Do you know where I can find it in Excel friendly format? This doesn't convert to Excel and I don't feel like cutting and pasting team by team hitter and pitcher grids.

you can't import that as a CSV?

you should be able to in excel. I'm at home and using open office in which of course I've had trouble with csv's...

edit, yeah just save it as a csv (comma separated values) and import into excel. I can email to you if you want

Wow, never knew what CSV files were. That was easy.

I'll finish uploading at intermission of the Hawks game. Will probably adjust my numbers to even weights on PECTOA/CHONE/MARCEL/James.

this will certainly happen.

however, it's also possible that on average you end up with worse projections than if you just used pecota

in a league like mine, the market almost acts as a quasi hedge against these outliers. ie pecota was way higher on Milledge than any other system last year...pfm had him at like $30 or something silly. but since I'm the only one (or 2) who uses pfm in my league, he would never have been bid up that high. It's good to be aware of these discrepancies...but I think averaging out different systems is a bit dangerous

I'm thinking about going back to the straight tier system and not worrying about exact $values. I mean, what's the difference between a guy who projects at $28.52 and a guy that's at $24.67? That's well within the margin of error for any given system, so why sweat it? It's even more foolish to sweat over guys who are $7.41 and $3.50, since you should be dealing in straight upside there anyway.

If I went back to the old school system, this is what it would look like for hitters;

Group A+, Superduperstar, $50+
Group A, Superstar, $35-$49
Group B+, Star, $25-$34
Group B, Above Average, $18-$24
Group C+, Average, $11-$17
Group C, Below Average, $6-$10
Group D, Fringe/Upside, $1-$5
Group F, Fringe, $1
Group P, Prospect

For 2009 catcher would look like;
Group A+-None
Group A-None
Group B+-Mauer, Martin, McCann
Group B-Soto
Group C+-Iannetta, Napoli, Doumit
Group C-Posada, Martinez, Wieters
Group D-Salty, Clement, Montero, Flores
Group F-Shoppach, Snyder, B.Molina, Teagarden, Towles
Group P-Santana, Salome, FLowers, Arencibia, Marson, Anderson

This way I'm focused on what I should really be focused on, and that's getting the cheapest player in a given tier. Of course I build the tiers based on the various projections and last season's numbers. I wouldn't put Joey Votto in with superstars Teixeira and Miggy even though PECOTA says. Knowing PECOTA likes him that much, I might raise him to group B instead of C+ (he was exactly $10 last year).

Alright, I trust Nate Silver's math, but couldn't find the article where he wrote about doing this..except for his 538 site, but that's not the same kind of combining projections we're talking about here. I can only say I'd be very surprised if the best can get better by, essentially, diluting it with inferior product. It just, fundamentally, does not make any sense whatsoever. and I'm talking about long-term and predictable results, not a few odd years having the benefit of hindsight.
It also seems that if combining with another projection system actually does improve PECOTA, they will have reverse-engineered to identify the specific/approximate cause and eliminated that potential for improvement.

peru, I like the idea of tiers and will probably use something like that...how'd it go for you when you used it before, and why did you stop using it?

I went looking for the article and couldn't find it either. but I literally re-read it this week, so I don't think I am misremembering

although it was based on 1 year of results.

figuring out whose projections are the best is nearly as difficult as creating projections in the first place, and I'm not sure it's in a lot of peoples best interests to go through th work to do it.

I stopped using it because in the old days, I'd look at a player and put him in a tier. No basis in fact or stats, just personal opinion. I semi-stopped once I started doing my own projections and using PECOTA because I was leaning on the cold hard numbers, although I've still visualized in terms of a tier system.

I think the utility of PECOTA and learning how to do my own spreadsheet has been huge. Now I know exactly what one type of performance is worth and how the positions relate to each other. I'll still do a spreadsheet, and likely I'll still have the projections in front of me. The only difference will be in how I assign the actual $values and the presentation.

Don't worry. I remember too. It was definitely reverse reverse-engineered.

Here are the articles your looking for.

