Risk Management

News, Injuries, Rosters - All baseball all the time!
OaktownSteve
Posts: 318
Joined: Mon Mar 28, 2011 9:55 am

Risk Management

Postby OaktownSteve » Thu Jan 23, 2014 3:03 pm

Started talking about this a little in the comments yesterday. This year so far there seems to be a perception that there is more risk in the draft. I think that from a risk management standpoint you could argue that there is always the exact same amount of total risk in every draft simply because it's a closed system and players who might be viewed as very high risk in a market with many safe options can seem like safer options in a market with few. That's probably splitting hairs though.

Any the who, I started thinking about what we really mean when we talk about risk in FBB. I started to explore the idea here a little: http://razzball.com/forums/viewtopic.php?f=33&t=36213. In each of the pairings I tried to either provide one safe player and one risky one or give two players of roughly equivalent risk. The idea being that who you chose would be a reflection of your own biases. In almost all cases, I think that the risk assigned to each player is a non-quantifiable perception of risk.

I unpacked the whole idea a little further and came up with a list of as many different kinds of risk perception as I could think of that an FBB might perceive. As you go down the list, see which flavors feel more or less relevant to you.

Injury:

Chronic: hamstrings and shoulders, see Matt Kemp
Acute/Sudden: often playing style related, see early Kinsler, Harper
Workload: pitchers mostly, see Verlander
Mechanical: See Chris Sale
Pitch type: See Baumgartner (38% SL)


Regression:
Return to previous, lower performance levels, see Chris Davis. Difficult to predict, see also Bautista

Character:
Off-field: see Puig
On-field: see also Puig

Contract:
Signed: Players who got paid, See Cano

Playing Time:
Platoon/handed splits: See Moss, Scheirholz
Blocked: See Oscar Tavares
Managerial: See Scioscia, Baker

Trade/City Change:
Relievers: Closers on marginal teams who could become set up men via Trade; see Huston Street
Ballpark: Players in premium ballparks who move or could move to less friendly venues, see again Cano

PEDs
Caught: players who could regress off the juice, see Braun, Cruz, also Melky (cautionary)
Not yet caught: see everybody

Weight:
See Sabbathia, Fielder, Panda

Age:
Young /short track record: See Profar, Wacha
Old and creaky: See Yankees

No US track record: See Abreu, Tanaka

Market:
Tout hype: risks over drafting a player you like because the has overly inflated that player...see sleepers so often named as sleepers on the web they no longer sleep

Volatility:
Players who have a good total career baseline but in any given year may perform way above or below their average: see Alex Rios

OaktownSteve
Posts: 318
Joined: Mon Mar 28, 2011 9:55 am

Re: Risk Management

Postby OaktownSteve » Thu Jan 23, 2014 3:10 pm

so the real debate I have with myself is whether or not these perceived risks have any value. Are they predictive in any way or are they simply narrative noise that gets in the way of other types of analysis that filters out these types of bias. The regression risk is probably the only one that can get actual numbers attached.

If the risk perception carries any weight, then I think you can approach your draft by trying to create a balanced risk portfolio and that may have some merit.

If, on the other hand if we overweight our perception of risk when that perception actually has no predictive power then it makes no sense to try to balance the portfolio because it's all just noise and we should go with a projection model and draft accordingly without any factoring for risk.

Beardawg
Sponsor
Sponsor
Posts: 574
Joined: Sun Apr 25, 2010 6:50 pm

Re: Risk Management

Postby Beardawg » Thu Jan 23, 2014 5:20 pm

Ron Schandler / Baseball HQ's Forecaster takes on risk through a reliability grade, looking only at health, experience and performance consistency.

http://www.baseballhq.com/content/ron-s ... aster-2014

Because it is a qualitative assessment, players get a letter grade A-F., along with grades/ratings 1-5 on other key skills. It's a nice little system that gives you a rough measure of risk, but in the context and on a similar scale with a player's other attributes. Their draft philosophy has generally been to spread the risk on offense, and go stars and scrubs with pitching.
10 team, keep 9 at auction $, 6x6 OPS,HLD

