Crowdsourcing Team Risk
In my Tuesday chat this week, a commenter asked me about the Mariners bullpen, which has already endured some key injuries this spring. I responded as such:
Six of 10 worry points is more than half the worry points. I can’t say for sure even how worrisome that is, because worry points are a thing I made up at 12:27 p.m. EDT on March 22, 2016, but I can say for sure that it’s more worrisome than a team whose bullpen deserves just four out of 10 worry points.
People seemed to like the worry-points concept, and a few commenters expressed a desire for a worry-point article, with a couple even giving me ideas. I didn’t use either of those specific ideas, but the silly worry-point concept did get me thinking about team risk, and how that can influence our perception of different ballclubs before the season begins.
Like, say you’ve got two teams, each projected for 85 wins. But one of those 85-win teams has one super-elite star that comprises a bunch of their value on his own, and therefore that team is just one injury away from losing a big chunk of those projected 85 wins. And not only that, but the supporting cast around that star is littered with players over 30, guys with past injury histories, and a couple guys coming off bad seasons who are being counted on to bounce back. The other projected 85-win team, on the other hand, is pretty clean — lots of guys in their 20s, little injury history, everyone is coming off pretty standard seasons.
On pure talent, both teams are projected to be equal, but that first team has worry points through the roof. It’s easy to see where things could go wrong. The other team feels safe. This could wildly change our (sub?)conscious perceptions of these otherwise equal-looking teams.
I’m kind of thinking about the Dodgers right now, but I’m also kind of not. I’m mostly wondering what the spread is? Every team has areas of concern. Some seemingly more than others, but then at the same time, every team has guys who feel as safe as can be that will inevitably become injured or underperform this year. This game is almost impossible to predict.
This is going to be one of those posts Jeff likes to run with a bunch of polls and then a crowdsourced follow-up, because the opinions of 1,000 of you on something none of us really know about are worth more than the opinions of one of me. Who am I to assign worry points to every team in baseball? The results of this will be interesting now, and they’ll be interesting later. You’ll all put your brains together, and you’ll determine a handful of teams that seem riskier than the rest, and a handful of teams that seem safer than the rest. And at the All-Star break and at the end of the season, it will be interesting to look back and see what happened to those risks. Did the risky teams succumb and/or underperform? Did the safe teams stay clean and/or overperform? Will there even be a noticeable spread in the results of this poll? Is any one team actually riskier than another? These are all questions I can’t answer on my own, but would like to see answered. I’m thankful not only to write for an audience, but a smart and engaging audience.
It’s time to dole out some worry points. Except, for the sake of clarity, I’ll be referring to them as risk points. I encourage you to leave comments below for use in the follow-up post as to why you voted the way you did for any particular team about which you felt strongly.
Here’s a non-exhaustive list of things to consider:
- This isn’t about how good or bad a team is. The Braves shouldn’t automatically be more risky than the Cubs just because they’re a worse baseball team. Try and think of each team’s amount of risk in a vacuum, relative to its own general skill level.
- How much of this team’s value is tied up in one or two players? Is this team an injury away from being decimated?
- How many injury risks are present on this team? How much is being counted on from the injury-risk guys? Are key players already banged up?
- What’s the depth look like on this particular team? Are there adequate replacements to fill in for the potential loss of some of the riskier players?
- What’s the average age of the players on this team? What’s the average age of the key players?
- Is this team counting on bouncebacks from key players, who were most recently disappointing?
- Is this team counting on key contributions from young players, who have proven very little?
- Does this team have a history of injury prevention? Injury susceptibility?
- Style of play?
- Any other things you can think of that might make a team risky! Risk is subjective, and so are its causes!
Angels
Astros
Athletics
Blue Jays
Braves
Brewers
Cardinals
Cubs
Diamondbacks
Dodgers
Giants
Indians
Mariners
Marlins
Mets
Nationals
Orioles
Padres
Phillies
Pirates
Rangers
Rays
Red Sox
Reds
Rockies
Royals
Tigers
Twins
White Sox
Yankees
August used to cover the Indians for MLB and ohio.com, but now he's here and thinks writing these in the third person is weird. So you can reach me on Twitter @AugustFG_ or e-mail at august.fagerstrom@fangraphs.com.
Interesting way to think about it…
I voted low risk on Cubs, only because I think Theo did a great job getting pitching depth and players who can play multiple positions….Cubs have Baez, Soler, Adam Warren, Cahill as backups on the roster who can fill-in for injuries
I voted low risk on Phillies, because they will be bad, they know it, we know it, they’re counting on it actually, and there is just little “risk” of them not being bad….
As a Cubs fan, I am devoting all 10 of my worry points to Jake Arietta bicycle accidents.
I went two points on the Phillies only because Herrera hasn’t played in over a week with his middle finger issue, and their offense is basically Franco and Herrera. Not much risk in pitching with depth at starting pitcher though no true ace and no one you’d really notice if they were injured in the pen.
