How sports TV rights work, and who actually gets paid

A league announces that it has sold its television rights. The figure is enormous, the word "record" appears in the first sentence, and the story is gone by the following morning. What almost never gets explained is what was actually sold: which matches, in which countries, on which devices, for how long, and what the buyer is now allowed to stop everybody else doing. That gap matters, because how sports TV rights work is the largest single force acting on every competition covered on this site. Wage bills, kick-off times, expansion into new markets, the shape of the calendar, the existence of entire leagues: all of it sits downstream of a contract negotiated in a room no supporter will ever see.

The contract is not complicated in principle. A competition controls the right to make and distribute moving pictures of its matches. It licenses slices of that right to buyers for a fixed period. Everything that is actually interesting lies in how the slices are cut, how the buyers are made to compete against each other, and what happens to the money once it arrives.

A rights package is not "the football"

Nobody buys the football. They buy a package, and a package is a precisely drafted bundle defined along several dimensions at once. Change any one of them and you have a different product at a different price.

Which matches. Rarely a fixed list. Usually a number of picks per round, with an order of selection. The holder of the first package chooses first in each round, the second package chooses from what is left, and so on down. Pick order is frequently worth more than pick count, because the gap between choosing first and choosing fourth is the gap between the fixture everybody wants and the one nobody would have scheduled deliberately.

Which windows. Slots, by day and time. A Sunday afternoon slot and a Monday night slot are separate products with separate audiences. A league will protect each window from its own other packages, so that the buyer receives genuine scarcity rather than a match competing with two others from the same competition.

Which territory. Rights are sold country by country, or in regional blocks, and the boundaries are legal rather than geographic. Territory definitions have to cope with overseas territories, shared-language markets and, increasingly, the awkward fact that an internet stream does not respect a border without deliberate engineering.

Which platform. Satellite, cable, terrestrial, IPTV, mobile, in-flight, hotel, and commercial premises such as pubs and bars. Commercial premises are almost always a separate licence at a separate price, which is why a pub subscription costs a multiple of a domestic one.

Which language. Commentary in one language is a different product from commentary in another, and in multilingual markets the same match can be licensed twice over.

Which type of coverage. Live is only the headline. Near-live, delayed, full-match replays, extended highlights, short-form clips, social clips with a length cap and a delay before posting, archive footage, still images, radio and audio commentary, in-stadium screens: each can be sold separately. Clip rights in particular have become a contested category, because they are how a competition reaches anybody who does not subscribe to anything.

For how long. Terms usually run three to five years. The length is itself a negotiated variable, and it is one of the few places where the two sides have genuinely opposed interests: a rights holder that expects prices to rise wants a short term, and a buyer that expects the same thing wants a long one.

Underneath all of this sits a question that decides who bears the largest fixed cost. Who produces the pictures? In most major competitions the league or a nominated host broadcaster produces a single world feed, controls camera positions, graphics and replays, and licenses that feed to everybody. Buyers add their own commentary, studio and branding on top. In other competitions the buyer produces its own coverage, which is far more expensive and gives it far more control. A tender that shifts production from the buyer to the league changes the value of every package in it, and the headline number will not tell you which way.

The last thing to understand about a package is that it is defined as much by what it prevents as by what it permits. Exclusivity is the product. A buyer is not really paying for the ability to show a match. It is paying for the certainty that nobody else may show it, in that territory, on that platform, in that window, for the length of the term.

Why the same sport is sold in pieces

If exclusivity is what buyers want, the obvious move for a seller is to offer total exclusivity to one buyer and collect the premium. Competitions tried exactly that. Two forces stopped them.

The first is competition law. A single buyer holding every live match of the dominant competition in a territory forecloses the whole pay-television market: no rival can build a subscription business without the content, so no rival enters, so the incumbent's price to consumers stops being disciplined by anything. European competition authorities examined joint selling arrangements repeatedly and reached broadly the same conclusion each time. Joint selling could continue, but only on conditions. The rights had to be split into several packages. The packages had to be genuinely attractive rather than one good lot and several token ones. The term had to be limited so the market was retested regularly. Unsold rights could not simply be withheld from the market. And in some cases no single buyer was permitted to hold every package.

