The ‘Free to Play’ Model for NBA Top Shot’s FastBreak Gaming

How changing one separate system can now make the game good for the collector economy

When NBA Top Shot released their game FastBreak, it was a breath of fresh air for users: finally we’d get to use our moments in a way we wanted, instead of “collect the random player who hits the last free throw of the third quarter in each game for a reward.” Finally we were being rewarded for holding good players instead of random ones (“it’s a bottleneck!”), and the predictive nature would prevent overpaying during price runups due to forced demand like with Flash Challenges. Users were happily posting free marketing on X/Twitter about their lineups and showing off their reward pack hits…

So if the game was so successful, why didn’t collectors see the values of moments we held go up since the start of the year? Why does the Market Cap continues to fall? And why did we see several high profile layoffs at Dapper due to waning revenue from pack demand and marketplace transactions?

On first glance it may seem odd that unlike Dapper’s NFL All Day product, the game was not built to consume floor moments for a shot at something better in packs:

It may also seem odd that as a collectibles platform there was no systematic advantage to collecting more of the players you were playing the game with, or holding more collectible common debuts over the cheapest floor common.

Both of these design decisions, since packs were being issued as rewards, made the game both a demand-limiter and dilutive by nature. Fun game maybe, but not great economic design. Shockingly though, I’ve come around to realizing these were not oversights; those were good and logical decisions on Dapper’s part within the system they had to operate. So what were those reasons not to originally incorporate burning or holding value?:

1. Undermining the new player-economy
2. The user distaste for TSS as a way of rewarding value

Part of what I assume Dapper liked about the FastBreak game design is that finally gaming utility would tie market prices for star players to a more logical one vs prices for bench players. So that logic holds that 1) if the game is requiring people to hold only one moment at their chosen tier of good players, while burning the worst performing players as the floor and 2) it is issuing more moments of both as the reward… then the inevitable conclusion is that this only helps churn the economy of non-star players, and star-players see their supply overrun demand. Okay, so no burning the floor.

How about the opposite then? Where users would burn the moments they’ve used as game pieces for a pack reward if we win. This would be burning star players good for the game for a pack that probably contains worse players. No one is going to do that, and that value proposition will drive zero new users to the site.

So why not tie game scoring to TSS? It seemed like such an obvious solution to the complaints of “give debuts more utility” and “give S1s more utility” and “give stacks more utility.” Without knowing it, all those users were complaining about the same thing: the game is detached from, and undermines holding, value. And Top Shot has an integrated and automated system to measure and reward value: TSS. So why wasn’t that used for FastBreak tie breaker scoring?

Because no one likes to play a game they know they can’t win.

And TSS, promising to users “your score reflects what you paid and can never go down,” has created several unintended consequences. For example:

I have been holding onto this Giannis moment for two years because it has lost so much value in the market. That gives me outsized utility versus any new spend. My $629.40 worth of TSS means me holding or burning that moment is worth the same as someone today buying or burning more than seventeen of this moment! A new user may say “So I can buy 16 of these to burn, and spend $595, and he’s going to win by burning ONE $35 moment?!” With TSS as the measure of utility… yes. So why would they join a platform like that, just to constantly feel like they’ll never get ahead?

Likewise, why am I as an existing user going to buy new moments that don’t carry this utility per spend? I’m not. I’m systematically rewarded instead to not spend any new money, waiting to extract my reward and suck liquidity out of the economy.

Add on to that, the issue that 2020 OGs didn’t even need to pay the prices I did for their TSS because their score was grandfathered in at the highest average sale price at the time of the May 2022 snapshot- not the cheap market prices during closed beta- and you have a system that makes the majority of users feel even more like it is not fair for them.

This is why so many users hate the idea of more things being rewarded specifically through TSS.

Top Shot decision makers are not stupid, I’m assuming they see this problem. So instead, the game gambled toward the logic that if prices of common moments were low enough, then a brand new user could start a lineup with $5 and this would draw the masses. It didn’t want to scare off users by making them compete using a system where they could never get ahead. So the game then limits floor demand and collectible demand on purpose to keep its entry prices low, while introducing new supply that dilutes the price of holding those game pieces further.

