3. Methodology
• Compute a new statistic “WAR+” to measure value to team
WAR+ = Projected WAR
(Salary/ Marginal Value)
Sum of annual WAR+ values for remaining years under contract
- Measure of added value over duration of contract
4. Aging Curves
• Compute Aging curves to breakdown three main components of WAR
–wRAA
–UZR (rSB and rPP for catchers)
–BsR
• Each aging curve broken down by standard deviation above/below the mean
– Account for deviation in
better/worse subsets
– 18 total aging curves
Highest Standard Deviation Group
3
6. Recalculating WAR
● Aging Curves extrapolated (1) wRAA (2) UZR (3) BsR through
2028
○ Based on 2016 original data
● League Adjustment, Positional Adjustment, Replacement Runs,
Runs Per Win kept constant
○ Not assuming position switch
7. Contract Projections
● Pre-Arbitration
○ Projections purely based on service time
● Arbitration
○ Arbitration model based on arbitration class and projected WAR totals
○ Assembled arbitration database on players from 2012-2016
● Contracts
○ Options assumed to be accepted given value
8. ● Marginal Value: Cost per year adjusted for inflation
○ Cost of win for 2017: $7,000,000
○ Inflation Percentage: 5%
Marginal Value
9. WAR / $
● “WAR+” to measure value to team over duration of contract
WAR+ = Projected WAR
(Salary / Marginal Value)
● Large contracts are more burdensome than lower WAR figures
10. Life of Contract WAR and $
• WAR through 2028
–Mike Trout (3rd)
–Large Contract + High Production
• Salary owed through duration of contract
–Giancarlo Stanton (1st)
–Miguel Cabrera (2nd)
–Joey Votto (3rd)
11. Risk Assessment
• Three Part Factor (Weighted on strength of the model)
–Injury Risk
–Age Risk
–Performance Risk
• Injury Risk
–DL days projection based on lag and two lags of annual DL days
• Age Risk
–Standard deviations of wRAA, UZR, BsR by age
• Performance Risk
–Standard deviations of annual WAR performance
–Rookies with 1 year of experience projected on average Yr 1 to 2 risk
12. Risk in Conjunction with Projections
• Applied Error Bars
–Upper and lower bounds of projection
• Acceptable level of risk varies by team situation
–Lower payroll teams cannot afford as much risk
• Lower bound of risk-associated projection used as final figures
–Worst possible likely outcome assumed