That last condition, the no-single-buyer rule, is the sharpest instrument in the set, and it is worth being precise about how it works. It does not stop the strongest bidder winning the biggest package. It stops the strongest bidder winning all of them, which means at least one meaningful package goes to somebody else at a price the strongest bidder was willing to beat. The seller loses the monopoly premium on that lot. Consumers gain a second service worth subscribing to. Whether that trade is worth making has been argued about for twenty years and will be argued about for twenty more.

The second force is arithmetic with nothing to do with regulators. Splitting a competition into lots raises the total price under most conditions, because it lets bidders who could never afford everything bid seriously for something. A broadcaster with a modest budget cannot compete for the whole of a major league. It can compete very hard for one package of Saturday matches. Every additional credible bidder in a lot raises the price of that lot, and the sum of several contested lots regularly beats one uncontested block.

There is also a risk argument that sellers rarely make out loud. A competition with one broadcast partner has one counterparty. If that partner has a bad year, loses a corporate parent, or simply decides that sport no longer fits its strategy, the competition's entire revenue is on the table at once. Three partners on staggered terms is a portfolio rather than a bet.

How sports TV rights work at auction: inside the tender

The sale itself is a formal procurement, and the formality is the point. A tender that looks improvised invites a legal challenge from the losers and a competition complaint from anybody who was kept out.

It begins long before any bidding. The rights holder decides the packaging: how many lots, what is in each, how the picks interleave, how long the term runs, what production arrangement applies. This is the most consequential decision in the entire process and it happens in private, usually with advisers who do this for several sports and know what each configuration is likely to fetch. Get the packaging wrong and no auction design will rescue it.

An invitation to tender then goes out. Interested parties are normally pre-qualified, because a rights holder does not want to open its commercial data to a party with no capacity to pay, and financial standing is tested at this stage rather than after somebody has won. Qualified bidders receive an information memorandum and access to a data room: audience and subscriber history where it exists, production specifications, the draft licence agreement, the fixture calendar, and the rules of the process itself.

Those rules matter more than outsiders assume. They set the deadline, the format of a valid bid, whether bids may be made conditional on winning another lot, whether a bidder may submit alternative bids for different combinations of packages, how many lots any one party may hold, and what the rights holder may do if the bids disappoint. Almost every tender reserves the right to reject all bids, to negotiate with any bidder, and to re-run the process. That reservation is not decoration. It is the seller's only defence against a soft round, and every bidder knows it is there.

1. 1The packaging is decidedThe rights holder fixes how many packages exist, what is in each, how picks are ordered, how long the term runs and who produces the pictures. Everything downstream is set here.
2. 2The invitation to tender goes outPotential buyers are pre-qualified on financial standing and given the process rules, the draft licence and the data room. Nobody bids on what they have not been allowed to examine.
3. 3Bidders build a value, not a priceEach buyer models what the package does to its own business: subscribers gained, subscribers retained, advertising sold, production cost carried. That number is the ceiling it will bid to.
4. 4Sealed bids land on the deadlineEvery bidder submits blind. Nobody sees a rival's number, which is exactly why an incumbent has to bid what the package is worth rather than one increment above the visible field.
5. 5The reserve is testedThe rights holder compares the bids against an undisclosed reserve price and against the rules on how many packages one buyer may hold. A lot below reserve can be withdrawn and re-offered later.
6. 6Contracts and guarantees are signedThe licence fixes the term, the payment schedule, the production obligations, the anti-piracy commitments and the consequences of failure on either side. Payment is staged across the term, not handed over up front.
7. 7The central pot is netted downCompetition costs come out first: production of the host feed, the competition's own operating costs, whatever it owes upwards, and any contractually retained fund. What remains is the distributable pool.
8. 8Clubs are paid on the formulaThe pool is divided by whatever rule the members previously agreed: equal shares, merit payments, facility fees, and any parachute or solidarity commitment. The formula, not the auction, decides who gets rich.

The generic shape of a competition rights sale. Individual competitions vary the order of steps four to six, and some run more than one bidding round.

