15 replies to this topic

[Impeccable]

Does anyone have opinions on Bill James and Sabermetrics books?  Have you read them?  Are they worth the read...etc???

[jsp2014]

I don't think Bill James has written anything recently. he's currently working for the Red Sox so I'd imagine anything he's working on is proprietary. it might be interesting from a historical perspective but most of that stuff is outdated. between fangraphs and blogs, you can learn all you need to know for free.

it's worth noting that the Bill James projections aren't really authored by Bill James. I think he originally helped create the projection system. his name is probably there to sell books at this point in time.

Edited by jsp2014, 26 December 2012 - 05:34 PM.

[SoundMaster]

I'm a huge proponent, and reader, of Ron Shandler's annual Baseball Forecaster (received my '13 copy just a week ago!).        His approach isn't necessarily unique (emphasizing skills and not 'surface stats' such as avg, hr, rbi, ERA, etc.) but the collection of all of this data in one-place is highly useful and easy.

[dzemens]

SoundMaster, on 26 December 2012 - 08:23 PM, said:

I'm a huge proponent, and reader, of Ron Shandler's annual Baseball Forecaster (received my '13 copy just a week ago!). His approach isn't necessarily unique (emphasizing skills and not 'surface stats' such as avg, hr, rbi, ERA, etc.) but the collection of all of this data in one-place is highly useful and easy.

Agreed, love the Baseball Forecaster, wife just got it for me for Christmas (only thing I asked for!). Its an invaluable tool, the tier system that is in the forecaster is pretty spot on as well.

[RotoRaysfan]

The Forecaster is a great read - ESPECIALLY if you don't know much about metrics - because it explains it all there.

Re: James, he's done a lot of great work in the past - as for his forecast system, though - if he loves a player, a MVP-like season is projected, and if he doesn't, scrub-level production.  It has always been a little off in that the reality lies in between, so his projections REALLY have to be taken with a grain of salt (as is the case with most systems, but with James, the difference between tout / down is HUGE).

EDIT: Interesting that James doesn't author the projections anymore.

[SoundMaster]

RotoRaysfan, on 27 December 2012 - 10:11 AM, said:

Re: James, he's done a lot of great work in the past - as for his forecast system, though - if he loves a player, a MVP-like season is projected, and if he doesn't, scrub-level production.  It has always been a little off in that the reality lies in between, so his projections REALLY have to be taken with a grain of salt (as is the case with most systems, but with James, the difference between tout / down is HUGE).

Out of curiosity, how much weight do the folks here place in pre-season projections, if any?         Chasing pre-season projected stats is all but a fool's crusade, it seems.  

While Shandler's forecaster indeed provides projected stats, their method of emphasizing underlying skills has proven for me to be far more effective.

[dzemens]

SoundMaster, on 27 December 2012 - 10:24 AM, said:

RotoRaysfan, on 27 December 2012 - 10:11 AM, said:

Re: James, he's done a lot of great work in the past - as for his forecast system, though - if he loves a player, a MVP-like season is projected, and if he doesn't, scrub-level production.  It has always been a little off in that the reality lies in between, so his projections REALLY have to be taken with a grain of salt (as is the case with most systems, but with James, the difference between tout / down is HUGE).

Out of curiosity, how much weight do the folks here place in pre-season projections, if any? Chasing pre-season projected stats is all but a fool's crusade, it seems.  

While Shandler's forecaster indeed provides projected stats, their method of emphasizing underlying skills has proven for me to be far more effective.

I think people may use it as a general idea, its always nice to see what someone thinks. I know personally I compile multiple projections into a master projection, tweaked for my own input as well. The method of how people get to the stats through underlying skills though is the important part, as anyone can throw numbers out there. But if you can understand the process and methodology behind it, it can become a useful tool.

[RotoRaysfan]

SoundMaster, on 27 December 2012 - 10:24 AM, said:

Out of curiosity, how much weight do the folks here place in pre-season projections, if any? Chasing pre-season projected stats is all but a fool's crusade, it seems.  

While Shandler's forecaster indeed provides projected stats, their method of emphasizing underlying skills has proven for me to be far more effective.

I would echo the sentiment that skills analysis is a more predictable method of future projection than preseason raw data projections, and more reliable.   The reason why tiering is so attractive is that by its nature, it lumps players of similar output ranges together - and so by its nature, accounts for the variability in projections.  

When owners can look at skills progression (which metrics are often better at detecting early changes than scouting is), the player's potential for growth, and on-field scouting (because scouting is still a useful tool, just as metrics are), and combine it with the ability to project for opportunity and PT, that's the multi-modal method that works best.  Projections are just a visual version - the problem is, as you & others mention, chasing those projections, well, that's less likely to succeed than relying on the multi-modal approach.

