Some of the new features you’ll see on these cards:
* ten-year projections, versus the seven-year projections previously offered.
* top 100 comparables, versus the top 20 previously offered.
* better integration with the rest of the site, including links to articles, chats, and roundtable mentions. Much more coming on this front.
* update: forgot to mention that tables are sortable by column. We’re not sure that this has any use with these particular tables, but since we’re going to be using it for a lot of other stuff, we figured we’d throw it on these and see if you all wanted it for anything. Just click a header to sort, and click again to reverse order.
Some stuff either doesn’t work yet or requires more explanation, including but not limited to:
* we have display issues with Internet Explorer, and I’m sure the cards are screwed up in interesting ways in other less mainstream browsing environments.
* search box performance is marginal. Try using the index to manually browse if you have trouble with it.
* similarity scores. We have them, but they’re very different in scale than they were previously, so we’re not displaying them. More on those next week.
* top comparables. The criteria for comps has been refined frequently over the past couple of months, and that explains the differences you see between the comps on the weighted means spreadsheet and cards and those in Baseball Prospectus 2010. More on that soon as well.
* multiple projections for 2010. Here’s the decoder ring: in the “Biographical” box, next to the hitter picture, is our best available PECOTA projection for this hitter, including statistics that our friends who play fantasy will be interested in. If the hitter is on a depth chart and has a playing time projection, we use those numbers (which should match his team’s depth chart). If the hitter is not, we use his 50o projection.
The 2010 projection in the “2010 Forecast” section of the card is park-adjusted, while the 2010 projection in the “Performance Forecast” (10-year projection) section of the card is not park-adjusted, which explains the difference between them. [update: we decided to make these both park-adjusted, because the additional information didn’t seem to merit the confusion having both would cause. Please be aware that in the “Performance Forecast” section, the 2010 projection is park-adjusted but later years are not, and let us know what you think in the comments.]
* weighted mean. We don’t have it (yet). We’re using the player’s 50o projection for this first pass.
* update: we got the “k bar” reversed in the player profile graph. We’ll fix that in the next run of the projections.
* you can’t log in and out of your Baseball Prospectus account from the cards (yet). Just head back to the front page to log in if you need to. This will be fixed soon.
Honestly, not sure. We'll have an ETA soon and try like the devil not to miss it. Now that we've made it through an entire card production run with all the new moving pieces we have, it should be quicker than the hitters were.
The comments from the annuals? You see through 2009, right?
We can't publish the player comments from Baseball Prospectus 2010 until 2011--our publisher quite reasonably would have an issue with that. I'd surely like to get more player-specific content on these cards, though, and we're talking over ways to do that.
There are comments from some years missing, and it's a goal of ours to get all those filled in in 2010 as well.
Another thing we'd like to do is add comment threads for each player card. We're looking at how to implement that.
In the year 2010, this separation between print and web content seems very antiquated. Give subscribers access to all the same content that's in the book, even if it costs a bit more; then people can decide whether they want to buy the book or the online version.
Better yet, include free access to the online version for people who buy the book (may need to adjust purchase price, of course). Magazines do this all the time - the print subscribers can access the articles and/or supplemental material online. I'm not in the publishing world, but this doesn't seem like it would be hard.
I want to second that. The idea that there's no way to get the current comments on the player cards is frustrating. It seems so logical to put them up there and figure out a way to make that work pay-structure wise rather than saying it doesn't work with the current way we're doing payment, so we'll post player comments a year late.
Please keep in mind that Wiley has exclusive rights to publish the comments, as well as the rest of Baseball Prospectus 2010, for a year. It's not a matter of us not being willing to publish them online--we simply do not have the right to do so.
I understand that, but they have those rights as a result of negotiation. If the results are illogical, and in they seem to me to be in this case, a better solution could probably be negotiated. For example, what if online subscribers could pay some additional price(less than the cost of the book) to be allowed to view current year player cards online, some percentage of which could go to Wiley. The current system doesn't make a ton of sense.
Keep in mind that the marginal costs are much higher for the print edition. Wiley is still paying for a ton of sunk costs even if they sell less books because people purchase the data online.