The sealed bid, and the bluff nobody gets to call

Most major sports rights are sold by sealed bid rather than open auction, and the choice is deliberate.

In an open ascending auction, bidders learn from one another. The incumbent watches the challenger drop out and stops one increment above it, paying not what the package is worth but what the second-best bidder could afford. That is excellent for buyers and poor for sellers whenever the field is thin, which in sports rights it usually is. The number of organisations capable of writing the cheque for a major package in any single territory can often be counted on one hand.

A sealed bid removes the information. Each bidder submits one number, blind, on a deadline, and it either wins or it does not. The incumbent is no longer bidding against a visible field. It is bidding against the possibility of a rival it cannot see, which pushes it towards its own true valuation rather than towards its challenger's budget.

This produces a specific and slightly cruel dynamic. The incumbent has more to lose than anybody else, because it has already built a subscriber base on the content and would have to explain to those subscribers why the thing they pay for has gone. Losing a package it currently holds is not a neutral outcome, it is a business event with a churn number attached. So the incumbent bids a premium reflecting the cost of losing rather than the value of winning, and everybody who designs these tenders knows it.

It also explains why rights holders work so hard to get new entrants into the room. A bidder with no realistic chance of winning is still valuable to the seller, because its presence changes what the incumbent believes it needs to bid. Competitions court technology companies, telecoms operators and international streamers for months before a tender opens, for precisely this reason. Whether those parties ever win is close to secondary.

The counterweight is the winner's curse. In a sealed bid for an asset of uncertain value, the winner is by definition the bidder who was most optimistic, and the most optimistic bidder is disproportionately likely to have been wrong. Sports rights buyers have run into this repeatedly: a package won at a price that assumed subscriber growth which did not arrive, then three or four years of contracted payments against disappointing returns, then a much more disciplined bid at renewal. The cycle is visible across several sports and several countries, and it is one of the main reasons rights values do not rise in a straight line.

Collective selling, individual selling, and why the same sport answers differently

There are only two ways to sell a competition's matches, and sports have split on the question in ways that look inconsistent until you see what is driving them.

Under collective selling, the league sells on behalf of all its members, signs one set of contracts, and distributes the proceeds by a formula the members agreed in advance. Under individual selling, each club sells its own home matches and keeps whatever it earns.

Collective selling is, on its face, a group of competitors agreeing not to compete on price. That is close to the textbook definition of a cartel, and it survives only where the law makes room for it. In the United States the accommodation is statutory: a 1961 act of Congress gave the major professional leagues a narrow antitrust exemption to pool the sale of their telecast rights, and that exemption is the legal foundation of the American model. In Europe the accommodation has been regulatory rather than statutory. Competition authorities accepted joint selling on the conditions described earlier, reasoning that a league product genuinely is a joint product and that a flatter distribution supports the competitive balance the product depends on.

The argument for collective selling is that a match is not made by one club. Nobody watches a team play by itself. The competition, its calendar and its jeopardy are what make the fixture valuable, and they are produced jointly by everybody in it. A flatter distribution keeps the bottom half of the table solvent, which keeps the fixtures worth watching, which keeps the rights worth buying. This is the same logic running through American league design generally, where a hard revenue-linked ceiling on spending exists precisely so that no club can convert a revenue advantage into a permanent sporting one, as the NFL's cap mechanics show in detail.

The argument against is that collective selling transfers money from the clubs that generate the attention to the clubs that do not, and that this is a tax on success wearing the costume of solidarity. Where the largest clubs have had the political power to force the question, they have usually won it, and leagues that sold individually for long stretches tended to end up with a small number of very rich clubs and a long tail that could not compete on the pitch or in the market. Several of those leagues have since moved back towards collective arrangements, having watched their own competitive balance decay in public.

There is a third pattern that neither label fits. Some competitions are owned outright by a company or a board rather than by their member clubs, so the question of collective selling never arises. The owner sells everything and pays the clubs whatever the participation agreement says. That structure is common in newer competitions, particularly across the global network of franchise T20 leagues, and it makes the broadcast contract and the league's business plan effectively the same document. It also removes any pretence that the clubs have a vote on how the money is split.

