F1 Rankings
Teammate comparison as the least-bad instrument in motorsport
Comparing two competitors in the same equipment is the closest motorsport gets to a controlled experiment, and its limitations are severe enough to require constant stating.
Top Ten Pitch — Ranked, argued, explained
F1 Rankings
Reliability, era and the arithmetic of a career record
A motorsport career total is shaped by how many events existed, how often machinery survived them and how many competitors were capable of winning.
Latest
Browse everything →
F1 Rankings
Why a points system is a statement about what racing should reward
Allocating points across finishing positions looks like bookkeeping and functions as a set of incentives that shapes how competitors approach a season.
F1 Rankings
Driver and machine: the attribution problem motorsport cannot solve
Every result in a machine sport is produced jointly by a competitor and equipment they did not build, and no available method fully separates the two contributions.
NBA Rankings
Positionless basketball and the collapse of positional ranking
Ranking by position assumed roles were stable categories, and as those categories dissolved the comparison frameworks built on them stopped describing anything real.
NBA Rankings
Team success as a player criterion, and the attribution problem
Championships are used constantly as evidence in individual comparisons, and the reasoning required to get from a team outcome to an individual verdict is rarely supplied.
NBA Rankings
Pace adjustment and why raw totals mislead across eras
The number of opportunities a game provides has changed substantially over time, and comparing accumulated totals without accounting for that compares schedules rather than players.
Top
All top →
Top
Which sports can be ranked objectively, and which never will be
The reliability of a ranking depends on the structure of the competition itself, and some sports simply do not generate the kind of evidence an ordering requires.
Top
Recency bias and the quiet editing of sporting memory
The competitors we rate highly are disproportionately the ones we watched clearly, and that availability effect shapes all-time lists far more than any deliberate argument does.
Top
Peak versus longevity: the choice that decides every greatest-of list
Almost every dispute about all-time standing reduces to whether the question is about the highest level ever reached or the total amount of excellence delivered.
FIFA
All fifa →
FIFA
Draw randomness versus competitive balance in tournament design
A tournament format is a set of decisions about how much chance to allow, and every mechanism that improves fairness removes some of the uncertainty audiences arrive for.
FIFA
Why friendly matches distort a rating system, and how weightings answer it
Matches played without stakes generate results that look like evidence and behave like noise, and every serious rating system has to decide what to do about them.
FIFA
The confederation problem: comparing teams that rarely meet
Rating systems depend on competitors playing each other, and international football is organised in a way that starves those systems of the connecting evidence they need.
Wimbledon
All wimbledon →
Wimbledon
Head-to-head records as evidence, and where the sample runs out
A direct meeting between two competitors feels like the strongest possible evidence, yet the number of meetings is almost always far too small to support the conclusions drawn.
Wimbledon
Points defence and the illusion of decline in a rankings table
A player can hold their level exactly and still fall many places, because the arithmetic of expiring points describes the calendar rather than the competitor.
Wimbledon
Seeding on grass: when a committee overrules the computer
Departing from a published ranking to seed an event looks like interference until the reasoning is examined, and then it looks like a correction for a known measurement error.
ICC
All icc →
ICC
Team ratings versus league tables, and why they disagree
A standings table and a rating system are answering different questions, and the fact that they routinely produce different orders is a feature of both designs.
ICC
Decay functions: how a rating forgets last season
Every rating has to decide how quickly old evidence stops mattering, and that single setting does more to shape the published order than the rest of the formula combined.
ICC
Format fragmentation and why one cricket ranking cannot hold
A sport played in several formats with different constraints produces skills that do not transfer cleanly, and a combined ranking would obscure more than it revealed.
Player Rankings
All player rankings →
Player Rankings
Era adjustment: what it fixes and what it cannot
Comparing competitors from different periods requires translating between conditions that changed in several directions at once, and no adjustment can recover what was never measured.
