I’m looking forward to having you help me make these the best depth charts available on the web.
Clay - not sure if this is the type of error you want to hear about or not, but Lastings Milledge is not getting any playing time for the Nationals, and is not in PFM.
Thanks.
Yes, looks like the entire Nationals output got squashed. A dollar says its related to being the last team alphabetically, and something cutting off before it got to all of them.
The easiest ones to miss are the ones who either sign to Japan or get picked up/dropped in minor league transactions. I guess the good news is I didn't have much planned for him anyway.
Great job - though I'm bummed to see the Cards projected to be sub-.500. Didn't see Jason Motte in the Cards' depth chart. Do you think that LaRussa will not use him significantly this year? I think he's a sleeper for the closer position in St. Louis.
What I'm waiting for is how long it will take for the turmoil to begin when more people realize that you've predicted a 79-83 record for the LA Angels. I can immediately understand why, but it'll be tough explaining that one to folks less familiar with what you and a few other have been doing over the last decade.
Clay if I am reading this correctly. You are saying that the Mets have done a pretty terrible job of building around a monstrous young core of Beltran, Wright, Reyes, and Santana. It would seem that they should be immensely successful if only they had decent options at LF, RF, 2B, C, S3, S4. How would you fix them?
How long will it take to update positional eligibility within the PFM? I want to avoid drafting B.J. Upton as a 2B, Carlos Guillen as a SS, etc.? The PFM is a great tool that really shows only the "right" number of players at each position, so I am eager to get its view of the draft-able player universe within each position.
One possible problem, however: I just ran a test with the PFM, using UPSIDE as the only statistical category. The hitters seemed on target but the pitchers were not. The pitcher with the best upside, according to this measurement, is Kameron Mickolio, with a 0.8 upside. Every other pitcher has an even smaller upside.
I did have something happen where he wasn't recognized when I typed in "Jamie Shields" out of habit. I'm thinking the same problem might have happened when the PFM loaded.
No, its an artifact. My top priority was on getting accurate rosters and then playing time estimates logged for every player; while I was trying to set batting orders as I went, there were times when I skipped that step.
Rock. On. Too bad I have some deadlines at the office or I might just have taken some sick time this afternoon...
Clay, kudos for getting version 1.0-2009 up and running, and I'm sure you'll get all the kinks worked out for version v1.1.
Any chance that v1.1 can include Fielding Percentage for the position players? My Yahoo fantasy league uses it to add "defensive balance," and I get killed every year in this stat... keeps me from winning my league, actually.
It doesn't look like any of the no-PECOTA players came through - and Moyer, surprisingly, was one of them. He's got to be running pretty short on the comps.
So how did this affect the projection for the phils? The man is probably going to pitch at least 80 innings this year. He could prevent 10 runs for them, he could prevent 30. Can we through in some dummy projection?
I don't think it affects the projection at all. If I understand it right - dangerous words to be sure - the player stats are part of the projection (because I know I put them there, and they were part of that summation).
The part of the program that collects the data that goes into the PFM isn't grabbing them, possibly because when I entered a "fake" line for the player I did not enter a value for every possible field, and I left out something that's actually important for _that_ process.
Wait, there is literally nothing there for Jamie Moyer. What did you use for Jamie Moyer to run a simulation? He must have been given some innings or they must have been given to a different player, no?
Also, I don't think the Phillies depth chart really reflects the usage patterns that they tend to use. If Shane Victorino gets 80% of the PA in CF, it will be that Werth shifts over when he's not available or hurt or whatever, like last year. I've never even heard of Jason Ellison and I can't imagine Chris Snelling staying healthy enough to run all the way to CF from the dugout, so it seems odd to zap a -9.4 VORP on them for backup CF when in reality, Jenkins, Stairs, Bruntlett, and Dobbs will probably cover all time that Werth, Victorino, and Ibanez are not playing. It seems like there are just CF thrown in there.
Adding a win or two for Moyer, and a win or so for this CF playing time issue, and the projection for the Phillies won't be as off as this one.