The distribution model a competition picks is not a detail bolted on afterwards. It shapes what kind of competition it becomes, which is the same argument running underneath the older question of open pyramids against closed franchise systems.

Where the money actually goes once the auction is over

The auction produces a number. Almost nobody receives that number.

The first deduction is the cost of producing the coverage, wherever the competition rather than the buyer is the producer. Host broadcast production of a full season is a substantial industrial operation: outside broadcast trucks, camera crews, technical staff, connectivity from every venue, graphics, replay systems, archive. It comes out of the top.

Then the competition's own costs: officials, technology, the disciplinary machinery, marketing, staff. Then whatever it owes upwards to a national governing body or an international federation. Then any contractually retained fund, which is usually where solidarity payments to the wider game and parachute payments to relegated clubs are held.

What is left is the distributable pool, and the formula for splitting it is where a competition reveals what it actually believes.

Four mechanisms do nearly all the work.

Equal shares. Every club receives the same amount regardless of anything that happens. This is the floor, it is predictable years in advance, and that predictability is why it functions as collateral: a club can borrow against a contracted equal share in a way it cannot borrow against prize money it might not win. In competitions where the equal share is large, a promoted club arrives with a genuine budget rather than an invitation to be relegated.

Merit payments. A ladder tied to final position. It rewards finishing high, and because finishing high correlates with having spent more, it compounds: this year's merit payment funds next year's squad. Every competition using a merit ladder is choosing, deliberately, to accelerate the separation between its top and its bottom. The only real question is by how much.

Facility fees. A payment for each live televised appearance. This one is more interesting than it looks, because it is the mechanism that buys clubs' consent to be messed about. A club asked to kick off at midday on a Sunday for the convenience of a foreign market has a concrete reason to agree, and the reason is the facility fee. It also, quietly, rewards being watchable rather than being good, which is not the same thing, and it penalises clubs broadcasters never select whatever their league position.

Parachute payments. Money paid to relegated clubs, declining over two or three years, so that relegation is not an immediate insolvency event. The intention is sound. Without them, clubs facing the drop would either refuse to invest at all or collapse when it happened, and neither is good for the competition they are leaving. The side effect is well documented and rarely denied: a recently relegated club drawing a parachute payment is far richer than the clubs it has joined, which distorts the division below and turns the top of that division into a queue of parachute recipients. Every competition that pays them is trading fairness in the lower division for stability in its own.

Sitting alongside all four is solidarity: payments down the pyramid to clubs that are not in the competition at all, and to the grassroots. Solidarity money is often presented as generosity. It is better understood as the price of political consent, paid by a competition that needs the rest of its sport to keep endorsing an arrangement in which it takes most of the money.

A worked example of the split, with entirely invented numbers

The arithmetic is easier to see than to describe. Everything in this section is an invented example. The league does not exist, the pool is a round number chosen because it divides cleanly, and none of it should be quoted as a real figure anywhere.

Take a competition of twenty clubs with a domestic pool of £1,000m for one season, after production and central costs. Its members have previously agreed the following split.

50%21%20%9%

• Equal shares, 20 clubs500m
• Merit ladder, 20 steps210m
• Facility fees, 200 live picks200m
• Solidarity and parachutes90m

A fictional twenty-club league with a round pool. Not a real competition and not a real figure. The proportions are chosen to make the arithmetic legible.

Show the numbers

Invented worked example: how a £1,000m pool is carved up

Item	Value
Equal shares, 20 clubs	500m
Merit ladder, 20 steps	210m
Facility fees, 200 live picks	200m
Solidarity and parachutes	90m

The equal share is £500m divided twenty ways, so £25m each, paid whether a club wins the title or finishes bottom.

The merit ladder is £210m across twenty positions, awarded in even steps: the champion takes twenty units, the runner-up nineteen, and so on down to a single unit for last place. The units total 210, so each unit is worth £1m. The champion collects £20m of merit money and the bottom club collects £1m.

The facility fee pot is £200m spread across 200 live domestic selections at £1m per appearance. Those selections are not distributed evenly, because broadcasters pick the matches they believe people will watch.