Looks like the bar height for strikeouts on the player profile bar graphs is the opposite of what it should be - i.e. it's quite low when the player doesn't K and quite high when they have a high tendency to do so.
Lemme see if I understand this. The "bio" projection is the one that PECOTA spits out based purely on comparable player performance while the "forecast" projection takes that and adjusts it for the park?
I feel like I should be able to figure this out on my own.
"Just when I think you couldn't possibly be any dumber, you go and do something like this… and totally redeem yourself!"
If the previous PECOTA confusion was necessary to get these awesome looking player cards up - it was completely worth it. Love how all the mentions of a player are tied in, too, even chats.
This was the best example I could find for the problem I have with the Performance Forecast. Every player - no matter their age or ability - is simply projected to put up the same year or worse and then gradually decline no matter what. Even if that's the safe prediction it seems pretty not-useful.
Gentlemen, let me be the next (of many) to congratulate you on bringing the best Player Card set-up ever. I love love love all the extra facets of the cards this year (the links to the chat comments as well as the pertinent articles are both absolutely fantastic additions). Will these links be updated throughout the year, and, if so, how often? Also, the overall presentation of the cards is splendid and much easier on the eyes than ever before. A hearty congrats.
However, my main issue comes with PECOTA itself. I still don't think all the kinks have been ironed out, and, as mentioned above, Justin Upton is a great example. It's pretty hard to believe that his 90th percentile projection in 2010 is actually significantly lower than his 90th percentile from last year's card given what he did in '09. And his Projected Forecast for the next 7 years suffers the same affliction. His forecast is worse across the board for all future seasons (and decliningly so) compared to last year's forecast. You guys did remember to tell PECOTA he wasn't 35 years old, right? :)
I should also note another example, just because I see it as equally egregious.
Chris Davis' 90th percentile this year has him at .307/.365/.645 with 52 homeruns (!) in 660 plate appearances. His 90th percentile last season was .287/.345/.559 (which was already pretty high).
I don't think we can look at any 2010 PECOTA projection material thus far and call it anything but broken (or beta-broken, if you prefer). I will, however, anxiously await the day (hopefully in the near future) when the system has been fixed and we can all have confidence in its projections once again.
I've heard rumblings of this nature, but I didn't pay close enough attention to last year's numbers to really lean one way or the other. Thanks for the head's up, though!
Agreed that this problem started last year (at least) but not with the implied(as I read it) conclusion that the underlying problem has been fixed for this year.
I can't find a promising young player who doesn't fit this profile in the cards. I understand having conservative projections and discounting out years, but I don't understand why they are projected to have a 2010-2012 peak, followed by decline.
And Montero's 2010 projection is spectacular; Montero's viewed to peak next year at a high level at the age of 20 and then decline.
This is a radical change in what projections look like for these sorts of players. There has been a substantial change to PECOTA. It isn't clear that the people who changed it know the changes made; the issues with all sorts of numbers are pretty substantial.
The SS/Sim numbers look a little off, too - Miguel Tejada oughtn't rank below Daric Barton for next year with the projections given.
The real problem here is that the cards are pretty, but broken in some untrackable way.
yeah, this is what I'm seeing as well. I haven't checked a lot of cards, but I haven't yet found anyone who's projected to have a better 2011 than 2010, which just seems kind of unlikely.
Yes. This looks like a total disaster that will not be fixed until next season. When the core of your offering is accuracy of projection, and you are experimenting aimlessly in March for the current season, it's a very, very bad sign. Glad I have not renewed my sub. yet.
Also, I checked the cards for Mets prospects Davis and Thole, and the cards say there are no articles that mention them. This is not true, as they are mentioned in Goldtein's top 11 Mets Prospects article. Both players also DECLINE every year from the present, which is hard to believe for decent prospects under the age of 23.
Overall, they look good. As a practical matter, I am not sure that making the "cards" approx. 20 browser pages long is necessary. The related articles, roundtables and chats should be separate links, rather than part of the player's main page.
