keyboard_arrow_uptop

A popular question I get around this time of year is along these lines:

The PFM ranks Steven Souza much higher than your bid limits/draft rankings do. I don’t understand. Why is it so different? What’s the deal? Huh, huh, huh, Huh, Huh, HUH?!?!?!?!?!?”

I have spent a lot of time—both in emails and in articles specifically devoted to the subject—explaining to our valued readers why the PFM is useful but why I won’t go into an auction strictly using PFM values as a tool. It occurred to me recently that while my reasons might be sound, it violates one of the simplest rules of writing: show, don’t tell. So instead of going through another long explanation of why I don’t use the PFM, I will use the PFM as a tool to construct a hypothetical Tout Wars NL team to see if this is practical or not.

The rules of this exercise are simple. The prices are from Baseball Prospectus’ PFM using a 12-team, NL-only, 5×5 league with OBP as of Monday, March 23rd. The inflation settings have been turned off and the SGP has been turned on. I will replace my results from the NL Tout Wars auction with the results of the PFM. I will populate my team using the Tout Wars prices plus one dollar, unless I purchased the player at Tout Wars, in which case I will use my price. The only other rule I have inserted into this exercise is that I will not purchase more than two closers at $10 or more. Even if I were using the PFM, common sense would tell me to stop bidding in this instance.

Scenario 1: Build around an ace pitcher

Pos

Player

Salary

PFM

C

Travis D'Arnaud

14

$16.63

C

Jose Lobaton

1

$1.63

1B

Scott Van Slyke

7

$10.44

2B

Javier Baez

11

$14.66

SS

Freddy Galvis

1

$7.27

3B

Chris Johnson

8

$12.27

CO

Tommy Medica

1

$3.04

MI

Eugenio Suarez

1

$1.63

OF

Billy Hamilton

26

$31.30

OF

Ben Revere

18

$20.66

OF

Yasmany Tomas

12

$18.18

OF

Wil Myers

17

$21.26

UT

Jon Jay

9

$12.11

SW

Khris Davis

16

$20.39

P

Clayton Kershaw

37

$39.18

P

Craig Kimbrel

22

$27.56

P

Adam Wainwright

19

$21.43

P

Matt Cain

9

$14.39

P

Jon Papelbon

14

$16.77

P

Michael Fiers

11

$14.52

P

Sergio Romo

4

$9.36

P

Travis Wood

1

$7.88

P

Tim Lincecum

1

$6.82

Total

260

$349.38

Players in bold are players I actually purchased in Tout Wars NL.

Just like I did at Tout Wars, I grabbed Kershaw for $37 with my first buy of the day. In this case, though, the PFM didn’t have me push for Paul Goldschmidt and Devon Mesoraco early, but had me grab Billy Hamilton and Craig Kimbrel instead. Despite no other players over $20, this model still pushed me into dollar derby fairly early. There is no doubt that the pitching for this team is going to dominate, but spending $118 on pitching will do that. The thing I find myself wondering is if this team can compete on offense, particularly with $44 out of $142 spent on two guys whose earnings are dominated by stolen bases.

Scenario 2: Balance, balance, balance

Pos

Player

Salary

PFM

C

Travis D'Arnaud

14

$16.63

C

Christian Bethancourt

3

$8.04

1B

Scott Van Slyke

7

$10.44

2B

Joe Panik

6

$8.68

SS

Javier Baez

11

$14.66

3B

Chris Johnson

8

$12.27

CO

Jose Uribe

8

$11.35

MI

Zack Cozart

4

$8.72

OF

Billy Hamilton

26

$31.30

OF

Ben Revere

18

$20.66

OF

Yasmany Tomas

12

$18.18

OF

Wil Myers

17

$21.26

UT

Jon Jay

9

$12.11

SW

Khris Davis

16

$20.39

P

Craig Kimbrel

22

$27.56

P

Adam Wainwright

19

$21.43

P

Matt Cain

9

$14.39

P

Jon Papelbon

14

$16.77

P

Michael Fiers

11

$14.52

P

Sergio Romo

4

$9.36

P

Jake Peavy

8

$10.07

P

Hyun-Jin Ryu

11

$14.32

P

LaTroy Hawkins

3

$7.47

Total

260

$350.58

In Scenario 2, I do what fantasy owners often do in a situation with fast bidding when a high-priced player is nominated early: chicken out. This is a bit of a cheat, but in an earlier version of the NL-only PFM, Kershaw was a shade below $39 ($38.74 or something like that) and I am only taking players who are a bargain of $2 or more. So in this scenario, the PFM is pushing me even further toward balance, as only two players crack the $20 barrier and a group of players ranging in salary from $3-11 replace Kershaw.

I manage to spend less on pitching in Scenario 2 but still manage to spend $101 on my staff. However, this is mitigated by the fact that I purchased 13 everyday players on offense and Scott Van Slyke. Oddly enough, I’m more nervous about the pitching with this squad than I am about the offense. The rotation seems like boom or bust to me; if Wainwright, Cain, Fiers, and Ryu are all healthy (or in Ryu’s case, relatively healthy) this staff could be ridiculously good, but with Ryu already out, I can’t afford any slippage from any of these pitchers.

It is worth noting that whether I construct a Stars and Scrubs team or a balanced team that the earnings for both teams are nearly the same. There is often a misguided belief that Stars and Scrubs is bad because it locks you out of better bargains later. While this can happen, the reality is that a) there is still value to be had with some of the $1 players at the bottom of the player pool and b) you cost yourself by passing up on bargains at the top end of the pool.

Obviously, the earning potential of a team you put together using your system is going to be tremendous. But both of these teams appear to be categorically imbalanced, with too much pitching and not enough non-speed offense. Can they win categorically?