Now take three clubs from that invented season.

• Equal share
• Merit payment
• Facility fees

Champion, 28 live picks25m20m28m

Tenth place, 10 picks25m11m10m

Bottom club, 4 picks25m1m4m

Fictional league, fictional finishing positions, fictional selection counts. Equal share £25m each, merit at £1m per ladder step, facility fees at £1m per live pick.

Show the numbers

Invented worked example: what three clubs collect from the same pool

Item	Equal share	Merit payment	Facility fees
Champion, 28 live picks	25m	20m	28m
Tenth place, 10 picks	25m	11m	10m
Bottom club, 4 picks	25m	1m	4m

The champion collects £73m. The tenth-placed club collects £46m. The bottom club collects £30m. The ratio between the top and the bottom of this invented league is a little over two and a half to one, which is far flatter than the ratio between their wage bills would be, and considerably steeper than an equal split would produce. The same £910m of distributed money spread evenly would hand every club £45.5m.

• 73mChampion's total
• 46mTenth place
• 30mBottom club
• 45.5mIf split entirely equally

All four figures are outputs of the fictional example above, stated in £m. No real competition is being described.

Relegation is the same arithmetic run backwards. Suppose the invented league pays a relegated club half of one equal share in its first season down, £12.5m, and a third of one in its second, roughly £8.3m, both drawn from the retained £90m. That club has arrived in the division below with more guaranteed broadcast income than any established club there earns, and it holds that advantage for two full years. Nobody designed that outcome deliberately. It falls out of a rule written to solve a different problem.

Three things are visible here that get lost in the coverage of real deals. The formula, not the auction, determines who gets rich. A club can raise its income by finishing higher, by being selected more often, or by both, and those are genuinely different behaviours to build a strategy around. And the equal share, the least glamorous line in the table, is the one that makes the competition financeable, because it is the only part of the number a club can promise to a lender. The specific case most readers will have seen argued about, the Premier League's own three-way split, is a variant of exactly this structure with different proportions.

Why domestic and international rights pulled apart

For most of the history of televised sport, international rights were an afterthought. They were sold in bulk to an agency that resold them territory by territory and returned a modest share. Domestic rights were the business. That relationship has changed sharply, and for several reasons at once.

Distribution stopped costing anything at the margin. Adding a territory to a satellite footprint used to require capacity, encryption and a local partner. Adding a territory to an internet service requires a rights clearance and a payment processor. The economics of selling into a small market improved enormously the moment the delivery cost approached zero.

Time zones turned from a problem into a product. A competition whose matches fall in the middle of the night at home may fall in prime time somewhere with a far larger population, and competitions have restructured their calendars accordingly. Kick-off times, tournament windows and the scheduling of finals now reflect audiences on other continents. This is not a conspiracy theory. It is what the facility fee exists to compensate.

Migration built audiences that no domestic market could supply. A sport with a large expatriate population has a paying audience in a dozen countries where it has no clubs and no history.

And domestic markets matured while international ones did not. In a country where the competition has been on television for thirty years, the number of households willing to pay for it is largely known, and growth has to come from price rather than volume. In a country where it is new, growth can come from both.

The strategic consequence is that competitions have become far more careful about how international rights are packaged. Selling them as one global block to a single streaming buyer is administratively simple and produces a clean headline, but it forfeits price discovery in every individual market and hands one counterparty control of the competition's entire overseas presence. Selling territory by territory means dozens of separate negotiations on staggered terms, which is slow and expensive and leaves the competition holding a portfolio rather than a bet. Most large competitions now do some of each.

There is also a distribution wrinkle that matters more than it should. International money is frequently split more equally than domestic money, sometimes entirely equally, because it was small when the formula was written and nobody bothered to fight over it. As overseas income has grown, that historical accident has become one of the largest equalising forces in some competitions, and the largest clubs have noticed. Arguments about how international money is divided are now among the most bitter in club sport, and they are arguments about a rule that was set casually decades ago by people who assumed it would never matter.

