David Lucey

redwallanalytics.comA Data-Driven Survey of MLB Franchise Management – Redwall Analytics

Published a data-driven survey of MLB franchise mgmt — 30 years of data, 8 analytical acts, and a full scorecard. Powered by {lahmanTools} + DuckDB + GitHub Copilot CLI. Thread 👇 redwallanalytics.com/posts/2026-04-19-a-data-driven-survey-of-mlb-franchise-management/ #BaseballAnalytics #rstats

14 Sept

4" of rain this morning and our storm drain on Roosevelt still surcharging this evening under heavy pressure. This spot already suffered a sinkhole collapse from sub-base washout last year. We need DPW line scoping and a real drainage upgrade. 🕳️⚠️ #OldGreenwich #GreenwichCT #Infrastructure #CTNews

redwallanalytics.comA Data-Driven Survey of MLB Franchise Management – Redwall Analytics

Act 7: Making the playoffs is not enough. Teams go deep when the WAR they banked from April through September actually shows up in October. Health and usable depth are the final edge. #MLB #Playoffs #BaseballAnalytics

Act 8: The complete franchise scorecard - percentile ranks across pipeline, trades, retention, dead money, injury luck, and playoff conversion. No franchise is clean across every dimension. github.com/luceydav/mlb-franchise-analysis | mlbfranchiseanalysis.netlify.app #rstats #DataVisualization

redwallanalytics.comA Data-Driven Survey of MLB Franchise Management – Redwall Analytics

Act 4: Dead money (salary paid to zero-WAR players) is a budget elasticity problem. Large markets absorb it. Small markets get crippled. One bad long-term deal can swallow an entire year of flexibility. #Sabermetrics #MLB #PayrollAnalysis

Act 6: Payroll explains ~60% of long-run playoff achievement (r = 0.77). Money matters — but it's incomplete. The gap between payroll and results is where front-office skill (or failure) lives. #MLB #BaseballData #Sabermetrics

redwallanalytics.comA Data-Driven Survey of MLB Franchise Management – Redwall Analytics

Act 1: Pipeline edge has two parts — rookie quality AND roster share. Most teams get one. Very few do both. Cardinals, Astros, Braves in top-right. Yankees surprisingly thin. #MLB #FrontOffice #BaseballAnalytics

Act 2: Trades are where clubs extend the pipeline or cut it off at the knees. Net losers rarely recover without elite draft luck. #MLB #BaseballAnalytics #Sabermetrics

redwallanalytics.comA Data-Driven Survey of MLB Franchise Management – Redwall Analytics

github.comGitHub - luceydav/lahmanTools: Tools for setting up a Lahman baseball data duckdb and interacting with mcpTools for setting up a Lahman baseball data duckdb and interacting with mcp - luceydav/lahmanTools

The analysis runs on {lahmanTools} — my R package built to extend the Lahman baseball DB with WAR, modern salary data, and a DuckDB backend. GitHub Copilot CLI did the heavy lifting on scaffolding. 📦 github.com/luceydav/lahmanTools #rstats #DuckDB #OpenSource

strongtowns.orgHousing-Ready CitiesThese six policies will help you make your town housing-ready.

21 Apr 2025

These are the kinds of things I think CT’s housing reform should start with over Work Live Ride and Fair Share ideas coming out of Hartford. www.strongtowns.org/housingready

16 Jan 2025

CT is discussing adding $545m to $2.36b of state education aid, half to shore up our 5 largest cities. I'm for it, but with reforms like greater competition via charters. Data shows significantly improved quality of education for Newark's families after reforms. public.tableau.com/app/profile/...

nytimes.comIs Micromanaging Teachers a Recipe for School Success? (Gift Article)A new superintendent is trying to improve Houston’s public schools through strictly structured teaching. The district says it’s working. Many educators and parents hate it.