Wait Clay. It really doesn't affect the projection at all? Moyer innings are certainly different than Kendrick innings. So if they aren't included, the projection would differ. Right?
This is from a Met's fan who just doesn't want false hopes to fail to be robust for the third damn time in a row.
I've been checking BP daily for PFM- now my pre-season keeper trading can begin in earnest! Well, it will once I know what PFM has to say about Lastings Milledge.
Quick question on the depth chart: how does your system handle closers and saves? As it stands now, even though Qualls was listed as the closer late in 2008, he is projected to have 6 saves on the depth chart, causing his value to plummet. Surely your system would not rely entirely on PECOTA, which would only use his 2008 saves total as its basis for 2009 saves.
Also, how do you handle uncertain starter circumstances? For example, your Reds CF is Willy Taveras, although the Reds would appear to have a much better option in Chris Dickerson. On what do you base your AB totals on those situations?
And, finally, my most surprising take on the Depth Charts: my Cubs are predicted to lead all of baseball in OBP. Eep!
The idea isn't to put together the best WARP team, but anticipate what the teams are likely to do. Taveras carries the "proven base stealer" and "defensive stud", for right ot wrong, and I think it will be up to him to play his way out of the lineup.
Any chance of including free agents (Dunn, Manny, etc) before they actually sign? I am sure there are park-effect and playing time issues, but a rough cut would be better than nothing.
Closer to the XMas mornings of my youth than any other day through the year. Thanks for all the hard work Clay. Long time fantasy participant, long-suffering O's fan here in Charm City. Interested in the fantasy league opening if not filled yet.
IS that a light at the end of this dark Os tunnel as represented by the PECOTAs for Markakis, Jones, Pie and Weiters?
Just checked on Josh Hamilton. His projections seem to be from 2008: 24 HR and .350 OBP. Or, is PECOTA really projecting that much of a decline? Seems odd that Hamilton is ranked well down (18th or 19th) when using PFM. Inputs were for R, SB, HR, RBI, TB, OBP. I trust you guys, but this one seems kind of low. Thanks.
Dustin Moseley's stat line on the Angels looks ridiculous to me. First, he is a "spot" starter with 38 starts. 15 Quality starts and only 110 innings. The 15 Quality starts needs at least 90 innings (six innings per), leaving only 20 innings for the remaining 23 starts........
I don't know if your PT%'s are calculated or manual but from someone who follows the Jays pretty closely both Thigpen and Adams have been DFA'd. I think its pretty much common knowledge that Snider will be the primary LF (assuming he starts with the Jays) and Lind will be the primary DH with the possibility of him seeing some games at first. Also as of right now Bautista could see some time at first vs. LHP assuming the Jays don't buy out his asinine contract.
Jays won't buy out Bautista's contract - they just signed it a couple of weeks ago. If they wanted to get rid of him, they would have non-tendered him. Also, the majority of his PT should come at third subbing for Rolen's inevitable injuries.
They have an option to buy out similar to Reed Johnson's last year. His deal is ridiculous .... in this market its at least 3 times his market value. You could buy him out for $400,000 and get a much better player for $2 mil in this market.
For the Pirates, all indications have been that Nyjer Morgan will start the season penciled in at LF and Steve Pearce will start in AAA. Pearce is almost certainly not going to get 70% of the AB in LF.
For the Pirates, all indications have been that Nyjer Morgan will start the season penciled in at LF and Steve Pearce will start in AAA. Pearce is almost certainly not going to get 70% of the AB in LF.
Any chance of adding Innings per start as a category? I know that would give weird numbers for the swing pitchers, but it would be straightforward for the full time starters. Also, maybe Secondary Average? It's a terrible stat, but one of my leagues uses it. (TB - H + BB + SB - CS) / AB
Maybe I'm reading this wrong, but it looks like the Nats will be leaving CF vacant for 90% of their games, with Elijah Dukes and Willie Harris splitting the remaining 10%
Has defense been included yet? I remember the Rays were projected by PECOTA to be about a .500 team this time last year. Then, in early March when defense was added, they went up to about 90 wins.