I disagree with you about the chats, roundtables, and articles. I think having them on the cards makes for a much simpler one-stop shopping style of browsing. Why should I have to open up multiple tabs in my browser to do what one page can easily do. This also makes it easier and quicker to access the info for mobile users, too. I abhor having to jump back and forth between pages on any mobile device.
In this day and age, I'm not sure why having more info on one "page" of the internet is a bad thing. If you want to print this stuff off, you can always chop the parameters to what you feel is the vital info and leave it at that.
Agreed; these cards are VERY long. I find it hard to justify carrying the comps from 20 out to 100. The Similarity Index (missing from the new cards btw...a real error in my opinion) usually got so low after 5 or so players that there wasn't much reason to even have 20.
Let me second the request for a player's age in the 10 year forecasts. Knowing what age they are for a given year is key for interpreting those long term forecasts...
I guess I could argue Davis having a higher 90% due to the additional age year, I don't think we're talking about the same issue as a young hitter like Upton projected not to grow.
My only point was that an awful lot of the 2010 PECOTA projections and forecasts don't pass the sniff test. Like at all. I think there's a lot more wrong here than just Justin Upton and "young hitters like [him]."
In my opinion, there's just no way that the PECOTA program of Nate Silver's era projects Davis with a 90th percentile SLG% of .645 and 52 homers after the year he just had. No way in a million years. That's what I mean about the sniff test. Upton's doesn't pass, but neither does Davis' for an entirely different reason. I like Chris Davis as much as just about anyone, but that's a loopy projection.
No. In original design there were in principle between 20 and 50 comps per player. And to accomplish that they sometimes needed to relax criteria or collapse categories for individuals who were relatively "incomparable."
One thing that seems to be missing from the cards is the overall Similarity Index, which was one way to assess whether the given individual had a good-fitting set of comps.
Adding in the chat listings is a great feature. There's a lot of good info in the chats that seems to get archived into the ether. Speaking of which, is there a way to search the chats?
Just looked at Longoria's card and noticed for the splits that his SLG and OBP vs RHP and LHP are all lower than his actual projected and eq projected OBP and SLG (not quite for EqSLG, but it still doesn't add up). Seems like something's off, though maybe it has something to do with translations or adjustments that I'm missing.
While I'm sure this has been answered, I don't think I quite understand. Any help would be appreciated.
Why is it that a player's projected playing time would exceed his 90th percentile PECOTA numbers? I'm thinking Rajai Davis here - projection of 551 PA and 41 SB, while the PECOTA 90th percentile shows 454 PA and 40 SB. Why would someone be projected to outperform his 90th percentile?
The playing time projections come from our Depth Charts, behind which a real person is looking at team context and estimating where a team will allocate its playing time.
PECOTA is projecting Rajai Davis' playing time based on his part-timer status over the last few years. Especially for anyone who has been riding a bench or otherwise missed time over the last couple of seasons, you'll want to use the Depth Chart playing time projections instead.
First off, terrific work on this - particularly linking all of the article chat comments about a player to the card. That's a very nice feature, IMO.
Secondly, will the playing time in the depth charts match the playing time in the PFM? It would be great to be able to get the depth chart playing time forecasts and even these 10 year forecasts in .csv, .xls or .sql (in a perfect world - linked to some playerID like MLBAMID) so that we can really play around with them.
I think the playing in the depth charts already matches the playing time in the PFM. I haven't verified this, but I always thought the playing time in the PFM came from the depth charts.
PECOTA has always tempered projections on rookies. Take a look at Chris Coghlan's card from last year. They're almost never fully slated for greatness because there's too high a risk they'll flame out. Remember when Delmon Young used to get Ken Griffey Jr. comparisons?
That is simply not true. No one thought any thing much of Coghlan going into last year, he wasn't a top 100 prospect and no one thought of him as RoY material starting the 2009 season.