C - W. Ramos
1 - M. Cabrera
2 - A. Mondessi
S - T. Story
3 - N. Arenado
CI - L. Voit
MI - M. Machado
O - C. Yelich
O - M. Brantley
O - J. Upton
O - A. Meadows
U - G. Hampson
BN - Franmil Reyes
BN - N. Senzel
BN - J. Bauers

SP - C. Carrasco
SP - M. Mikolas
SP - H. Ryu
SP - S. Matz
SP - M. Harvey
SP - M. Strahm

RP - A. Miller
RP - S. Doolittle
RP - J. Hader
RP - S. Dominguez
RP - R. Presley
RP - C. Devenski
RP - T. Rogers
RP - D. Castillo
RP - D. Betances

OaktownSteve
Posts: 318
Joined: Mon Mar 28, 2011 9:55 am

Re: Risk Management

Postby OaktownSteve » Thu Jan 23, 2014 5:58 pm

the question is does Shandler's risk measure prove to have any predictive power over time? I would guess that you'd have to look at a group of players over/under performance relative to an independent prediction model like Oliver or Steamer and then cross reference that against their preseason risk scores to see if they were predictive of performance. Otherwise it's just another piece of noise that happens to be narrative hiding in a number cloak. Do you know if Shandler produces any metrics for the utility of the measure?

User avatar
Simply Fred
Sponsor
Sponsor
Posts: 402
Joined: Fri Mar 06, 2009 9:15 am

Re: Risk Management

Postby Simply Fred » Thu Jan 23, 2014 6:21 pm

I have a bias that I believe is supported by track record and is predictive...regarding--injuries.

http://razzball.com/forums/viewtopic.php?f=33&t=31503

OaktownSteve
Posts: 318
Joined: Mon Mar 28, 2011 9:55 am

Re: Risk Management

Postby OaktownSteve » Fri Jan 24, 2014 8:00 am

Hey Fred, no offense, but off your list of guys to avoid in 2013 Byrd, Longoria, Ellsbury, Werth, Utley, Ortiz, Cuddyer and Napoli were some of the best values in FBB last year.

User avatar
Simply Fred
Sponsor
Sponsor
Posts: 402
Joined: Fri Mar 06, 2009 9:15 am

Re: Risk Management

Postby Simply Fred » Fri Jan 24, 2014 12:51 pm

risk didn't come close to returning reward for two years. did for the third. i'll stick by the two years. you run with the risk. we'll take stock end of season... :)

Beardawg
Sponsor
Sponsor
Posts: 574
Joined: Sun Apr 25, 2010 6:50 pm

Re: Risk Management

Postby Beardawg » Fri Jan 24, 2014 2:54 pm

OTS, Shandler has done some stats on DL days as a leading indicator, finding that players injured in one year are more likely to be injured in a subsequent year, and that previously injured players tend to spend longer on the DL. So, "A" level players have fewer than 30 DL days over the past 5 years, while "F" grade players have more than 120.

Regarding consistency, again, they've run stats to show that that you need about a three year trend to project into a fourth. Anything less and the predictive value of any consistency of a trend starts falling off.

I don't want to plagerize the Forecaster, so I'll just say that it is well worth the $25 they ask for it.
10 team, keep 9 at auction $, 6x6 OPS,HLD

C - W. Ramos
1 - M. Cabrera
2 - A. Mondessi
S - T. Story
3 - N. Arenado
CI - L. Voit
MI - M. Machado
O - C. Yelich
O - M. Brantley
O - J. Upton
O - A. Meadows
U - G. Hampson
BN - Franmil Reyes
BN - N. Senzel
BN - J. Bauers

SP - C. Carrasco
SP - M. Mikolas
SP - H. Ryu
SP - S. Matz
SP - M. Harvey
SP - M. Strahm

RP - A. Miller
RP - S. Doolittle
RP - J. Hader
RP - S. Dominguez
RP - R. Presley
RP - C. Devenski
RP - T. Rogers
RP - D. Castillo
RP - D. Betances


Return to “Baseball - General Player Discussion”

Who is online

Users browsing this forum: No registered users and 17 guests