I would also say to up the aid to any struggling urban district who would allow a Houston-like intervention. www.nytimes.com/2025/02/03/u...

bit.lyAmateur Urbanist Critique of Work-Live-Ride using 360k Fairfield County Parcels - Redwall Analytics

Finally wrote up why #connecticut #wlr doesn't hit spot for me, though recognize the urgency to get more #housing. I gathered Fairfield County 360 parcels with State's new GIS data, and showed why "90% of parcels are 1+ acre" talking points are misleading regarding MNR stations. bit.ly/42HLLa2

And here are counts of 1 acre plus lots moving again moving progressively from stations.

I tried to show why WLR looked so difficult from a data and topology perspective from my Town of Greenwich. I hope this Legislative round can involve more compromise on both sides. I would love to see my town become Transit hub.

bit.lyAmateur Urbanist Critique of Work-Live-Ride using 360k Fairfield County Parcels - Redwall Analytics

Here are average residential lot sizes and percentage residential stepping out progressively from 26 stations.

16 Dec 2024

It turns out that slicing a docx based on pattern match is a deceptively challenging problem not easily solved by an LLM, definitely not SO. I would gladly trade my working R #officer solution for anyone’s Python version or even sympathetic words. #rstats #python #python-docx

stackoverflow.comHow to subset text from a word docx AFTER a matching phraseI would like to subset text from an original word docx ("original.docx") into a new word docx ("desired.docx"), AFTER the match of the phrase "Drop Text Before Here", ...

21 Dec 2024

Nothing against Python here, this was not an easy problem and needed help to solve in R, but it was possible. Don’t think it can be solved with python-docx without much more advanced parsing of XML if at all. stackoverflow.com/questions/71...

3 Jan 2025

To finalize, solution is to run your R function from Python with rpy2. The word document object is zipped xml, so murder to copy subset of text, tables, images and formats cleanly without breaking links. R’s officer package “cursor”functions abstracts xml navigation for people like me.

16 Dec 2024

@hrbrmstr.mastodon.social.ap.brid.gy

21 Dec 2024

Update of progress, this is possible with Python’s win32com library, essentially opening every docx & saving after pattern, but slow for my 1000 slices/wk and breaks for small number with images. Wondering if I can port some R functions into my new Python workspace in case any thoughts out there!

David Lucey reposted

hrbrmstr 🇺🇦 🇬🇱 🇨🇦🇧🇪 4

10 Dec 2024

#RStats {dbi.table}: Database Queries Using 'data.table' Syntax Query database tables over a 'DBI' connection using 'data.table' syntax. Attach database schemas to the search path. Automatically merge using FK constraints. Supports DuckDB! Looks interesting! […]

mastodon.socialOriginal post on mastodon.social

28 Nov 2024

Homebrew upgraded to 3.13 along with all my packages,many relying on python-packaging 3.12, which got unlinked and now can’t be symlinked. This is probably bad, but had pyenv 3.12 virtualenv set to global and would like to stay with 3.12. Any advice/resources on how to untangle this mess welcome!

16 Nov 2024

Updated one of my older side #rstats #shiny projects tracking property sales in #ct for new data from Socrata. Over 25 years, price/assessment for lowest ⅓ of sales values is often lower, suggesting that lower may be persistently over-assessed vs higher priced properties in many towns.

luceyda.shinyapps.ioConnecticut Property Selling Prices vs Assessment Values over Four Revaluation Cycles

16 Nov 2024

CT's assessments are not alone having these patterns, notably Cook County, IL. App has 800k sales coded "arms length" out of 1.1m transactions. I have posted previously, but haven't had real feedback about why this occurs, so curious what others think?https://luceyda.shinyapps.io/ct_real_assess/

16 Nov 2024

For example, the lowest ⅓ of sales prices in Hartford shown in purple. luceyda.shinyapps.io/ct_real_asse...

2024 NYS state assessment data from May has been released today, here are ELA results for the NYC larger charter networks, DOE and all other charters ("charters")..

and for context, the number of 3-8 grader taking the ELA tests by subgroup at each network.

I'm biased working at CCS with four schools (about to be six) in the So Brx these last 3y, but many things are said about charters which seem off base to me. I read one particularly misinformed discussion on Reddit only today. Posting just in case this year's repeat data is of interest to some!

2024 NYS state assessment data from May has been released today, here are ELA results for the NYC larger charter networks, DOE and all other charters ("charters")..