The miscalculation Dapper made in this design was that people would come to the platform for cheap fun with cheap moments (“accessible onramp for new users” in business speak), neglecting the fact that the masses that came to the platform in 2021 did so because they thought holding moments could be valuable and would go up over time but left when it appeared that wasn’t true. Instead with TSS and the game itself limiting demand for new moments, the rewards for playing that game and winning became unsellable and thus worthless commons and are trending in the wrong direction even for the lucky few who pull rares+.

The Fix

So how can the popularity of FastBreak be retained without the slow damage to the economy that it causes? How can it reward existing users without alienating potential new ones to spend money here, and that those existing users are hoping for to drive increasing demand value for moments they hold?

We need to fix the value measurement system so that it can be used without fear of its consequences. This will also then let NFL All Day decision makers feel comfortable expanding this TSS-style (but not TSS) measurement system to it’s own FastBreak style game, team leaderboards, and player leaderboards.

Rolling Median Sale Price (ASP) in the player held is the new TSS.

I know I know, another new scoring model? Baller Score → Collector Score → Top Shot Score → Now Median Sale Price? Rugged again, right? Wrong. If the current model we have is turning away new users, and it’s not encouraging new spend… does it really matter if you keep all your TSS points forever if those two never improve? I’ll show how switching to rolling Median Sale Price vs original spend will improve both.

In the earlier customer point of friction the hypothetical new user said “So I can buy 16 of these to burn, and spend $595, and they’re going to win by burning ONE $35 moment?!” That’s the giveaway: $35 is the value I am holding. That is the value I’d be locking, and that is the value I’d be burning. But how exactly do you quantify value?

Let’s play out those earlier points of user friction, but now using rolling median value held instead of TSS:
1) Priority queues or hard gating for higher-tier pack access: If a 2020 OG, a user who bought in at the 2021 market peak, and a 2024 new user all hold one Joker TSD, they all hold the same median sale value ($172.02 today via otmnft.com).
-They should be on a level playing field for the utility of access to higher level packs for proving they can own moments vs just flipping pack contents. Because if they chose not to exercise that utility, they could all three sell that Joker TSD for the same $172 value instead.
2) Burning: If a 2020 OG, a 2021 market peak user, and a 2024 new user all burn one Joker TSD for a burn leaderboard they are all contributing equally to the same value reduction of Joker’s existing supply: one less Joker debut valued at $172.02 in the marketplace.
-They should be on a level playing field for the reward, with ties broken only by who placed it for burning first. Because if they chose not to exercise that utility, they could all three sell that Joker TSD for the same $172 value instead.
3) Locking: If a 2020 OG, a 2021 market peak user, and a 2024 new user all lock one Joker TSD today they are all contributing equally to the temporary removal of existing Joker moment supply: one less Joker debut valued at $172.02 in the marketplace.
-They should be on a level playing field on his leaderboard, tie broken only by who locked it first. Because if they chose not to exercise that utility, they could all three sell that Joker TSD for the same $172 value instead.

Now the last one: If FastBreak gives a bonus to value you hold in a player so the game were to become economically sustainable- and a 2020 OG, a 2021 market peak user, and a 2024 new user all hold one Joker TSD they are playing- they should be on a level playing field for FastBreak score because they all hold the same true value of Joker moments.

I think Dapper sees this and it’s the reason that when NFL All Day finally got burn leaderboards, the ranking was tied to ASP- not spend amount:

“But you said Median Sale Price above. Why not Average Sale Price? Everybody knows what average means.”

Not so fast. All Day does it average all-time sale price. So if one type of moment used to be valued a lot higher, like Series 1 debuts, but now the market looks at those as commons out of /10,000 and values them just as highly as a Series 2023 uncommon out of /1500… tell me which user is burning more in market value to be rewarded and how their service to the economy should be measured?

Average Sale Price all-time says user A burning an Elijah Mitchell is burning $31 in value but user B burning a C.J. Stroud is burning only $9 in value. We are not doing any great service by patting ourselves on the back for burning a bunch of Mitchells, and encouraging people to stockpile the moments that have lost the most value if they want utility.

OTM gets it right. They want to show “what is this moment actually selling for today?” which is what should matter to All Day too, so they calculate the Mitchell as an average sale price of $2.04 and the Stroud as an average sale price of $13.82 which is much closer to its actual current market value. The key difference is the rolling period:

Top Shot uses sales volume to bracket the rolling period of time it will count average sales for on its site to calculate a current ASP, but NFL All Day does not. And so All Day over-incentivizes burning things that have lost value over things that currently have value, just in a different way that TSS does for Top Shot.