Since you want us to disregard projected wins, or mean wins, what about tying this data into the spread of the projected wins, or the standard deviation for wins? Can you analyze team projections to see if this data is able to predict the teams that have higher standard deviations for their projected wins?
Yeah it’s a lot easier to rank the bad teams as less risky since they don’t have very many good players.
Also, I would love to see team worry points broken down into Bullpen, Rotation, OF Defense, IF Defense, OF offense and IF offense, or some categories like how your initial worry point model allocated 6/10 to the SEA bullpen.
But that is too many polls.
“But that is too many polls.”
BLASPHEMY!!!
I’m of two minds about the Dodgers. I think they are certainly more susceptible to injury than the average team, but that is largely by design. What they have also done is acquire of lot of options, knowing that they will likely get a good handful of injuries. They are obviously one of the deeper teams, but even that is not endless. With Anderson, Bolsinger, and Ryu unavailable, and Urias and De Leon pretty much ruled out to start the season, we’re still seeing some weakness at the back of the rotation.
I might settle on medium risk for them, except that, if something were to happen to Kershaw, that would probably automatically knock them out as division favorites. Are the Giants equally susceptible in that way? Losing Posey would probably be just as disastrous for them, but just by not being a pitcher, Posey is less likely to get a season-ending injury.
Catcher vulnerability has decreased with the new rules but catchers are still one foul-tip away from the DL.
The giants are nowhere near as risky as the angels but they’re not really that much safer than the Dodgers.
Cueto and Shark are also somewhat risky for them.
Let’s have a look at the Rockies.
How much of this team’s value is tied up in one or two players? Is this team an injury away from being decimated? While I think WAR is dinging Blackmon and Gonzalez too much, Arenado’s on a different tier even if you give those guys 1 1/2 more WAR each. He’s also projected for a win+ over every pitcher.
How many injury risks are present on this team? How much is being counted on from the injury-risk guys? Are key players already banged up? Ignoring Jose Reyes, Carlos Gonzalez’ injury history isn’t great. Jordan Lyles and Tyler Anderson keep having weird non-baseball related injuries. The team is counting on Jon Gray and Jeff Hoffman to develop, and those guys are pitchers, so automatically their injury risk is pretty high. Future seasons really depend on these two guys in particular, so them getting hurt would not only blow up any chance for improvement in 2016, but multiple years after 2016 though that is irrelevant to the question asked. What’s going on with Tyler Matzek is a bummer.
What’s the depth look like on this particular team? The fourth outfielder is quite possibly Ryan Raburn. Their shorstops are either going to court or have a combined 58 MLB plate appearances. Eddie Butler totally forgot how to strike guys out sometime between 2013 and 2014 and Chris Rusin is a viable rotation option if someone goes down.
Is this team counting on bouncebacks from key players, who were most recently disappointing? /checks Cargo’s first half stats again, vomits, Boone Logan is the personification of disappointment (and decisions by bad managers), again, whatever happened to Eddie Butler that doesn’t look fixed.
Is this team counting on key contributions from young players, who have proven very little? Gray, Story/Cristhian Adames, Hoffman (the team weirdly thinks it can compete, so he’ll probably be up in September if not earlier), Butler and Arenado are both only 25, Tom Murphy. A good number, actually. I’m a believer in Chad Bettis but he has only one good season.
Any other things you can think of that might make a team risky! They play at altitude, and that’s probably not great for pitchers who are already fragile bunnies on the whole to start with, though I have no evidence to back this up.
“This isn’t about how good or bad a team is.” Guys, it’s even in bold and everything. The Rockies can have a season torpedoed way more easily than your standard 69 win team.
…but if the Rockies season is “torpedoed” and they win, say, 62 games rather than 69, does it matter? If you underperform expectations by seven wins at the 69-win point on the curve, it doesn’t matter much, you weren’t going to the postseason anyway. If you’re expected to win 90 and win 83, that’s a *HUGE* difference.
“This isn’t about how good or bad a team is. The Braves shouldn’t automatically be more risky than the Cubs just because they’re a worse baseball team. Try and think of each team’s amount of risk in a vacuum, relative to its own general skill level”
Isn’t this a contradictory statement? “Risk” implies that a team has something to lose if players underperform or succumb to injury. The Braves playoff odds already stand at 0%, so based on that, I would classify them as one of the lowest risk teams in MLB this season.
I think the point is risk of hitting their baseline wins, not risk of missing the playoffs.
I gave the Cubs and White Sox both a 3. The community has given them a 2 and 4, respectively.
The Cubs are the better team, obviously, but I don’t think they are a safer bet to hit 96 wins than the White Sox are to hit 80. Maybe that’s not the right way to go about it, but I think that is how people are voting; the six teams voted less risky than the Cubs are the six bottom-feeders (i.e. the teams below the Royals in the projected standings).
I understand the optimism. I love the Cubs rotation depth- Travis Wood, Adam Warren, and Trevor Cahill would all probably start for the White Sox. The Sox need to lure Buehrle out of “retirement” to just get to five starters. And if Sale gets hurt, they are in big trouble.