Newer competitions get to write the rule with full knowledge of what it does. The rapid growth of the women's game is partly a story about rights that were previously bundled in as a free extra being unbundled and sold on their own terms, a shift covered in more detail in the piece on how women's cricket built its commercial base.

Listed events: the matches a competition is not allowed to sell freely

In the United Kingdom, some sporting events cannot simply be sold to the highest bidder. The regime is statutory, it has been in place in its modern form since the 1990s, and it works by restricting exclusivity rather than by setting prices.

The government designates a list of events of national significance, split into two groups. Group A events must be available live to a qualifying free-to-air service. Group B events may be shown live on pay television, provided adequate secondary coverage, meaning highlights or delayed transmission, is made available to a qualifying free-to-air broadcaster.

The word "qualifying" carries the mechanism. A qualifying service is one that is genuinely free at the point of use and reaches a high proportion of the population. The list of channels meeting that test is short. That is the entire design: it is not enough for an event to appear somewhere for free, it has to appear on a service that almost everybody can already receive.

The regime does not force anybody to buy anything, and it does not set a price. It restricts what a rights holder may sell, by requiring that the rights be offered to qualifying services on fair and reasonable terms, with the regulator able to withhold consent for arrangements falling outside the rules. If no free-to-air broadcaster wants the event at a reasonable price, the event still may not be sold exclusively to a pay operator.

The financial effect is straightforward, and it is the reason the argument never ends. Exclusivity is the product. Listing an event removes exclusivity. Removing exclusivity removes the premium. A listed event is worth materially less to its rights holder than the identical unlisted event would be, and the difference is a transfer from the sport's governing body to the viewing public. Governing bodies argue that the transfer costs them money they would otherwise spend developing the sport at grassroots level. Broadcasters that qualify argue these are national occasions rather than commercial assets. Both positions are honestly held and neither can be settled by evidence, which is why the composition of the list is fought over every time it is reviewed.

Anti-siphoning in Australia works on the buyer, not the seller

Australia reached a similar destination by a different route, and the difference in mechanism is worth understanding because it produces different failures.

The Australian anti-siphoning regime places its restriction on the acquirer rather than on the seller. Listed events cannot be acquired by a subscription broadcaster unless a free-to-air broadcaster has first had the opportunity to acquire them. The obligation sits in the pay operator's licence conditions. No free-to-air broadcaster is required to buy anything, and the sport is not required to sell at any particular price.

Two features follow from that design. Events come off the list after a defined period before they take place, so that rights nobody free-to-air actually wanted do not go unshown. And, far more significantly, the regime has no effect on any distributor that is not a licensed subscription broadcaster.

That second point became the central weakness of the Australian rules for years. The legislation was written around subscription television licences at a time when there was no other way to charge people for pictures. Streaming services that were not licensed subscription broadcasters sat outside the regime's reach, which meant a listed event could in principle move behind a paywall the rules had never contemplated. The regime has since been the subject of sustained legislative attention aimed at bringing online services within scope, and the general lesson travels well beyond Australia: a rule that names a technology rather than a behaviour will be obsolete within about a decade of being written.

For readers in both countries the practical point is the same. A listed or anti-siphoned event is one where the sport has been told it cannot capture the full market value of what it owns, and the amount being given up rises every time pay television and streaming grow relative to free-to-air.

Why the United States has no equivalent, and what it has instead

American readers looking for the equivalent list will not find one, and the reasons are structural rather than accidental.

Content mandates of this kind sit awkwardly with a constitutional tradition that is deeply reluctant to have government direct what private broadcasters must carry. The public interest obligations that do attach to American broadcast licences concern spectrum, localism and access. They have never been read as a power to stop particular sporting events migrating to pay platforms.

The intervention the United States actually made was of a completely different kind. Rather than protecting free-to-air access by restricting sales, Congress protected pooled selling by granting an exemption from antitrust law. The 1961 statute permitted the major leagues to sell their telecast rights collectively, something that would otherwise have been an obvious agreement among competitors, and it attached conditions of its own concerning the protection of local attendance. The bargain struck was between the leagues and the antitrust laws, not between the leagues and the viewing public.