Look at some real prospects from last year. Travis Snider and Colby Rasmus were Top 10 prospects. Snider was projected to post a .268 EqA in 2009 and peak at .285 in 2012. Rasmus was projected at .267 in 2009, peaking at .276 in 2011. Now look at Jason Heyward from this year, also a top 10 OF prospect. He is projected for .282 for 2010, and then he steadily declines for the next 10 years. That's ludicrous and completely different from Snider and Rasmus. The fact that Heyward is younger than either Snider or Rasmus makes the projection even more ridiculous.
The whole point of PECOTA is for it to NOT hedge it's bets. The Heyward projection is useless. It says he'll be good this year, gee that's news. Then he'll be ok in the future and almost play as much as a regular. But of course his actual numbers will be ok he just won't play full-time for some reason. If it was really projecting that he would suck that would be one thing, but instead it seems to think he'll be good, but won't go out on a limb at all to project any growth. Actually, I don't think PECOTA has bothered to even climb the tree. It's like one of those long range National Weather Service Forecasts, 33% chance of normal temps, 33% chance of above normal temps, 33% chance of below normal temps. Gee thanks, that really nails it down.
Heyward vs. Snider and Rasmus? I'd just note that Heyward's not all that comparable; Heyward's really a better prospect (accentuating your point). BP has Heyward hitting 96 HR's in the next 10 years.
I'd take the over and give 2-1 odds, thanks. And even that's a sucker bet.
Plus look at Wieters projections from last year, .320 EqA's as far as the eye could see, with improvement from a .320 to .328 EqA and an average of 30 HRs per year.
If this isn't a bug in PECOTA, it's a really crappy feature that renders it worthless.
Above are the EQA/TAv projections from last year's player card compared to his Beta card for this year... keeping in mind the 2009 version shows weighted means and the 2010 is apparently showing the 50% lines. I mean, even the 50% PECOTA should be increasing over the next several years.
A couple silly thoughts... a) Does using the 50% projection cause downward feedback in subsequent years? Or, b) are the additional (and less successful) comps being weighted improperly, which brings down the career projection of the youngest stars-to-be?
Side note and compliment to everyone at BP involved in this endeavor: I think these new cards are going to be rock-solid once these kinks get worked out! They look awesome!
I've nothing but praise for the awesome look. But your "once these kinks get worked out" is consistent with my considerable concern that I can't comfortably count on the PECOTA cards to provide me with the substance I expect.
On the cover of this year's BP Book are the words "featuring Nate Silver's Deadly Accurate PECOTA Projections." And it kinda sucks when I'm not certain I believe those words.
That's understandable. My confidence is a bit shaken, too. But I'm also confident that we've got a dynamic and qualified crew at BP working on a fix.
I keep checking back hoping/expecting to see another announcement to the effect of "The n-teenth exponent of the such-and-such algorithm had a typo in this year's expansion of the comparable players. Beta release 2.0 is being compiled now, and BP's confidence is high that the underperformance of younger players will be rectified with the correction."
I would also have expected someone, anyone, to at least do a spot check on the PECOTAs before publishing them. 5 minutes into browsing some big name players and it's obvious that they are broken. It's also obvious that nobody at BP did that 5 minute check.
Just to pick a random example, Troy Glaus gets 430 PA on the Bravesdepth chart (333 at 1b and 97 at DH), and 430 on his player card but 411 in the PFM. They don't seem to match up.
I will agree that Upton seems too conservative. But PECOTA is a comparables system. A simplistic example: take his top 10 comparables by TAv, the median is between Kaline's 280 and JaunGone's 299. So Upton's 281 projection isn't way off. If you expand to the top 20, then the median TAv is 274.5, so... Upton could be Kaline/Griffey, Gonzalez/Sierra, or Brunansky/Francouer or egads Junior Felix. Sometimes things go well, sometimes they don't.
I'm staring at Gordon Bechham's PECOTA card, trying to make sense of what I should use to project future value. The 10-year forecast screams "mediocrity," and while the "Projected Playing Time" looks a bit more optimistic, it gives me nothing to project beyond one year.
Am I the only one in the crowd who is scratching his head while trying to make use of these? After all, if PECOTA is telling me that I can bank on 10 seasons of so-so performance from Beckham, what's all the hype about? And if the answer is "believe the hype, and don't accept the numbers on face value," then why need I PECOTA?