Good point about defense. That was a huge lift for the Rays. Question, it would seem the AL East has a 'strength of schedule' adjustment to it's Wins / Losses, as the Jays Pathag would be higher at that Run Differential.
At the moment it looks like he only has PECOTA by proxy. Being that the participants enter a set of ranked players, I think BP might get short-changed if the PFM isn't used, and I'd really like to see how it stacks up.
Can someone remind me if postseason is factored into PECOTA? I'm wondering about Upton's projection of 16 HR's, just over double what he did in the playoffs.
I'm pretty sure PECOTA does not use post-season. But it does (since last year, at least) weight the second-half of the previous season a bit more heavily than the first half (something like 55-45 IIRC).
The play time projections are taking a beating from Mariners fans in the know. Ichiro as 525 PA? He hasn't had under 700 since forever. (Of course, PECOTA can't figure Ichiro, so whatever.)
Also, are there any defensive metrics that go into calculating RA for each team, or is that based solely on the pitcher's RA and play time estimates? (This is another one thing M's fans don't like, as we're keen on the OF defense that we're bringing to the table next year.)
The amount of people who didnt read the comments above are stunning. I found the first fifteen "Player X on my favorite team doesnt have the appropriate amount of playing time!" comments funny. The next hundred are just irritating. RTFA, people.
Clay is already aware of these problems and it says so above. Please stop.
A pretty minor thing, but Alex Cora will likely get the bulk of the time subbing in for Luis Castillo, not Argenis Reyes. It is, in fact, the reason the Mets brought in Cora in the first place.
I would guess it's probably Castillo-55, Cora-30, Reyes-10, Anderson-5.
I'll take a look later today when defense is factored in. My guess....Rays move up, Yankees move down, Phillies move behind Braves, Cardinals fall further, Brewers fall further, Angels move up, but still behind the Athletics.
I'm curious as to whether the model handles positional battles / depth correctly, or whether it could be improved, perhaps dramatically (project for next year).
Team A has one SS with a PECOTA projection of a 750 OPS. Team B has an identical player, but also a second guy with a 730 projection. You reasonably assign a 50 / 50 PT weight to the latter pair, even though (as you explain) in some universes the actual split will be 90 / 10 or 10 / 90 depending on who wins the job in ST.
I'm guessing that, in your model, team A ends up getting better projection from SS (750) than team B (740). In the real world, of course, the opposite is true, because often the team will pick the player having the better season (or likely destined to, based on physical condition). Significantly often, the guy who wins the job will outperform his PECOTA mean projection, and (perhaps more importantly) the winner is much less likely to grossly underperform. No matter who wins the job in ST, if he's at his 10% projection in mid-May, the other guy is likely to get a shot.
The proper way to do this would be with a season simulator that has both a real and random component for variation about a player's mean projection and which dynamically adjusts PT. The simpler way would be to just use the 60% (or 75% or whatever) PECOTA projection for players at deep positions.
Interesting idea, but keep in mind that the projections are actually major league equivalents. There are plenty of players in the PECOTA spreadsheet who may draw 500 or more plate appearances but may not get a single one at the big-league level even if their equivalent is expected to outdo the big club's regular.
The full, complex methodology would be smart enough to handle this and more.
For instance, you project the MLB 1B to hit .274 / .395 / .582 and the blocked AAA 1B to hit .259 / .341 / .508. In certain runs of the simulator, the MLB guy gets hurt and puts up something like a .207 / .360 / .352 through the end of June before being shut down. In those runs, the simulator calls up the kid from AAA and maybe in some of them he hits .288 / .356 / .567. Voila, extra accuracy! But without such a methodology, you are not factoring in the way having a Ryan Howard in AAA protects you from the unexpected collapse of a Jim Thome.
Even the simpler methodology I suggested would do a better job than simply averaging PT. If the Yankees suffer an injury to one of their top 5, the replacement will be whoever is pitching best among Hughes, Aceves, and Kennedy. So whoever gets more starts than projected from that group (at the expense of the others) will also be likely outperforming his PECOTA mean projection. This puts a team like the Yankees in a better position than one with just one or two extra viable rotation candidates.