But most of the Cubs optimism doesn’t come from depth, it’s from youth. (As an aside, the Astros and Twins are next in low-risk teams.) In my mind, second-year players are the riskiest players there are (may be a tad of a hyperbole, but one season doesn’t establish a player). I think Kris Bryant is pretty good, but he is projected to be the fifth best batter next year according to WAR. I can easily see Kyle Schwarber going the way of Wil Myers. Addison Russell could easily have a Brett Lawrie-type of career. Javier Baez and Jorge Soler will actually be third year players, and they are considered the less established ones. Plus, all five of these guys strike out like 30% of the time, and Baez was the only one to get significant action in AAA before being called up.
Then you have the Sox. I don’t see any of their top hitters as that big of risks. So that leaves you with the unheralded starters of Jimmy Rollins, Melky Cabrera, and Avisail Garcia. If J-Roll doesn’t bounce-back, then you go with Tyler Saladino, or Tim Anderson, or even Carlos Sanchez; not a big loss. There aren’t any good in-house replacements for Cabrera and Garcia, but they aren’t even projected to total 1 WAR as a pair, so if they don’t pan out, the White Sox shouldn’t- in theory- drop more than one win off their pace.
Actually the cubs hype is built arobithepth. Baez and soler are benc. Pieces that have solid floors while also having huge upside. Same for the pitching side, Warren actually projects as an above average starter. Then you have wood who if starting projects as average as 7th starter. The floor is so high is why they are low risk. Comparing schwarber to Myers is odd as schwarber walks a ton and clearly has more power and no injury history unlike Myers. Russell as a Lawrie comp is crazy, Lawrie hit a ton to have his good first year. Russell is a great as that they don’t need to hit. The only risk the cubs truly have is in arrietta because of innings total last year, zobrist because of age might not be able to handle second, and Montero because schwarber nor Contreras are ready to catch full time meaning Ross or fedorwizc get too many starts. Btw Frazier isa Big risk, he was terrible in the second half. Austin Jackson is on a 3 yr decline, they have terrible catcher d, Last that pitching is super thin, both starters and in the pen.
“Actually the cubs hype is built arobithepth. Baez and soler are benc.”
Did you run this through the Serbian-French-Vietnamese language shredder?
Risk is probability times the consequences. Perhaps an easier way to get at this is to attack each variable independently and then combine. JZ posts projected days missed in his MASH at some point, maybe use that and apply it to projected WAR, and add them for each team to come up with a probability of failure or injury (you could even monte carlo it). Then, look at the consequences, which here is usually games lost below projected wins with consideration for how close teams are to making the playoffs. The you can plot the results and see how teams stack up. Asking people to calculate this all at once and combine it in their heads is very difficult, as you stated.
Do you know of any good sources writing about quantifying risk? Baseball or economic.
The methodology I am familiar with is specific to engineering but the general concepts can be applied to almost anything. These are just two of many sources, I am sure that other fields have their own references:
Von Thun, J.L. (1985), “Application of Statistical Data from Dam Failures and Accidents to Risk-Based Decision Analysis on Existing Dams,” Bureau of Reclamation, Denver, CO.
Whitman, R.V. (1984). “Evaluating calculated risk in geotechnical engineering.” J. Geotechnical Engineering, ASCE, Reston, Virginia, Vol. 110 (2): 145-188.
I like that approach.
The main problem I see with this current drill is that it isn’t granular enough. There’s simply too many factors going into risk for a simple 1-5 ranking. For starters, it isn’t enough to look at the number of “risky” players, the position they play and the role they fill in the team’s makeup matters. Losing a journeyman starting catcher can be far deadlier to a team than losing a journeyman corner or a third or fourth pitcher.
FWIW, I would prefer to see something that starts with the positional rankings, applies an injury/underperformance risk factor per position to represent depth, and then a weight factor per position to reflect the importance of that position to *that* team. For pitching, I’d like to see the five starters rated individually, plus the bullpen as a unit.
Sabermetrics has advanced enough that risk assessment doesn’t have to be gut feelings or general perception. Crowd sourcing is fun, yes. But how about some hard numbers to play with?
Great, now risk doesn’t even look like a real word anymore.
Interestingly, at the 85-win mark, more risk points are actually a good thing. You’d rather be the 85-win team that could win 95 and make the playoffs or win 75 and bomb than be the 85-win team that’s confined to the narrow 81-89 band and much less likely to advance.
For all teams who aren’t frontrunners, risk points = opportunity.
What you’re talking about is volatility, but I think the idea of this is to look at only the team’s potential downside
Sounds like there is a suggestion that a stars and scrubs model is inherently more risky than a balanced team. While it is true that if the Angels lose Mike Trout they are toast. However, losses of virtually any other player will not be very problematic. Trading for, or swapping in minor league talent isn’t much of a challenge when you only need to replace a 0.5 to 2.0 WAR player. The Angels could afford to lose multiple players and assuming that none are named Trout, the impact will be negligible.