In the 7 year projection Felix Pie last year was projected to peak at around .261 EqA or between 6 or 7 VORP.
This year, in the 10 year projection Pie is projected for a .261 TAv for 2013, but that equates to 11 VORP.
How does the same EqA or TAv result in VORP's that are 50% different ? I was going to guess playing time, but the lower playing time is associated with the higher VORP's so that doesn't really make sense.
I'm beginning to think of PECOTA like the guy with one hand in boiling water and one hand in ice, on average, he's comfortable.
PECOTA takes these 10 year projections out of comps, but no one really ends up being half one sort of dramatic career path and half another sort of dramatic career path. They tend to follow one or the other. On average, PECOTA's long term comp are comfortable, but in reality, they are a mix of extremes that don't seem to translate well.
So let's take a look at how PECOTA projects the top five hitting prospects in baseball to "grow" over the next five years [TAv taken from 10-year forecast]:
Pedro Alvarez (23 years old):
2010: .266
2011: .260
2012: .269
2013: .258
2014: .259
So, basically none of the top five prospects in baseball are projected to improve over the next five years. Apparently, each has already peaked as a mediocre MLB regular. Anyone who has used PECOTA projections over the years will understand how massively different these projections look than those of years past. They (Pease et al.) have essentially diluted the informational content out of prospect projecions to the point where all major prospects are projected to follow an eerily similar career path.
In short, this is worse than New Coke. Someone has significantly changed the algorithm (intentional or not), and there is no documentation of what has changed or why. There is simply no way to trust any of the PECOTA projections for this season -- esp. those of prospects. This is extremely unfortunate as long-term projections were the last remaining competitive advantage BP had over competing forecast services (for data forecasts, not editorial content). A full article on this debacle (not another "Unfiltered" side-note) is warranted.
I am not trying to bash BP as much as I am expressing my personal disappointment at not having source for accurate long-term prospect projections for the first season in a very long time. I honestly don't know of anyone else who takes a numerical approach to evaluating minor leaguers. If anyone does, please post.
This is why I said (above) that if PECOTA is broken, and the current staff is not sure how it's broken (or how to fix it), it's potentially a website- and brand-endangering problem. Baseball Prospectus has tremendous statistical analysis and great editorial content, and will continue to get my business for that reason, regardless of PECOTA's problems. But PECOTA, and the ability to reliably predict the course of most players' (and most teams') next and future seasons is the bedrock that gives BP an advantage in the marketplace. And whether I like it or not (and to be clear, I don't -- I'm rooting for, not against, BP), if PECOTA is broken and neither current nor former staffers are able to fix it, BP is in serious trouble. I'm sure they know that.
I was going to say that I almost feel bad for BP in that one drawback of having such a passionate customer base is that we clamor for things like PECOTA and then when you give us a Beta version we pick it apart. However, then I looked at my calendar. It's just not unrealistic for people to expect to have PECOTA in their hands by now, the regular season ended 5 months ago, and spring training games start in a few days. It is disconcerting not that while we may have a cool new format for PECOTA cards, the actual content is still not correct at this late stage. The other thing that bugged me is why were the Top 11 lists so late this year ? The book has been out for weeks, but we still haven't seen a Top 100 Prospect article. Typically this arrived just before the book, or coincided with the release of the book. I have my copy of the book now so it's not a huge deal, but when you start to see timelines slipping, and questionable data, you start to wonder if the numbers in the book are worth anything. It pains me to say that because I've been a long-time subscriber to the site and purchaser of the books. I come to BP for the superior content. I don't need pretty web pages. Give me a spreadsheet with the PECOTA card data, and the usual high quality writing on a timely basis, and I'd be VERY happy. If I want flashy, crappy content I can get that at any mainstream sports site. Keep your eye on the ball BP.
I'm also in the "get it right before trying to make it better" camp. And I have often asked - but have never seen an answer - why the PECOTA's seem to get later each season. If the foundation of PECOTA is historical data, why aren't the PECOTA's released closer to the end of the previous season than they are to the beginning of the upcoming season?