For hitters http://www.baseballprospectus.com/unfiltered/?p=564 which says to weigh your projections with 5 parts PECOTA, 4 parts ZIPS and 2 parts Marcel

And for pitchers http://www.baseballprospectus.com/unfiltered/?p=569 which says to weigh your projections with 4 parts PECOTA and 3 parts CHONE.

I play in a H2H keeper league with weird categories. It has 9 hitters categories but only 7 pitcher categories. The pitcher categories are innings pitched, wins, shutouts, saves, strikeouts, era, and whip. There are 22 roster spots with 2 SP slots, 2 RP slots and 3 P slots.

In this format is it worth it to draft any relievers on my team? Its roster space that could be going towards hitters since there are more hitter categories or could be going towards starters to make up ground in innings pitched, wins, shutouts, and strikeouts. The fact that there are 2 slots for RP seems to be a little problem with this strat, but not a big deal since there are daily roster changes.

Yes I know these rules are dumb and hopefully I can change some before the season starts but if not I would like to know which direction to go.

Keepers were announced in my league today (13 team, 5 keeper, 5x5). I have the 4th pick in our snake draft. (I kept Wright, BJ Upton, Morneau, Hart, and Furcal)

Best hitters out there are:

Adrian Gonzalez
Chipper Jones
Carlos Pena
Alex Rios
Matt Kemp
V. Martinez
Dye

Pitchers:
Lee
Liriano
Shields
Kazmir
King Felix
E. Santana

Edit: Price and Joba are also available if that changes anything.

Couple questions:

The league is structured so that you get 5 keeper every year, no restrictions. Weiters is unowned. Is it worth drafting him 4th overall, over the players listed here? I'm thinking no, but I doubt he'd be there coming back at 23, and I don't want to regret not drafting him for years to come.

Who would you take first? (I'm looking to take a hitter most likely, I'm assuming gonzalez will be gone)

Thanks for finding these..even a few years data would be unreliable, and in these articles he's analyzing one year, and with the benefit of hindsight. You can't accurately predict which combo will improve upon PECOTA for 2009, and subjectively combining projections like this is terrible from both a time-management and probable results standpoint. I just don't get it.

I'm in a league like this, too, but mine's 10x7 w/ QS and BB taking the place of your CGSO and IP, roto, and has 25 roster spots and an IP max of 1300, so RP still has significant value. In your league, depending on the number of teams and number of offensive starters, I wouldn't touch RP, and would target, instead, likely SP w/ RP eligibility. Punting a category is not typically a good idea, but your league's not typical.
Another decent strategy, especially if there are more than 12 teams in your league, would be to go after a couple dominant RP and moderate, undervalued SP and stream more starts.

I love leagues like this (except the tie-machine of CGSO, wtf?), the more unusual the settings and cats, the better for those who know how it affects player values.

How screwed am I in my main keeper league? I have to keep 8 players. Stats are R, HR, RBI, SB, K, AVG & OPS / W, L, SV, K, ERA, WHIP, K:BB

Garrett Atkins (Col - 1B,3B)
Jason Bay (Bos - OF)
Carlos Beltrán (NYM - OF)
Carlos Delgado (NYM - 1B)
Travis Hafner (Cle - 1B)
Kelly Johnson (Atl - 2B)
Paul Konerko (CWS - 1B)
Jed Lowrie (Bos - 3B,SS)
Víctor Martínez (Cle - C,1B)
Hideki Matsui (NYY - OF)
Gary Sheffield (Det - OF)
Miguel Tejada (Hou - SS)
Joey Votto (Cin - 1B,OF)
Vernon Wells (Tor - OF)

Dave Bush (Mil - SP)
Cole Hamels (Phi - SP)
Rich Harden (ChC - SP)
Félix Hernández (Sea - SP)
Trevor Hoffman (Mil - RP)
Jim Johnson (Bal - RP)
Pedro Martínez (NYM - SP)
Mike Mussina (NYY - SP)
Brad Penny (Bos - SP)
Fernando Rodney (Det - RP)
Jeff Samardzija (ChC - RP)
Carlos Zambrano (ChC - SP)

this team would have crushed like 3 years ago

I agree with fifield. Try and dominate the counting stats. Once you figure out what you need in mid July, you can adjust your roster accordingly.