One thing projections are great for, along with ADP - it gives a general sense of where player values are - now, to be clear, it matters more what your 9/11/15/19 other league mates think.   But if the public is all-in on a player - well, that's useful info to have.  But again, it only takes one other owner to disagree with the public version and go with yours to create a bidding war in auction, or to jump 1-2 rounds earlier than expected as your safe target in a draft.  But projections can give owners a place to start on public perception (which is why ADP is another generally useful tool, with cautions on using it as bible, for the same reasons).

[jb_power]

RotoRaysfan, on 27 December 2012 - 10:48 AM, said:

SoundMaster, on 27 December 2012 - 10:24 AM, said:

Out of curiosity, how much weight do the folks here place in pre-season projections, if any? Chasing pre-season projected stats is all but a fool's crusade, it seems.  

While Shandler's forecaster indeed provides projected stats, their method of emphasizing underlying skills has proven for me to be far more effective.

I would echo the sentiment that skills analysis is a more predictable method of future projection than preseason raw data projections, and more reliable.   The reason why tiering is so attractive is that by its nature, it lumps players of similar output ranges together - and so by its nature, accounts for the variability in projections.  

When owners can look at skills progression (which metrics are often better at detecting early changes than scouting is), the player's potential for growth, and on-field scouting (because scouting is still a useful tool, just as metrics are), and combine it with the ability to project for opportunity and PT, that's the multi-modal method that works best.  Projections are just a visual version - the problem is, as you & others mention, chasing those projections, well, that's less likely to succeed than relying on the multi-modal approach.

One thing projections are great for, along with ADP - it gives a general sense of where player values are - now, to be clear, it matters more what your 9/11/15/19 other league mates think.   But if the public is all-in on a player - well, that's useful info to have.  But again, it only takes one other owner to disagree with the public version and go with yours to create a bidding war in auction, or to jump 1-2 rounds earlier than expected as your safe target in a draft.  But projections can give owners a place to start on public perception (which is why ADP is another generally useful tool, with cautions on using it as bible, for the same reasons).

  This is the exact reason why I pay very little attention to ADP and put more weight on metrics, skills and projections when I compile my tiered rankings.

[rdj3video]

Just my two cents...The Bill James' Historical Baseball Abstract (2001) is a must read as is Whatever Happened To The Hall of Fame - Baseball, Cooperstown, and the Politics of Glory (1995). These are more for historical purposes, but provide some backbone for some of his sabermetric principles. If you care about Win Shares then the book titled Win Shares (2002) is a complex, dorky read about how to figure out Win Shares, but it includes several essays comparing and contrasting players. A lot of people are caught up with WAR, but Win Shares has its place in my view. Bill James' yearly Handbooks have all the stats in them, but the essays and analysis are kept to a minimum in each section of the book. I do know The Hardball Times has put out entertaining annuals the last 7-8 years. They no longer include a ton of charts, graphs, and stats, but have lots of essays on all kinds of things tied to sabermetrics, team reviews, and the business of baseball.

[FouLLine]

jsp2014, on 26 December 2012 - 05:34 PM, said:

I don't think Bill James has written anything recently. he's currently working for the Red Sox so I'd imagine anything he's working on is proprietary.

Makes perfect sense.

[tucker26]

FouLLine, on 29 December 2012 - 06:26 PM, said:

jsp2014, on 26 December 2012 - 05:34 PM, said:

I don't think Bill James has written anything recently. he's currently working for the Red Sox so I'd imagine anything he's working on is proprietary.

Makes perfect sense.

Wait, what? Outrage. Next your gonna tell me Bret Saberhagen isn't doing math anymore either...

[PRoSPx]

tucker26, on 29 December 2012 - 10:29 PM, said:

FouLLine, on 29 December 2012 - 06:26 PM, said:

jsp2014, on 26 December 2012 - 05:34 PM, said:

I don't think Bill James has written anything recently. he's currently working for the Red Sox so I'd imagine anything he's working on is proprietary.

Makes perfect sense.

Wait, what? Outrage. Next your gonna tell me Bret Saberhagen isn't doing math anymore either...

Bill James is certainly doing math still.  He just isn't publishing new material.  Why would The Red Sox pay him to do research for them if he was going to make all his research public.

[Impeccable]

SoundMaster, on 26 December 2012 - 08:23 PM, said:

I'm a huge proponent, and reader, of Ron Shandler's annual Baseball Forecaster (received my '13 copy just a week ago!). His approach isn't necessarily unique (emphasizing skills and not 'surface stats' such as avg, hr, rbi, ERA, etc.) but the collection of all of this data in one-place is highly useful and easy.

Just got this book in the mail from Amazon.  Can't put it down.  Thanks for the recommendation.

[ballfan4141]

ok what did this bill james predict for last year?

trout?

edwin ecarnaction?

alex rios?

[umphrey]

The Bill James projections are pretty terrible.  They are always extremely optimistic.  Sometimes they project just random things, like a career year for a guy in his mid 30s.  I like ZiPs the best, and Steamer is good too.  Marcel is a good one to just evaluate the last 3 years of performance without asking your projection system to predict the future too much.