Ideas in the queue. Tell me which one you want tackled next.
Does the sim match real-life extremes for contact hitters in extreme parks (e.g. Ashburn: 0 HR at Astrodome, 14 HR at GRF in a coded league)? Validate engine vs. real life and test whether building for your park by hitter type is exploitable.
Vs diminishing returns: This one asks whether the engine is right (and whether you can exploit it). Diminishing returns is a separate question: even if the engine is additive, the value of +1 HR depends on baseline—so valuation and “build for your park” both touch the same theme but from different angles.
Can we validate the Flood study through our analysis process and reproduce its outcomes? The fielding bands and data page uses Flood (and Dooley, willibphx, Nick Flynn) and asks for datapoints to refine the alignment chart.
Given what we know about park effects, there’s a diminishing return on benefit per player. A +1 HR for a 30+ HR hitter doesn’t move the needle much; a +1 HR for a player who normally hits 4 is a big percentage edge. Quantify marginal value of park boost by baseline power.
Separate from park × hitter type: This is about valuation (how much is +X HR worth for this player type?), not about whether the sim over-credits contact types. Same “build for your park” theme, different question.
Do some players have more variance than others? If we can quantify variance (e.g. season-to-season or sim-run-to-sim-run), we could flag high-variance vs low-variance profiles and see if that informs draft or roster strategy.
We have a lot of salary history on players across leagues and seasons. Is there something we can do with it to determine value for “this cycle” analysis—e.g. typical salary paths by profile, market shifts, or over/under pay relative to performance? Open-ended; needs scoping.
Document and build a visual analysis workflow plus a player dashboard along the lines board user knip described: graph all players by adjusted OPS (or weighted adjOPS favouring OBP) vs salary, fit a trend line to get a formula (salary × value + intercept ≈ replacement value), then derive % above average and units above average. Filter by a minimum % above average and OBP floor (e.g. .300) to get a shortlist for team building. For pitchers, use R/9 IP (and in no-DH leagues, factor in P batting). Document the method and expose it as a dashboard so others can use the same visual/value approach.
Baseline comparison: This simpler approach can serve as a baseline when we implement our more complex tool (defensive value + park effects). Compare rankings and dollar values between the two to see what difference the extra complexity actually makes.
Pick one (or suggest your own) and submit. You’ll get a quick confirmation.
Or reply on the Imagine Sports board or at Buy me a coffee.