RRosternomics

TTS — Trade Timing Score

How closely a GM’s trade strategy matched what the roster needed at the time. A 0–100 measure of trade timing and team-need identification.

What it measures

TTS asks whether a GM pursued the right objective at the right time. It does not grade who eventually won the trade. It compares the team’s needs immediately before the deal with the kind of value the deal was designed to acquire.

Every trade can blend four objectives: Immediate production, Championship leverage, Future controlled production, and Flexibility. The team also receives a continuous need profile across those same four dimensions.

How need and strategy are identified

The Need Mix starts with the team's published pre-trade playoff probability. When those odds are unavailable, TTS estimates them from the prior completed roster's WAR using only seasons that had already happened. Payroll relative to the league determines how strongly financial flexibility matters.

The assets establish how much of the trade is aimed at present production, controlled future production, or salary relief. The team's actual competitive opportunity then divides present production between Immediate and Championship. Acquiring a star does not automatically become a championship strategy when the acquiring club has little chance to contend. A controlled star can also count toward Future because his production extends into the next roster foundation, even when prospects were surrendered.

The formula

Let (N_k) be the team’s share of need (k), and (S_k) the trade’s strategy share. Both vectors sum to one. Their total-variation distance is:

\[ D=\frac{1}{2}\sum_k |S_k-N_k| \] \[ \text{trade alignment}=1-D \]

The career or tenure score weights each trade by its materiality and strategic commitment:

\[ \text{TTS}=100\times\frac{\sum_i \text{alignment}_i\times\text{exposure}_i}{\sum_i \text{exposure}_i} \]

Units

Strategic-alignment points on a 0–100 scale. A 70 means the GM’s trade strategies matched roughly 70% of the team’s measured needs after accounting for trades that blend objectives. It is not a probability or WAR total.

How to read it

Worked example — Brian Cashman trades for Juan Soto (2023)

What the Yankees needed: with a 75% pre-trade playoff probability, the model assigned 40.0% Immediate production, 37.9% Championship leverage, 2.7% Future value, and 19.4% Financial flexibility.

What the trade pursued: the package was measured as 54.7% Immediate, 31.4% Championship, and 13.9% Future. Soto's remaining control gives the deal some future-foundation value, but it remains primarily a present-focused move.

1. TTS adds the overlap in all four dimensions: \(40.0+31.4+2.7+0=\mathbf{74.1}\).

2. The Yankees' side of the Soto trade therefore receives a 74.1 TTS.

In plain English: New York needed impact talent for a roster positioned to contend, and Cashman traded for exactly that. The score does not say who won the trade or grade Soto's later performance; it grades whether the objective fit the Yankees at the time.

Validation

TTS repeats across alternating career trades at r = 0.538, corresponding to an estimated full-career reliability of 0.700. The outcome evidence is objective-specific: Immediate strategy predicts same-season playoff success when Immediate is the main need (adjusted r = 0.233). Modern Future strategy has a positive but currently inconclusive relationship with five-year playoff frequency (adjusted r = 0.127); RFA provides the stronger test of whether the future value acquired was effective.

Extending the generic forecast to ten years does not change the conclusion: TTS correlates -0.031 with ten-year playoff frequency, and using average win growth instead produces -0.084. Within modern Future-need seasons, ten-year TTS is mildly positive (0.064) but inconclusive. TTS should therefore be read as a repeatable process diagnostic, not a standalone forecast of franchise results.

Built from trade-side decisions since 1985. Values are credited to the executive and franchise responsible for the trade; short ranges are descriptive rather than full-career reliability estimates.