Rounding. Unlike the Adjusted Standings, there are no decimal places on the wins in the depth charts, and at any given time it's possible a few might be getting lost when figures get rounded up or down.
I always wondered if there was a way to extend these depth charts down into each organization, at least to AAA and AA, if not A ball. I was thinking of that when I was looking over the Rangers organization issue at first base, and couldn't figure out who would go where. It looked like
Chad Tracy AAA
Mitchel Moreland AA
Ian Gac High-A
Justin Smoak Low-A
but if they wanted to challenge Smoak by sending him to High-A, I don't see how they can, as Ian hasn't really conquered High-A to the extent that he wouldn't bomb out in AA and while Moreland certainly earned the promotion, he could be played at outfield or DH. But if you leave the lower levels alone and send Tracy back to AA then there's really no one to play first at AAA unless you want to use Chis Shelton or another castoff there. Not that AAA first baseman are hard to find. I guess the big question is how confident are you in Chris Davis' ability to stay healthy. I personally think he won't have much trouble with that but its got to be some kind of concern.
Also Clay: You mention the "three decimal places... check." I'm assuming you are suggesting this is fixed but I'm still getting just two decimal places in stats like OBP, SLG, AVG...
These aren't usable right now. I just downloaded AL- and NL-only versions. The total salaries come ina round $3120 (check), but the number of absolute sucks is too high to write-off as insignificant.
Jamie Moyer's $$ is over $40, and who is this Micheal Madsen coming in as the top player?
Depending on how I put in pitchers (6SP, 4RP as opposed to 5SP, 2P, 3RP, etc) some pitchers disappear completely - those listed as swingmen (some of whom will NOT be swingmen). I also occasionally get a few nobodies with crappy stats listed at $20-30 players.
Question for anyone. I only use PFM for initial values - I don't play with it at all during the draft. I use Disposition moderate, PA-1, no inflation (of course), and SGP-1. How do people feel about PA-2 or PA-3? Other settings?
Clay I know you said that you'll be updating these as the season progresses. When you publish an update are you going to be telling us which teams you've updated? Or just that something somewhere is new? For now, of course, there is wholesale cleanup, but once things stabilize I'd expect the changes to be more piecemeal. Thanks!
It's nice that it's out there, but the PFM is unusable right now. Projected Saves are an absolute joke and that throws off the whole value for pitchers. The Padres are projected with 19 saves as a team. How is that possible, even with uncertainty over who will get the saves?
I appreciate the work, but we're basically beta testers at this point. Granted, there's still time to fix this, but I think it was a mistake to release this with this many bugs.
This may have been said earlier, but I read no indications that Antonelli will be the primary 2B in San Diego. In fact, Eckstein was signed for that spot with Rodriguez (sigh) apparently still at SS.
Ok, so PFM was updated today, and it now shows Holliday as an Oakland A. I simply don't believe the projection. How is a guy with a career OPS outside of Coors just barely north of .800 going to hit .327 in Oakland, one of the worst hitter's parks in the tougher league??
New to PFM: I see at least one person saying he doesn't use inflation - other opinions on this? I thought it seemed useful, but do folks not find it so?
104 comments have been left for this post.
Clay - not sure if this is the type of error you want to hear about or not, but Lastings Milledge is not getting any playing time for the Nationals, and is not in PFM.
Thanks.
Dovetail on Dschmitz3, Josh Willingham isn't up either. The Nats depth chart in general looks a bit thin (< 100% PT at each position).
That said, I'm as giddy as a schoolgirl over these. Thanks, Clay!
Also I'm pretty confident that - as happy as this would make me - the Nats won't score 786 runs.
Yes, looks like the entire Nationals output got squashed. A dollar says its related to being the last team alphabetically, and something cutting off before it got to all of them.
I think you'd be right since there are no hitters listed that appear after Kearns alphabetically.