For me the hope of getting useful PECOTA's in time for drafting in 2010 is essentially gone. But I hope that DURING the 2010 season the PECOTA algorithms are fixed in time to get something useful - and much earlier - in 2011.
For one-year projections, CHONE (using WAR and RAR as surrogates for VORP) looks like an improvement on PECOTA, even when it was "working" over the last two years. ZiPS has, at least eventually, 25/50/75 percentile forecasts. Neither of these, however, dig as deep into the minors as PECOTA.
On the further downside, CHONE doesn't adjust for real-life projected PT (although that information can be grabbed from other sources). And, sadly, I haven't seen anything numerical anywhere (I've been looking A LOT over the past couple of weeks as this debacle has unfolded) that could stand in the place of UPSIDE, much less any of the cool extra stuff on the cards. So, for this year at least, we're flying blind.
Looking at Everth Cabrera, I'm really confused. The "Projected Playing Time" performance forecast is better than his 90th percentile for 2010 on the same card?
Gotta say, something is VERY VERY broken here. I'm not usually one to demand customer support, but with many of us more then just baseball fans(fantasy owners), I simply cannot use these current projects for 2010 or beyond. There is simply no explicable way on earth an older, worse hitter who had a bad year like Chris Davis can have a higher 90th percentile score then a young, all-world super-stud Justin Upton who is coming off another improving season and by almost every expert account is bound for an amazing, potentially HOF career. This just simply can't be.
As a loyal(5+ year) customer, I have six days left before my subscription ends. I'll be very honest, if there isn't some kind of announcement of something like: 'PECOTA PROBLEM: We're aware of it and a half-dozen BP folks are working 'round the clock to get Pecota fixed, STAY TUNED!!'...
Sadly, not subscribing -- if enough people do it -- is the only way to force BP to treat this problem with sufficient resources and a true sense of urgency. I have been a subscriber for eight years, and have seen plenty of incrementally positive and negative changes over the years at BP. But what has happened with PECOTA since Silver left is nothing short of negligent.
I'd like to hear, "we've called Nate Silver, and though he is very busy, he has agreed to step in and fix the problems sometime this week." Nate used to wrestle with PECOTA data as if he were smarter than the system itself (oh wait, he was, and is...).
I'm in this boat as well. If PECOTA was never truly translated properly (even in '09 as alluded to earlier in the comments) from Silver's spreadsheets to the current program BP is using, I'd have to think getting his mind involved as soon as possible to fix this issue would be the top priority. Also, as mentioned above, I am absolutely rooting for BP to fix this. PECOTA and BP are one of the pillars of my baseball belief system at this point, and my world is crumbling piece by piece as I search more and more PECOTA projections this year.
By the way, Bob, I love the image of Silver actually wrestling with his computer at 4am with smashed Red Bull cans and half-eaten Hot Pockets littered about as collateral damage. Good thoughts, sir.
That's the first step. Do these things to defeat your demons:
Step 1: I see that I have no power over BROKOTA. Me and my buddies ground PECOTA into a fine powder, and snorted 'til it was gone. I must stop using PECOTA in its current form, and stop pretending sitewide that it's delicious in its white powdery form.
Step 2: I recognize that there is a power greater than myself that could restore PECOTA to sanity and could stop me from breaking stuff.
Step 3: I am turning my will over to Nate as I understand him.
Step 4: I will make a searching and fearless inventory of our PECOTA failings. I (and to the extent I can persuade them, my fellow BROKOTA addicts) will stop saying that individualized problems with virtually every number don't affect the final product.
Step 5: I will admit to Nate, to myself, and to others the exact nature of PECOTA's current failings and of our failings in busting it up.
Step 6: I am ready to have Nate remove all of the defects so generated.
Step 7: I will ask Nate to tell me how to stop this in the future.
Step 8: I will make a list of all the subscribers I have harmed, and be willing to make amends.
Step 9: I will make amends.
Step 10: I will continue in the future to actually watch for obvious errors pointed out by smart members, and admit errors early, rather than after the pitchforks and torches are already out.