Table 3: PFM Tout Wars NL Team Point Totals versus Gianella Point Totals

Team

HR

RBI

SB

OBP

Runs

W

SV

K

ERA

WHIP

Points

Finish

PFM1

11

8

12

1

10

12

12

12

12

12

102

1st

PFM2

12

12

12

1

12

8

12

9

12

12

102

1st

MG

12

10

8

2

10

12

8.5

12

12

12

98.5

1st

Somehow, both of these PFM teams manage to project well on offense despite a high percentage of dollars being pushed toward pitching. This is particularly true for the PFM2 team, which loses some pitching points in wins and strikeouts, but still finishes first in saves, ERA, and WHIP. The PFM1 team is pretty strong on the offensive projection relatively speaking, given that it spends only $142 on offense.

I couldn’t help including myself in this exercise. The MG row is how the PFM projects my team to do. Without using the PFM as a tool during my auction, I still managed to buy a 98.5 point team (out of a possible 120 points), at least according to the PFM. In other words, despite the fact that I wasn’t working off of the PFM, I still came within 3.5 points of two optimal PFM scenarios, at least in terms of point totals.

It’s easy to pat myself on the back and take a victory lap about how great I am compared to the PFM, but this hasn’t always been the case for me in my Tout Wars auctions. In 2014, my auction was decidedly mediocre, at least by the PFM’s standards.

Table 4: PFM Projected Standings, Tout Wars NL 2014

Team

R

HR

RBI

SB

OBP

W

SV

ERA

WHIP

SO

Points

Actual

Diff

Carty

12

12

12

3

9

12

2

5

6

12

85

75

-10

Trachtman

10

11

11

8

11

3.5

3

8

10

1.5

77

86

9

Zola

8

7.5

9

10

2

6.5

12

6

5

9

75

55

-20

Cockcroft

5

7.5

7

1

10

6.5

11

9

9

6

72

84.5

12.5

Gardner

7

9

8

6

7

9.5

7

7

4

7

71.5

85

13.5

Gianella

11

10

10

12

4

1.5

9

2

1

4

64.5

53

-11.5

Walton

9

4

5

9

8

9.5

4

3

7

5

63.5

47.5

-16

McCaffrey

3

1

1

7

1

11

10

10

8

10

62

57.5

-4.5

Hertz

4

6

6

2

3

5

8

12

12

3

61

79

18

Melnick

1

3

3

5

6

8

1

11

11

11

60

57.5

-2.5

Wilderman

6

5

4

11

5

3.5

5

4

3

8

54.5

50

-4.5

Kreutzer

2

2

2

4

12

1.5

6

1

2

1.5

34

50

16

Earlier this winter, I looked at how the PFM did with individual predictions versus the expert market and came to the conclusion that it was close to a coin flip. Here, the PFM is “graded” in regard to how it predicted the final finish based on the auction rosters versus how the auction rosters actually looked.

In one sense, the PFM seems to be all over the place. It only comes close on three teams—McCaffrey, Melnick, and Wilderman’s—and doesn’t predict the auction winner correctly. Derek Carty has the best projected team coming out of the auction by the PFM’s projections and comes in fifth.

However, there isn’t anything too crazy here, and the baselines are fairly reliable in some cases. Kreutzer’s team was in trouble coming out of the auction and Trachtman, Cockcroft, and Gardner were all solid contenders throughout the entire season. In my case, the PFM only “messed up” because it couldn’t predict that Jim Henderson and Jose Veras would lose their ninth-inning gigs five minutes after the season started. The PFM predicted what my cockeyed optimism couldn’t coming into 2014: that while I had a solid offensive core, I was going to have a lot of trouble on the pitching side based upon my stupid buys. And it turns out that—like an action hero in a B-level Western—I had more trouble than I could handle. Despite some decent in-season pickups, I dug such a deep hole for myself coming out of the auction that I wasn’t able to recover.

This all is interesting to look at, but the danger of the PFM (or any other projection system) is that you can use it to delude yourself into believing that you have purchased the best team in the history of fantasy baseball and become too passive once the season starts. A number of the projection systems out there are saying that I had a good auction in Tout Wars this year. ESPN’s system agrees with the PFM that I had the best auction. Steamer has me “finishing” in second while Baseball HQ’s projections put me in third, only three points behind the eventual “winner.”

But at the risk of stating the obvious, none of this matters much. The projection system tells me that I build a solid foundation for my team. How this plays out during the season is what matters, particularly given that the projections are anything but foolproof.

You need to be logged in to comment. Login or Subscribe
swarmee
3/30
I thought Tristan Cockroft won in 2014; you have him second in points.
swarmee
3/30
I mean third.
MikeGianella
3/30
These are the "draft standings": where teams would have finished with their auctioned teams and no moves
playertbnl
3/30
I made thie mistake of relying on PAM (for the first time... without really understanding it) in a 15 team mixed league draft last week. But I got locked into the dollar derby way too early amd missed out on a LOT of my pitching targets. Needless to say, it didn't go nearly as well for me as it did for Mike.
ravenight
3/31
I don't see a clear argument here for why its bad to rely on pure PFM (provided you do the normal adjustments for injury and what you know about your league mates). The only actual downside you list is about getting cocky once it tells you your team is amazing. So, what are the actual problems with PFM that you are trying to correct by making your own limits? Do your limits start from the PFM and then adjust in specific ways you find helpful? Are those ways actually helpful or do you just distrust the machines?
mikeduin
3/31
To echo ravenight's comment, I agree -- I appreciated this piece but didn't feel like it answered the Souza question at the top, for example.