“Okay, so rolling Average Sales Price it is then!”

Hahaa not so fast. This isn’t without its problems and possibilities for manipulation too.

Here’s a hypothetical: burn any Jayson Tatum moments to earn a new Tatum moment… what is the value of this moment burned vs his others?

This moment has the volume to put it into Top Shot’s bracket of ‘last 10 transactions’ average sale price. But since one user paid $350 for serial #75 within those last 10 purchases, everyone who holds this moment is now showing it’s worth twice the low ask price of $24 and almost 4x the median sale price of $16. The correct answer is option #3- Median Sale Price over the rolling period. Anything else allows users of low liquidity moments to make one large transaction and gain outsized utility. It is just not possible to manipulate median sale price that way at enough scale. It can even still be called TSS, but that calculation running underneath it needs to change.

So TSS needs to be changed to measure Rolling Median Sale Price of moments held because that is the true value of the moments the user is holding. It also needs to have this measure used as the tiebreaker for FastBreak scores:

In this scenario, maybe I’m a new user and I nailed it and Jonas Valanciunas and Nick Claxton go off for more rebounds and blocks than any other players that night. I could then get away with owning just one $1 common of each player. But if the more chalky picks go off as we expect them to- for that I’m going to want to have collected more than the minimum to win, like holding Wemby’s Top Shot debut next year instead of his cheaper 8k common when that comes out. Because the way the game is currently constructed, this utility incentive when minting new moments undermines the long-term collector economy the game is supposed to support.

By making this change to do the short-term unpopular thing of changing the utility measure system again, and now tying it to rolling median sale price of moments held, we can finally have a game that:

1. Still does not require anyone to spend more than $5 to set up a lineup and start playing (and win).
2. Lets a savvy new user play a lineup most people are not using, and possibly beat the field (the top scoring lineup wins, total value held is only the tiebreaker). As a bonus this removes the incentive to ‘copy off my homework’ and incentivizes making more interesting lineup plays.
3. Doesn’t force users to burn anything, since burning stars for bench players has no user value proposition and burning bench players to earn stars does little economic service.
4. Subtly encourages users to spend a little more on quality moments from players they think they’ll be using often,
   -Which stops the game from capping demand when there are more available moments of star players than gamers needing them
   -And prevents the lack of burning from being dilutive
5. Provides logical and automated utility to things the market already support, like debut moments or ceiling moments at each tier or holding stacks of a moment, without having the game diminish the value of holding them.
6. Retains the tier-based utility for number of plays, while adding a bit of game theory: a user who holds a lot of value in common debuts is going to tiebreak to a higher place for that player’s score, but they still only get one chance to use that player where a user holding cheaper rares will get to try using that player again.

**Bonus points if the UI shows the image of the player’s highest ASP-valued moment you hold in the game screen, rather than just a SoRare-style stock image completely detached from moment collectibles**

This essentially makes FastBreak now follow a proven successful business model: Free-to-Play Gaming (well, cheap to play anyway). Those models entice users with a great experience and value proposition for very little upfront cost: the path to success in the game is there for cheap if you’re good at choosing the players who will have outsized performances that other users won’t expect. Then slowly grinding out common pack rewards for profit or to build up a bigger collection for the points multiplier.

Only in time as the user gets more serious about the game, does a free-to-play game prompt them with things like “Sure, you can grind out that new armor or tool. But you can also buy 50 gems if you want to speed up the clock.” You can’t keep a game free with no additional revenue streams if it takes developer hours to build and run. And likewise we can’t have a functioning economy here with a game that does not create any additional incentive for collecting more, just now with a measurement system that has had its inconsistencies ironed out and now works to reward and encourage new collecting from OGs, peak users, and new users alike.

Integrating The Game Into The NFL Schedule

NFL All Day does not yet have a game similar to NBA Top Shot’s FastBreak. Some of that I chalk up to the fact that they saw the dilution issues FastBreak created but also saw the ones TSS created. So without having worked out a scoring system that fixes those issues, they were holding off.

But it’s also that the NFL doesn’t work like the NBA, there aren’t games going on every day. And on days like Monday Night where there is only one matchup, you obviously can’t have requirements like “set a lineup of 5 players who will throw for the most touchdowns.” There will be two players max and it won’t be tough to pick which ones.