Player Rankings
Availability as a ranking criterion rather than an excuse
Time on the field is usually treated as a caveat attached to a rating, when a strong case exists for treating it as part of the quality being rated.
Player Rankings
The replacement-level idea, and what it quietly assumes
Measuring a competitor against a freely available alternative is one of the most useful ideas in sports analysis, and its usefulness rests on an assumption that is often false.
Club Rankings
All club rankings →
Club Rankings
Trophy counts versus sustained level as the basis of a club order
A cabinet of honours and a record of consistent performance measure different things, and choosing between them decides almost every argument about which club ranks higher.
Club Rankings
Cross-league comparison and the calibration problem
Placing clubs from different domestic competitions on one scale requires bridging evidence that only continental fixtures provide, and there is never enough of it.
Club Rankings
Elo for clubs: rating a squad that is never the same twice
Pairwise rating systems were designed for individuals whose ability changes slowly, and applying them to clubs requires accepting that the rated entity keeps being replaced.
Stadium Rankings
All stadium rankings →
Stadium Rankings
Heritage against comfort in ranking a venue
Older venues score badly on almost every measurable property and keep appearing near the top of published lists, which reveals what those lists are actually valuing.
Stadium Rankings
Capacity, sightlines and access: three axes that rarely agree
The properties that make a venue good for watching, good for reaching and good for filling pull against each other in the design, and no building optimises all three.
Stadium Rankings
Atmosphere as a measurable quantity, and where measurement stops
Crowd noise can be quantified with instruments, and the numbers those instruments produce describe something narrower than what supporters mean by atmosphere.
Top 10 Goals
All top 10 goals →
Top 10 Goals
Judging technique when the outcome is already known
Every assessment of a memorable moment is made after seeing that it worked, and knowing the result changes how the decision leading to it is evaluated.
Top 10 Goals
Context weighting: the same strike in two different matches
An identical piece of execution can be trivial or decisive depending on when it arrives, and how much a ranking should care about that is the central unresolved question.
Top 10 Goals
Why footage availability decides which goals get remembered
The pool of moments any list can draw from is limited by what was recorded, from which angles and in what quality, and that limitation is not distributed evenly.
Transfer Deals
All transfer deals →
Transfer Deals
Success criteria for a signing, set before the outcome is known
Recruitment decisions are almost always judged against a standard invented afterwards, and writing the standard down in advance changes what the assessment can show.
Transfer Deals
Fee inflation and why cross-era transfer comparison fails
Comparing the scale of moves across decades requires a common unit, and every candidate for that unit changes the answer in a different direction.
Transfer Deals
Value, need and opportunity cost in judging a signing
Assessing a recruitment decision requires three separate questions, and public discussion almost always collapses them into a single argument about whether somebody is good.
NBA Rankings
All nba rankings →NBA Rankings
Positionless basketball and the collapse of positional ranking
Ranking by position assumed roles were stable categories, and as those categories dissolved the comparison frameworks built on them stopped describing anything real.
NBA Rankings
Team success as a player criterion, and the attribution problem
Championships are used constantly as evidence in individual comparisons, and the reasoning required to get from a team outcome to an individual verdict is rarely supplied.
NBA Rankings
Pace adjustment and why raw totals mislead across eras
The number of opportunities a game provides has changed substantially over time, and comparing accumulated totals without accounting for that compares schedules rather than players.
F1 Rankings
All f1 rankings →
F1 Rankings
Reliability, era and the arithmetic of a career record
A motorsport career total is shaped by how many events existed, how often machinery survived them and how many competitors were capable of winning.
F1 Rankings
Teammate comparison as the least-bad instrument in motorsport
Comparing two competitors in the same equipment is the closest motorsport gets to a controlled experiment, and its limitations are severe enough to require constant stating.
F1 Rankings
Why a points system is a statement about what racing should reward
Allocating points across finishing positions looks like bookkeeping and functions as a set of incentives that shapes how competitors approach a season.