For decades the outcome resembled a listed events regime anyway, because the buyers were the free-to-air networks. National broadcast television was where the money and the audience were, so that is where the sport went, and no statute was needed to produce the result. As cable matured and then streaming arrived, the sport followed the money onto platforms that require payment, and there was no statutory backstop to slow it down. The migration happened event by event and sport by sport, with no single moment at which anybody had to decide whether it should happen at all.

Whether that is better or worse depends entirely on what you think broadcasting rules are for. The American approach maximises the revenue flowing into the sport and leaves access to the market. The British and Australian approach caps the revenue on a small number of designated occasions and guarantees access to those. Both are internally coherent. Neither is a compromise the other side would accept.

What changed when the buyer became a streamer

A traditional broadcaster and a streaming service look like the same customer. They are not, and the difference has quietly rewritten what a package is worth.

A broadcaster funded by advertising is buying an audience. Its currency is reach: how many people watch, in what demographic, for how long, and what an advertiser will pay per thousand of them. Live sport is the last reliable source of a large simultaneous audience that cannot be skipped, which is why it commands a premium no drama series can match. Under that model the most valuable package is the one containing the matches with the largest audiences, and a mid-week fixture between two mid-table clubs is close to worthless.

A subscription streamer is buying something else entirely. Its business is measured in acquisitions, and far more importantly in churn: the proportion of subscribers cancelling each month. Sport is unusually good at suppressing churn, because a season is a calendar rather than a library. Eight or nine months of weekly appointments means a subscriber waiting for next weekend does not cancel this weekend.

Once the metric is visible, the behaviour makes sense. A streamer will pay for volume, including matches a broadcaster would consider unattractive, because a full season of something is worth more to a retention model than eight rounds of something excellent. It will value continuity across the whole competition above owning the marquee fixture. It will pay a large premium for exclusivity, because there is no advertising market to share and a subscriber who can watch the same match elsewhere is not a subscriber. And it will bid differently in year four of a term than it did in year one, because by then it knows precisely which subscribers arrived for the sport and how many of them stayed.

That last point is the most under-appreciated change in the whole market. A traditional broadcaster bid on panel-based audience estimates that told it, imperfectly, how many people watched. A streamer bids on its own server logs, which tell it who signed up on the day the fixtures were announced, what else they watched, when they stopped watching, and what they cancelled after. It knows the marginal value of the package to its own business with a confidence no broadcaster ever had. Better information does not automatically mean higher bids. In several cases it has meant lower and considerably more disciplined ones.

There are costs to the shift that competitions are still working through. Streaming services do not have universal reach, so a competition selling everything to one loses its shop window and, with it, the casual viewer who becomes next decade's paying subscriber. Production expectations rise, because subscribers expect multiple camera feeds, alternative commentary and integrated data. Global packages tempt a competition into signing away every territory at once for the sake of a single clean negotiation. And the technical demands of live streaming at scale are unforgiving in a way broadcast never was: a satellite feed does not fall over because too many people wanted to watch the same moment.

Piracy, and why enforcement is priced rather than solved

Every rights buyer knows that the product it has just bought exclusivity in is available for nothing, at reasonable quality, within a minute of kick-off. This is not a secret and it is not treated as one during negotiations.

The economics are brutally simple. The marginal cost of restreaming a match is close to zero. The pirate pays no rights fee, carries no production cost, and has no obligations to anybody. The legitimate rights holder therefore cannot compete on price and must compete on reliability, quality, latency and convenience, which is a real advantage but a narrow one.

Enforcement in the more sophisticated jurisdictions has stopped trying to remove illegal streams permanently and now targets the only window that matters. Courts in several countries grant blocking orders operating dynamically during live events, requiring internet providers to block server addresses identified in real time while the match is on, with the block lifting afterwards. The design concedes the point openly. The aim is not eradication, it is to make the illegal option unreliable during the ninety minutes when reliability is exactly what the viewer wants. A stream that dies at half time sends some proportion of its audience to a paid service, and that proportion is the entire return on the enforcement budget.

