Risk Ratings: Using Risk and Relative Value to Help Your Fantasy Baseball Draft

Will Carroll@injuryexpertSports Injuries Lead WriterFebruary 26, 2014

Feb 23, 2014; Peoria, AZ, USA; Seattle Mariners second baseman Robinson Cano (22) warms up during camp at Peoria Sports Complex. Mandatory Credit: Rick Scuteri-USA TODAY Sports
Rick Scuteri-USA TODAY Sports

So you saw who the riskiest players were yesterday, but the question I always get from fantasy players is how they use the Risk Ratings to adjust their own draft boards. It's a key question and one that can be a real differentiator. It doesn't matter what ratings you use or what projections. The spread isn't significant among them as composite rankings like what Cory Schwartz does at Fantasy 411 shows. 

The difference comes in two ways: the "all things equal" picks and avoiding or at least managing risk. This is where the Risk Ratings come in very handy, both as a general and specific adjustment. At its most basic, the Risk Ratings should simply be a note to be sure that you have reasonable adjustments.

If you think Matt Kemp is going to get 600 plate appearances this year, I have a bridge to sell you. Setting his draft position or auction value based on unrealistic expectations is one of the quickest ways to screw up your fantasy team.

There is a slight difference between adjusting for injury and adjusting for risk. It's subtle, but important. Risk has to be taken as a class, while trying to predict injury is much more difficult. It's relatively easy, but artless, to say that a player who has an elbow injury is likely to have another elbow injury. This is true and proven, but it fails to differentiate the two basic types of injuries (traumatic and chronic) or to associate any context to the injury.

Video Play Button
Videos you might like

Adjusting for risk is a better solution overall, though in any specific case, it can be problematic. Knowing that Matt Kemp—who will be my default player example since he has the highest Risk Rating this season and because he has both a chronic and a traumatic injury—is coming back from a pair of surgeries could predict more injuries, but knowing that he was coming off a shoulder injury last year told you nothing about his ankle. 

To understand how you should have valued Kemp, think about insurance. In the simplest of terms, it's a pooling of risk. If you imagine 100 houses in your neighborhood, all of relatively equal value, quality and ownership, they would all have similar risk. Rates to insure them against fire and theft would be roughly equal. However, this would tell you nothing about which house was most likely to catch fire.

In fact, insurance companies don't care. They care that the rates they charge and the incidence of claims match up, giving them a profit. This is why insurance companies are scared so much by mass catastrophes like tornadoes and hurricanes, which have led to the process of reinsurance. (If you're interested, reinsurance is one of the most fascinating of insurance topics, though most will find it a cure for insomnia.) 

For fantasy baseball, the use of risk should be to mitigate and manage the possible losses. Even though I can't tell you which player is going to hurt their elbow, blow out their knee or any other specific injury, neither can anyone else, including the teams. Loading up with injured players certainly doesn't make you more likely to win, even if only half the players you draft actually get injured. In large numbers, the Risk Ratings will come very close to actual performance. 

For players in the "red" ratings in my proprietary system, they have an injury risk that is above 45 percent. That's very nearly a coin flip. The simplest of probabilities knowledge will remind you that the odds of any particular flip is 50 percent, but that they are not connected. A run of "heads" is improbable, but entirely possible.

A study done by the team that originated the STEAMER projections showed that my Risk Ratings did have a particular value that could adjust projections. I think that those numbers are a bit low, however, and for simplicity, I suggest that red ratings have their projections discounted by 10 percent. Yellow ratings should be held steady, and green ratings should be given a 5 percent premium. It involves a bit of work, but the round numbers should help, and a spreadsheet makes it easier still.

That's the simple part, but I believe that any serious fantasy player has to take another step. For years in fantasy football, the orthodoxy has been to take a running back first and often second. The reasoning, first articulated by ESPN's Matthew Berry, was that the difference between the first and 24th running back in terms of points was vastly different than the gap between the first and 12th quarterback. (Both of these assume standard rules and a 12-team league.) 

That gap in production has always existed in fantasy baseball as well, but it wasn't well understood. Positional scarcity is a term that is tossed around but seldom quantified. Joe Pisapia, the author of the Fantasy Baseball Black Book, has taken it a step further and introduced a concept called Relative Positional Value (or RPV). 

Stick with me through a little bit of technical stuff, and you'll have a powerful tool and strategy to pair with your adjusted projections. Pisapia calculates RPV with a simple series of calculations:

The formula to determine the RPV or the percentage in which a player is better than the fantasy league average is: (Individual Player Point Value − Fantasy League Average of the Position) ÷ Fantasy League Average of the Position = RPV

Pisapia has calculated that the best possible players at each position, such as Paul Goldschmidt (+25), Robinson Cano (+23) and Clayton Kershaw (+25), add a value to your team above and beyond the mere points value. 

Once you have this calculated, a secondary shift has to come in. The drop-off for each position has to be calculated out, and the player at each position with the biggest drop-off to the next available same position player should be selected. 

To give you an example, the difference between the first- and second-best catcher is small, just a +5 (Carlos Santana at +31 and Buster Posey at +26). The difference at first base is more significant at +10 (Paul Goldschmidt at +25, Joey Votto at +15). I shouldn't have to explain why Goldschmidt is better in both RPV and in pure points. 

Of course, you could always just buy Pisapia's book, where he handles much of the hard work for you, but there's a value to understanding the calculations as well. Using RPV along with the Risk Ratings puts you even further ahead of the competition. 

Pisapia explains what RPV does very well in this statement: 

The basic crux of what [RPV does] is relying on reality of actual production more than projections, which can be very misleading if you rely mostly on them. RPV then weighs that player production with the bi-product being an overall picture of how valuable each position as a whole truly is, depending upon the depth and style of your league.

There are a lot of projections out there, from the sublime to the ridiculous. As with most sabermetrics, the steps forward are small, and the differences few. Even using a consensus projection has advantages, though it means you will miss some outliers. As long as you have a valid projection system to base from, using RPV and risk adjustments will make you significantly better than if you happen to have picked a projection system that's down a bit this year against your competition. 

The great part about fantasy sports is that anyone can play and have fun. That said, it takes a bit more work to be good at it. Using the techniques here will make your next team great. 

slash iconYour sports. Delivered.

Enjoy our content? Join our newsletter to get the latest in sports news delivered straight to your inbox!