1. Kemp (not close)
2. Santana
3. Wieters
4. Price

I think you can justify Wieters #2 if you have confidence in your ability to field a rotation later on. In that case you could move Chipper to #4.


Bay, Beltran, Johnson and Votto look alright as hitters. Hamels, Harden, and Hernandez are the foundation of a pretty sweet rotation. That last spot kinda sucks, but I think you just take Zambrano and are done drafting starting pitchers. Maybe you can trade Big-Z for an above average closer to give you a little more balance. Then you target a lot of high ceiling offense in the draft. Those keepers aren't optimal, but they are by no means terrible.

Depends, how many teams in your league? I'd probably at least keep Hamels, Harden, Felix, and Zambrano and move one of them for value elsewhere (prob MI if possible) if you can.

What do you guys think about a separate "2+2 community fantasy baseball projections" thread? There has been a lot of discussion in this thread, but it seems like it would be cool to have one uber-projection spreadsheet for discussion/debate, that everyone could download and manipulate for their leagues. Would depend of course on whether people are willing to share their projections.

well, the makers of pecota may not love it, and a lot of us use it either exclusively or partially

marcel/chone/a few others are freely available though

Come on, Nate is 2+2 alum, he won't mind!

That is a good point though, I guess I just take it for granted that everyone subscribes to BP. What could still possibly work is a thread for specific projections, where maybe each day someone throws out a name (Delmon Young?), and people post their gut projection with rationale, which can be based on existing forecasting, an aggregate, etc.

12 Teams 25 players per roster. Offense tends be valued much more highly then pitching. Hafner, Atkins and Vmart likely have much more value trading wise than Felix or Zambrano in this league.

The problem is those guys suck more, so if you get stuck, it hurts. I factor in potnetial trade value when valuing a player, but much like when I call out guys at the end of the draft, I don't keep anybody I wouldn't want to get stuck with.

How does hafner have any trade value whatsoever?

Can we not taint this thread too with that debate? I've been talking with a few people in my league and they don't love my team nearly as much as I do. Most of them overvalue starting pitching and I have none.

11 team, Auction, Roto, 7 X 7 (K, OPS, L & HLD). My pre-draft lineup:
C-Brian McCann
1B-Lance Berkman
2B-Kelly Johnson
SS-Open
3B-Chris Davis
OF-Ryan Braun
OF-Jacoby Ellsbury
OF-Chris B. Young (to be upgraded)
UTIL-Alex Gordon (to be upgraded)

SP-Aaron Harrang
SP-Johnny Cueto
SP-Jeremy Bonderman
SP-Dustin McGowan
SP-Ian Snell
RP-Joakim Soria
RP-Huston Street

I also have David Price and Carlos Carrasco who I can bring up, but are eligible to stay on my minor league roster all year and save 1 year of service time.

Salary wise I'm in good shape. I can very realistically add S.Drew, Vlad (I need another low K guy to make up for some of my hackers) & Bradley on offense. 4-5 SP like Randy Johnson, Vazquez, the other Chris Young, Nolasco, etc. A couple good relievers like Balfour, Putz & Devine and start the year with arguably the best team. Am I way off?

Somebody is sure to say I need a #1 starter. The top pitchers available who I expect to go for ridic amounts are Hamels, Halladay, Oswalt, E.Santana, Lackey, Lowe, Lee, Dempster, Volquez.

This team looks light on superstar talent overall. No problemo with the pitching. Those guys are collectively pretty bad, but there is plenty of upside and you've already got a couple of closers. Of the pitchers listed, I'd say Santana is the best target. FIf you miss on him, you might want to jump on Halladay just to stabalize everything. My guess is, in an 11 team league, you are going to be in the upper half, but not the best.

thoughts on BJ Upton? he's another guy who projection systems might have trouble with b/c they dont know he played with a busted shoulder last year.

I think a 20/30 year should be right in his wheelhouse, anyways