But using the FastBreak style system for a different stat each week, like “This week set a lineup of the [wide receivers] who will score the most using 10 points per yard and 6 points per touchdown” creates too many issues; we’d either lose the utility of repeat uses for holding moments at higher tiers (yearly utility? So no reason to hold a common all year after you’ve used it?) or it would need to use the tier as the multiplier and again we’d lose the tie to ASP value for users who have high value commons or stacks of them.

Assuming the NFL doesn’t have exclusive licenses with all of the fantasy football platforms out there, that’s the model Dapper should go with for NFL All Day. For example: “Set a lineup of players who will score the most fantasy points this week using one QB, two RBs, two WRs, one TE, one DL, one LB, one secondary, one kicker. 4 week run. Can use commons one week during that run, rares two weeks, legendaries all four weeks. Highest scores win packs, tiebreaker by rolling median sale price of moments held in the player used.”

Integrating Into The Sports NFT 2.0 model

So how do FastBreak pack rewards work with the Sports NFT 2.0 Model if there are no longer tons of moments users haven’t asked for being minted as filler to chase the one thing worth a lot more than the pack price?

The same way that burn leaderboards work by holding back 10% of minted supply to power them: weekly packs are held back as rewards for FastBreak during that week. The Sports NFT 2.0 Model ensures these moments have actual user demand and allows greater speed to market in them being released, making these packs much better rewards for playing the game than gamble packs mostly filled with moments that don’t have demand:

The number of people earning packs per night should be set to a good enough level where the game is engaging enough people, but not handing out packs easily to everyone who tries. Maybe that’s #1 highest score per night gets a legendary pack (7 per week’s release cycle), with #2–4 earning rare packs (21 per week), and #5-10 earning common packs (42 per week). Maybe that’s not enough people being rewarded and those spots should be 1–2 | 3-10 | 11–30 each night. Whatever those optimal levels are, it’s flexible.

But no more rewarding users for doing the bare minimum to get over the threshold, diluting everyone who plays and those who don’t. This is the way the game stays fun, and at whatever level a user is comfortable committing, without slowly diluting the economy for both players and non-players alike.

With this approach every single part of the Sports NFT 2.0 Model supports the other modes of play, instead of undermining them:
-Vote to Mint: current and popular high-performing players likely to be ones voted in for mints, making series supply likely the one highly demanded for FastBreak
-Burn to Earn: Popular players voted in have their supply burned down more often, opening up burns to commons reduces their available supply making ones held for FastBreak more valuable over time
-Locking: Accessible leaderboard rewards users can count on for every player prevents Dapper from needing to mint lots no one wants to fulfill the license requirements for gamble packs full of stuff no one is asking for. This gives the moments in packs real demand and prevents FastBreak prizes earned from mostly being stuff no one wanted. Tiered rewards of higher placement on these leaderboards giving more scarce parallels drives greater than minimum locking behavior, but keeping those “Common” prevents handouts to the top that drastically effect new users’ ability to compete in FastBreak.
-FastBreak changed to exclusively competition with scores tiebroken by current market value held in the player played drives further demand for moments of good active players, removing dilution fears from repeat mintings.

Closing

Cheap game pieces for easy rewards did not bring the masses. Falling values drove away users. Rising values is the only thing that will cause them to return. It can’t just be 5,000 active users collecting the minimum $1 moments of players and waiting for packs of unsellable $1 moments in return or to dilute each other into oblivion. In an ecosystem where 60k commons exist as available game pieces, even if the game became so popular that we 10x the active user base it would still make no material impact on the economy.

Success in the game needs to be tied to value so increased demand drives moment values for all users. The tie to value measured just needs to be corrected, updating TSS to measure rolling median sale price of moments held so its’ previous fundamental unfairness doesn’t turn away new users and new spend knowing it can never catch up.

Users threatening to leave Top Shot if Dapper takes away that unfair utility driving away new users and new spend is terrorism; it is threatening the entire health and growth opportunity of the product to try to protect that user’s personal gain. Don’t be swayed by the argument that it is about consistency, it is an argument driven by that user’s expectation to receive far more in utility than how much the market (all of us) is actually valuing their moments. The irony being that if existing users are the only ones here, we all have that overspend and each user is not getting ahead of the pack by retaining the TSS system- merely cutting off our nose to spite our face. And if existing users are all being over-rewarded to hold TSS in old moments that have lost value instead of buying new, where is the demand coming from to buy whatever new moment that user flips their inflated TSS for?