Two things follow that rarely get stated plainly. Piracy caps rights values, because sophisticated bidders discount their bids by the leakage they expect, and the discount is larger in markets where enforcement is weak or where the legal service is priced beyond what most households will pay. And enforcement is a negotiated contract term rather than a public good: the licence specifies who funds monitoring and litigation, what the competition guarantees about protecting the exclusivity it has sold, and occasionally what happens to the fee if it fails to. A buyer complaining publicly about piracy is usually positioning for the next renewal.

The uncomfortable corollary is that fragmentation makes it worse. A viewer needing four subscriptions to follow one competition has a strong incentive to find a single illegal source carrying all of it. The package splitting competition authorities require in order to protect consumers has a cost that shows up in exactly this way, and the two policy goals have never been properly reconciled by anybody.

When the value falls, the ratchet only turns one way

Rights values do not only rise. Individual packages have sold for less than the previous cycle in several sports and several countries, and the pattern of what follows is consistent enough to describe in advance.

The problem is asymmetry. On the way up, revenue arrives before costs adjust, and clubs enjoy a season or two of unusual comfort before wages catch up. On the way down, costs are contracted and revenue is not. Player contracts run for years and are frequently guaranteed. Stadium debt is amortised over decades and does not care what the broadcast market did this cycle, which is why the way grounds are financed can turn a revenue dip into an existential problem. Staff, infrastructure and academies are all fixed in the short run. A competition whose rights income falls ten per cent does not have a ten per cent problem. It has a problem concentrated entirely in whatever part of its cost base it can still cut, which is usually the part it least wants to touch.

Competitions have a standard toolkit for a soft market, and every tool in it costs something.

They shorten the term, so the market is retested sooner, at the price of giving buyers less certainty and therefore accepting a lower price now. They add packages to widen the field, at the price of fragmenting the audience further. They take production in-house to make the packages cheaper to buy, at the price of carrying a large operating cost themselves. They accept hybrid deals with a lower minimum guarantee plus a share of the buyer's revenue, converting a fixed income into a variable one and taking back risk they used to sell. Or they launch their own direct-to-consumer service and become the distributor, which keeps the whole margin and requires them to fund marketing, technology and customer service out of money they no longer receive up front.

The reason the ratchet effectively runs one way is that revenue-linked cost controls exist in some sports and not in others. Where spending limits are pegged to revenue, as under the NFL's cap or under the Premier League's profitability and sustainability rules, a fall in income mechanically lowers what clubs are permitted to spend, and the system self-corrects with a lag. Where no such link exists, clubs carry on spending against an income that has already gone, and the correction arrives as insolvency rather than as a rule.

The clubs that fail first are never the biggest. They are the ones whose entire business model assumed the distribution would keep growing, which describes most of the middle of most divisions.

How sports TV rights work in practice: reading the announcement

Most of what gets published on the day of a rights announcement is unusable, because the headline number has been engineered to be quoted rather than compared. Six questions turn it back into information.

Is the figure per season, or for the whole term? A four-year deal reported as a single total will always sound larger than the three-year deal it replaced. Divide by the years before comparing anything to anything.

Is it domestic, international, or both combined? Competitions increasingly report a combined figure, which is fine until you try to set it against a previous cycle that reported the two separately.

What is actually in the packages, and has it changed? More live matches for more money is not growth. A rise in the number of matches sold, a change in pick order, or the addition of clip and archive rights can move a total substantially with no change in the price of anything.

Who produces the pictures? If production has moved from the buyer to the competition, the competition has taken on a large cost and the gross number is no longer comparable with the previous one. If it has moved the other way, the reverse applies.

Is it a guarantee, or a guarantee plus a share? A minimum guarantee with revenue sharing above it is a different instrument from a fixed fee, and only the guaranteed part is money a competition can plan around.

And who is missing from the announcement? A tender that ended with one buyer holding everything, or with a package quietly withdrawn and re-offered months later, tells you more about the state of the market than any figure in the press release does.

Ask those six and the annual ritual becomes readable. The number is a headline. The packaging is the strategy. The distribution formula is the sport's real constitution, written in money rather than in rules, and it decides far more about what a competition will look like in ten years than anything that happens on the pitch.

More explanations of how the money moves, across every sport covered here, are collected in the multi-sport archive.