Step 11: I will ask Nate for his will, and have his will be done.
Step 12: I will try to carry on this devotion to honesty and transparency, so others can emulate my behavior.
Follow the twelve steps to serenity. The guys at New Coke, Crystal Pepsi, the Apple Lisa, and the IBM PC Jr. went, and it all got better. We're rooting for you, and the recovery program is necessary.
My guess is that the problem lies not with a different model than Nate's, but with the different data set being input into the model as compared to previously. I think they included a lot more minor leaguers (most of whom never had major league careers) and that is somehow throwing the prospect predictions out of whack.
Very disturbing - especially considering the imminence of the season. Comment posts in Unfiltered are not enough of a response and do not hearken confidence in this subscriber that the severity of the problem is understood and garnering the attention it should.
I bought the book for the first time this year - turns out to be an expensive, un-attractive decorative item for the confidence I feel in the data therein. And please do not patronize me once again with the rote platitude "The data in the book is good, trust us." Or focus on the book comments as the foundational value of that product - how many of those comments focus on the data or tell a story around the data? Are you offering refunds, or perhaps re-issuing the volume once PECOTA has been upgraded from useless?
Any reference to PECOTA data, until you've fixed the problem, should be banned site-wide. If Normandin wants to establish his own fantasy rankings, more power to him. Structuring that analysis around flawed PECOTA severely limits any credibility his analysis may otherwise have provided.
Rephrasing the question asked above - why the terrible strategic planning in allowing this problem to first surface so close to Spring Training and prime fantasy baseball draft season? Who made the determination to completely overhaul the cornerstone of the BP franchise and publish the book based on data emanating from a system that had not been vetted?
I sincerely hope you all address this issue with the vigilance it deserves. Apologizing to the folks I have turned on to BP over the last couple of years is not a favorably anticipated endeavor.
Other readers have rated this comment below the viewing threshold. Click here to view anyway.
I'd rather have these Beta projections than nothing at all. Projecting someone to get a 3.74 ERA instead of a 3.89 ERA isn't the important part. Projecting a 3.74 vs. someone else with a 4.09 is the important part.
Just so I am clear, is the main issue with the 10 year forecasts or PECOTA as a whole? If the problem is greater than the forecasts what other problems areas are people finding?
How about almost every established star's 90th percentile being batsh*t insane compared to last year's? Heck, even many young solid players have some pretty crazy 90th percentile projections. I'm talking about stuff where PECOTA projected them with 100+ points of extra slugging % this year compared to last year's PECOTA projections. There isn't a single stat from PECOTA that I trust right now on anyone's card. And that stinks.
As was mentioned above, the number of PAs in the depth charts do not match the number of PAs in the PFM spreadsheet. After some cursory sleuthing, it seems that the counting stats on the depth charts are in error.
Using the PFM spreadsheet, if one calculates Prince Fielder's OBP from the projected H, BB, HBP and PA (ignoring sacrifices), one gets 0.409, which is very close to the stated OBP of 0.408. Since the depth charts don't display H, BB or HBP, you have no way of knowing whether they've been appropriately scaled for the 667 PAs listed in the depth charts.
However both the depth charts and PFM spreadsheet project the same number of HRs (41). And this DOES seem like an appropriate number of HRs for the PFM spreadsheet. Computing SLG from the raw data from that sheet, one gets 0.583 which is consistent with the stated 0.584.
So it seems like the counting stats on the depth charts need to be roundly scaled up. (For example, Fielder should be at 0.285/93/43/115 rather than 0.285/89/41/110.) Perhaps I'm reading too much into this, but maybe this is part of the reason why the projected team batting lines still don't exactly jibe with the projected runs scored.
The book seems to match the PECOTA cards. The problem seems to be with the 10 year forecasts more than with the 2010 projections. Of course that doesn't mean the 2010 numbers are correct, but I don't think anyone has really questioned them yet, and obviously people have been pouring over the data. Even if the projections in the book have to be changed, it's worth getting for the player comments and team articles.
118 comments have been left for this post.
They look excellent, gentlemen. Outstanding work.